diff options
Diffstat (limited to 'fs')
361 files changed, 11066 insertions, 7438 deletions
diff --git a/fs/9p/acl.c b/fs/9p/acl.c index 02a2cf6..5154552 100644 --- a/fs/9p/acl.c +++ b/fs/9p/acl.c @@ -21,8 +21,8 @@ #include <linux/posix_acl_xattr.h> #include "xattr.h" #include "acl.h" -#include "v9fs_vfs.h" #include "v9fs.h" +#include "v9fs_vfs.h" static struct posix_acl *__v9fs_get_acl(struct p9_fid *fid, char *name) { @@ -59,7 +59,8 @@ int v9fs_get_acl(struct inode *inode, struct p9_fid *fid) struct v9fs_session_info *v9ses; v9ses = v9fs_inode2v9ses(inode); - if ((v9ses->flags & V9FS_ACCESS_MASK) != V9FS_ACCESS_CLIENT) { + if (((v9ses->flags & V9FS_ACCESS_MASK) != V9FS_ACCESS_CLIENT) || + ((v9ses->flags & V9FS_ACL_MASK) != V9FS_POSIX_ACL)) { set_cached_acl(inode, ACL_TYPE_DEFAULT, NULL); set_cached_acl(inode, ACL_TYPE_ACCESS, NULL); return 0; @@ -71,11 +72,15 @@ int v9fs_get_acl(struct inode *inode, struct p9_fid *fid) if (!IS_ERR(dacl) && !IS_ERR(pacl)) { set_cached_acl(inode, ACL_TYPE_DEFAULT, dacl); set_cached_acl(inode, ACL_TYPE_ACCESS, pacl); - posix_acl_release(dacl); - posix_acl_release(pacl); } else retval = -EIO; + if (!IS_ERR(dacl)) + posix_acl_release(dacl); + + if (!IS_ERR(pacl)) + posix_acl_release(pacl); + return retval; } @@ -100,9 +105,10 @@ int v9fs_check_acl(struct inode *inode, int mask, unsigned int flags) return -ECHILD; v9ses = v9fs_inode2v9ses(inode); - if ((v9ses->flags & V9FS_ACCESS_MASK) != V9FS_ACCESS_CLIENT) { + if (((v9ses->flags & V9FS_ACCESS_MASK) != V9FS_ACCESS_CLIENT) || + ((v9ses->flags & V9FS_ACL_MASK) != V9FS_POSIX_ACL)) { /* - * On access = client mode get the acl + * On access = client and acl = on mode get the acl * values from the server */ return 0; @@ -128,6 +134,10 @@ static int v9fs_set_acl(struct dentry *dentry, int type, struct posix_acl *acl) struct inode *inode = dentry->d_inode; set_cached_acl(inode, type, acl); + + if (!acl) + return 0; + /* Set a setxattr request to server */ size = posix_acl_xattr_size(acl->a_count); buffer = kmalloc(size, GFP_KERNEL); @@ -177,10 +187,8 @@ int v9fs_acl_chmod(struct dentry *dentry) int v9fs_set_create_acl(struct dentry *dentry, struct posix_acl *dpacl, struct posix_acl *pacl) { - if (dpacl) - v9fs_set_acl(dentry, ACL_TYPE_DEFAULT, dpacl); - if (pacl) - v9fs_set_acl(dentry, ACL_TYPE_ACCESS, pacl); + v9fs_set_acl(dentry, ACL_TYPE_DEFAULT, dpacl); + v9fs_set_acl(dentry, ACL_TYPE_ACCESS, pacl); posix_acl_release(dpacl); posix_acl_release(pacl); return 0; diff --git a/fs/9p/cache.c b/fs/9p/cache.c index 0dbe0d1..5b335c5 100644 --- a/fs/9p/cache.c +++ b/fs/9p/cache.c @@ -33,67 +33,11 @@ #define CACHETAG_LEN 11 -struct kmem_cache *vcookie_cache; - struct fscache_netfs v9fs_cache_netfs = { .name = "9p", .version = 0, }; -static void init_once(void *foo) -{ - struct v9fs_cookie *vcookie = (struct v9fs_cookie *) foo; - vcookie->fscache = NULL; - vcookie->qid = NULL; - inode_init_once(&vcookie->inode); -} - -/** - * v9fs_init_vcookiecache - initialize a cache for vcookies to maintain - * vcookie to inode mapping - * - * Returns 0 on success. - */ - -static int v9fs_init_vcookiecache(void) -{ - vcookie_cache = kmem_cache_create("vcookie_cache", - sizeof(struct v9fs_cookie), - 0, (SLAB_RECLAIM_ACCOUNT| - SLAB_MEM_SPREAD), - init_once); - if (!vcookie_cache) - return -ENOMEM; - - return 0; -} - -/** - * v9fs_destroy_vcookiecache - destroy the cache of vcookies - * - */ - -static void v9fs_destroy_vcookiecache(void) -{ - kmem_cache_destroy(vcookie_cache); -} - -int __v9fs_cache_register(void) -{ - int ret; - ret = v9fs_init_vcookiecache(); - if (ret < 0) - return ret; - - return fscache_register_netfs(&v9fs_cache_netfs); -} - -void __v9fs_cache_unregister(void) -{ - v9fs_destroy_vcookiecache(); - fscache_unregister_netfs(&v9fs_cache_netfs); -} - /** * v9fs_random_cachetag - Generate a random tag to be associated * with a new cache session. @@ -133,9 +77,9 @@ static uint16_t v9fs_cache_session_get_key(const void *cookie_netfs_data, } const struct fscache_cookie_def v9fs_cache_session_index_def = { - .name = "9P.session", - .type = FSCACHE_COOKIE_TYPE_INDEX, - .get_key = v9fs_cache_session_get_key, + .name = "9P.session", + .type = FSCACHE_COOKIE_TYPE_INDEX, + .get_key = v9fs_cache_session_get_key, }; void v9fs_cache_session_get_cookie(struct v9fs_session_info *v9ses) @@ -163,33 +107,33 @@ void v9fs_cache_session_put_cookie(struct v9fs_session_info *v9ses) static uint16_t v9fs_cache_inode_get_key(const void *cookie_netfs_data, void *buffer, uint16_t bufmax) { - const struct v9fs_cookie *vcookie = cookie_netfs_data; - memcpy(buffer, &vcookie->qid->path, sizeof(vcookie->qid->path)); - - P9_DPRINTK(P9_DEBUG_FSC, "inode %p get key %llu", &vcookie->inode, - vcookie->qid->path); - return sizeof(vcookie->qid->path); + const struct v9fs_inode *v9inode = cookie_netfs_data; + memcpy(buffer, &v9inode->fscache_key->path, + sizeof(v9inode->fscache_key->path)); + P9_DPRINTK(P9_DEBUG_FSC, "inode %p get key %llu", &v9inode->vfs_inode, + v9inode->fscache_key->path); + return sizeof(v9inode->fscache_key->path); } static void v9fs_cache_inode_get_attr(const void *cookie_netfs_data, uint64_t *size) { - const struct v9fs_cookie *vcookie = cookie_netfs_data; - *size = i_size_read(&vcookie->inode); + const struct v9fs_inode *v9inode = cookie_netfs_data; + *size = i_size_read(&v9inode->vfs_inode); - P9_DPRINTK(P9_DEBUG_FSC, "inode %p get attr %llu", &vcookie->inode, + P9_DPRINTK(P9_DEBUG_FSC, "inode %p get attr %llu", &v9inode->vfs_inode, *size); } static uint16_t v9fs_cache_inode_get_aux(const void *cookie_netfs_data, void *buffer, uint16_t buflen) { - const struct v9fs_cookie *vcookie = cookie_netfs_data; - memcpy(buffer, &vcookie->qid->version, sizeof(vcookie->qid->version)); - - P9_DPRINTK(P9_DEBUG_FSC, "inode %p get aux %u", &vcookie->inode, - vcookie->qid->version); - return sizeof(vcookie->qid->version); + const struct v9fs_inode *v9inode = cookie_netfs_data; + memcpy(buffer, &v9inode->fscache_key->version, + sizeof(v9inode->fscache_key->version)); + P9_DPRINTK(P9_DEBUG_FSC, "inode %p get aux %u", &v9inode->vfs_inode, + v9inode->fscache_key->version); + return sizeof(v9inode->fscache_key->version); } static enum @@ -197,13 +141,13 @@ fscache_checkaux v9fs_cache_inode_check_aux(void *cookie_netfs_data, const void *buffer, uint16_t buflen) { - const struct v9fs_cookie *vcookie = cookie_netfs_data; + const struct v9fs_inode *v9inode = cookie_netfs_data; - if (buflen != sizeof(vcookie->qid->version)) + if (buflen != sizeof(v9inode->fscache_key->version)) return FSCACHE_CHECKAUX_OBSOLETE; - if (memcmp(buffer, &vcookie->qid->version, - sizeof(vcookie->qid->version))) + if (memcmp(buffer, &v9inode->fscache_key->version, + sizeof(v9inode->fscache_key->version))) return FSCACHE_CHECKAUX_OBSOLETE; return FSCACHE_CHECKAUX_OKAY; @@ -211,7 +155,7 @@ fscache_checkaux v9fs_cache_inode_check_aux(void *cookie_netfs_data, static void v9fs_cache_inode_now_uncached(void *cookie_netfs_data) { - struct v9fs_cookie *vcookie = cookie_netfs_data; + struct v9fs_inode *v9inode = cookie_netfs_data; struct pagevec pvec; pgoff_t first; int loop, nr_pages; @@ -220,7 +164,7 @@ static void v9fs_cache_inode_now_uncached(void *cookie_netfs_data) first = 0; for (;;) { - nr_pages = pagevec_lookup(&pvec, vcookie->inode.i_mapping, + nr_pages = pagevec_lookup(&pvec, v9inode->vfs_inode.i_mapping, first, PAGEVEC_SIZE - pagevec_count(&pvec)); if (!nr_pages) @@ -249,115 +193,114 @@ const struct fscache_cookie_def v9fs_cache_inode_index_def = { void v9fs_cache_inode_get_cookie(struct inode *inode) { - struct v9fs_cookie *vcookie; + struct v9fs_inode *v9inode; struct v9fs_session_info *v9ses; if (!S_ISREG(inode->i_mode)) return; - vcookie = v9fs_inode2cookie(inode); - if (vcookie->fscache) + v9inode = V9FS_I(inode); + if (v9inode->fscache) return; v9ses = v9fs_inode2v9ses(inode); - vcookie->fscache = fscache_acquire_cookie(v9ses->fscache, + v9inode->fscache = fscache_acquire_cookie(v9ses->fscache, &v9fs_cache_inode_index_def, - vcookie); + v9inode); P9_DPRINTK(P9_DEBUG_FSC, "inode %p get cookie %p", inode, - vcookie->fscache); + v9inode->fscache); } void v9fs_cache_inode_put_cookie(struct inode *inode) { - struct v9fs_cookie *vcookie = v9fs_inode2cookie(inode); + struct v9fs_inode *v9inode = V9FS_I(inode); - if (!vcookie->fscache) + if (!v9inode->fscache) return; P9_DPRINTK(P9_DEBUG_FSC, "inode %p put cookie %p", inode, - vcookie->fscache); + v9inode->fscache); - fscache_relinquish_cookie(vcookie->fscache, 0); - vcookie->fscache = NULL; + fscache_relinquish_cookie(v9inode->fscache, 0); + v9inode->fscache = NULL; } void v9fs_cache_inode_flush_cookie(struct inode *inode) { - struct v9fs_cookie *vcookie = v9fs_inode2cookie(inode); + struct v9fs_inode *v9inode = V9FS_I(inode); - if (!vcookie->fscache) + if (!v9inode->fscache) return; P9_DPRINTK(P9_DEBUG_FSC, "inode %p flush cookie %p", inode, - vcookie->fscache); + v9inode->fscache); - fscache_relinquish_cookie(vcookie->fscache, 1); - vcookie->fscache = NULL; + fscache_relinquish_cookie(v9inode->fscache, 1); + v9inode->fscache = NULL; } void v9fs_cache_inode_set_cookie(struct inode *inode, struct file *filp) { - struct v9fs_cookie *vcookie = v9fs_inode2cookie(inode); + struct v9fs_inode *v9inode = V9FS_I(inode); struct p9_fid *fid; - if (!vcookie->fscache) + if (!v9inode->fscache) return; - spin_lock(&vcookie->lock); + spin_lock(&v9inode->fscache_lock); fid = filp->private_data; if ((filp->f_flags & O_ACCMODE) != O_RDONLY) v9fs_cache_inode_flush_cookie(inode); else v9fs_cache_inode_get_cookie(inode); - spin_unlock(&vcookie->lock); + spin_unlock(&v9inode->fscache_lock); } void v9fs_cache_inode_reset_cookie(struct inode *inode) { - struct v9fs_cookie *vcookie = v9fs_inode2cookie(inode); + struct v9fs_inode *v9inode = V9FS_I(inode); struct v9fs_session_info *v9ses; struct fscache_cookie *old; - if (!vcookie->fscache) + if (!v9inode->fscache) return; - old = vcookie->fscache; + old = v9inode->fscache; - spin_lock(&vcookie->lock); - fscache_relinquish_cookie(vcookie->fscache, 1); + spin_lock(&v9inode->fscache_lock); + fscache_relinquish_cookie(v9inode->fscache, 1); v9ses = v9fs_inode2v9ses(inode); - vcookie->fscache = fscache_acquire_cookie(v9ses->fscache, + v9inode->fscache = fscache_acquire_cookie(v9ses->fscache, &v9fs_cache_inode_index_def, - vcookie); - + v9inode); P9_DPRINTK(P9_DEBUG_FSC, "inode %p revalidating cookie old %p new %p", - inode, old, vcookie->fscache); + inode, old, v9inode->fscache); - spin_unlock(&vcookie->lock); + spin_unlock(&v9inode->fscache_lock); } int __v9fs_fscache_release_page(struct page *page, gfp_t gfp) { struct inode *inode = page->mapping->host; - struct v9fs_cookie *vcookie = v9fs_inode2cookie(inode); + struct v9fs_inode *v9inode = V9FS_I(inode); - BUG_ON(!vcookie->fscache); + BUG_ON(!v9inode->fscache); - return fscache_maybe_release_page(vcookie->fscache, page, gfp); + return fscache_maybe_release_page(v9inode->fscache, page, gfp); } void __v9fs_fscache_invalidate_page(struct page *page) { struct inode *inode = page->mapping->host; - struct v9fs_cookie *vcookie = v9fs_inode2cookie(inode); + struct v9fs_inode *v9inode = V9FS_I(inode); - BUG_ON(!vcookie->fscache); + BUG_ON(!v9inode->fscache); if (PageFsCache(page)) { - fscache_wait_on_page_write(vcookie->fscache, page); + fscache_wait_on_page_write(v9inode->fscache, page); BUG_ON(!PageLocked(page)); - fscache_uncache_page(vcookie->fscache, page); + fscache_uncache_page(v9inode->fscache, page); } } @@ -380,13 +323,13 @@ static void v9fs_vfs_readpage_complete(struct page *page, void *data, int __v9fs_readpage_from_fscache(struct inode *inode, struct page *page) { int ret; - const struct v9fs_cookie *vcookie = v9fs_inode2cookie(inode); + const struct v9fs_inode *v9inode = V9FS_I(inode); P9_DPRINTK(P9_DEBUG_FSC, "inode %p page %p", inode, page); - if (!vcookie->fscache) + if (!v9inode->fscache) return -ENOBUFS; - ret = fscache_read_or_alloc_page(vcookie->fscache, + ret = fscache_read_or_alloc_page(v9inode->fscache, page, v9fs_vfs_readpage_complete, NULL, @@ -418,13 +361,13 @@ int __v9fs_readpages_from_fscache(struct inode *inode, unsigned *nr_pages) { int ret; - const struct v9fs_cookie *vcookie = v9fs_inode2cookie(inode); + const struct v9fs_inode *v9inode = V9FS_I(inode); P9_DPRINTK(P9_DEBUG_FSC, "inode %p pages %u", inode, *nr_pages); - if (!vcookie->fscache) + if (!v9inode->fscache) return -ENOBUFS; - ret = fscache_read_or_alloc_pages(vcookie->fscache, + ret = fscache_read_or_alloc_pages(v9inode->fscache, mapping, pages, nr_pages, v9fs_vfs_readpage_complete, NULL, @@ -453,11 +396,22 @@ int __v9fs_readpages_from_fscache(struct inode *inode, void __v9fs_readpage_to_fscache(struct inode *inode, struct page *page) { int ret; - const struct v9fs_cookie *vcookie = v9fs_inode2cookie(inode); + const struct v9fs_inode *v9inode = V9FS_I(inode); P9_DPRINTK(P9_DEBUG_FSC, "inode %p page %p", inode, page); - ret = fscache_write_page(vcookie->fscache, page, GFP_KERNEL); + ret = fscache_write_page(v9inode->fscache, page, GFP_KERNEL); P9_DPRINTK(P9_DEBUG_FSC, "ret = %d", ret); if (ret != 0) v9fs_uncache_page(inode, page); } + +/* + * wait for a page to complete writing to the cache + */ +void __v9fs_fscache_wait_on_page_write(struct inode *inode, struct page *page) +{ + const struct v9fs_inode *v9inode = V9FS_I(inode); + P9_DPRINTK(P9_DEBUG_FSC, "inode %p page %p", inode, page); + if (PageFsCache(page)) + fscache_wait_on_page_write(v9inode->fscache, page); +} diff --git a/fs/9p/cache.h b/fs/9p/cache.h index a94192b..049507a 100644 --- a/fs/9p/cache.h +++ b/fs/9p/cache.h @@ -25,20 +25,6 @@ #include <linux/fscache.h> #include <linux/spinlock.h> -extern struct kmem_cache *vcookie_cache; - -struct v9fs_cookie { - spinlock_t lock; - struct inode inode; - struct fscache_cookie *fscache; - struct p9_qid *qid; -}; - -static inline struct v9fs_cookie *v9fs_inode2cookie(const struct inode *inode) -{ - return container_of(inode, struct v9fs_cookie, inode); -} - extern struct fscache_netfs v9fs_cache_netfs; extern const struct fscache_cookie_def v9fs_cache_session_index_def; extern const struct fscache_cookie_def v9fs_cache_inode_index_def; @@ -64,23 +50,8 @@ extern int __v9fs_readpages_from_fscache(struct inode *inode, struct list_head *pages, unsigned *nr_pages); extern void __v9fs_readpage_to_fscache(struct inode *inode, struct page *page); - - -/** - * v9fs_cache_register - Register v9fs file system with the cache - */ -static inline int v9fs_cache_register(void) -{ - return __v9fs_cache_register(); -} - -/** - * v9fs_cache_unregister - Unregister v9fs from the cache - */ -static inline void v9fs_cache_unregister(void) -{ - __v9fs_cache_unregister(); -} +extern void __v9fs_fscache_wait_on_page_write(struct inode *inode, + struct page *page); static inline int v9fs_fscache_release_page(struct page *page, gfp_t gfp) @@ -117,28 +88,27 @@ static inline void v9fs_readpage_to_fscache(struct inode *inode, static inline void v9fs_uncache_page(struct inode *inode, struct page *page) { - struct v9fs_cookie *vcookie = v9fs_inode2cookie(inode); - fscache_uncache_page(vcookie->fscache, page); + struct v9fs_inode *v9inode = V9FS_I(inode); + fscache_uncache_page(v9inode->fscache, page); BUG_ON(PageFsCache(page)); } -static inline void v9fs_vcookie_set_qid(struct inode *inode, +static inline void v9fs_fscache_set_key(struct inode *inode, struct p9_qid *qid) { - struct v9fs_cookie *vcookie = v9fs_inode2cookie(inode); - spin_lock(&vcookie->lock); - vcookie->qid = qid; - spin_unlock(&vcookie->lock); + struct v9fs_inode *v9inode = V9FS_I(inode); + spin_lock(&v9inode->fscache_lock); + v9inode->fscache_key = qid; + spin_unlock(&v9inode->fscache_lock); } -#else /* CONFIG_9P_FSCACHE */ - -static inline int v9fs_cache_register(void) +static inline void v9fs_fscache_wait_on_page_write(struct inode *inode, + struct page *page) { - return 1; + return __v9fs_fscache_wait_on_page_write(inode, page); } -static inline void v9fs_cache_unregister(void) {} +#else /* CONFIG_9P_FSCACHE */ static inline int v9fs_fscache_release_page(struct page *page, gfp_t gfp) { @@ -168,9 +138,11 @@ static inline void v9fs_readpage_to_fscache(struct inode *inode, static inline void v9fs_uncache_page(struct inode *inode, struct page *page) {} -static inline void v9fs_vcookie_set_qid(struct inode *inode, - struct p9_qid *qid) -{} +static inline void v9fs_fscache_wait_on_page_write(struct inode *inode, + struct page *page) +{ + return; +} #endif /* CONFIG_9P_FSCACHE */ #endif /* _9P_CACHE_H */ diff --git a/fs/9p/fid.c b/fs/9p/fid.c index b00223c..cd63e00 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -125,46 +125,17 @@ err_out: return -ENOMEM; } -/** - * v9fs_fid_lookup - lookup for a fid, try to walk if not found - * @dentry: dentry to look for fid in - * - * Look for a fid in the specified dentry for the current user. - * If no fid is found, try to create one walking from a fid from the parent - * dentry (if it has one), or the root dentry. If the user haven't accessed - * the fs yet, attach now and walk from the root. - */ - -struct p9_fid *v9fs_fid_lookup(struct dentry *dentry) +static struct p9_fid *v9fs_fid_lookup_with_uid(struct dentry *dentry, + uid_t uid, int any) { - int i, n, l, clone, any, access; - u32 uid; - struct p9_fid *fid, *old_fid = NULL; struct dentry *ds; - struct v9fs_session_info *v9ses; char **wnames, *uname; + int i, n, l, clone, access; + struct v9fs_session_info *v9ses; + struct p9_fid *fid, *old_fid = NULL; v9ses = v9fs_inode2v9ses(dentry->d_inode); access = v9ses->flags & V9FS_ACCESS_MASK; - switch (access) { - case V9FS_ACCESS_SINGLE: - case V9FS_ACCESS_USER: - case V9FS_ACCESS_CLIENT: - uid = current_fsuid(); - any = 0; - break; - - case V9FS_ACCESS_ANY: - uid = v9ses->uid; - any = 1; - break; - - default: - uid = ~0; - any = 0; - break; - } - fid = v9fs_fid_find(dentry, uid, any); if (fid) return fid; @@ -250,6 +221,45 @@ err_out: return fid; } +/** + * v9fs_fid_lookup - lookup for a fid, try to walk if not found + * @dentry: dentry to look for fid in + * + * Look for a fid in the specified dentry for the current user. + * If no fid is found, try to create one walking from a fid from the parent + * dentry (if it has one), or the root dentry. If the user haven't accessed + * the fs yet, attach now and walk from the root. + */ + +struct p9_fid *v9fs_fid_lookup(struct dentry *dentry) +{ + uid_t uid; + int any, access; + struct v9fs_session_info *v9ses; + + v9ses = v9fs_inode2v9ses(dentry->d_inode); + access = v9ses->flags & V9FS_ACCESS_MASK; + switch (access) { + case V9FS_ACCESS_SINGLE: + case V9FS_ACCESS_USER: + case V9FS_ACCESS_CLIENT: + uid = current_fsuid(); + any = 0; + break; + + case V9FS_ACCESS_ANY: + uid = v9ses->uid; + any = 1; + break; + + default: + uid = ~0; + any = 0; + break; + } + return v9fs_fid_lookup_with_uid(dentry, uid, any); +} + struct p9_fid *v9fs_fid_clone(struct dentry *dentry) { struct p9_fid *fid, *ret; @@ -261,3 +271,39 @@ struct p9_fid *v9fs_fid_clone(struct dentry *dentry) ret = p9_client_walk(fid, 0, NULL, 1); return ret; } + +static struct p9_fid *v9fs_fid_clone_with_uid(struct dentry *dentry, uid_t uid) +{ + struct p9_fid *fid, *ret; + + fid = v9fs_fid_lookup_with_uid(dentry, uid, 0); + if (IS_ERR(fid)) + return fid; + + ret = p9_client_walk(fid, 0, NULL, 1); + return ret; +} + +struct p9_fid *v9fs_writeback_fid(struct dentry *dentry) +{ + int err; + struct p9_fid *fid; + + fid = v9fs_fid_clone_with_uid(dentry, 0); + if (IS_ERR(fid)) + goto error_out; + /* + * writeback fid will only be used to write back the + * dirty pages. We always request for the open fid in read-write + * mode so that a partial page write which result in page + * read can work. + */ + err = p9_client_open(fid, O_RDWR); + if (err < 0) { + p9_client_clunk(fid); + fid = ERR_PTR(err); + goto error_out; + } +error_out: + return fid; +} diff --git a/fs/9p/fid.h b/fs/9p/fid.h index c3bbd6a..bb0b6e7 100644 --- a/fs/9p/fid.h +++ b/fs/9p/fid.h @@ -19,7 +19,8 @@ * Boston, MA 02111-1301 USA * */ - +#ifndef FS_9P_FID_H +#define FS_9P_FID_H #include <linux/list.h> /** @@ -45,3 +46,5 @@ struct v9fs_dentry { struct p9_fid *v9fs_fid_lookup(struct dentry *dentry); struct p9_fid *v9fs_fid_clone(struct dentry *dentry); int v9fs_fid_add(struct dentry *dentry, struct p9_fid *fid); +struct p9_fid *v9fs_writeback_fid(struct dentry *dentry); +#endif diff --git a/fs/9p/v9fs.c b/fs/9p/v9fs.c index 2f77cd3..c82b017 100644 --- a/fs/9p/v9fs.c +++ b/fs/9p/v9fs.c @@ -39,6 +39,7 @@ static DEFINE_SPINLOCK(v9fs_sessionlist_lock); static LIST_HEAD(v9fs_sessionlist); +struct kmem_cache *v9fs_inode_cache; /* * Option Parsing (code inspired by NFS code) @@ -55,7 +56,7 @@ enum { /* Cache options */ Opt_cache_loose, Opt_fscache, /* Access options */ - Opt_access, + Opt_access, Opt_posixacl, /* Error token */ Opt_err }; @@ -73,6 +74,7 @@ static const match_table_t tokens = { {Opt_fscache, "fscache"}, {Opt_cachetag, "cachetag=%s"}, {Opt_access, "access=%s"}, + {Opt_posixacl, "posixacl"}, {Opt_err, NULL} }; @@ -194,15 +196,7 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts) else if (strcmp(s, "any") == 0) v9ses->flags |= V9FS_ACCESS_ANY; else if (strcmp(s, "client") == 0) { -#ifdef CONFIG_9P_FS_POSIX_ACL v9ses->flags |= V9FS_ACCESS_CLIENT; -#else - P9_DPRINTK(P9_DEBUG_ERROR, - "access=client option not supported\n"); - kfree(s); - ret = -EINVAL; - goto free_and_return; -#endif } else { v9ses->flags |= V9FS_ACCESS_SINGLE; v9ses->uid = simple_strtoul(s, &e, 10); @@ -212,6 +206,16 @@ static int v9fs_parse_options(struct v9fs_session_info *v9ses, char *opts) kfree(s); break; + case Opt_posixacl: +#ifdef CONFIG_9P_FS_POSIX_ACL + v9ses->flags |= V9FS_POSIX_ACL; +#else + P9_DPRINTK(P9_DEBUG_ERROR, + "Not defined CONFIG_9P_FS_POSIX_ACL. " + "Ignoring posixacl option\n"); +#endif + break; + default: continue; } @@ -260,19 +264,12 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, list_add(&v9ses->slist, &v9fs_sessionlist); spin_unlock(&v9fs_sessionlist_lock); - v9ses->flags = V9FS_ACCESS_USER; strcpy(v9ses->uname, V9FS_DEFUSER); strcpy(v9ses->aname, V9FS_DEFANAME); v9ses->uid = ~0; v9ses->dfltuid = V9FS_DEFUID; v9ses->dfltgid = V9FS_DEFGID; - rc = v9fs_parse_options(v9ses, data); - if (rc < 0) { - retval = rc; - goto error; - } - v9ses->clnt = p9_client_create(dev_name, data); if (IS_ERR(v9ses->clnt)) { retval = PTR_ERR(v9ses->clnt); @@ -281,10 +278,20 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, goto error; } - if (p9_is_proto_dotl(v9ses->clnt)) + v9ses->flags = V9FS_ACCESS_USER; + + if (p9_is_proto_dotl(v9ses->clnt)) { + v9ses->flags = V9FS_ACCESS_CLIENT; v9ses->flags |= V9FS_PROTO_2000L; - else if (p9_is_proto_dotu(v9ses->clnt)) + } else if (p9_is_proto_dotu(v9ses->clnt)) { v9ses->flags |= V9FS_PROTO_2000U; + } + + rc = v9fs_parse_options(v9ses, data); + if (rc < 0) { + retval = rc; + goto error; + } v9ses->maxdata = v9ses->clnt->msize - P9_IOHDRSZ; @@ -306,6 +313,14 @@ struct p9_fid *v9fs_session_init(struct v9fs_session_info *v9ses, v9ses->flags |= V9FS_ACCESS_ANY; v9ses->uid = ~0; } + if (!v9fs_proto_dotl(v9ses) || + !((v9ses->flags & V9FS_ACCESS_MASK) == V9FS_ACCESS_CLIENT)) { + /* + * We support ACL checks on clinet only if the protocol is + * 9P2000.L and access is V9FS_ACCESS_CLIENT. + */ + v9ses->flags &= ~V9FS_ACL_MASK; + } fid = p9_client_attach(v9ses->clnt, NULL, v9ses->uname, ~0, v9ses->aname); @@ -467,6 +482,63 @@ static void v9fs_sysfs_cleanup(void) kobject_put(v9fs_kobj); } +static void v9fs_inode_init_once(void *foo) +{ + struct v9fs_inode *v9inode = (struct v9fs_inode *)foo; +#ifdef CONFIG_9P_FSCACHE + v9inode->fscache = NULL; + v9inode->fscache_key = NULL; +#endif + inode_init_once(&v9inode->vfs_inode); +} + +/** + * v9fs_init_inode_cache - initialize a cache for 9P + * Returns 0 on success. + */ +static int v9fs_init_inode_cache(void) +{ + v9fs_inode_cache = kmem_cache_create("v9fs_inode_cache", + sizeof(struct v9fs_inode), + 0, (SLAB_RECLAIM_ACCOUNT| + SLAB_MEM_SPREAD), + v9fs_inode_init_once); + if (!v9fs_inode_cache) + return -ENOMEM; + + return 0; +} + +/** + * v9fs_destroy_inode_cache - destroy the cache of 9P inode + * + */ +static void v9fs_destroy_inode_cache(void) +{ + kmem_cache_destroy(v9fs_inode_cache); +} + +static int v9fs_cache_register(void) +{ + int ret; + ret = v9fs_init_inode_cache(); + if (ret < 0) + return ret; +#ifdef CONFIG_9P_FSCACHE + return fscache_register_netfs(&v9fs_cache_netfs); +#else + return ret; +#endif +} + +static void v9fs_cache_unregister(void) +{ + v9fs_destroy_inode_cache(); +#ifdef CONFIG_9P_FSCACHE + fscache_unregister_netfs(&v9fs_cache_netfs); +#endif +} + /** * init_v9fs - Initialize module * diff --git a/fs/9p/v9fs.h b/fs/9p/v9fs.h index c4b5d88..bd8496d 100644 --- a/fs/9p/v9fs.h +++ b/fs/9p/v9fs.h @@ -20,6 +20,9 @@ * Boston, MA 02111-1301 USA * */ +#ifndef FS_9P_V9FS_H +#define FS_9P_V9FS_H + #include <linux/backing-dev.h> /** @@ -28,8 +31,10 @@ * @V9FS_PROTO_2000L: whether or not to use 9P2000.l extensions * @V9FS_ACCESS_SINGLE: only the mounting user can access the hierarchy * @V9FS_ACCESS_USER: a new attach will be issued for every user (default) + * @V9FS_ACCESS_CLIENT: Just like user, but access check is performed on client. * @V9FS_ACCESS_ANY: use a single attach for all users * @V9FS_ACCESS_MASK: bit mask of different ACCESS options + * @V9FS_POSIX_ACL: POSIX ACLs are enforced * * Session flags reflect options selected by users at mount time */ @@ -37,13 +42,15 @@ V9FS_ACCESS_USER | \ V9FS_ACCESS_CLIENT) #define V9FS_ACCESS_MASK V9FS_ACCESS_ANY +#define V9FS_ACL_MASK V9FS_POSIX_ACL enum p9_session_flags { V9FS_PROTO_2000U = 0x01, V9FS_PROTO_2000L = 0x02, V9FS_ACCESS_SINGLE = 0x04, V9FS_ACCESS_USER = 0x08, - V9FS_ACCESS_CLIENT = 0x10 + V9FS_ACCESS_CLIENT = 0x10, + V9FS_POSIX_ACL = 0x20 }; /* possible values of ->cache */ @@ -109,8 +116,28 @@ struct v9fs_session_info { struct list_head slist; /* list of sessions registered with v9fs */ struct backing_dev_info bdi; struct rw_semaphore rename_sem; + struct p9_fid *root_fid; /* Used for file system sync */ +}; + +/* cache_validity flags */ +#define V9FS_INO_INVALID_ATTR 0x01 + +struct v9fs_inode { +#ifdef CONFIG_9P_FSCACHE + spinlock_t fscache_lock; + struct fscache_cookie *fscache; + struct p9_qid *fscache_key; +#endif + unsigned int cache_validity; + struct p9_fid *writeback_fid; + struct inode vfs_inode; }; +static inline struct v9fs_inode *V9FS_I(const struct inode *inode) +{ + return container_of(inode, struct v9fs_inode, vfs_inode); +} + struct p9_fid *v9fs_session_init(struct v9fs_session_info *, const char *, char *); extern void v9fs_session_close(struct v9fs_session_info *v9ses); @@ -124,16 +151,15 @@ extern int v9fs_vfs_rename(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry); extern void v9fs_vfs_put_link(struct dentry *dentry, struct nameidata *nd, void *p); -extern struct inode *v9fs_inode(struct v9fs_session_info *v9ses, - struct p9_fid *fid, - struct super_block *sb); - +extern struct inode *v9fs_inode_from_fid(struct v9fs_session_info *v9ses, + struct p9_fid *fid, + struct super_block *sb); extern const struct inode_operations v9fs_dir_inode_operations_dotl; extern const struct inode_operations v9fs_file_inode_operations_dotl; extern const struct inode_operations v9fs_symlink_inode_operations_dotl; -extern struct inode *v9fs_inode_dotl(struct v9fs_session_info *v9ses, - struct p9_fid *fid, - struct super_block *sb); +extern struct inode *v9fs_inode_from_fid_dotl(struct v9fs_session_info *v9ses, + struct p9_fid *fid, + struct super_block *sb); /* other default globals */ #define V9FS_PORT 564 @@ -158,7 +184,7 @@ static inline int v9fs_proto_dotl(struct v9fs_session_info *v9ses) } /** - * v9fs_inode_from_fid - Helper routine to populate an inode by + * v9fs_get_inode_from_fid - Helper routine to populate an inode by * issuing a attribute request * @v9ses: session information * @fid: fid to issue attribute request for @@ -166,11 +192,12 @@ static inline int v9fs_proto_dotl(struct v9fs_session_info *v9ses) * */ static inline struct inode * -v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid, - struct super_block *sb) +v9fs_get_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid, + struct super_block *sb) { if (v9fs_proto_dotl(v9ses)) - return v9fs_inode_dotl(v9ses, fid, sb); + return v9fs_inode_from_fid_dotl(v9ses, fid, sb); else - return v9fs_inode(v9ses, fid, sb); + return v9fs_inode_from_fid(v9ses, fid, sb); } +#endif diff --git a/fs/9p/v9fs_vfs.h b/fs/9p/v9fs_vfs.h index b789f8e..4014160 100644 --- a/fs/9p/v9fs_vfs.h +++ b/fs/9p/v9fs_vfs.h @@ -20,6 +20,8 @@ * Boston, MA 02111-1301 USA * */ +#ifndef FS_9P_V9FS_VFS_H +#define FS_9P_V9FS_VFS_H /* plan9 semantics are that created files are implicitly opened. * But linux semantics are that you call create, then open. @@ -36,6 +38,7 @@ * unlink calls remove, which is an implicit clunk. So we have to track * that kind of thing so that we don't try to clunk a dead fid. */ +#define P9_LOCK_TIMEOUT (30*HZ) extern struct file_system_type v9fs_fs_type; extern const struct address_space_operations v9fs_addr_operations; @@ -45,13 +48,15 @@ extern const struct file_operations v9fs_dir_operations; extern const struct file_operations v9fs_dir_operations_dotl; extern const struct dentry_operations v9fs_dentry_operations; extern const struct dentry_operations v9fs_cached_dentry_operations; +extern const struct file_operations v9fs_cached_file_operations; +extern const struct file_operations v9fs_cached_file_operations_dotl; +extern struct kmem_cache *v9fs_inode_cache; -#ifdef CONFIG_9P_FSCACHE struct inode *v9fs_alloc_inode(struct super_block *sb); void v9fs_destroy_inode(struct inode *inode); -#endif - struct inode *v9fs_get_inode(struct super_block *sb, int mode); +int v9fs_init_inode(struct v9fs_session_info *v9ses, + struct inode *inode, int mode); void v9fs_evict_inode(struct inode *inode); ino_t v9fs_qid2ino(struct p9_qid *qid); void v9fs_stat2inode(struct p9_wstat *, struct inode *, struct super_block *); @@ -62,8 +67,19 @@ void v9fs_inode2stat(struct inode *inode, struct p9_wstat *stat); int v9fs_uflags2omode(int uflags, int extended); ssize_t v9fs_file_readn(struct file *, char *, char __user *, u32, u64); +ssize_t v9fs_fid_readn(struct p9_fid *, char *, char __user *, u32, u64); void v9fs_blank_wstat(struct p9_wstat *wstat); int v9fs_vfs_setattr_dotl(struct dentry *, struct iattr *); int v9fs_file_fsync_dotl(struct file *filp, int datasync); - -#define P9_LOCK_TIMEOUT (30*HZ) +ssize_t v9fs_file_write_internal(struct inode *, struct p9_fid *, + const char __user *, size_t, loff_t *, int); +int v9fs_refresh_inode(struct p9_fid *fid, struct inode *inode); +int v9fs_refresh_inode_dotl(struct p9_fid *fid, struct inode *inode); +static inline void v9fs_invalidate_inode_attr(struct inode *inode) +{ + struct v9fs_inode *v9inode; + v9inode = V9FS_I(inode); + v9inode->cache_validity |= V9FS_INO_INVALID_ATTR; + return; +} +#endif diff --git a/fs/9p/vfs_addr.c b/fs/9p/vfs_addr.c index b7f2a8e..2524e4c 100644 --- a/fs/9p/vfs_addr.c +++ b/fs/9p/vfs_addr.c @@ -39,16 +39,16 @@ #include "v9fs.h" #include "v9fs_vfs.h" #include "cache.h" +#include "fid.h" /** - * v9fs_vfs_readpage - read an entire page in from 9P + * v9fs_fid_readpage - read an entire page in from 9P * - * @filp: file being read + * @fid: fid being read * @page: structure to page * */ - -static int v9fs_vfs_readpage(struct file *filp, struct page *page) +static int v9fs_fid_readpage(struct p9_fid *fid, struct page *page) { int retval; loff_t offset; @@ -67,7 +67,7 @@ static int v9fs_vfs_readpage(struct file *filp, struct page *page) buffer = kmap(page); offset = page_offset(page); - retval = v9fs_file_readn(filp, buffer, NULL, PAGE_CACHE_SIZE, offset); + retval = v9fs_fid_readn(fid, buffer, NULL, PAGE_CACHE_SIZE, offset); if (retval < 0) { v9fs_uncache_page(inode, page); goto done; @@ -87,6 +87,19 @@ done: } /** + * v9fs_vfs_readpage - read an entire page in from 9P + * + * @filp: file being read + * @page: structure to page + * + */ + +static int v9fs_vfs_readpage(struct file *filp, struct page *page) +{ + return v9fs_fid_readpage(filp->private_data, page); +} + +/** * v9fs_vfs_readpages - read a set of pages from 9P * * @filp: file being read @@ -124,7 +137,6 @@ static int v9fs_release_page(struct page *page, gfp_t gfp) { if (PagePrivate(page)) return 0; - return v9fs_fscache_release_page(page, gfp); } @@ -137,20 +149,89 @@ static int v9fs_release_page(struct page *page, gfp_t gfp) static void v9fs_invalidate_page(struct page *page, unsigned long offset) { + /* + * If called with zero offset, we should release + * the private state assocated with the page + */ if (offset == 0) v9fs_fscache_invalidate_page(page); } +static int v9fs_vfs_writepage_locked(struct page *page) +{ + char *buffer; + int retval, len; + loff_t offset, size; + mm_segment_t old_fs; + struct v9fs_inode *v9inode; + struct inode *inode = page->mapping->host; + + v9inode = V9FS_I(inode); + size = i_size_read(inode); + if (page->index == size >> PAGE_CACHE_SHIFT) + len = size & ~PAGE_CACHE_MASK; + else + len = PAGE_CACHE_SIZE; + + set_page_writeback(page); + + buffer = kmap(page); + offset = page_offset(page); + + old_fs = get_fs(); + set_fs(get_ds()); + /* We should have writeback_fid always set */ + BUG_ON(!v9inode->writeback_fid); + + retval = v9fs_file_write_internal(inode, + v9inode->writeback_fid, + (__force const char __user *)buffer, + len, &offset, 0); + if (retval > 0) + retval = 0; + + set_fs(old_fs); + kunmap(page); + end_page_writeback(page); + return retval; +} + +static int v9fs_vfs_writepage(struct page *page, struct writeback_control *wbc) +{ + int retval; + + retval = v9fs_vfs_writepage_locked(page); + if (retval < 0) { + if (retval == -EAGAIN) { + redirty_page_for_writepage(wbc, page); + retval = 0; + } else { + SetPageError(page); + mapping_set_error(page->mapping, retval); + } + } else + retval = 0; + + unlock_page(page); + return retval; +} + /** * v9fs_launder_page - Writeback a dirty page - * Since the writes go directly to the server, we simply return a 0 - * here to indicate success. - * * Returns 0 on success. */ static int v9fs_launder_page(struct page *page) { + int retval; + struct inode *inode = page->mapping->host; + + v9fs_fscache_wait_on_page_write(inode, page); + if (clear_page_dirty_for_io(page)) { + retval = v9fs_vfs_writepage_locked(page); + if (retval) + return retval; + } return 0; } @@ -173,9 +254,15 @@ static int v9fs_launder_page(struct page *page) * with an error. * */ -ssize_t v9fs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, - loff_t pos, unsigned long nr_segs) +static ssize_t +v9fs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, + loff_t pos, unsigned long nr_segs) { + /* + * FIXME + * Now that we do caching with cache mode enabled, We need + * to support direct IO + */ P9_DPRINTK(P9_DEBUG_VFS, "v9fs_direct_IO: v9fs_direct_IO (%s) " "off/no(%lld/%lu) EINVAL\n", iocb->ki_filp->f_path.dentry->d_name.name, @@ -183,11 +270,84 @@ ssize_t v9fs_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, return -EINVAL; } + +static int v9fs_write_begin(struct file *filp, struct address_space *mapping, + loff_t pos, unsigned len, unsigned flags, + struct page **pagep, void **fsdata) +{ + int retval = 0; + struct page *page; + struct v9fs_inode *v9inode; + pgoff_t index = pos >> PAGE_CACHE_SHIFT; + struct inode *inode = mapping->host; + + v9inode = V9FS_I(inode); +start: + page = grab_cache_page_write_begin(mapping, index, flags); + if (!page) { + retval = -ENOMEM; + goto out; + } + BUG_ON(!v9inode->writeback_fid); + if (PageUptodate(page)) + goto out; + + if (len == PAGE_CACHE_SIZE) + goto out; + + retval = v9fs_fid_readpage(v9inode->writeback_fid, page); + page_cache_release(page); + if (!retval) + goto start; +out: + *pagep = page; + return retval; +} + +static int v9fs_write_end(struct file *filp, struct address_space *mapping, + loff_t pos, unsigned len, unsigned copied, + struct page *page, void *fsdata) +{ + loff_t last_pos = pos + copied; + struct inode *inode = page->mapping->host; + + if (unlikely(copied < len)) { + /* + * zero out the rest of the area + */ + unsigned from = pos & (PAGE_CACHE_SIZE - 1); + + zero_user(page, from + copied, len - copied); + flush_dcache_page(page); + } + + if (!PageUptodate(page)) + SetPageUptodate(page); + /* + * No need to use i_size_read() here, the i_size + * cannot change under us because we hold the i_mutex. + */ + if (last_pos > inode->i_size) { + inode_add_bytes(inode, last_pos - inode->i_size); + i_size_write(inode, last_pos); + } + set_page_dirty(page); + unlock_page(page); + page_cache_release(page); + + return copied; +} + + const struct address_space_operations v9fs_addr_operations = { - .readpage = v9fs_vfs_readpage, - .readpages = v9fs_vfs_readpages, - .releasepage = v9fs_release_page, - .invalidatepage = v9fs_invalidate_page, - .launder_page = v9fs_launder_page, - .direct_IO = v9fs_direct_IO, + .readpage = v9fs_vfs_readpage, + .readpages = v9fs_vfs_readpages, + .set_page_dirty = __set_page_dirty_nobuffers, + .writepage = v9fs_vfs_writepage, + .write_begin = v9fs_write_begin, + .write_end = v9fs_write_end, + .releasepage = v9fs_release_page, + .invalidatepage = v9fs_invalidate_page, + .launder_page = v9fs_launder_page, + .direct_IO = v9fs_direct_IO, }; diff --git a/fs/9p/vfs_dentry.c b/fs/9p/vfs_dentry.c index 233b7d4..b6a3b9f 100644 --- a/fs/9p/vfs_dentry.c +++ b/fs/9p/vfs_dentry.c @@ -63,20 +63,15 @@ static int v9fs_dentry_delete(const struct dentry *dentry) * v9fs_cached_dentry_delete - called when dentry refcount equals 0 * @dentry: dentry in question * - * Only return 1 if our inode is invalid. Only non-synthetic files - * (ones without mtime == 0) should be calling this function. - * */ - static int v9fs_cached_dentry_delete(const struct dentry *dentry) { - struct inode *inode = dentry->d_inode; - P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p)\n", dentry->d_name.name, - dentry); + P9_DPRINTK(P9_DEBUG_VFS, " dentry: %s (%p)\n", + dentry->d_name.name, dentry); - if(!inode) + /* Don't cache negative dentries */ + if (!dentry->d_inode) return 1; - return 0; } @@ -105,7 +100,41 @@ static void v9fs_dentry_release(struct dentry *dentry) } } +static int v9fs_lookup_revalidate(struct dentry *dentry, struct nameidata *nd) +{ + struct p9_fid *fid; + struct inode *inode; + struct v9fs_inode *v9inode; + + if (nd->flags & LOOKUP_RCU) + return -ECHILD; + + inode = dentry->d_inode; + if (!inode) + goto out_valid; + + v9inode = V9FS_I(inode); + if (v9inode->cache_validity & V9FS_INO_INVALID_ATTR) { + int retval; + struct v9fs_session_info *v9ses; + fid = v9fs_fid_lookup(dentry); + if (IS_ERR(fid)) + return PTR_ERR(fid); + + v9ses = v9fs_inode2v9ses(inode); + if (v9fs_proto_dotl(v9ses)) + retval = v9fs_refresh_inode_dotl(fid, inode); + else + retval = v9fs_refresh_inode(fid, inode); + if (retval <= 0) + return retval; + } +out_valid: + return 1; +} + const struct dentry_operations v9fs_cached_dentry_operations = { + .d_revalidate = v9fs_lookup_revalidate, .d_delete = v9fs_cached_dentry_delete, .d_release = v9fs_dentry_release, }; diff --git a/fs/9p/vfs_dir.c b/fs/9p/vfs_dir.c index b84ebe8..9c2bdda 100644 --- a/fs/9p/vfs_dir.c +++ b/fs/9p/vfs_dir.c @@ -295,7 +295,6 @@ int v9fs_dir_release(struct inode *inode, struct file *filp) P9_DPRINTK(P9_DEBUG_VFS, "v9fs_dir_release: inode: %p filp: %p fid: %d\n", inode, filp, fid ? fid->fid : -1); - filemap_write_and_wait(inode->i_mapping); if (fid) p9_client_clunk(fid); return 0; diff --git a/fs/9p/vfs_file.c b/fs/9p/vfs_file.c index 240c306..78bcb97 100644 --- a/fs/9p/vfs_file.c +++ b/fs/9p/vfs_file.c @@ -44,8 +44,7 @@ #include "fid.h" #include "cache.h" -static const struct file_operations v9fs_cached_file_operations; -static const struct file_operations v9fs_cached_file_operations_dotl; +static const struct vm_operations_struct v9fs_file_vm_ops; /** * v9fs_file_open - open a file (or directory) @@ -57,11 +56,13 @@ static const struct file_operations v9fs_cached_file_operations_dotl; int v9fs_file_open(struct inode *inode, struct file *file) { int err; + struct v9fs_inode *v9inode; struct v9fs_session_info *v9ses; struct p9_fid *fid; int omode; P9_DPRINTK(P9_DEBUG_VFS, "inode: %p file: %p\n", inode, file); + v9inode = V9FS_I(inode); v9ses = v9fs_inode2v9ses(inode); if (v9fs_proto_dotl(v9ses)) omode = file->f_flags; @@ -89,20 +90,30 @@ int v9fs_file_open(struct inode *inode, struct file *file) } file->private_data = fid; - if ((fid->qid.version) && (v9ses->cache)) { - P9_DPRINTK(P9_DEBUG_VFS, "cached"); - /* enable cached file options */ - if(file->f_op == &v9fs_file_operations) - file->f_op = &v9fs_cached_file_operations; - else if (file->f_op == &v9fs_file_operations_dotl) - file->f_op = &v9fs_cached_file_operations_dotl; - + if (v9ses->cache && !v9inode->writeback_fid) { + /* + * clone a fid and add it to writeback_fid + * we do it during open time instead of + * page dirty time via write_begin/page_mkwrite + * because we want write after unlink usecase + * to work. + */ + fid = v9fs_writeback_fid(file->f_path.dentry); + if (IS_ERR(fid)) { + err = PTR_ERR(fid); + goto out_error; + } + v9inode->writeback_fid = (void *) fid; + } #ifdef CONFIG_9P_FSCACHE + if (v9ses->cache) v9fs_cache_inode_set_cookie(inode, file); #endif - } - return 0; +out_error: + p9_client_clunk(file->private_data); + file->private_data = NULL; + return err; } /** @@ -335,25 +346,22 @@ out_err: } /** - * v9fs_file_readn - read from a file - * @filp: file pointer to read + * v9fs_fid_readn - read from a fid + * @fid: fid to read * @data: data buffer to read data into * @udata: user data buffer to read data into * @count: size of buffer * @offset: offset at which to read data * */ - ssize_t -v9fs_file_readn(struct file *filp, char *data, char __user *udata, u32 count, +v9fs_fid_readn(struct p9_fid *fid, char *data, char __user *udata, u32 count, u64 offset) { int n, total, size; - struct p9_fid *fid = filp->private_data; P9_DPRINTK(P9_DEBUG_VFS, "fid %d offset %llu count %d\n", fid->fid, - (long long unsigned) offset, count); - + (long long unsigned) offset, count); n = 0; total = 0; size = fid->iounit ? fid->iounit : fid->clnt->msize - P9_IOHDRSZ; @@ -379,6 +387,22 @@ v9fs_file_readn(struct file *filp, char *data, char __user *udata, u32 count, } /** + * v9fs_file_readn - read from a file + * @filp: file pointer to read + * @data: data buffer to read data into + * @udata: user data buffer to read data into + * @count: size of buffer + * @offset: offset at which to read data + * + */ +ssize_t +v9fs_file_readn(struct file *filp, char *data, char __user *udata, u32 count, + u64 offset) +{ + return v9fs_fid_readn(filp->private_data, data, udata, count, offset); +} + +/** * v9fs_file_read - read from a file * @filp: file pointer to read * @udata: user data buffer to read data into @@ -410,45 +434,22 @@ v9fs_file_read(struct file *filp, char __user *udata, size_t count, return ret; } -/** - * v9fs_file_write - write to a file - * @filp: file pointer to write - * @data: data buffer to write data from - * @count: size of buffer - * @offset: offset at which to write data - * - */ - -static ssize_t -v9fs_file_write(struct file *filp, const char __user * data, - size_t count, loff_t * offset) +ssize_t +v9fs_file_write_internal(struct inode *inode, struct p9_fid *fid, + const char __user *data, size_t count, + loff_t *offset, int invalidate) { - ssize_t retval; - size_t total = 0; int n; - struct p9_fid *fid; + loff_t i_size; + size_t total = 0; struct p9_client *clnt; - struct inode *inode = filp->f_path.dentry->d_inode; loff_t origin = *offset; unsigned long pg_start, pg_end; P9_DPRINTK(P9_DEBUG_VFS, "data %p count %d offset %x\n", data, (int)count, (int)*offset); - fid = filp->private_data; clnt = fid->clnt; - - retval = generic_write_checks(filp, &origin, &count, 0); - if (retval) - goto out; - - retval = -EINVAL; - if ((ssize_t) count < 0) - goto out; - retval = 0; - if (!count) - goto out; - do { n = p9_client_write(fid, NULL, data+total, origin+total, count); if (n <= 0) @@ -457,25 +458,60 @@ v9fs_file_write(struct file *filp, const char __user * data, total += n; } while (count > 0); - if (total > 0) { + if (invalidate && (total > 0)) { pg_start = origin >> PAGE_CACHE_SHIFT; pg_end = (origin + total - 1) >> PAGE_CACHE_SHIFT; if (inode->i_mapping && inode->i_mapping->nrpages) invalidate_inode_pages2_range(inode->i_mapping, pg_start, pg_end); *offset += total; - i_size_write(inode, i_size_read(inode) + total); - inode->i_blocks = (i_size_read(inode) + 512 - 1) >> 9; + i_size = i_size_read(inode); + if (*offset > i_size) { + inode_add_bytes(inode, *offset - i_size); + i_size_write(inode, *offset); + } } - if (n < 0) - retval = n; - else - retval = total; + return n; + + return total; +} + +/** + * v9fs_file_write - write to a file + * @filp: file pointer to write + * @data: data buffer to write data from + * @count: size of buffer + * @offset: offset at which to write data + * + */ +static ssize_t +v9fs_file_write(struct file *filp, const char __user * data, + size_t count, loff_t *offset) +{ + ssize_t retval = 0; + loff_t origin = *offset; + + + retval = generic_write_checks(filp, &origin, &count, 0); + if (retval) + goto out; + + retval = -EINVAL; + if ((ssize_t) count < 0) + goto out; + retval = 0; + if (!count) + goto out; + + return v9fs_file_write_internal(filp->f_path.dentry->d_inode, + filp->private_data, + data, count, offset, 1); out: return retval; } + static int v9fs_file_fsync(struct file *filp, int datasync) { struct p9_fid *fid; @@ -505,28 +541,182 @@ int v9fs_file_fsync_dotl(struct file *filp, int datasync) return retval; } -static const struct file_operations v9fs_cached_file_operations = { +static int +v9fs_file_mmap(struct file *file, struct vm_area_struct *vma) +{ + int retval; + + retval = generic_file_mmap(file, vma); + if (!retval) + vma->vm_ops = &v9fs_file_vm_ops; + + return retval; +} + +static int +v9fs_vm_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) +{ + struct v9fs_inode *v9inode; + struct page *page = vmf->page; + struct file *filp = vma->vm_file; + struct inode *inode = filp->f_path.dentry->d_inode; + + + P9_DPRINTK(P9_DEBUG_VFS, "page %p fid %lx\n", + page, (unsigned long)filp->private_data); + + v9inode = V9FS_I(inode); + /* make sure the cache has finished storing the page */ + v9fs_fscache_wait_on_page_write(inode, page); + BUG_ON(!v9inode->writeback_fid); + lock_page(page); + if (page->mapping != inode->i_mapping) + goto out_unlock; + + return VM_FAULT_LOCKED; +out_unlock: + unlock_page(page); + return VM_FAULT_NOPAGE; +} + +static ssize_t +v9fs_direct_read(struct file *filp, char __user *udata, size_t count, + loff_t *offsetp) +{ + loff_t size, offset; + struct inode *inode; + struct address_space *mapping; + + offset = *offsetp; + mapping = filp->f_mapping; + inode = mapping->host; + if (!count) + return 0; + size = i_size_read(inode); + if (offset < size) + filemap_write_and_wait_range(mapping, offset, + offset + count - 1); + + return v9fs_file_read(filp, udata, count, offsetp); +} + +/** + * v9fs_cached_file_read - read from a file + * @filp: file pointer to read + * @udata: user data buffer to read data into + * @count: size of buffer + * @offset: offset at which to read data + * + */ +static ssize_t +v9fs_cached_file_read(struct file *filp, char __user *data, size_t count, + loff_t *offset) +{ + if (filp->f_flags & O_DIRECT) + return v9fs_direct_read(filp, data, count, offset); + return do_sync_read(filp, data, count, offset); +} + +static ssize_t +v9fs_direct_write(struct file *filp, const char __user * data, + size_t count, loff_t *offsetp) +{ + loff_t offset; + ssize_t retval; + struct inode *inode; + struct address_space *mapping; + + offset = *offsetp; + mapping = filp->f_mapping; + inode = mapping->host; + if (!count) + return 0; + + mutex_lock(&inode->i_mutex); + retval = filemap_write_and_wait_range(mapping, offset, + offset + count - 1); + if (retval) + goto err_out; + /* + * After a write we want buffered reads to be sure to go to disk to get + * the new data. We invalidate clean cached page from the region we're + * about to write. We do this *before* the write so that if we fail + * here we fall back to buffered write + */ + if (mapping->nrpages) { + pgoff_t pg_start = offset >> PAGE_CACHE_SHIFT; + pgoff_t pg_end = (offset + count - 1) >> PAGE_CACHE_SHIFT; + + retval = invalidate_inode_pages2_range(mapping, + pg_start, pg_end); + /* + * If a page can not be invalidated, fall back + * to buffered write. + */ + if (retval) { + if (retval == -EBUSY) + goto buff_write; + goto err_out; + } + } + retval = v9fs_file_write(filp, data, count, offsetp); +err_out: + mutex_unlock(&inode->i_mutex); + return retval; + +buff_write: + mutex_unlock(&inode->i_mutex); + return do_sync_write(filp, data, count, offsetp); +} + +/** + * v9fs_cached_file_write - write to a file + * @filp: file pointer to write + * @data: data buffer to write data from + * @count: size of buffer + * @offset: offset at which to write data + * + */ +static ssize_t +v9fs_cached_file_write(struct file *filp, const char __user * data, + size_t count, loff_t *offset) +{ + + if (filp->f_flags & O_DIRECT) + return v9fs_direct_write(filp, data, count, offset); + return do_sync_write(filp, data, count, offset); +} + +static const struct vm_operations_struct v9fs_file_vm_ops = { + .fault = filemap_fault, + .page_mkwrite = v9fs_vm_page_mkwrite, +}; + + +const struct file_operations v9fs_cached_file_operations = { .llseek = generic_file_llseek, - .read = do_sync_read, + .read = v9fs_cached_file_read, + .write = v9fs_cached_file_write, .aio_read = generic_file_aio_read, - .write = v9fs_file_write, + .aio_write = generic_file_aio_write, .open = v9fs_file_open, .release = v9fs_dir_release, .lock = v9fs_file_lock, - .mmap = generic_file_readonly_mmap, + .mmap = v9fs_file_mmap, .fsync = v9fs_file_fsync, }; -static const struct file_operations v9fs_cached_file_operations_dotl = { +const struct file_operations v9fs_cached_file_operations_dotl = { .llseek = generic_file_llseek, - .read = do_sync_read, + .read = v9fs_cached_file_read, + .write = v9fs_cached_file_write, .aio_read = generic_file_aio_read, - .write = v9fs_file_write, + .aio_write = generic_file_aio_write, .open = v9fs_file_open, .release = v9fs_dir_release, .lock = v9fs_file_lock_dotl, .flock = v9fs_file_flock_dotl, - .mmap = generic_file_readonly_mmap, + .mmap = v9fs_file_mmap, .fsync = v9fs_file_fsync_dotl, }; diff --git a/fs/9p/vfs_inode.c b/fs/9p/vfs_inode.c index b76a40b..8a2c232 100644 --- a/fs/9p/vfs_inode.c +++ b/fs/9p/vfs_inode.c @@ -203,26 +203,25 @@ v9fs_blank_wstat(struct p9_wstat *wstat) wstat->extension = NULL; } -#ifdef CONFIG_9P_FSCACHE /** * v9fs_alloc_inode - helper function to allocate an inode - * This callback is executed before setting up the inode so that we - * can associate a vcookie with each inode. * */ - struct inode *v9fs_alloc_inode(struct super_block *sb) { - struct v9fs_cookie *vcookie; - vcookie = (struct v9fs_cookie *)kmem_cache_alloc(vcookie_cache, - GFP_KERNEL); - if (!vcookie) + struct v9fs_inode *v9inode; + v9inode = (struct v9fs_inode *)kmem_cache_alloc(v9fs_inode_cache, + GFP_KERNEL); + if (!v9inode) return NULL; - - vcookie->fscache = NULL; - vcookie->qid = NULL; - spin_lock_init(&vcookie->lock); - return &vcookie->inode; +#ifdef CONFIG_9P_FSCACHE + v9inode->fscache = NULL; + v9inode->fscache_key = NULL; + spin_lock_init(&v9inode->fscache_lock); +#endif + v9inode->writeback_fid = NULL; + v9inode->cache_validity = 0; + return &v9inode->vfs_inode; } /** @@ -234,35 +233,18 @@ static void v9fs_i_callback(struct rcu_head *head) { struct inode *inode = container_of(head, struct inode, i_rcu); INIT_LIST_HEAD(&inode->i_dentry); - kmem_cache_free(vcookie_cache, v9fs_inode2cookie(inode)); + kmem_cache_free(v9fs_inode_cache, V9FS_I(inode)); } void v9fs_destroy_inode(struct inode *inode) { call_rcu(&inode->i_rcu, v9fs_i_callback); } -#endif -/** - * v9fs_get_inode - helper function to setup an inode - * @sb: superblock - * @mode: mode to setup inode with - * - */ - -struct inode *v9fs_get_inode(struct super_block *sb, int mode) +int v9fs_init_inode(struct v9fs_session_info *v9ses, + struct inode *inode, int mode) { - int err; - struct inode *inode; - struct v9fs_session_info *v9ses = sb->s_fs_info; - - P9_DPRINTK(P9_DEBUG_VFS, "super block: %p mode: %o\n", sb, mode); - - inode = new_inode(sb); - if (!inode) { - P9_EPRINTK(KERN_WARNING, "Problem allocating inode\n"); - return ERR_PTR(-ENOMEM); - } + int err = 0; inode_init_owner(inode, NULL, mode); inode->i_blocks = 0; @@ -292,14 +274,20 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode) case S_IFREG: if (v9fs_proto_dotl(v9ses)) { inode->i_op = &v9fs_file_inode_operations_dotl; - inode->i_fop = &v9fs_file_operations_dotl; + if (v9ses->cache) + inode->i_fop = + &v9fs_cached_file_operations_dotl; + else + inode->i_fop = &v9fs_file_operations_dotl; } else { inode->i_op = &v9fs_file_inode_operations; - inode->i_fop = &v9fs_file_operations; + if (v9ses->cache) + inode->i_fop = &v9fs_cached_file_operations; + else + inode->i_fop = &v9fs_file_operations; } break; - case S_IFLNK: if (!v9fs_proto_dotu(v9ses) && !v9fs_proto_dotl(v9ses)) { P9_DPRINTK(P9_DEBUG_ERROR, "extended modes used with " @@ -335,12 +323,37 @@ struct inode *v9fs_get_inode(struct super_block *sb, int mode) err = -EINVAL; goto error; } +error: + return err; - return inode; +} -error: - iput(inode); - return ERR_PTR(err); +/** + * v9fs_get_inode - helper function to setup an inode + * @sb: superblock + * @mode: mode to setup inode with + * + */ + +struct inode *v9fs_get_inode(struct super_block *sb, int mode) +{ + int err; + struct inode *inode; + struct v9fs_session_info *v9ses = sb->s_fs_info; + + P9_DPRINTK(P9_DEBUG_VFS, "super block: %p mode: %o\n", sb, mode); + + inode = new_inode(sb); + if (!inode) { + P9_EPRINTK(KERN_WARNING, "Problem allocating inode\n"); + return ERR_PTR(-ENOMEM); + } + err = v9fs_init_inode(v9ses, inode, mode); + if (err) { + iput(inode); + return ERR_PTR(err); + } + return inode; } /* @@ -403,6 +416,8 @@ error: */ void v9fs_evict_inode(struct inode *inode) { + struct v9fs_inode *v9inode = V9FS_I(inode); + truncate_inode_pages(inode->i_mapping, 0); end_writeback(inode); filemap_fdatawrite(inode->i_mapping); @@ -410,41 +425,67 @@ void v9fs_evict_inode(struct inode *inode) #ifdef CONFIG_9P_FSCACHE v9fs_cache_inode_put_cookie(inode); #endif + /* clunk the fid stashed in writeback_fid */ + if (v9inode->writeback_fid) { + p9_client_clunk(v9inode->writeback_fid); + v9inode->writeback_fid = NULL; + } } -struct inode * -v9fs_inode(struct v9fs_session_info *v9ses, struct p9_fid *fid, - struct super_block *sb) +static struct inode *v9fs_qid_iget(struct super_block *sb, + struct p9_qid *qid, + struct p9_wstat *st) { - int err, umode; - struct inode *ret = NULL; - struct p9_wstat *st; - - st = p9_client_stat(fid); - if (IS_ERR(st)) - return ERR_CAST(st); + int retval, umode; + unsigned long i_ino; + struct inode *inode; + struct v9fs_session_info *v9ses = sb->s_fs_info; + i_ino = v9fs_qid2ino(qid); + inode = iget_locked(sb, i_ino); + if (!inode) + return ERR_PTR(-ENOMEM); + if (!(inode->i_state & I_NEW)) + return inode; + /* + * initialize the inode with the stat info + * FIXME!! we may need support for stale inodes + * later. + */ umode = p9mode2unixmode(v9ses, st->mode); - ret = v9fs_get_inode(sb, umode); - if (IS_ERR(ret)) { - err = PTR_ERR(ret); + retval = v9fs_init_inode(v9ses, inode, umode); + if (retval) goto error; - } - - v9fs_stat2inode(st, ret, sb); - ret->i_ino = v9fs_qid2ino(&st->qid); + v9fs_stat2inode(st, inode, sb); #ifdef CONFIG_9P_FSCACHE - v9fs_vcookie_set_qid(ret, &st->qid); - v9fs_cache_inode_get_cookie(ret); + v9fs_fscache_set_key(inode, &st->qid); + v9fs_cache_inode_get_cookie(inode); #endif - p9stat_free(st); - kfree(st); - return ret; + unlock_new_inode(inode); + return inode; error: + unlock_new_inode(inode); + iput(inode); + return ERR_PTR(retval); + +} + +struct inode * +v9fs_inode_from_fid(struct v9fs_session_info *v9ses, struct p9_fid *fid, + struct super_block *sb) +{ + struct p9_wstat *st; + struct inode *inode = NULL; + + st = p9_client_stat(fid); + if (IS_ERR(st)) + return ERR_CAST(st); + + inode = v9fs_qid_iget(sb, &st->qid, st); p9stat_free(st); kfree(st); - return ERR_PTR(err); + return inode; } /** @@ -458,8 +499,8 @@ error: static int v9fs_remove(struct inode *dir, struct dentry *file, int rmdir) { int retval; - struct inode *file_inode; struct p9_fid *v9fid; + struct inode *file_inode; P9_DPRINTK(P9_DEBUG_VFS, "inode: %p dentry: %p rmdir: %d\n", dir, file, rmdir); @@ -470,8 +511,20 @@ static int v9fs_remove(struct inode *dir, struct dentry *file, int rmdir) return PTR_ERR(v9fid); retval = p9_client_remove(v9fid); - if (!retval) - drop_nlink(file_inode); + if (!retval) { + /* + * directories on unlink should have zero + * link count + */ + if (rmdir) { + clear_nlink(file_inode); + drop_nlink(dir); + } else + drop_nlink(file_inode); + + v9fs_invalidate_inode_attr(file_inode); + v9fs_invalidate_inode_attr(dir); + } return retval; } @@ -531,7 +584,7 @@ v9fs_create(struct v9fs_session_info *v9ses, struct inode *dir, } /* instantiate inode and assign the unopened fid to the dentry */ - inode = v9fs_inode_from_fid(v9ses, fid, dir->i_sb); + inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb); if (IS_ERR(inode)) { err = PTR_ERR(inode); P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", err); @@ -570,9 +623,10 @@ v9fs_vfs_create(struct inode *dir, struct dentry *dentry, int mode, int err; u32 perm; int flags; - struct v9fs_session_info *v9ses; - struct p9_fid *fid; struct file *filp; + struct v9fs_inode *v9inode; + struct v9fs_session_info *v9ses; + struct p9_fid *fid, *inode_fid; err = 0; fid = NULL; @@ -592,8 +646,25 @@ v9fs_vfs_create(struct inode *dir, struct dentry *dentry, int mode, goto error; } + v9fs_invalidate_inode_attr(dir); /* if we are opening a file, assign the open fid to the file */ if (nd && nd->flags & LOOKUP_OPEN) { + v9inode = V9FS_I(dentry->d_inode); + if (v9ses->cache && !v9inode->writeback_fid) { + /* + * clone a fid and add it to writeback_fid + * we do it during open time instead of + * page dirty time via write_begin/page_mkwrite + * because we want write after unlink usecase + * to work. + */ + inode_fid = v9fs_writeback_fid(dentry); + if (IS_ERR(inode_fid)) { + err = PTR_ERR(inode_fid); + goto error; + } + v9inode->writeback_fid = (void *) inode_fid; + } filp = lookup_instantiate_filp(nd, dentry, generic_file_open); if (IS_ERR(filp)) { err = PTR_ERR(filp); @@ -601,6 +672,10 @@ v9fs_vfs_create(struct inode *dir, struct dentry *dentry, int mode, } filp->private_data = fid; +#ifdef CONFIG_9P_FSCACHE + if (v9ses->cache) + v9fs_cache_inode_set_cookie(dentry->d_inode, filp); +#endif } else p9_client_clunk(fid); @@ -625,8 +700,8 @@ static int v9fs_vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) { int err; u32 perm; - struct v9fs_session_info *v9ses; struct p9_fid *fid; + struct v9fs_session_info *v9ses; P9_DPRINTK(P9_DEBUG_VFS, "name %s\n", dentry->d_name.name); err = 0; @@ -636,6 +711,9 @@ static int v9fs_vfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) if (IS_ERR(fid)) { err = PTR_ERR(fid); fid = NULL; + } else { + inc_nlink(dir); + v9fs_invalidate_inode_attr(dir); } if (fid) @@ -687,7 +765,7 @@ struct dentry *v9fs_vfs_lookup(struct inode *dir, struct dentry *dentry, return ERR_PTR(result); } - inode = v9fs_inode_from_fid(v9ses, fid, dir->i_sb); + inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb); if (IS_ERR(inode)) { result = PTR_ERR(inode); inode = NULL; @@ -747,17 +825,19 @@ int v9fs_vfs_rename(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry) { + int retval; struct inode *old_inode; + struct inode *new_inode; struct v9fs_session_info *v9ses; struct p9_fid *oldfid; struct p9_fid *olddirfid; struct p9_fid *newdirfid; struct p9_wstat wstat; - int retval; P9_DPRINTK(P9_DEBUG_VFS, "\n"); retval = 0; old_inode = old_dentry->d_inode; + new_inode = new_dentry->d_inode; v9ses = v9fs_inode2v9ses(old_inode); oldfid = v9fs_fid_lookup(old_dentry); if (IS_ERR(oldfid)) @@ -798,9 +878,30 @@ v9fs_vfs_rename(struct inode *old_dir, struct dentry *old_dentry, retval = p9_client_wstat(oldfid, &wstat); clunk_newdir: - if (!retval) + if (!retval) { + if (new_inode) { + if (S_ISDIR(new_inode->i_mode)) + clear_nlink(new_inode); + else + drop_nlink(new_inode); + /* + * Work around vfs rename rehash bug with + * FS_RENAME_DOES_D_MOVE + */ + v9fs_invalidate_inode_attr(new_inode); + } + if (S_ISDIR(old_inode->i_mode)) { + if (!new_inode) + inc_nlink(new_dir); + drop_nlink(old_dir); + } + v9fs_invalidate_inode_attr(old_inode); + v9fs_invalidate_inode_attr(old_dir); + v9fs_invalidate_inode_attr(new_dir); + /* successful rename */ d_move(old_dentry, new_dentry); + } up_write(&v9ses->rename_sem); p9_client_clunk(newdirfid); @@ -831,9 +932,10 @@ v9fs_vfs_getattr(struct vfsmount *mnt, struct dentry *dentry, P9_DPRINTK(P9_DEBUG_VFS, "dentry: %p\n", dentry); err = -EPERM; v9ses = v9fs_inode2v9ses(dentry->d_inode); - if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) - return simple_getattr(mnt, dentry, stat); - + if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) { + generic_fillattr(dentry->d_inode, stat); + return 0; + } fid = v9fs_fid_lookup(dentry); if (IS_ERR(fid)) return PTR_ERR(fid); @@ -891,17 +993,20 @@ static int v9fs_vfs_setattr(struct dentry *dentry, struct iattr *iattr) if (iattr->ia_valid & ATTR_GID) wstat.n_gid = iattr->ia_gid; } - - retval = p9_client_wstat(fid, &wstat); - if (retval < 0) - return retval; - if ((iattr->ia_valid & ATTR_SIZE) && iattr->ia_size != i_size_read(dentry->d_inode)) { retval = vmtruncate(dentry->d_inode, iattr->ia_size); if (retval) return retval; } + /* Write all dirty data */ + if (S_ISREG(dentry->d_inode->i_mode)) + filemap_write_and_wait(dentry->d_inode->i_mapping); + + retval = p9_client_wstat(fid, &wstat); + if (retval < 0) + return retval; + v9fs_invalidate_inode_attr(dentry->d_inode); setattr_copy(dentry->d_inode, iattr); mark_inode_dirty(dentry->d_inode); @@ -924,6 +1029,7 @@ v9fs_stat2inode(struct p9_wstat *stat, struct inode *inode, char tag_name[14]; unsigned int i_nlink; struct v9fs_session_info *v9ses = sb->s_fs_info; + struct v9fs_inode *v9inode = V9FS_I(inode); inode->i_nlink = 1; @@ -983,6 +1089,7 @@ v9fs_stat2inode(struct p9_wstat *stat, struct inode *inode, /* not real number of blocks, but 512 byte ones ... */ inode->i_blocks = (i_size_read(inode) + 512 - 1) >> 9; + v9inode->cache_validity &= ~V9FS_INO_INVALID_ATTR; } /** @@ -1115,8 +1222,8 @@ static int v9fs_vfs_mkspecial(struct inode *dir, struct dentry *dentry, int mode, const char *extension) { u32 perm; - struct v9fs_session_info *v9ses; struct p9_fid *fid; + struct v9fs_session_info *v9ses; v9ses = v9fs_inode2v9ses(dir); if (!v9fs_proto_dotu(v9ses)) { @@ -1130,6 +1237,7 @@ static int v9fs_vfs_mkspecial(struct inode *dir, struct dentry *dentry, if (IS_ERR(fid)) return PTR_ERR(fid); + v9fs_invalidate_inode_attr(dir); p9_client_clunk(fid); return 0; } @@ -1166,8 +1274,8 @@ v9fs_vfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry) { int retval; - struct p9_fid *oldfid; char *name; + struct p9_fid *oldfid; P9_DPRINTK(P9_DEBUG_VFS, " %lu,%s,%s\n", dir->i_ino, dentry->d_name.name, @@ -1186,7 +1294,10 @@ v9fs_vfs_link(struct dentry *old_dentry, struct inode *dir, sprintf(name, "%d\n", oldfid->fid); retval = v9fs_vfs_mkspecial(dir, dentry, P9_DMLINK, name); __putname(name); - + if (!retval) { + v9fs_refresh_inode(oldfid, old_dentry->d_inode); + v9fs_invalidate_inode_attr(dir); + } clunk_fid: p9_client_clunk(oldfid); return retval; @@ -1237,6 +1348,32 @@ v9fs_vfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t rdev) return retval; } +int v9fs_refresh_inode(struct p9_fid *fid, struct inode *inode) +{ + loff_t i_size; + struct p9_wstat *st; + struct v9fs_session_info *v9ses; + + v9ses = v9fs_inode2v9ses(inode); + st = p9_client_stat(fid); + if (IS_ERR(st)) + return PTR_ERR(st); + + spin_lock(&inode->i_lock); + /* + * We don't want to refresh inode->i_size, + * because we may have cached data + */ + i_size = inode->i_size; + v9fs_stat2inode(st, inode, inode->i_sb); + if (v9ses->cache) + inode->i_size = i_size; + spin_unlock(&inode->i_lock); + p9stat_free(st); + kfree(st); + return 0; +} + static const struct inode_operations v9fs_dir_inode_operations_dotu = { .create = v9fs_vfs_create, .lookup = v9fs_vfs_lookup, diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index fe3ffa9..67c138e 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -86,40 +86,63 @@ static struct dentry *v9fs_dentry_from_dir_inode(struct inode *inode) return dentry; } +static struct inode *v9fs_qid_iget_dotl(struct super_block *sb, + struct p9_qid *qid, + struct p9_fid *fid, + struct p9_stat_dotl *st) +{ + int retval; + unsigned long i_ino; + struct inode *inode; + struct v9fs_session_info *v9ses = sb->s_fs_info; + + i_ino = v9fs_qid2ino(qid); + inode = iget_locked(sb, i_ino); + if (!inode) + return ERR_PTR(-ENOMEM); + if (!(inode->i_state & I_NEW)) + return inode; + /* + * initialize the inode with the stat info + * FIXME!! we may need support for stale inodes + * later. + */ + retval = v9fs_init_inode(v9ses, inode, st->st_mode); + if (retval) + goto error; + + v9fs_stat2inode_dotl(st, inode); +#ifdef CONFIG_9P_FSCACHE + v9fs_fscache_set_key(inode, &st->qid); + v9fs_cache_inode_get_cookie(inode); +#endif + retval = v9fs_get_acl(inode, fid); + if (retval) + goto error; + + unlock_new_inode(inode); + return inode; +error: + unlock_new_inode(inode); + iput(inode); + return ERR_PTR(retval); + +} + struct inode * -v9fs_inode_dotl(struct v9fs_session_info *v9ses, struct p9_fid *fid, - struct super_block *sb) +v9fs_inode_from_fid_dotl(struct v9fs_session_info *v9ses, struct p9_fid *fid, + struct super_block *sb) { - struct inode *ret = NULL; - int err; struct p9_stat_dotl *st; + struct inode *inode = NULL; st = p9_client_getattr_dotl(fid, P9_STATS_BASIC); if (IS_ERR(st)) return ERR_CAST(st); - ret = v9fs_get_inode(sb, st->st_mode); - if (IS_ERR(ret)) { - err = PTR_ERR(ret); - goto error; - } - - v9fs_stat2inode_dotl(st, ret); - ret->i_ino = v9fs_qid2ino(&st->qid); -#ifdef CONFIG_9P_FSCACHE - v9fs_vcookie_set_qid(ret, &st->qid); - v9fs_cache_inode_get_cookie(ret); -#endif - err = v9fs_get_acl(ret, fid); - if (err) { - iput(ret); - goto error; - } - kfree(st); - return ret; -error: + inode = v9fs_qid_iget_dotl(sb, &st->qid, fid, st); kfree(st); - return ERR_PTR(err); + return inode; } /** @@ -136,16 +159,17 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode, struct nameidata *nd) { int err = 0; - char *name = NULL; gid_t gid; int flags; mode_t mode; - struct v9fs_session_info *v9ses; - struct p9_fid *fid = NULL; - struct p9_fid *dfid, *ofid; + char *name = NULL; struct file *filp; struct p9_qid qid; struct inode *inode; + struct p9_fid *fid = NULL; + struct v9fs_inode *v9inode; + struct p9_fid *dfid, *ofid, *inode_fid; + struct v9fs_session_info *v9ses; struct posix_acl *pacl = NULL, *dacl = NULL; v9ses = v9fs_inode2v9ses(dir); @@ -196,6 +220,7 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode, err); goto error; } + v9fs_invalidate_inode_attr(dir); /* instantiate inode and assign the unopened fid to the dentry */ fid = p9_client_walk(dfid, 1, &name, 1); @@ -205,7 +230,7 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode, fid = NULL; goto error; } - inode = v9fs_inode_from_fid(v9ses, fid, dir->i_sb); + inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb); if (IS_ERR(inode)) { err = PTR_ERR(inode); P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", err); @@ -219,6 +244,22 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode, /* Now set the ACL based on the default value */ v9fs_set_create_acl(dentry, dacl, pacl); + v9inode = V9FS_I(inode); + if (v9ses->cache && !v9inode->writeback_fid) { + /* + * clone a fid and add it to writeback_fid + * we do it during open time instead of + * page dirty time via write_begin/page_mkwrite + * because we want write after unlink usecase + * to work. + */ + inode_fid = v9fs_writeback_fid(dentry); + if (IS_ERR(inode_fid)) { + err = PTR_ERR(inode_fid); + goto error; + } + v9inode->writeback_fid = (void *) inode_fid; + } /* Since we are opening a file, assign the open fid to the file */ filp = lookup_instantiate_filp(nd, dentry, generic_file_open); if (IS_ERR(filp)) { @@ -226,6 +267,10 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode, return PTR_ERR(filp); } filp->private_data = ofid; +#ifdef CONFIG_9P_FSCACHE + if (v9ses->cache) + v9fs_cache_inode_set_cookie(inode, filp); +#endif return 0; error: @@ -300,7 +345,7 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir, goto error; } - inode = v9fs_inode_from_fid(v9ses, fid, dir->i_sb); + inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb); if (IS_ERR(inode)) { err = PTR_ERR(inode); P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", @@ -327,7 +372,8 @@ static int v9fs_vfs_mkdir_dotl(struct inode *dir, } /* Now set the ACL based on the default value */ v9fs_set_create_acl(dentry, dacl, pacl); - + inc_nlink(dir); + v9fs_invalidate_inode_attr(dir); error: if (fid) p9_client_clunk(fid); @@ -346,9 +392,10 @@ v9fs_vfs_getattr_dotl(struct vfsmount *mnt, struct dentry *dentry, P9_DPRINTK(P9_DEBUG_VFS, "dentry: %p\n", dentry); err = -EPERM; v9ses = v9fs_inode2v9ses(dentry->d_inode); - if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) - return simple_getattr(mnt, dentry, stat); - + if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) { + generic_fillattr(dentry->d_inode, stat); + return 0; + } fid = v9fs_fid_lookup(dentry); if (IS_ERR(fid)) return PTR_ERR(fid); @@ -406,16 +453,20 @@ int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr) if (IS_ERR(fid)) return PTR_ERR(fid); - retval = p9_client_setattr(fid, &p9attr); - if (retval < 0) - return retval; - if ((iattr->ia_valid & ATTR_SIZE) && iattr->ia_size != i_size_read(dentry->d_inode)) { retval = vmtruncate(dentry->d_inode, iattr->ia_size); if (retval) return retval; } + /* Write all dirty data */ + if (S_ISREG(dentry->d_inode->i_mode)) + filemap_write_and_wait(dentry->d_inode->i_mapping); + + retval = p9_client_setattr(fid, &p9attr); + if (retval < 0) + return retval; + v9fs_invalidate_inode_attr(dentry->d_inode); setattr_copy(dentry->d_inode, iattr); mark_inode_dirty(dentry->d_inode); @@ -439,6 +490,7 @@ int v9fs_vfs_setattr_dotl(struct dentry *dentry, struct iattr *iattr) void v9fs_stat2inode_dotl(struct p9_stat_dotl *stat, struct inode *inode) { + struct v9fs_inode *v9inode = V9FS_I(inode); if ((stat->st_result_mask & P9_STATS_BASIC) == P9_STATS_BASIC) { inode->i_atime.tv_sec = stat->st_atime_sec; @@ -497,20 +549,21 @@ v9fs_stat2inode_dotl(struct p9_stat_dotl *stat, struct inode *inode) /* Currently we don't support P9_STATS_BTIME and P9_STATS_DATA_VERSION * because the inode structure does not have fields for them. */ + v9inode->cache_validity &= ~V9FS_INO_INVALID_ATTR; } static int v9fs_vfs_symlink_dotl(struct inode *dir, struct dentry *dentry, const char *symname) { - struct v9fs_session_info *v9ses; - struct p9_fid *dfid; - struct p9_fid *fid = NULL; - struct inode *inode; - struct p9_qid qid; - char *name; int err; gid_t gid; + char *name; + struct p9_qid qid; + struct inode *inode; + struct p9_fid *dfid; + struct p9_fid *fid = NULL; + struct v9fs_session_info *v9ses; name = (char *) dentry->d_name.name; P9_DPRINTK(P9_DEBUG_VFS, "v9fs_vfs_symlink_dotl : %lu,%s,%s\n", @@ -534,6 +587,7 @@ v9fs_vfs_symlink_dotl(struct inode *dir, struct dentry *dentry, goto error; } + v9fs_invalidate_inode_attr(dir); if (v9ses->cache) { /* Now walk from the parent so we can get an unopened fid. */ fid = p9_client_walk(dfid, 1, &name, 1); @@ -546,7 +600,7 @@ v9fs_vfs_symlink_dotl(struct inode *dir, struct dentry *dentry, } /* instantiate inode and assign the unopened fid to dentry */ - inode = v9fs_inode_from_fid(v9ses, fid, dir->i_sb); + inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb); if (IS_ERR(inode)) { err = PTR_ERR(inode); P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", @@ -588,10 +642,10 @@ v9fs_vfs_link_dotl(struct dentry *old_dentry, struct inode *dir, struct dentry *dentry) { int err; - struct p9_fid *dfid, *oldfid; char *name; - struct v9fs_session_info *v9ses; struct dentry *dir_dentry; + struct p9_fid *dfid, *oldfid; + struct v9fs_session_info *v9ses; P9_DPRINTK(P9_DEBUG_VFS, "dir ino: %lu, old_name: %s, new_name: %s\n", dir->i_ino, old_dentry->d_name.name, @@ -616,29 +670,17 @@ v9fs_vfs_link_dotl(struct dentry *old_dentry, struct inode *dir, return err; } + v9fs_invalidate_inode_attr(dir); if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) { /* Get the latest stat info from server. */ struct p9_fid *fid; - struct p9_stat_dotl *st; - fid = v9fs_fid_lookup(old_dentry); if (IS_ERR(fid)) return PTR_ERR(fid); - st = p9_client_getattr_dotl(fid, P9_STATS_BASIC); - if (IS_ERR(st)) - return PTR_ERR(st); - - v9fs_stat2inode_dotl(st, old_dentry->d_inode); - - kfree(st); - } else { - /* Caching disabled. No need to get upto date stat info. - * This dentry will be released immediately. So, just hold the - * inode - */ - ihold(old_dentry->d_inode); + v9fs_refresh_inode_dotl(fid, old_dentry->d_inode); } + ihold(old_dentry->d_inode); d_instantiate(dentry, old_dentry->d_inode); return err; @@ -657,12 +699,12 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, int omode, dev_t rdev) { int err; + gid_t gid; char *name; mode_t mode; struct v9fs_session_info *v9ses; struct p9_fid *fid = NULL, *dfid = NULL; struct inode *inode; - gid_t gid; struct p9_qid qid; struct dentry *dir_dentry; struct posix_acl *dacl = NULL, *pacl = NULL; @@ -699,6 +741,7 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, int omode, if (err < 0) goto error; + v9fs_invalidate_inode_attr(dir); /* instantiate inode and assign the unopened fid to the dentry */ if (v9ses->cache == CACHE_LOOSE || v9ses->cache == CACHE_FSCACHE) { fid = p9_client_walk(dfid, 1, &name, 1); @@ -710,7 +753,7 @@ v9fs_vfs_mknod_dotl(struct inode *dir, struct dentry *dentry, int omode, goto error; } - inode = v9fs_inode_from_fid(v9ses, fid, dir->i_sb); + inode = v9fs_get_inode_from_fid(v9ses, fid, dir->i_sb); if (IS_ERR(inode)) { err = PTR_ERR(inode); P9_DPRINTK(P9_DEBUG_VFS, "inode creation failed %d\n", @@ -782,6 +825,31 @@ ndset: return NULL; } +int v9fs_refresh_inode_dotl(struct p9_fid *fid, struct inode *inode) +{ + loff_t i_size; + struct p9_stat_dotl *st; + struct v9fs_session_info *v9ses; + + v9ses = v9fs_inode2v9ses(inode); + st = p9_client_getattr_dotl(fid, P9_STATS_ALL); + if (IS_ERR(st)) + return PTR_ERR(st); + + spin_lock(&inode->i_lock); + /* + * We don't want to refresh inode->i_size, + * because we may have cached data + */ + i_size = inode->i_size; + v9fs_stat2inode_dotl(st, inode); + if (v9ses->cache) + inode->i_size = i_size; + spin_unlock(&inode->i_lock); + kfree(st); + return 0; +} + const struct inode_operations v9fs_dir_inode_operations_dotl = { .create = v9fs_vfs_create_dotl, .lookup = v9fs_vfs_lookup, diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index dbaabe3..09fd08d 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -86,12 +86,15 @@ v9fs_fill_super(struct super_block *sb, struct v9fs_session_info *v9ses, } else sb->s_op = &v9fs_super_ops; sb->s_bdi = &v9ses->bdi; + if (v9ses->cache) + sb->s_bdi->ra_pages = (VM_MAX_READAHEAD * 1024)/PAGE_CACHE_SIZE; - sb->s_flags = flags | MS_ACTIVE | MS_SYNCHRONOUS | MS_DIRSYNC | - MS_NOATIME; + sb->s_flags = flags | MS_ACTIVE | MS_DIRSYNC | MS_NOATIME; + if (!v9ses->cache) + sb->s_flags |= MS_SYNCHRONOUS; #ifdef CONFIG_9P_FS_POSIX_ACL - if ((v9ses->flags & V9FS_ACCESS_MASK) == V9FS_ACCESS_CLIENT) + if ((v9ses->flags & V9FS_ACL_MASK) == V9FS_POSIX_ACL) sb->s_flags |= MS_POSIXACL; #endif @@ -151,7 +154,6 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, retval = PTR_ERR(inode); goto release_sb; } - root = d_alloc_root(inode); if (!root) { iput(inode); @@ -166,7 +168,7 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, retval = PTR_ERR(st); goto release_sb; } - + root->d_inode->i_ino = v9fs_qid2ino(&st->qid); v9fs_stat2inode_dotl(st, root->d_inode); kfree(st); } else { @@ -183,10 +185,21 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, p9stat_free(st); kfree(st); } + v9fs_fid_add(root, fid); retval = v9fs_get_acl(inode, fid); if (retval) goto release_sb; - v9fs_fid_add(root, fid); + /* + * Add the root fid to session info. This is used + * for file system sync. We want a cloned fid here + * so that we can do a sync_filesystem after a + * shrink_dcache_for_umount + */ + v9ses->root_fid = v9fs_fid_clone(root); + if (IS_ERR(v9ses->root_fid)) { + retval = PTR_ERR(v9ses->root_fid); + goto release_sb; + } P9_DPRINTK(P9_DEBUG_VFS, " simple set mount, return 0\n"); return dget(sb->s_root); @@ -197,15 +210,11 @@ close_session: v9fs_session_close(v9ses); kfree(v9ses); return ERR_PTR(retval); - release_sb: /* - * we will do the session_close and root dentry release - * in the below call. But we need to clunk fid, because we haven't - * attached the fid to dentry so it won't get clunked - * automatically. + * we will do the session_close and root dentry + * release in the below call. */ - p9_client_clunk(fid); deactivate_locked_super(sb); return ERR_PTR(retval); } @@ -223,7 +232,7 @@ static void v9fs_kill_super(struct super_block *s) P9_DPRINTK(P9_DEBUG_VFS, " %p\n", s); kill_anon_super(s); - + p9_client_clunk(v9ses->root_fid); v9fs_session_cancel(v9ses); v9fs_session_close(v9ses); kfree(v9ses); @@ -276,11 +285,31 @@ done: return res; } +static int v9fs_sync_fs(struct super_block *sb, int wait) +{ + struct v9fs_session_info *v9ses = sb->s_fs_info; + + P9_DPRINTK(P9_DEBUG_VFS, "v9fs_sync_fs: super_block %p\n", sb); + return p9_client_sync_fs(v9ses->root_fid); +} + +static int v9fs_drop_inode(struct inode *inode) +{ + struct v9fs_session_info *v9ses; + v9ses = v9fs_inode2v9ses(inode); + if (v9ses->cache) + return generic_drop_inode(inode); + /* + * in case of non cached mode always drop the + * the inode because we want the inode attribute + * to always match that on the server. + */ + return 1; +} + static const struct super_operations v9fs_super_ops = { -#ifdef CONFIG_9P_FSCACHE .alloc_inode = v9fs_alloc_inode, .destroy_inode = v9fs_destroy_inode, -#endif .statfs = simple_statfs, .evict_inode = v9fs_evict_inode, .show_options = generic_show_options, @@ -288,11 +317,11 @@ static const struct super_operations v9fs_super_ops = { }; static const struct super_operations v9fs_super_ops_dotl = { -#ifdef CONFIG_9P_FSCACHE .alloc_inode = v9fs_alloc_inode, .destroy_inode = v9fs_destroy_inode, -#endif + .sync_fs = v9fs_sync_fs, .statfs = v9fs_statfs, + .drop_inode = v9fs_drop_inode, .evict_inode = v9fs_evict_inode, .show_options = generic_show_options, .umount_begin = v9fs_umount_begin, @@ -303,5 +332,5 @@ struct file_system_type v9fs_fs_type = { .mount = v9fs_mount, .kill_sb = v9fs_kill_super, .owner = THIS_MODULE, - .fs_flags = FS_RENAME_DOES_D_MOVE, + .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT, }; @@ -47,10 +47,10 @@ config FS_POSIX_ACL def_bool n config EXPORTFS - tristate + bool config FILE_LOCKING - bool "Enable POSIX file locking API" if EMBEDDED + bool "Enable POSIX file locking API" if EXPERT default y help This option enables standard file locking support, required @@ -187,6 +187,7 @@ source "fs/omfs/Kconfig" source "fs/hpfs/Kconfig" source "fs/qnx4/Kconfig" source "fs/romfs/Kconfig" +source "fs/pstore/Kconfig" source "fs/sysv/Kconfig" source "fs/ufs/Kconfig" source "fs/exofs/Kconfig" diff --git a/fs/Makefile b/fs/Makefile index a7f7cef..fb68c2b 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -48,6 +48,8 @@ obj-$(CONFIG_FS_POSIX_ACL) += posix_acl.o xattr_acl.o obj-$(CONFIG_NFS_COMMON) += nfs_common/ obj-$(CONFIG_GENERIC_ACL) += generic_acl.o +obj-$(CONFIG_FHANDLE) += fhandle.o + obj-y += quota/ obj-$(CONFIG_PROC_FS) += proc/ @@ -121,3 +123,4 @@ obj-$(CONFIG_BTRFS_FS) += btrfs/ obj-$(CONFIG_GFS2_FS) += gfs2/ obj-$(CONFIG_EXOFS_FS) += exofs/ obj-$(CONFIG_CEPH_FS) += ceph/ +obj-$(CONFIG_PSTORE) += pstore/ diff --git a/fs/adfs/Kconfig b/fs/adfs/Kconfig index 1dd5f34..e55182a 100644 --- a/fs/adfs/Kconfig +++ b/fs/adfs/Kconfig @@ -1,7 +1,6 @@ config ADFS_FS tristate "ADFS file system support (EXPERIMENTAL)" depends on BLOCK && EXPERIMENTAL - depends on BKL # need to fix help The Acorn Disc Filing System is the standard file system of the RiscOS operating system which runs on Acorn's ARM-based Risc PC diff --git a/fs/adfs/dir.c b/fs/adfs/dir.c index 3b4a764..3d83075a 100644 --- a/fs/adfs/dir.c +++ b/fs/adfs/dir.c @@ -9,7 +9,6 @@ * * Common directory handling for ADFS */ -#include <linux/smp_lock.h> #include "adfs.h" /* @@ -27,8 +26,6 @@ adfs_readdir(struct file *filp, void *dirent, filldir_t filldir) struct adfs_dir dir; int ret = 0; - lock_kernel(); - if (filp->f_pos >> 32) goto out; @@ -70,7 +67,6 @@ free_out: ops->free(&dir); out: - unlock_kernel(); return ret; } @@ -276,7 +272,6 @@ adfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) struct object_info obj; int error; - lock_kernel(); error = adfs_dir_lookup_byname(dir, &dentry->d_name, &obj); if (error == 0) { error = -EACCES; @@ -288,7 +283,6 @@ adfs_lookup(struct inode *dir, struct dentry *dentry, struct nameidata *nd) if (inode) error = 0; } - unlock_kernel(); d_add(dentry, inode); return ERR_PTR(error); } diff --git a/fs/adfs/inode.c b/fs/adfs/inode.c index 65794b8..09fe401 100644 --- a/fs/adfs/inode.c +++ b/fs/adfs/inode.c @@ -7,7 +7,6 @@ * it under the terms of the GNU General Public License version 2 as * published by the Free Software Foundation. */ -#include <linux/smp_lock.h> #include <linux/buffer_head.h> #include <linux/writeback.h> #include "adfs.h" @@ -316,8 +315,6 @@ adfs_notify_change(struct dentry *dentry, struct iattr *attr) unsigned int ia_valid = attr->ia_valid; int error; - lock_kernel(); - error = inode_change_ok(inode, attr); /* @@ -359,7 +356,6 @@ adfs_notify_change(struct dentry *dentry, struct iattr *attr) if (ia_valid & (ATTR_SIZE | ATTR_MTIME | ATTR_MODE)) mark_inode_dirty(inode); out: - unlock_kernel(); return error; } @@ -374,7 +370,6 @@ int adfs_write_inode(struct inode *inode, struct writeback_control *wbc) struct object_info obj; int ret; - lock_kernel(); obj.file_id = inode->i_ino; obj.name_len = 0; obj.parent_id = ADFS_I(inode)->parent_id; @@ -384,6 +379,5 @@ int adfs_write_inode(struct inode *inode, struct writeback_control *wbc) obj.size = inode->i_size; ret = adfs_dir_update(sb, &obj, wbc->sync_mode == WB_SYNC_ALL); - unlock_kernel(); return ret; } diff --git a/fs/adfs/super.c b/fs/adfs/super.c index 2d79540..06d7388 100644 --- a/fs/adfs/super.c +++ b/fs/adfs/super.c @@ -14,7 +14,6 @@ #include <linux/mount.h> #include <linux/seq_file.h> #include <linux/slab.h> -#include <linux/smp_lock.h> #include <linux/statfs.h> #include "adfs.h" #include "dir_f.h" @@ -120,15 +119,11 @@ static void adfs_put_super(struct super_block *sb) int i; struct adfs_sb_info *asb = ADFS_SB(sb); - lock_kernel(); - for (i = 0; i < asb->s_map_size; i++) brelse(asb->s_map[i].dm_bh); kfree(asb->s_map); kfree(asb); sb->s_fs_info = NULL; - - unlock_kernel(); } static int adfs_show_options(struct seq_file *seq, struct vfsmount *mnt) @@ -359,15 +354,11 @@ static int adfs_fill_super(struct super_block *sb, void *data, int silent) struct adfs_sb_info *asb; struct inode *root; - lock_kernel(); - sb->s_flags |= MS_NODIRATIME; asb = kzalloc(sizeof(*asb), GFP_KERNEL); - if (!asb) { - unlock_kernel(); + if (!asb) return -ENOMEM; - } sb->s_fs_info = asb; /* set default options */ @@ -485,7 +476,6 @@ static int adfs_fill_super(struct super_block *sb, void *data, int silent) adfs_error(sb, "get root inode failed\n"); goto error; } - unlock_kernel(); return 0; error_free_bh: @@ -493,7 +483,6 @@ error_free_bh: error: sb->s_fs_info = NULL; kfree(asb); - unlock_kernel(); return -EINVAL; } diff --git a/fs/affs/Makefile b/fs/affs/Makefile index b2c4f54..3988b4a 100644 --- a/fs/affs/Makefile +++ b/fs/affs/Makefile @@ -2,7 +2,7 @@ # Makefile for the Linux affs filesystem routines. # -#EXTRA_CFLAGS=-DDEBUG=1 +#ccflags-y := -DDEBUG=1 obj-$(CONFIG_AFFS_FS) += affs.o diff --git a/fs/afs/write.c b/fs/afs/write.c index 15690bb..789b3af 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -140,6 +140,7 @@ int afs_write_begin(struct file *file, struct address_space *mapping, candidate->first = candidate->last = index; candidate->offset_first = from; candidate->to_last = to; + INIT_LIST_HEAD(&candidate->link); candidate->usage = 1; candidate->state = AFS_WBACK_PENDING; init_waitqueue_head(&candidate->waitq); @@ -85,7 +85,7 @@ static int __init aio_setup(void) kiocb_cachep = KMEM_CACHE(kiocb, SLAB_HWCACHE_ALIGN|SLAB_PANIC); kioctx_cachep = KMEM_CACHE(kioctx,SLAB_HWCACHE_ALIGN|SLAB_PANIC); - aio_wq = create_workqueue("aio"); + aio_wq = alloc_workqueue("aio", 0, 1); /* used to limit concurrency */ abe_pool = mempool_create_kmalloc_pool(1, sizeof(struct aio_batch_entry)); BUG_ON(!aio_wq || !abe_pool); @@ -239,15 +239,23 @@ static void __put_ioctx(struct kioctx *ctx) call_rcu(&ctx->rcu_head, ctx_rcu_free); } -#define get_ioctx(kioctx) do { \ - BUG_ON(atomic_read(&(kioctx)->users) <= 0); \ - atomic_inc(&(kioctx)->users); \ -} while (0) -#define put_ioctx(kioctx) do { \ - BUG_ON(atomic_read(&(kioctx)->users) <= 0); \ - if (unlikely(atomic_dec_and_test(&(kioctx)->users))) \ - __put_ioctx(kioctx); \ -} while (0) +static inline void get_ioctx(struct kioctx *kioctx) +{ + BUG_ON(atomic_read(&kioctx->users) <= 0); + atomic_inc(&kioctx->users); +} + +static inline int try_get_ioctx(struct kioctx *kioctx) +{ + return atomic_inc_not_zero(&kioctx->users); +} + +static inline void put_ioctx(struct kioctx *kioctx) +{ + BUG_ON(atomic_read(&kioctx->users) <= 0); + if (unlikely(atomic_dec_and_test(&kioctx->users))) + __put_ioctx(kioctx); +} /* ioctx_alloc * Allocates and initializes an ioctx. Returns an ERR_PTR if it failed. @@ -569,7 +577,7 @@ static int __aio_put_req(struct kioctx *ctx, struct kiocb *req) spin_lock(&fput_lock); list_add(&req->ki_list, &fput_head); spin_unlock(&fput_lock); - queue_work(aio_wq, &fput_work); + schedule_work(&fput_work); } else { req->ki_filp = NULL; really_put_req(ctx, req); @@ -601,8 +609,13 @@ static struct kioctx *lookup_ioctx(unsigned long ctx_id) rcu_read_lock(); hlist_for_each_entry_rcu(ctx, n, &mm->ioctx_list, list) { - if (ctx->user_id == ctx_id && !ctx->dead) { - get_ioctx(ctx); + /* + * RCU protects us against accessing freed memory but + * we have to be careful not to get a reference when the + * reference count already dropped to 0 (ctx->dead test + * is unreliable because of races). + */ + if (ctx->user_id == ctx_id && !ctx->dead && try_get_ioctx(ctx)){ ret = ctx; break; } @@ -1629,6 +1642,23 @@ static int io_submit_one(struct kioctx *ctx, struct iocb __user *user_iocb, goto out_put_req; spin_lock_irq(&ctx->ctx_lock); + /* + * We could have raced with io_destroy() and are currently holding a + * reference to ctx which should be destroyed. We cannot submit IO + * since ctx gets freed as soon as io_submit() puts its reference. The + * check here is reliable: io_destroy() sets ctx->dead before waiting + * for outstanding IO and the barrier between these two is realized by + * unlock of mm->ioctx_lock and lock of ctx->ctx_lock. Analogously we + * increment ctx->reqs_active before checking for ctx->dead and the + * barrier is realized by unlock and lock of ctx->ctx_lock. Thus if we + * don't see ctx->dead set here, io_destroy() waits for our IO to + * finish. + */ + if (ctx->dead) { + spin_unlock_irq(&ctx->ctx_lock); + ret = -EINVAL; + goto out_put_req; + } aio_run_iocb(req); if (!list_empty(&ctx->run_list)) { /* drain the run list */ diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c index 014e7ab..e6f84d2 100644 --- a/fs/autofs4/root.c +++ b/fs/autofs4/root.c @@ -36,7 +36,7 @@ static long autofs4_root_compat_ioctl(struct file *,unsigned int,unsigned long); static int autofs4_dir_open(struct inode *inode, struct file *file); static struct dentry *autofs4_lookup(struct inode *,struct dentry *, struct nameidata *); static struct vfsmount *autofs4_d_automount(struct path *); -static int autofs4_d_manage(struct dentry *, bool, bool); +static int autofs4_d_manage(struct dentry *, bool); static void autofs4_dentry_release(struct dentry *); const struct file_operations autofs4_root_operations = { @@ -446,7 +446,7 @@ done: return NULL; } -int autofs4_d_manage(struct dentry *dentry, bool mounting_here, bool rcu_walk) +int autofs4_d_manage(struct dentry *dentry, bool rcu_walk) { struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb); @@ -454,7 +454,7 @@ int autofs4_d_manage(struct dentry *dentry, bool mounting_here, bool rcu_walk) dentry, dentry->d_name.len, dentry->d_name.name); /* The daemon never waits. */ - if (autofs4_oz_mode(sbi) || mounting_here) { + if (autofs4_oz_mode(sbi)) { if (!d_mountpoint(dentry)) return -EISDIR; return 0; diff --git a/fs/bfs/dir.c b/fs/bfs/dir.c index 685ecff..b14cebf 100644 --- a/fs/bfs/dir.c +++ b/fs/bfs/dir.c @@ -97,7 +97,7 @@ static int bfs_create(struct inode *dir, struct dentry *dentry, int mode, if (!inode) return -ENOSPC; mutex_lock(&info->bfs_lock); - ino = find_first_zero_bit(info->si_imap, info->si_lasti); + ino = find_first_zero_bit(info->si_imap, info->si_lasti + 1); if (ino > info->si_lasti) { mutex_unlock(&info->bfs_lock); iput(inode); diff --git a/fs/block_dev.c b/fs/block_dev.c index 333a7bb..8892870 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -873,6 +873,11 @@ int bd_link_disk_holder(struct block_device *bdev, struct gendisk *disk) ret = add_symlink(bdev->bd_part->holder_dir, &disk_to_dev(disk)->kobj); if (ret) goto out_del; + /* + * bdev could be deleted beneath us which would implicitly destroy + * the holder directory. Hold on to it. + */ + kobject_get(bdev->bd_part->holder_dir); list_add(&holder->list, &bdev->bd_holder_disks); goto out_unlock; @@ -909,6 +914,7 @@ void bd_unlink_disk_holder(struct block_device *bdev, struct gendisk *disk) del_symlink(disk->slave_dir, &part_to_dev(bdev->bd_part)->kobj); del_symlink(bdev->bd_part->holder_dir, &disk_to_dev(disk)->kobj); + kobject_put(bdev->bd_part->holder_dir); list_del_init(&holder->list); kfree(holder); } @@ -922,14 +928,15 @@ EXPORT_SYMBOL_GPL(bd_unlink_disk_holder); * flush_disk - invalidates all buffer-cache entries on a disk * * @bdev: struct block device to be flushed + * @kill_dirty: flag to guide handling of dirty inodes * * Invalidates all buffer-cache entries on a disk. It should be called * when a disk has been changed -- either by a media change or online * resize. */ -static void flush_disk(struct block_device *bdev) +static void flush_disk(struct block_device *bdev, bool kill_dirty) { - if (__invalidate_device(bdev)) { + if (__invalidate_device(bdev, kill_dirty)) { char name[BDEVNAME_SIZE] = ""; if (bdev->bd_disk) @@ -966,7 +973,7 @@ void check_disk_size_change(struct gendisk *disk, struct block_device *bdev) "%s: detected capacity change from %lld to %lld\n", name, bdev_size, disk_size); i_size_write(bdev->bd_inode, disk_size); - flush_disk(bdev); + flush_disk(bdev, false); } } EXPORT_SYMBOL(check_disk_size_change); @@ -1019,7 +1026,7 @@ int check_disk_change(struct block_device *bdev) if (!(events & DISK_EVENT_MEDIA_CHANGE)) return 0; - flush_disk(bdev); + flush_disk(bdev, true); if (bdops->revalidate_disk) bdops->revalidate_disk(bdev->bd_disk); return 1; @@ -1215,12 +1222,6 @@ int blkdev_get(struct block_device *bdev, fmode_t mode, void *holder) res = __blkdev_get(bdev, mode, 0); - /* __blkdev_get() may alter read only status, check it afterwards */ - if (!res && (mode & FMODE_WRITE) && bdev_read_only(bdev)) { - __blkdev_put(bdev, mode, 0); - res = -EACCES; - } - if (whole) { /* finish claiming */ mutex_lock(&bdev->bd_mutex); @@ -1298,6 +1299,11 @@ struct block_device *blkdev_get_by_path(const char *path, fmode_t mode, if (err) return ERR_PTR(err); + if ((mode & FMODE_WRITE) && bdev_read_only(bdev)) { + blkdev_put(bdev, mode); + return ERR_PTR(-EACCES); + } + return bdev; } EXPORT_SYMBOL(blkdev_get_by_path); @@ -1601,7 +1607,7 @@ fail: } EXPORT_SYMBOL(lookup_bdev); -int __invalidate_device(struct block_device *bdev) +int __invalidate_device(struct block_device *bdev, bool kill_dirty) { struct super_block *sb = get_super(bdev); int res = 0; @@ -1614,7 +1620,7 @@ int __invalidate_device(struct block_device *bdev) * hold). */ shrink_dcache_sb(sb); - res = invalidate_inodes(sb); + res = invalidate_inodes(sb, kill_dirty); drop_super(sb); } invalidate_bdev(bdev); diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c index 15b5ca2..9c94934 100644 --- a/fs/btrfs/acl.c +++ b/fs/btrfs/acl.c @@ -37,6 +37,9 @@ static struct posix_acl *btrfs_get_acl(struct inode *inode, int type) char *value = NULL; struct posix_acl *acl; + if (!IS_POSIXACL(inode)) + return NULL; + acl = get_cached_acl(inode, type); if (acl != ACL_NOT_CACHED) return acl; @@ -84,6 +87,9 @@ static int btrfs_xattr_acl_get(struct dentry *dentry, const char *name, struct posix_acl *acl; int ret = 0; + if (!IS_POSIXACL(dentry->d_inode)) + return -EOPNOTSUPP; + acl = btrfs_get_acl(dentry->d_inode, type); if (IS_ERR(acl)) diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c index f745287..4d2110e 100644 --- a/fs/btrfs/compression.c +++ b/fs/btrfs/compression.c @@ -562,7 +562,7 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, u64 em_len; u64 em_start; struct extent_map *em; - int ret; + int ret = -ENOMEM; u32 *sums; tree = &BTRFS_I(inode)->io_tree; @@ -577,6 +577,9 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, compressed_len = em->block_len; cb = kmalloc(compressed_bio_size(root, compressed_len), GFP_NOFS); + if (!cb) + goto out; + atomic_set(&cb->pending_bios, 0); cb->errors = 0; cb->inode = inode; @@ -597,13 +600,18 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, nr_pages = (compressed_len + PAGE_CACHE_SIZE - 1) / PAGE_CACHE_SIZE; - cb->compressed_pages = kmalloc(sizeof(struct page *) * nr_pages, + cb->compressed_pages = kzalloc(sizeof(struct page *) * nr_pages, GFP_NOFS); + if (!cb->compressed_pages) + goto fail1; + bdev = BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev; for (page_index = 0; page_index < nr_pages; page_index++) { cb->compressed_pages[page_index] = alloc_page(GFP_NOFS | __GFP_HIGHMEM); + if (!cb->compressed_pages[page_index]) + goto fail2; } cb->nr_pages = nr_pages; @@ -614,6 +622,8 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, cb->len = uncompressed_len; comp_bio = compressed_bio_alloc(bdev, cur_disk_byte, GFP_NOFS); + if (!comp_bio) + goto fail2; comp_bio->bi_private = cb; comp_bio->bi_end_io = end_compressed_bio_read; atomic_inc(&cb->pending_bios); @@ -681,6 +691,17 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, bio_put(comp_bio); return 0; + +fail2: + for (page_index = 0; page_index < nr_pages; page_index++) + free_page((unsigned long)cb->compressed_pages[page_index]); + + kfree(cb->compressed_pages); +fail1: + kfree(cb); +out: + free_extent_map(em); + return ret; } static struct list_head comp_idle_workspace[BTRFS_COMPRESS_TYPES]; @@ -900,7 +921,7 @@ int btrfs_decompress(int type, unsigned char *data_in, struct page *dest_page, return ret; } -void __exit btrfs_exit_compress(void) +void btrfs_exit_compress(void) { free_workspaces(); } diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 2c98b3a..7f78cc7 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -729,6 +729,15 @@ struct btrfs_space_info { u64 disk_total; /* total bytes on disk, takes mirrors into account */ + /* + * we bump reservation progress every time we decrement + * bytes_reserved. This way people waiting for reservations + * know something good has happened and they can check + * for progress. The number here isn't to be trusted, it + * just shows reclaim activity + */ + unsigned long reservation_progress; + int full; /* indicates that we cannot allocate any more chunks for this space */ int force_alloc; /* set if we need to force a chunk alloc for @@ -1254,6 +1263,7 @@ struct btrfs_root { #define BTRFS_MOUNT_SPACE_CACHE (1 << 12) #define BTRFS_MOUNT_CLEAR_CACHE (1 << 13) #define BTRFS_MOUNT_USER_SUBVOL_RM_ALLOWED (1 << 14) +#define BTRFS_MOUNT_ENOSPC_DEBUG (1 << 15) #define btrfs_clear_opt(o, opt) ((o) &= ~BTRFS_MOUNT_##opt) #define btrfs_set_opt(o, opt) ((o) |= BTRFS_MOUNT_##opt) @@ -2218,6 +2228,8 @@ int btrfs_error_unpin_extent_range(struct btrfs_root *root, u64 start, u64 end); int btrfs_error_discard_extent(struct btrfs_root *root, u64 bytenr, u64 num_bytes); +int btrfs_force_chunk_alloc(struct btrfs_trans_handle *trans, + struct btrfs_root *root, u64 type); /* ctree.c */ int btrfs_bin_search(struct extent_buffer *eb, struct btrfs_key *key, diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index b531c36..100b07f 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -359,10 +359,14 @@ static int csum_dirty_buffer(struct btrfs_root *root, struct page *page) tree = &BTRFS_I(page->mapping->host)->io_tree; - if (page->private == EXTENT_PAGE_PRIVATE) + if (page->private == EXTENT_PAGE_PRIVATE) { + WARN_ON(1); goto out; - if (!page->private) + } + if (!page->private) { + WARN_ON(1); goto out; + } len = page->private >> 2; WARN_ON(len == 0); @@ -1550,6 +1554,7 @@ static int transaction_kthread(void *arg) spin_unlock(&root->fs_info->new_trans_lock); trans = btrfs_join_transaction(root, 1); + BUG_ON(IS_ERR(trans)); if (transid == trans->transid) { ret = btrfs_commit_transaction(trans, root); BUG_ON(ret); @@ -2453,10 +2458,14 @@ int btrfs_commit_super(struct btrfs_root *root) up_write(&root->fs_info->cleanup_work_sem); trans = btrfs_join_transaction(root, 1); + if (IS_ERR(trans)) + return PTR_ERR(trans); ret = btrfs_commit_transaction(trans, root); BUG_ON(ret); /* run commit again to drop the original snapshot */ trans = btrfs_join_transaction(root, 1); + if (IS_ERR(trans)) + return PTR_ERR(trans); btrfs_commit_transaction(trans, root); ret = btrfs_write_and_wait_transaction(NULL, root); BUG_ON(ret); @@ -2484,7 +2493,7 @@ int close_ctree(struct btrfs_root *root) * ERROR state on disk. * * 2. when btrfs flips readonly just in btrfs_commit_super, - * and in such case, btrfs cannnot write sb via btrfs_commit_super, + * and in such case, btrfs cannot write sb via btrfs_commit_super, * and since fs_state has been set BTRFS_SUPER_FLAG_ERROR flag, * btrfs will cleanup all FS resources first and write sb then. */ @@ -2554,6 +2563,8 @@ int close_ctree(struct btrfs_root *root) kfree(fs_info->chunk_root); kfree(fs_info->dev_root); kfree(fs_info->csum_root); + kfree(fs_info); + return 0; } diff --git a/fs/btrfs/export.c b/fs/btrfs/export.c index 9786963..b4ffad8 100644 --- a/fs/btrfs/export.c +++ b/fs/btrfs/export.c @@ -21,9 +21,13 @@ static int btrfs_encode_fh(struct dentry *dentry, u32 *fh, int *max_len, int len = *max_len; int type; - if ((len < BTRFS_FID_SIZE_NON_CONNECTABLE) || - (connectable && len < BTRFS_FID_SIZE_CONNECTABLE)) + if (connectable && (len < BTRFS_FID_SIZE_CONNECTABLE)) { + *max_len = BTRFS_FID_SIZE_CONNECTABLE; return 255; + } else if (len < BTRFS_FID_SIZE_NON_CONNECTABLE) { + *max_len = BTRFS_FID_SIZE_NON_CONNECTABLE; + return 255; + } len = BTRFS_FID_SIZE_NON_CONNECTABLE; type = FILEID_BTRFS_WITHOUT_PARENT; @@ -171,6 +175,8 @@ static struct dentry *btrfs_get_parent(struct dentry *child) int ret; path = btrfs_alloc_path(); + if (!path) + return ERR_PTR(-ENOMEM); if (dir->i_ino == BTRFS_FIRST_FREE_OBJECTID) { key.objectid = root->root_key.objectid; diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index b552693..7b3089b 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -320,11 +320,6 @@ static int caching_kthread(void *data) if (!path) return -ENOMEM; - exclude_super_stripes(extent_root, block_group); - spin_lock(&block_group->space_info->lock); - block_group->space_info->bytes_readonly += block_group->bytes_super; - spin_unlock(&block_group->space_info->lock); - last = max_t(u64, block_group->key.objectid, BTRFS_SUPER_INFO_OFFSET); /* @@ -467,8 +462,10 @@ static int cache_block_group(struct btrfs_block_group_cache *cache, cache->cached = BTRFS_CACHE_NO; } spin_unlock(&cache->lock); - if (ret == 1) + if (ret == 1) { + free_excluded_extents(fs_info->extent_root, cache); return 0; + } } if (load_cache_only) @@ -3344,21 +3341,24 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans, u64 reserved; u64 max_reclaim; u64 reclaimed = 0; - int pause = 1; + long time_left; int nr_pages = (2 * 1024 * 1024) >> PAGE_CACHE_SHIFT; + int loops = 0; + unsigned long progress; block_rsv = &root->fs_info->delalloc_block_rsv; space_info = block_rsv->space_info; smp_mb(); reserved = space_info->bytes_reserved; + progress = space_info->reservation_progress; if (reserved == 0) return 0; max_reclaim = min(reserved, to_reclaim); - while (1) { + while (loops < 1024) { /* have the flusher threads jump in and do some IO */ smp_mb(); nr_pages = min_t(unsigned long, nr_pages, @@ -3371,17 +3371,31 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans, reserved = space_info->bytes_reserved; spin_unlock(&space_info->lock); + loops++; + if (reserved == 0 || reclaimed >= max_reclaim) break; if (trans && trans->transaction->blocked) return -EAGAIN; - __set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(pause); - pause <<= 1; - if (pause > HZ / 10) - pause = HZ / 10; + time_left = schedule_timeout_interruptible(1); + + /* We were interrupted, exit */ + if (time_left) + break; + + /* we've kicked the IO a few times, if anything has been freed, + * exit. There is no sense in looping here for a long time + * when we really need to commit the transaction, or there are + * just too many writers without enough free space + */ + + if (loops > 3) { + smp_mb(); + if (progress != space_info->reservation_progress) + break; + } } return reclaimed >= to_reclaim; @@ -3588,10 +3602,23 @@ void block_rsv_release_bytes(struct btrfs_block_rsv *block_rsv, if (num_bytes > 0) { if (dest) { - block_rsv_add_bytes(dest, num_bytes, 0); - } else { + spin_lock(&dest->lock); + if (!dest->full) { + u64 bytes_to_add; + + bytes_to_add = dest->size - dest->reserved; + bytes_to_add = min(num_bytes, bytes_to_add); + dest->reserved += bytes_to_add; + if (dest->reserved >= dest->size) + dest->full = 1; + num_bytes -= bytes_to_add; + } + spin_unlock(&dest->lock); + } + if (num_bytes) { spin_lock(&space_info->lock); space_info->bytes_reserved -= num_bytes; + space_info->reservation_progress++; spin_unlock(&space_info->lock); } } @@ -3824,6 +3851,7 @@ static void update_global_block_rsv(struct btrfs_fs_info *fs_info) if (block_rsv->reserved >= block_rsv->size) { num_bytes = block_rsv->reserved - block_rsv->size; sinfo->bytes_reserved -= num_bytes; + sinfo->reservation_progress++; block_rsv->reserved = block_rsv->size; block_rsv->full = 1; } @@ -3985,7 +4013,6 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) to_reserve = 0; } spin_unlock(&BTRFS_I(inode)->accounting_lock); - to_reserve += calc_csum_metadata_size(inode, num_bytes); ret = reserve_metadata_bytes(NULL, root, block_rsv, to_reserve, 1); if (ret) @@ -4012,6 +4039,7 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes) num_bytes = ALIGN(num_bytes, root->sectorsize); atomic_dec(&BTRFS_I(inode)->outstanding_extents); + WARN_ON(atomic_read(&BTRFS_I(inode)->outstanding_extents) < 0); spin_lock(&BTRFS_I(inode)->accounting_lock); nr_extents = atomic_read(&BTRFS_I(inode)->outstanding_extents); @@ -4112,6 +4140,7 @@ static int update_block_group(struct btrfs_trans_handle *trans, btrfs_set_block_group_used(&cache->item, old_val); cache->reserved -= num_bytes; cache->space_info->bytes_reserved -= num_bytes; + cache->space_info->reservation_progress++; cache->space_info->bytes_used += num_bytes; cache->space_info->disk_used += num_bytes * factor; spin_unlock(&cache->lock); @@ -4163,6 +4192,7 @@ static int pin_down_extent(struct btrfs_root *root, if (reserved) { cache->reserved -= num_bytes; cache->space_info->bytes_reserved -= num_bytes; + cache->space_info->reservation_progress++; } spin_unlock(&cache->lock); spin_unlock(&cache->space_info->lock); @@ -4213,6 +4243,7 @@ static int update_reserved_bytes(struct btrfs_block_group_cache *cache, space_info->bytes_readonly += num_bytes; cache->reserved -= num_bytes; space_info->bytes_reserved -= num_bytes; + space_info->reservation_progress++; } spin_unlock(&cache->lock); spin_unlock(&space_info->lock); @@ -4691,6 +4722,7 @@ void btrfs_free_tree_block(struct btrfs_trans_handle *trans, if (ret) { spin_lock(&cache->space_info->lock); cache->space_info->bytes_reserved -= buf->len; + cache->space_info->reservation_progress++; spin_unlock(&cache->space_info->lock); } goto out; @@ -5355,7 +5387,7 @@ again: num_bytes, data, 1); goto again; } - if (ret == -ENOSPC) { + if (ret == -ENOSPC && btrfs_test_opt(root, ENOSPC_DEBUG)) { struct btrfs_space_info *sinfo; sinfo = __find_space_info(root->fs_info, data); @@ -5633,6 +5665,7 @@ use_block_rsv(struct btrfs_trans_handle *trans, struct btrfs_root *root, u32 blocksize) { struct btrfs_block_rsv *block_rsv; + struct btrfs_block_rsv *global_rsv = &root->fs_info->global_block_rsv; int ret; block_rsv = get_block_rsv(trans, root); @@ -5640,14 +5673,39 @@ use_block_rsv(struct btrfs_trans_handle *trans, if (block_rsv->size == 0) { ret = reserve_metadata_bytes(trans, root, block_rsv, blocksize, 0); - if (ret) + /* + * If we couldn't reserve metadata bytes try and use some from + * the global reserve. + */ + if (ret && block_rsv != global_rsv) { + ret = block_rsv_use_bytes(global_rsv, blocksize); + if (!ret) + return global_rsv; + return ERR_PTR(ret); + } else if (ret) { return ERR_PTR(ret); + } return block_rsv; } ret = block_rsv_use_bytes(block_rsv, blocksize); if (!ret) return block_rsv; + if (ret) { + WARN_ON(1); + ret = reserve_metadata_bytes(trans, root, block_rsv, blocksize, + 0); + if (!ret) { + spin_lock(&block_rsv->lock); + block_rsv->size += blocksize; + spin_unlock(&block_rsv->lock); + return block_rsv; + } else if (ret && block_rsv != global_rsv) { + ret = block_rsv_use_bytes(global_rsv, blocksize); + if (!ret) + return global_rsv; + } + } return ERR_PTR(-ENOSPC); } @@ -6221,6 +6279,8 @@ int btrfs_drop_snapshot(struct btrfs_root *root, BUG_ON(!wc); trans = btrfs_start_transaction(tree_root, 0); + BUG_ON(IS_ERR(trans)); + if (block_rsv) trans->block_rsv = block_rsv; @@ -6318,6 +6378,7 @@ int btrfs_drop_snapshot(struct btrfs_root *root, btrfs_end_transaction_throttle(trans, tree_root); trans = btrfs_start_transaction(tree_root, 0); + BUG_ON(IS_ERR(trans)); if (block_rsv) trans->block_rsv = block_rsv; } @@ -6446,6 +6507,8 @@ static noinline int relocate_inode_pages(struct inode *inode, u64 start, int ret = 0; ra = kzalloc(sizeof(*ra), GFP_NOFS); + if (!ra) + return -ENOMEM; mutex_lock(&inode->i_mutex); first_index = start >> PAGE_CACHE_SHIFT; @@ -6531,7 +6594,7 @@ static noinline int relocate_data_extent(struct inode *reloc_inode, u64 end = start + extent_key->offset - 1; em = alloc_extent_map(GFP_NOFS); - BUG_ON(!em || IS_ERR(em)); + BUG_ON(!em); em->start = start; em->len = extent_key->offset; @@ -7477,7 +7540,7 @@ int btrfs_drop_dead_reloc_roots(struct btrfs_root *root) BUG_ON(reloc_root->commit_root != NULL); while (1) { trans = btrfs_join_transaction(root, 1); - BUG_ON(!trans); + BUG_ON(IS_ERR(trans)); mutex_lock(&root->fs_info->drop_mutex); ret = btrfs_drop_snapshot(trans, reloc_root); @@ -7535,7 +7598,7 @@ int btrfs_cleanup_reloc_trees(struct btrfs_root *root) if (found) { trans = btrfs_start_transaction(root, 1); - BUG_ON(!trans); + BUG_ON(IS_ERR(trans)); ret = btrfs_commit_transaction(trans, root); BUG_ON(ret); } @@ -7779,7 +7842,7 @@ static noinline int relocate_one_extent(struct btrfs_root *extent_root, trans = btrfs_start_transaction(extent_root, 1); - BUG_ON(!trans); + BUG_ON(IS_ERR(trans)); if (extent_key->objectid == 0) { ret = del_extent_zero(trans, extent_root, path, extent_key); @@ -8013,6 +8076,13 @@ out: return ret; } +int btrfs_force_chunk_alloc(struct btrfs_trans_handle *trans, + struct btrfs_root *root, u64 type) +{ + u64 alloc_flags = get_alloc_profile(root, type); + return do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, 1); +} + /* * helper to account the unused space of all the readonly block group in the * list. takes mirrors into account. @@ -8270,6 +8340,13 @@ int btrfs_free_block_groups(struct btrfs_fs_info *info) if (block_group->cached == BTRFS_CACHE_STARTED) wait_block_group_cache_done(block_group); + /* + * We haven't cached this block group, which means we could + * possibly have excluded extents on this block group. + */ + if (block_group->cached == BTRFS_CACHE_NO) + free_excluded_extents(info->extent_root, block_group); + btrfs_remove_free_space_cache(block_group); btrfs_put_block_group(block_group); @@ -8385,6 +8462,13 @@ int btrfs_read_block_groups(struct btrfs_root *root) cache->sectorsize = root->sectorsize; /* + * We need to exclude the super stripes now so that the space + * info has super bytes accounted for, otherwise we'll think + * we have more space than we actually do. + */ + exclude_super_stripes(root, cache); + + /* * check for two cases, either we are full, and therefore * don't need to bother with the caching work since we won't * find any space, or we are empty, and we can just add all @@ -8392,12 +8476,10 @@ int btrfs_read_block_groups(struct btrfs_root *root) * time, particularly in the full case. */ if (found_key.offset == btrfs_block_group_used(&cache->item)) { - exclude_super_stripes(root, cache); cache->last_byte_to_unpin = (u64)-1; cache->cached = BTRFS_CACHE_FINISHED; free_excluded_extents(root, cache); } else if (btrfs_block_group_used(&cache->item) == 0) { - exclude_super_stripes(root, cache); cache->last_byte_to_unpin = (u64)-1; cache->cached = BTRFS_CACHE_FINISHED; add_new_free_space(cache, root->fs_info, diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 2e993cf..714adc4 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -1433,12 +1433,13 @@ int extent_clear_unlock_delalloc(struct inode *inode, */ u64 count_range_bits(struct extent_io_tree *tree, u64 *start, u64 search_end, u64 max_bytes, - unsigned long bits) + unsigned long bits, int contig) { struct rb_node *node; struct extent_state *state; u64 cur_start = *start; u64 total_bytes = 0; + u64 last = 0; int found = 0; if (search_end <= cur_start) { @@ -1463,7 +1464,9 @@ u64 count_range_bits(struct extent_io_tree *tree, state = rb_entry(node, struct extent_state, rb_node); if (state->start > search_end) break; - if (state->end >= cur_start && (state->state & bits)) { + if (contig && found && state->start > last + 1) + break; + if (state->end >= cur_start && (state->state & bits) == bits) { total_bytes += min(search_end, state->end) + 1 - max(cur_start, state->start); if (total_bytes >= max_bytes) @@ -1472,6 +1475,9 @@ u64 count_range_bits(struct extent_io_tree *tree, *start = state->start; found = 1; } + last = state->end; + } else if (contig && found) { + break; } node = rb_next(node); if (!node) @@ -1865,7 +1871,7 @@ static int submit_one_bio(int rw, struct bio *bio, int mirror_num, bio_get(bio); if (tree->ops && tree->ops->submit_bio_hook) - tree->ops->submit_bio_hook(page->mapping->host, rw, bio, + ret = tree->ops->submit_bio_hook(page->mapping->host, rw, bio, mirror_num, bio_flags, start); else submit_bio(rw, bio); @@ -1920,6 +1926,8 @@ static int submit_extent_page(int rw, struct extent_io_tree *tree, nr = bio_get_nr_vecs(bdev); bio = btrfs_bio_alloc(bdev, sector, nr, GFP_NOFS | __GFP_HIGH); + if (!bio) + return -ENOMEM; bio_add_page(bio, page, page_size, offset); bio->bi_end_io = end_io_func; @@ -1944,6 +1952,7 @@ void set_page_extent_mapped(struct page *page) static void set_page_extent_head(struct page *page, unsigned long len) { + WARN_ON(!PagePrivate(page)); set_page_private(page, EXTENT_PAGE_PRIVATE_FIRST_PAGE | len << 2); } @@ -2126,7 +2135,7 @@ int extent_read_full_page(struct extent_io_tree *tree, struct page *page, ret = __extent_read_full_page(tree, page, get_extent, &bio, 0, &bio_flags); if (bio) - submit_one_bio(READ, bio, 0, bio_flags); + ret = submit_one_bio(READ, bio, 0, bio_flags); return ret; } @@ -2819,9 +2828,17 @@ int try_release_extent_state(struct extent_map_tree *map, * at this point we can safely clear everything except the * locked bit and the nodatasum bit */ - clear_extent_bit(tree, start, end, + ret = clear_extent_bit(tree, start, end, ~(EXTENT_LOCKED | EXTENT_NODATASUM), 0, 0, NULL, mask); + + /* if clear_extent_bit failed for enomem reasons, + * we can't allow the release to continue. + */ + if (ret < 0) + ret = 0; + else + ret = 1; } return ret; } @@ -2901,6 +2918,46 @@ out: return sector; } +/* + * helper function for fiemap, which doesn't want to see any holes. + * This maps until we find something past 'last' + */ +static struct extent_map *get_extent_skip_holes(struct inode *inode, + u64 offset, + u64 last, + get_extent_t *get_extent) +{ + u64 sectorsize = BTRFS_I(inode)->root->sectorsize; + struct extent_map *em; + u64 len; + + if (offset >= last) + return NULL; + + while(1) { + len = last - offset; + if (len == 0) + break; + len = (len + sectorsize - 1) & ~(sectorsize - 1); + em = get_extent(inode, NULL, 0, offset, len, 0); + if (!em || IS_ERR(em)) + return em; + + /* if this isn't a hole return it */ + if (!test_bit(EXTENT_FLAG_VACANCY, &em->flags) && + em->block_start != EXTENT_MAP_HOLE) { + return em; + } + + /* this is a hole, advance to the next extent */ + offset = extent_map_end(em); + free_extent_map(em); + if (offset >= last) + break; + } + return NULL; +} + int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len, get_extent_t *get_extent) { @@ -2910,16 +2967,19 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, u32 flags = 0; u32 found_type; u64 last; + u64 last_for_get_extent = 0; u64 disko = 0; + u64 isize = i_size_read(inode); struct btrfs_key found_key; struct extent_map *em = NULL; struct extent_state *cached_state = NULL; struct btrfs_path *path; struct btrfs_file_extent_item *item; int end = 0; - u64 em_start = 0, em_len = 0; + u64 em_start = 0; + u64 em_len = 0; + u64 em_end = 0; unsigned long emflags; - int hole = 0; if (len == 0) return -EINVAL; @@ -2929,6 +2989,10 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, return -ENOMEM; path->leave_spinning = 1; + /* + * lookup the last file extent. We're not using i_size here + * because there might be preallocation past i_size + */ ret = btrfs_lookup_file_extent(NULL, BTRFS_I(inode)->root, path, inode->i_ino, -1, 0); if (ret < 0) { @@ -2942,18 +3006,38 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]); found_type = btrfs_key_type(&found_key); - /* No extents, just return */ + /* No extents, but there might be delalloc bits */ if (found_key.objectid != inode->i_ino || found_type != BTRFS_EXTENT_DATA_KEY) { - btrfs_free_path(path); - return 0; + /* have to trust i_size as the end */ + last = (u64)-1; + last_for_get_extent = isize; + } else { + /* + * remember the start of the last extent. There are a + * bunch of different factors that go into the length of the + * extent, so its much less complex to remember where it started + */ + last = found_key.offset; + last_for_get_extent = last + 1; } - last = found_key.offset; btrfs_free_path(path); + /* + * we might have some extents allocated but more delalloc past those + * extents. so, we trust isize unless the start of the last extent is + * beyond isize + */ + if (last < isize) { + last = (u64)-1; + last_for_get_extent = isize; + } + lock_extent_bits(&BTRFS_I(inode)->io_tree, start, start + len, 0, &cached_state, GFP_NOFS); - em = get_extent(inode, NULL, 0, off, max - off, 0); + + em = get_extent_skip_holes(inode, off, last_for_get_extent, + get_extent); if (!em) goto out; if (IS_ERR(em)) { @@ -2962,22 +3046,38 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, } while (!end) { - hole = 0; - off = em->start + em->len; - if (off >= max) - end = 1; + u64 offset_in_extent; - if (em->block_start == EXTENT_MAP_HOLE) { - hole = 1; - goto next; - } + /* break if the extent we found is outside the range */ + if (em->start >= max || extent_map_end(em) < off) + break; - em_start = em->start; - em_len = em->len; + /* + * get_extent may return an extent that starts before our + * requested range. We have to make sure the ranges + * we return to fiemap always move forward and don't + * overlap, so adjust the offsets here + */ + em_start = max(em->start, off); + /* + * record the offset from the start of the extent + * for adjusting the disk offset below + */ + offset_in_extent = em_start - em->start; + em_end = extent_map_end(em); + em_len = em_end - em_start; + emflags = em->flags; disko = 0; flags = 0; + /* + * bump off for our next call to get_extent + */ + off = extent_map_end(em); + if (off >= max) + end = 1; + if (em->block_start == EXTENT_MAP_LAST_BYTE) { end = 1; flags |= FIEMAP_EXTENT_LAST; @@ -2988,42 +3088,34 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, flags |= (FIEMAP_EXTENT_DELALLOC | FIEMAP_EXTENT_UNKNOWN); } else { - disko = em->block_start; + disko = em->block_start + offset_in_extent; } if (test_bit(EXTENT_FLAG_COMPRESSED, &em->flags)) flags |= FIEMAP_EXTENT_ENCODED; -next: - emflags = em->flags; free_extent_map(em); em = NULL; - if (!end) { - em = get_extent(inode, NULL, 0, off, max - off, 0); - if (!em) - goto out; - if (IS_ERR(em)) { - ret = PTR_ERR(em); - goto out; - } - emflags = em->flags; - } - - if (test_bit(EXTENT_FLAG_VACANCY, &emflags)) { + if ((em_start >= last) || em_len == (u64)-1 || + (last == (u64)-1 && isize <= em_end)) { flags |= FIEMAP_EXTENT_LAST; end = 1; } - if (em_start == last) { + /* now scan forward to see if this is really the last extent. */ + em = get_extent_skip_holes(inode, off, last_for_get_extent, + get_extent); + if (IS_ERR(em)) { + ret = PTR_ERR(em); + goto out; + } + if (!em) { flags |= FIEMAP_EXTENT_LAST; end = 1; } - - if (!hole) { - ret = fiemap_fill_next_extent(fieinfo, em_start, disko, - em_len, flags); - if (ret) - goto out_free; - } + ret = fiemap_fill_next_extent(fieinfo, em_start, disko, + em_len, flags); + if (ret) + goto out_free; } out_free: free_extent_map(em); @@ -3192,7 +3284,13 @@ struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree, } if (!PageUptodate(p)) uptodate = 0; - unlock_page(p); + + /* + * see below about how we avoid a nasty race with release page + * and why we unlock later + */ + if (i != 0) + unlock_page(p); } if (uptodate) set_bit(EXTENT_BUFFER_UPTODATE, &eb->bflags); @@ -3216,9 +3314,26 @@ struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree, atomic_inc(&eb->refs); spin_unlock(&tree->buffer_lock); radix_tree_preload_end(); + + /* + * there is a race where release page may have + * tried to find this extent buffer in the radix + * but failed. It will tell the VM it is safe to + * reclaim the, and it will clear the page private bit. + * We must make sure to set the page private bit properly + * after the extent buffer is in the radix tree so + * it doesn't get lost + */ + set_page_extent_mapped(eb->first_page); + set_page_extent_head(eb->first_page, eb->len); + if (!page0) + unlock_page(eb->first_page); return eb; free_eb: + if (eb->first_page && !page0) + unlock_page(eb->first_page); + if (!atomic_dec_and_test(&eb->refs)) return exists; btrfs_release_extent_buffer(eb); @@ -3269,10 +3384,11 @@ int clear_extent_buffer_dirty(struct extent_io_tree *tree, continue; lock_page(page); + WARN_ON(!PagePrivate(page)); + + set_page_extent_mapped(page); if (i == 0) set_page_extent_head(page, eb->len); - else - set_page_private(page, EXTENT_PAGE_PRIVATE); clear_page_dirty_for_io(page); spin_lock_irq(&page->mapping->tree_lock); @@ -3462,6 +3578,13 @@ int read_extent_buffer_pages(struct extent_io_tree *tree, for (i = start_i; i < num_pages; i++) { page = extent_buffer_page(eb, i); + + WARN_ON(!PagePrivate(page)); + + set_page_extent_mapped(page); + if (i == 0) + set_page_extent_head(page, eb->len); + if (inc_all_pages) page_cache_get(page); if (!PageUptodate(page)) { diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index 7083cfa..9318dfe 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -191,7 +191,7 @@ void extent_io_exit(void); u64 count_range_bits(struct extent_io_tree *tree, u64 *start, u64 search_end, - u64 max_bytes, unsigned long bits); + u64 max_bytes, unsigned long bits, int contig); void free_extent_state(struct extent_state *state); int test_range_bit(struct extent_io_tree *tree, u64 start, u64 end, diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c index b0e1fce..2b6c12e 100644 --- a/fs/btrfs/extent_map.c +++ b/fs/btrfs/extent_map.c @@ -51,8 +51,8 @@ struct extent_map *alloc_extent_map(gfp_t mask) { struct extent_map *em; em = kmem_cache_alloc(extent_map_cache, mask); - if (!em || IS_ERR(em)) - return em; + if (!em) + return NULL; em->in_tree = 0; em->flags = 0; em->compress_type = BTRFS_COMPRESS_NONE; diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c index a562a25..4f19a3e 100644 --- a/fs/btrfs/file-item.c +++ b/fs/btrfs/file-item.c @@ -536,6 +536,8 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans, root = root->fs_info->csum_root; path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; while (1) { key.objectid = BTRFS_EXTENT_CSUM_OBJECTID; @@ -548,7 +550,10 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans, if (path->slots[0] == 0) goto out; path->slots[0]--; + } else if (ret < 0) { + goto out; } + leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index c800d58..f447b78 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -70,6 +70,19 @@ static noinline int btrfs_copy_from_user(loff_t pos, int num_pages, /* Flush processor's dcache for this page */ flush_dcache_page(page); + + /* + * if we get a partial write, we can end up with + * partially up to date pages. These add + * a lot of complexity, so make sure they don't + * happen by forcing this copy to be retried. + * + * The rest of the btrfs_file_write code will fall + * back to page at a time copies after we return 0. + */ + if (!PageUptodate(page) && copied < count) + copied = 0; + iov_iter_advance(i, copied); write_bytes -= copied; total_copied += copied; @@ -186,6 +199,7 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end, split = alloc_extent_map(GFP_NOFS); if (!split2) split2 = alloc_extent_map(GFP_NOFS); + BUG_ON(!split || !split2); write_lock(&em_tree->lock); em = lookup_extent_mapping(em_tree, start, len); @@ -762,6 +776,27 @@ out: } /* + * on error we return an unlocked page and the error value + * on success we return a locked page and 0 + */ +static int prepare_uptodate_page(struct page *page, u64 pos) +{ + int ret = 0; + + if ((pos & (PAGE_CACHE_SIZE - 1)) && !PageUptodate(page)) { + ret = btrfs_readpage(NULL, page); + if (ret) + return ret; + lock_page(page); + if (!PageUptodate(page)) { + unlock_page(page); + return -EIO; + } + } + return 0; +} + +/* * this gets pages into the page cache and locks them down, it also properly * waits for data=ordered extents to finish before allowing the pages to be * modified. @@ -776,6 +811,7 @@ static noinline int prepare_pages(struct btrfs_root *root, struct file *file, unsigned long index = pos >> PAGE_CACHE_SHIFT; struct inode *inode = fdentry(file)->d_inode; int err = 0; + int faili = 0; u64 start_pos; u64 last_pos; @@ -793,11 +829,24 @@ again: for (i = 0; i < num_pages; i++) { pages[i] = grab_cache_page(inode->i_mapping, index + i); if (!pages[i]) { + faili = i - 1; err = -ENOMEM; - BUG_ON(1); + goto fail; + } + + if (i == 0) + err = prepare_uptodate_page(pages[i], pos); + if (i == num_pages - 1) + err = prepare_uptodate_page(pages[i], + pos + write_bytes); + if (err) { + page_cache_release(pages[i]); + faili = i - 1; + goto fail; } wait_on_page_writeback(pages[i]); } + err = 0; if (start_pos < inode->i_size) { struct btrfs_ordered_extent *ordered; lock_extent_bits(&BTRFS_I(inode)->io_tree, @@ -837,6 +886,14 @@ again: WARN_ON(!PageLocked(pages[i])); } return 0; +fail: + while (faili >= 0) { + unlock_page(pages[faili]); + page_cache_release(pages[faili]); + faili--; + } + return err; + } static ssize_t btrfs_file_aio_write(struct kiocb *iocb, @@ -846,7 +903,6 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb, struct file *file = iocb->ki_filp; struct inode *inode = fdentry(file)->d_inode; struct btrfs_root *root = BTRFS_I(inode)->root; - struct page *pinned[2]; struct page **pages = NULL; struct iov_iter i; loff_t *ppos = &iocb->ki_pos; @@ -867,9 +923,6 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb, will_write = ((file->f_flags & O_DSYNC) || IS_SYNC(inode) || (file->f_flags & O_DIRECT)); - pinned[0] = NULL; - pinned[1] = NULL; - start_pos = pos; vfs_check_frozen(inode->i_sb, SB_FREEZE_WRITE); @@ -946,6 +999,10 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb, PAGE_CACHE_SIZE, PAGE_CACHE_SIZE / (sizeof(struct page *))); pages = kmalloc(nrptrs * sizeof(struct page *), GFP_KERNEL); + if (!pages) { + ret = -ENOMEM; + goto out; + } /* generic_write_checks can change our pos */ start_pos = pos; @@ -953,39 +1010,13 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb, first_index = pos >> PAGE_CACHE_SHIFT; last_index = (pos + iov_iter_count(&i)) >> PAGE_CACHE_SHIFT; - /* - * there are lots of better ways to do this, but this code - * makes sure the first and last page in the file range are - * up to date and ready for cow - */ - if ((pos & (PAGE_CACHE_SIZE - 1))) { - pinned[0] = grab_cache_page(inode->i_mapping, first_index); - if (!PageUptodate(pinned[0])) { - ret = btrfs_readpage(NULL, pinned[0]); - BUG_ON(ret); - wait_on_page_locked(pinned[0]); - } else { - unlock_page(pinned[0]); - } - } - if ((pos + iov_iter_count(&i)) & (PAGE_CACHE_SIZE - 1)) { - pinned[1] = grab_cache_page(inode->i_mapping, last_index); - if (!PageUptodate(pinned[1])) { - ret = btrfs_readpage(NULL, pinned[1]); - BUG_ON(ret); - wait_on_page_locked(pinned[1]); - } else { - unlock_page(pinned[1]); - } - } - while (iov_iter_count(&i) > 0) { size_t offset = pos & (PAGE_CACHE_SIZE - 1); size_t write_bytes = min(iov_iter_count(&i), nrptrs * (size_t)PAGE_CACHE_SIZE - offset); - size_t num_pages = (write_bytes + PAGE_CACHE_SIZE - 1) >> - PAGE_CACHE_SHIFT; + size_t num_pages = (write_bytes + offset + + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; WARN_ON(num_pages > nrptrs); memset(pages, 0, sizeof(struct page *) * nrptrs); @@ -1015,8 +1046,20 @@ static ssize_t btrfs_file_aio_write(struct kiocb *iocb, copied = btrfs_copy_from_user(pos, num_pages, write_bytes, pages, &i); - dirty_pages = (copied + PAGE_CACHE_SIZE - 1) >> - PAGE_CACHE_SHIFT; + + /* + * if we have trouble faulting in the pages, fall + * back to one page at a time + */ + if (copied < write_bytes) + nrptrs = 1; + + if (copied == 0) + dirty_pages = 0; + else + dirty_pages = (copied + offset + + PAGE_CACHE_SIZE - 1) >> + PAGE_CACHE_SHIFT; if (num_pages > dirty_pages) { if (copied > 0) @@ -1060,10 +1103,6 @@ out: err = ret; kfree(pages); - if (pinned[0]) - page_cache_release(pinned[0]); - if (pinned[1]) - page_cache_release(pinned[1]); *ppos = pos; /* diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index 60d6842..a039065 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -987,11 +987,18 @@ tree_search_offset(struct btrfs_block_group_cache *block_group, return entry; } -static void unlink_free_space(struct btrfs_block_group_cache *block_group, - struct btrfs_free_space *info) +static inline void +__unlink_free_space(struct btrfs_block_group_cache *block_group, + struct btrfs_free_space *info) { rb_erase(&info->offset_index, &block_group->free_space_offset); block_group->free_extents--; +} + +static void unlink_free_space(struct btrfs_block_group_cache *block_group, + struct btrfs_free_space *info) +{ + __unlink_free_space(block_group, info); block_group->free_space -= info->bytes; } @@ -1016,14 +1023,18 @@ static void recalculate_thresholds(struct btrfs_block_group_cache *block_group) u64 max_bytes; u64 bitmap_bytes; u64 extent_bytes; + u64 size = block_group->key.offset; /* * The goal is to keep the total amount of memory used per 1gb of space * at or below 32k, so we need to adjust how much memory we allow to be * used by extent based free space tracking */ - max_bytes = MAX_CACHE_BYTES_PER_GIG * - (div64_u64(block_group->key.offset, 1024 * 1024 * 1024)); + if (size < 1024 * 1024 * 1024) + max_bytes = MAX_CACHE_BYTES_PER_GIG; + else + max_bytes = MAX_CACHE_BYTES_PER_GIG * + div64_u64(size, 1024 * 1024 * 1024); /* * we want to account for 1 more bitmap than what we have so we can make @@ -1171,6 +1182,16 @@ static void add_new_bitmap(struct btrfs_block_group_cache *block_group, recalculate_thresholds(block_group); } +static void free_bitmap(struct btrfs_block_group_cache *block_group, + struct btrfs_free_space *bitmap_info) +{ + unlink_free_space(block_group, bitmap_info); + kfree(bitmap_info->bitmap); + kfree(bitmap_info); + block_group->total_bitmaps--; + recalculate_thresholds(block_group); +} + static noinline int remove_from_bitmap(struct btrfs_block_group_cache *block_group, struct btrfs_free_space *bitmap_info, u64 *offset, u64 *bytes) @@ -1195,6 +1216,7 @@ again: */ search_start = *offset; search_bytes = *bytes; + search_bytes = min(search_bytes, end - search_start + 1); ret = search_bitmap(block_group, bitmap_info, &search_start, &search_bytes); BUG_ON(ret < 0 || search_start != *offset); @@ -1211,13 +1233,8 @@ again: if (*bytes) { struct rb_node *next = rb_next(&bitmap_info->offset_index); - if (!bitmap_info->bytes) { - unlink_free_space(block_group, bitmap_info); - kfree(bitmap_info->bitmap); - kfree(bitmap_info); - block_group->total_bitmaps--; - recalculate_thresholds(block_group); - } + if (!bitmap_info->bytes) + free_bitmap(block_group, bitmap_info); /* * no entry after this bitmap, but we still have bytes to @@ -1250,13 +1267,8 @@ again: return -EAGAIN; goto again; - } else if (!bitmap_info->bytes) { - unlink_free_space(block_group, bitmap_info); - kfree(bitmap_info->bitmap); - kfree(bitmap_info); - block_group->total_bitmaps--; - recalculate_thresholds(block_group); - } + } else if (!bitmap_info->bytes) + free_bitmap(block_group, bitmap_info); return 0; } @@ -1359,22 +1371,14 @@ out: return ret; } -int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, - u64 offset, u64 bytes) +bool try_merge_free_space(struct btrfs_block_group_cache *block_group, + struct btrfs_free_space *info, bool update_stat) { - struct btrfs_free_space *right_info = NULL; - struct btrfs_free_space *left_info = NULL; - struct btrfs_free_space *info = NULL; - int ret = 0; - - info = kzalloc(sizeof(struct btrfs_free_space), GFP_NOFS); - if (!info) - return -ENOMEM; - - info->offset = offset; - info->bytes = bytes; - - spin_lock(&block_group->tree_lock); + struct btrfs_free_space *left_info; + struct btrfs_free_space *right_info; + bool merged = false; + u64 offset = info->offset; + u64 bytes = info->bytes; /* * first we want to see if there is free space adjacent to the range we @@ -1388,37 +1392,62 @@ int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, else left_info = tree_search_offset(block_group, offset - 1, 0, 0); - /* - * If there was no extent directly to the left or right of this new - * extent then we know we're going to have to allocate a new extent, so - * before we do that see if we need to drop this into a bitmap - */ - if ((!left_info || left_info->bitmap) && - (!right_info || right_info->bitmap)) { - ret = insert_into_bitmap(block_group, info); - - if (ret < 0) { - goto out; - } else if (ret) { - ret = 0; - goto out; - } - } - if (right_info && !right_info->bitmap) { - unlink_free_space(block_group, right_info); + if (update_stat) + unlink_free_space(block_group, right_info); + else + __unlink_free_space(block_group, right_info); info->bytes += right_info->bytes; kfree(right_info); + merged = true; } if (left_info && !left_info->bitmap && left_info->offset + left_info->bytes == offset) { - unlink_free_space(block_group, left_info); + if (update_stat) + unlink_free_space(block_group, left_info); + else + __unlink_free_space(block_group, left_info); info->offset = left_info->offset; info->bytes += left_info->bytes; kfree(left_info); + merged = true; } + return merged; +} + +int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, + u64 offset, u64 bytes) +{ + struct btrfs_free_space *info; + int ret = 0; + + info = kzalloc(sizeof(struct btrfs_free_space), GFP_NOFS); + if (!info) + return -ENOMEM; + + info->offset = offset; + info->bytes = bytes; + + spin_lock(&block_group->tree_lock); + + if (try_merge_free_space(block_group, info, true)) + goto link; + + /* + * There was no extent directly to the left or right of this new + * extent then we know we're going to have to allocate a new extent, so + * before we do that see if we need to drop this into a bitmap + */ + ret = insert_into_bitmap(block_group, info); + if (ret < 0) { + goto out; + } else if (ret) { + ret = 0; + goto out; + } +link: ret = link_free_space(block_group, info); if (ret) kfree(info); @@ -1621,6 +1650,7 @@ __btrfs_return_cluster_to_free_space( node = rb_next(&entry->offset_index); rb_erase(&entry->offset_index, &cluster->root); BUG_ON(entry->bitmap); + try_merge_free_space(block_group, entry, false); tree_insert_offset(&block_group->free_space_offset, entry->offset, &entry->offset_index, 0); } @@ -1685,13 +1715,8 @@ u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group, ret = offset; if (entry->bitmap) { bitmap_clear_bits(block_group, entry, offset, bytes); - if (!entry->bytes) { - unlink_free_space(block_group, entry); - kfree(entry->bitmap); - kfree(entry); - block_group->total_bitmaps--; - recalculate_thresholds(block_group); - } + if (!entry->bytes) + free_bitmap(block_group, entry); } else { unlink_free_space(block_group, entry); entry->offset += bytes; @@ -1789,6 +1814,8 @@ static u64 btrfs_alloc_from_bitmap(struct btrfs_block_group_cache *block_group, ret = search_start; bitmap_clear_bits(block_group, entry, ret, bytes); + if (entry->bytes == 0) + free_bitmap(block_group, entry); out: spin_unlock(&cluster->lock); spin_unlock(&block_group->tree_lock); @@ -1842,15 +1869,26 @@ u64 btrfs_alloc_from_cluster(struct btrfs_block_group_cache *block_group, entry->offset += bytes; entry->bytes -= bytes; - if (entry->bytes == 0) { + if (entry->bytes == 0) rb_erase(&entry->offset_index, &cluster->root); - kfree(entry); - } break; } out: spin_unlock(&cluster->lock); + if (!ret) + return 0; + + spin_lock(&block_group->tree_lock); + + block_group->free_space -= bytes; + if (entry->bytes == 0) { + block_group->free_extents--; + kfree(entry); + } + + spin_unlock(&block_group->tree_lock); + return ret; } diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 160b55b..512c3d1 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -90,13 +90,14 @@ static noinline int cow_file_range(struct inode *inode, unsigned long *nr_written, int unlock); static int btrfs_init_inode_security(struct btrfs_trans_handle *trans, - struct inode *inode, struct inode *dir) + struct inode *inode, struct inode *dir, + const struct qstr *qstr) { int err; err = btrfs_init_acl(trans, inode, dir); if (!err) - err = btrfs_xattr_security_init(trans, inode, dir); + err = btrfs_xattr_security_init(trans, inode, dir, qstr); return err; } @@ -416,7 +417,7 @@ again: } if (start == 0) { trans = btrfs_join_transaction(root, 1); - BUG_ON(!trans); + BUG_ON(IS_ERR(trans)); btrfs_set_trans_block_group(trans, inode); trans->block_rsv = &root->fs_info->delalloc_block_rsv; @@ -612,6 +613,7 @@ retry: GFP_NOFS); trans = btrfs_join_transaction(root, 1); + BUG_ON(IS_ERR(trans)); ret = btrfs_reserve_extent(trans, root, async_extent->compressed_size, async_extent->compressed_size, @@ -643,6 +645,7 @@ retry: async_extent->ram_size - 1, 0); em = alloc_extent_map(GFP_NOFS); + BUG_ON(!em); em->start = async_extent->start; em->len = async_extent->ram_size; em->orig_start = em->start; @@ -771,7 +774,7 @@ static noinline int cow_file_range(struct inode *inode, BUG_ON(root == root->fs_info->tree_root); trans = btrfs_join_transaction(root, 1); - BUG_ON(!trans); + BUG_ON(IS_ERR(trans)); btrfs_set_trans_block_group(trans, inode); trans->block_rsv = &root->fs_info->delalloc_block_rsv; @@ -819,6 +822,7 @@ static noinline int cow_file_range(struct inode *inode, BUG_ON(ret); em = alloc_extent_map(GFP_NOFS); + BUG_ON(!em); em->start = start; em->orig_start = em->start; ram_size = ins.offset; @@ -1049,7 +1053,7 @@ static noinline int run_delalloc_nocow(struct inode *inode, } else { trans = btrfs_join_transaction(root, 1); } - BUG_ON(!trans); + BUG_ON(IS_ERR(trans)); cow_start = (u64)-1; cur_offset = start; @@ -1168,6 +1172,7 @@ out_check: struct extent_map_tree *em_tree; em_tree = &BTRFS_I(inode)->extent_tree; em = alloc_extent_map(GFP_NOFS); + BUG_ON(!em); em->start = cur_offset; em->orig_start = em->start; em->len = num_bytes; @@ -1557,6 +1562,7 @@ out: out_page: unlock_page(page); page_cache_release(page); + kfree(fixup); } /* @@ -1703,7 +1709,7 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end) trans = btrfs_join_transaction_nolock(root, 1); else trans = btrfs_join_transaction(root, 1); - BUG_ON(!trans); + BUG_ON(IS_ERR(trans)); btrfs_set_trans_block_group(trans, inode); trans->block_rsv = &root->fs_info->delalloc_block_rsv; ret = btrfs_update_inode(trans, root, inode); @@ -1720,6 +1726,7 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end) trans = btrfs_join_transaction_nolock(root, 1); else trans = btrfs_join_transaction(root, 1); + BUG_ON(IS_ERR(trans)); btrfs_set_trans_block_group(trans, inode); trans->block_rsv = &root->fs_info->delalloc_block_rsv; @@ -1907,7 +1914,7 @@ static int btrfs_clean_io_failures(struct inode *inode, u64 start) private = 0; if (count_range_bits(&BTRFS_I(inode)->io_failure_tree, &private, - (u64)-1, 1, EXTENT_DIRTY)) { + (u64)-1, 1, EXTENT_DIRTY, 0)) { ret = get_state_private(&BTRFS_I(inode)->io_failure_tree, start, &private_failure); if (ret == 0) { @@ -2354,6 +2361,7 @@ void btrfs_orphan_cleanup(struct btrfs_root *root) */ if (is_bad_inode(inode)) { trans = btrfs_start_transaction(root, 0); + BUG_ON(IS_ERR(trans)); btrfs_orphan_del(trans, inode); btrfs_end_transaction(trans, root); iput(inode); @@ -2381,6 +2389,7 @@ void btrfs_orphan_cleanup(struct btrfs_root *root) if (root->orphan_block_rsv || root->orphan_item_inserted) { trans = btrfs_join_transaction(root, 1); + BUG_ON(IS_ERR(trans)); btrfs_end_transaction(trans, root); } @@ -2641,7 +2650,7 @@ int btrfs_unlink_inode(struct btrfs_trans_handle *trans, path = btrfs_alloc_path(); if (!path) { ret = -ENOMEM; - goto err; + goto out; } path->leave_spinning = 1; @@ -2714,9 +2723,10 @@ static int check_path_shared(struct btrfs_root *root, struct extent_buffer *eb; int level; u64 refs = 1; - int uninitialized_var(ret); for (level = 0; level < BTRFS_MAX_LEVEL; level++) { + int ret; + if (!path->nodes[level]) break; eb = path->nodes[level]; @@ -2727,7 +2737,7 @@ static int check_path_shared(struct btrfs_root *root, if (refs > 1) return 1; } - return ret; /* XXX callers? */ + return 0; } /* @@ -4134,7 +4144,7 @@ struct inode *btrfs_lookup_dentry(struct inode *dir, struct dentry *dentry) } srcu_read_unlock(&root->fs_info->subvol_srcu, index); - if (root != sub_root) { + if (!IS_ERR(inode) && root != sub_root) { down_read(&root->fs_info->cleanup_work_sem); if (!(inode->i_sb->s_flags & MS_RDONLY)) btrfs_orphan_cleanup(sub_root); @@ -4347,6 +4357,8 @@ int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc) trans = btrfs_join_transaction_nolock(root, 1); else trans = btrfs_join_transaction(root, 1); + if (IS_ERR(trans)) + return PTR_ERR(trans); btrfs_set_trans_block_group(trans, inode); if (nolock) ret = btrfs_end_transaction_nolock(trans, root); @@ -4372,6 +4384,7 @@ void btrfs_dirty_inode(struct inode *inode) return; trans = btrfs_join_transaction(root, 1); + BUG_ON(IS_ERR(trans)); btrfs_set_trans_block_group(trans, inode); ret = btrfs_update_inode(trans, root, inode); @@ -4692,7 +4705,7 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry, if (IS_ERR(inode)) goto out_unlock; - err = btrfs_init_inode_security(trans, inode, dir); + err = btrfs_init_inode_security(trans, inode, dir, &dentry->d_name); if (err) { drop_inode = 1; goto out_unlock; @@ -4753,7 +4766,7 @@ static int btrfs_create(struct inode *dir, struct dentry *dentry, if (IS_ERR(inode)) goto out_unlock; - err = btrfs_init_inode_security(trans, inode, dir); + err = btrfs_init_inode_security(trans, inode, dir, &dentry->d_name); if (err) { drop_inode = 1; goto out_unlock; @@ -4794,9 +4807,6 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir, int err; int drop_inode = 0; - if (inode->i_nlink == 0) - return -ENOENT; - /* do not allow sys_link's with other subvols of the same device */ if (root->objectid != BTRFS_I(inode)->root->objectid) return -EPERM; @@ -4809,10 +4819,11 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir, goto fail; /* - * 1 item for inode ref + * 2 items for inode and inode ref * 2 items for dir items + * 1 item for parent inode */ - trans = btrfs_start_transaction(root, 3); + trans = btrfs_start_transaction(root, 5); if (IS_ERR(trans)) { err = PTR_ERR(trans); goto fail; @@ -4881,7 +4892,7 @@ static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) drop_on_err = 1; - err = btrfs_init_inode_security(trans, inode, dir); + err = btrfs_init_inode_security(trans, inode, dir, &dentry->d_name); if (err) goto out_fail; @@ -5176,6 +5187,8 @@ again: em = NULL; btrfs_release_path(root, path); trans = btrfs_join_transaction(root, 1); + if (IS_ERR(trans)) + return ERR_CAST(trans); goto again; } map = kmap(page); @@ -5266,6 +5279,128 @@ out: return em; } +struct extent_map *btrfs_get_extent_fiemap(struct inode *inode, struct page *page, + size_t pg_offset, u64 start, u64 len, + int create) +{ + struct extent_map *em; + struct extent_map *hole_em = NULL; + u64 range_start = start; + u64 end; + u64 found; + u64 found_end; + int err = 0; + + em = btrfs_get_extent(inode, page, pg_offset, start, len, create); + if (IS_ERR(em)) + return em; + if (em) { + /* + * if our em maps to a hole, there might + * actually be delalloc bytes behind it + */ + if (em->block_start != EXTENT_MAP_HOLE) + return em; + else + hole_em = em; + } + + /* check to see if we've wrapped (len == -1 or similar) */ + end = start + len; + if (end < start) + end = (u64)-1; + else + end -= 1; + + em = NULL; + + /* ok, we didn't find anything, lets look for delalloc */ + found = count_range_bits(&BTRFS_I(inode)->io_tree, &range_start, + end, len, EXTENT_DELALLOC, 1); + found_end = range_start + found; + if (found_end < range_start) + found_end = (u64)-1; + + /* + * we didn't find anything useful, return + * the original results from get_extent() + */ + if (range_start > end || found_end <= start) { + em = hole_em; + hole_em = NULL; + goto out; + } + + /* adjust the range_start to make sure it doesn't + * go backwards from the start they passed in + */ + range_start = max(start,range_start); + found = found_end - range_start; + + if (found > 0) { + u64 hole_start = start; + u64 hole_len = len; + + em = alloc_extent_map(GFP_NOFS); + if (!em) { + err = -ENOMEM; + goto out; + } + /* + * when btrfs_get_extent can't find anything it + * returns one huge hole + * + * make sure what it found really fits our range, and + * adjust to make sure it is based on the start from + * the caller + */ + if (hole_em) { + u64 calc_end = extent_map_end(hole_em); + + if (calc_end <= start || (hole_em->start > end)) { + free_extent_map(hole_em); + hole_em = NULL; + } else { + hole_start = max(hole_em->start, start); + hole_len = calc_end - hole_start; + } + } + em->bdev = NULL; + if (hole_em && range_start > hole_start) { + /* our hole starts before our delalloc, so we + * have to return just the parts of the hole + * that go until the delalloc starts + */ + em->len = min(hole_len, + range_start - hole_start); + em->start = hole_start; + em->orig_start = hole_start; + /* + * don't adjust block start at all, + * it is fixed at EXTENT_MAP_HOLE + */ + em->block_start = hole_em->block_start; + em->block_len = hole_len; + } else { + em->start = range_start; + em->len = found; + em->orig_start = range_start; + em->block_start = EXTENT_MAP_DELALLOC; + em->block_len = found; + } + } else if (hole_em) { + return hole_em; + } +out: + + free_extent_map(hole_em); + if (err) { + free_extent_map(em); + return ERR_PTR(err); + } + return em; +} + static struct extent_map *btrfs_new_extent_direct(struct inode *inode, u64 start, u64 len) { @@ -5280,8 +5415,8 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode, btrfs_drop_extent_cache(inode, start, start + len - 1, 0); trans = btrfs_join_transaction(root, 0); - if (!trans) - return ERR_PTR(-ENOMEM); + if (IS_ERR(trans)) + return ERR_CAST(trans); trans->block_rsv = &root->fs_info->delalloc_block_rsv; @@ -5505,7 +5640,7 @@ static int btrfs_get_blocks_direct(struct inode *inode, sector_t iblock, * while we look for nocow cross refs */ trans = btrfs_join_transaction(root, 0); - if (!trans) + if (IS_ERR(trans)) goto must_cow; if (can_nocow_odirect(trans, inode, start, len) == 1) { @@ -5640,7 +5775,7 @@ again: BUG_ON(!ordered); trans = btrfs_join_transaction(root, 1); - if (!trans) { + if (IS_ERR(trans)) { err = -ENOMEM; goto out; } @@ -5920,6 +6055,7 @@ static void btrfs_submit_direct(int rw, struct bio *bio, struct inode *inode, if (!skip_sum) { dip->csums = kmalloc(sizeof(u32) * bio->bi_vcnt, GFP_NOFS); if (!dip->csums) { + kfree(dip); ret = -ENOMEM; goto free_ordered; } @@ -6088,7 +6224,7 @@ out: static int btrfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len) { - return extent_fiemap(inode, fieinfo, start, len, btrfs_get_extent); + return extent_fiemap(inode, fieinfo, start, len, btrfs_get_extent_fiemap); } int btrfs_readpage(struct file *file, struct page *page) @@ -6968,7 +7104,7 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry, if (IS_ERR(inode)) goto out_unlock; - err = btrfs_init_inode_security(trans, inode, dir); + err = btrfs_init_inode_security(trans, inode, dir, &dentry->d_name); if (err) { drop_inode = 1; goto out_unlock; diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index a506a22..5fdb2ab 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -203,7 +203,7 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg) trans = btrfs_join_transaction(root, 1); - BUG_ON(!trans); + BUG_ON(IS_ERR(trans)); ret = btrfs_update_inode(trans, root, inode); BUG_ON(ret); @@ -907,6 +907,10 @@ static noinline int btrfs_ioctl_resize(struct btrfs_root *root, if (new_size > old_size) { trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) { + ret = PTR_ERR(trans); + goto out_unlock; + } ret = btrfs_grow_device(trans, device, new_size); btrfs_commit_transaction(trans, root); } else { @@ -1067,12 +1071,15 @@ static noinline int btrfs_ioctl_subvol_setflags(struct file *file, if (copy_from_user(&flags, arg, sizeof(flags))) return -EFAULT; - if (flags & ~BTRFS_SUBVOL_CREATE_ASYNC) + if (flags & BTRFS_SUBVOL_CREATE_ASYNC) return -EINVAL; if (flags & ~BTRFS_SUBVOL_RDONLY) return -EOPNOTSUPP; + if (!is_owner_or_cap(inode)) + return -EACCES; + down_write(&root->fs_info->subvol_sem); /* nothing to do */ @@ -1093,7 +1100,7 @@ static noinline int btrfs_ioctl_subvol_setflags(struct file *file, goto out_reset; } - ret = btrfs_update_root(trans, root, + ret = btrfs_update_root(trans, root->fs_info->tree_root, &root->root_key, &root->root_item); btrfs_commit_transaction(trans, root); @@ -1898,7 +1905,10 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, memcpy(&new_key, &key, sizeof(new_key)); new_key.objectid = inode->i_ino; - new_key.offset = key.offset + destoff - off; + if (off <= key.offset) + new_key.offset = key.offset + destoff - off; + else + new_key.offset = destoff; trans = btrfs_start_transaction(root, 1); if (IS_ERR(trans)) { @@ -2082,7 +2092,7 @@ static long btrfs_ioctl_trans_start(struct file *file) ret = -ENOMEM; trans = btrfs_start_ioctl_transaction(root, 0); - if (!trans) + if (IS_ERR(trans)) goto out_drop; file->private_data = trans; @@ -2138,9 +2148,9 @@ static long btrfs_ioctl_default_subvol(struct file *file, void __user *argp) path->leave_spinning = 1; trans = btrfs_start_transaction(root, 1); - if (!trans) { + if (IS_ERR(trans)) { btrfs_free_path(path); - return -ENOMEM; + return PTR_ERR(trans); } dir_id = btrfs_super_root_dir(&root->fs_info->super_copy); @@ -2201,7 +2211,7 @@ long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg) int num_types = 4; int alloc_size; int ret = 0; - int slot_count = 0; + u64 slot_count = 0; int i, c; if (copy_from_user(&space_args, @@ -2240,7 +2250,7 @@ long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg) goto out; } - slot_count = min_t(int, space_args.space_slots, slot_count); + slot_count = min_t(u64, space_args.space_slots, slot_count); alloc_size = sizeof(*dest) * slot_count; @@ -2260,6 +2270,9 @@ long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg) for (i = 0; i < num_types; i++) { struct btrfs_space_info *tmp; + if (!slot_count) + break; + info = NULL; rcu_read_lock(); list_for_each_entry_rcu(tmp, &root->fs_info->space_info, @@ -2281,7 +2294,10 @@ long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg) memcpy(dest, &space, sizeof(space)); dest++; space_args.total_spaces++; + slot_count--; } + if (!slot_count) + break; } up_read(&info->groups_sem); } @@ -2334,6 +2350,8 @@ static noinline long btrfs_ioctl_start_sync(struct file *file, void __user *argp u64 transid; trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) + return PTR_ERR(trans); transid = trans->transid; btrfs_commit_transaction_async(trans, root, 0); diff --git a/fs/btrfs/lzo.c b/fs/btrfs/lzo.c index cc9b450..a178f5e 100644 --- a/fs/btrfs/lzo.c +++ b/fs/btrfs/lzo.c @@ -280,6 +280,7 @@ static int lzo_decompress_biovec(struct list_head *ws, unsigned long tot_out; unsigned long tot_len; char *buf; + bool may_late_unmap, need_unmap; data_in = kmap(pages_in[0]); tot_len = read_compress_length(data_in); @@ -300,11 +301,13 @@ static int lzo_decompress_biovec(struct list_head *ws, tot_in += in_len; working_bytes = in_len; + may_late_unmap = need_unmap = false; /* fast path: avoid using the working buffer */ if (in_page_bytes_left >= in_len) { buf = data_in + in_offset; bytes = in_len; + may_late_unmap = true; goto cont; } @@ -329,14 +332,17 @@ cont: if (working_bytes == 0 && tot_in >= tot_len) break; - kunmap(pages_in[page_in_index]); - page_in_index++; - if (page_in_index >= total_pages_in) { + if (page_in_index + 1 >= total_pages_in) { ret = -1; - data_in = NULL; goto done; } - data_in = kmap(pages_in[page_in_index]); + + if (may_late_unmap) + need_unmap = true; + else + kunmap(pages_in[page_in_index]); + + data_in = kmap(pages_in[++page_in_index]); in_page_bytes_left = PAGE_CACHE_SIZE; in_offset = 0; @@ -346,6 +352,8 @@ cont: out_len = lzo1x_worst_compress(PAGE_CACHE_SIZE); ret = lzo1x_decompress_safe(buf, in_len, workspace->buf, &out_len); + if (need_unmap) + kunmap(pages_in[page_in_index - 1]); if (ret != LZO_E_OK) { printk(KERN_WARNING "btrfs decompress failed\n"); ret = -1; @@ -363,8 +371,7 @@ cont: break; } done: - if (data_in) - kunmap(pages_in[page_in_index]); + kunmap(pages_in[page_in_index]); return ret; } diff --git a/fs/btrfs/ordered-data.c b/fs/btrfs/ordered-data.c index 2b61e1d..083a554 100644 --- a/fs/btrfs/ordered-data.c +++ b/fs/btrfs/ordered-data.c @@ -141,7 +141,7 @@ static inline struct rb_node *tree_search(struct btrfs_ordered_inode_tree *tree, u64 file_offset) { struct rb_root *root = &tree->tree; - struct rb_node *prev; + struct rb_node *prev = NULL; struct rb_node *ret; struct btrfs_ordered_extent *entry; diff --git a/fs/btrfs/print-tree.c b/fs/btrfs/print-tree.c index 0d126be..fb2605d 100644 --- a/fs/btrfs/print-tree.c +++ b/fs/btrfs/print-tree.c @@ -260,6 +260,7 @@ void btrfs_print_leaf(struct btrfs_root *root, struct extent_buffer *l) #else BUG(); #endif + break; case BTRFS_BLOCK_GROUP_ITEM_KEY: bi = btrfs_item_ptr(l, i, struct btrfs_block_group_item); diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 045c9c2..31ade58 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -1157,6 +1157,7 @@ static int clone_backref_node(struct btrfs_trans_handle *trans, new_node->bytenr = dest->node->start; new_node->level = node->level; new_node->lowest = node->lowest; + new_node->checked = 1; new_node->root = dest; if (!node->lowest) { @@ -2028,6 +2029,7 @@ static noinline_for_stack int merge_reloc_root(struct reloc_control *rc, while (1) { trans = btrfs_start_transaction(root, 0); + BUG_ON(IS_ERR(trans)); trans->block_rsv = rc->block_rsv; ret = btrfs_block_rsv_check(trans, root, rc->block_rsv, @@ -2147,6 +2149,12 @@ again: } trans = btrfs_join_transaction(rc->extent_root, 1); + if (IS_ERR(trans)) { + if (!err) + btrfs_block_rsv_release(rc->extent_root, + rc->block_rsv, num_bytes); + return PTR_ERR(trans); + } if (!err) { if (num_bytes != rc->merging_rsv_size) { @@ -3222,6 +3230,7 @@ truncate: trans = btrfs_join_transaction(root, 0); if (IS_ERR(trans)) { btrfs_free_path(path); + ret = PTR_ERR(trans); goto out; } @@ -3628,6 +3637,7 @@ int prepare_to_relocate(struct reloc_control *rc) set_reloc_control(rc); trans = btrfs_join_transaction(rc->extent_root, 1); + BUG_ON(IS_ERR(trans)); btrfs_commit_transaction(trans, rc->extent_root); return 0; } @@ -3644,6 +3654,7 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc) u32 item_size; int ret; int err = 0; + int progress = 0; path = btrfs_alloc_path(); if (!path) @@ -3656,8 +3667,10 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc) } while (1) { + progress++; trans = btrfs_start_transaction(rc->extent_root, 0); - + BUG_ON(IS_ERR(trans)); +restart: if (update_backref_cache(trans, &rc->backref_cache)) { btrfs_end_transaction(trans, rc->extent_root); continue; @@ -3770,6 +3783,15 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc) } } } + if (trans && progress && err == -ENOSPC) { + ret = btrfs_force_chunk_alloc(trans, rc->extent_root, + rc->block_group->flags); + if (ret == 0) { + err = 0; + progress = 0; + goto restart; + } + } btrfs_release_path(rc->extent_root, path); clear_extent_bits(&rc->processed_blocks, 0, (u64)-1, EXTENT_DIRTY, @@ -3804,7 +3826,10 @@ static noinline_for_stack int relocate_block_group(struct reloc_control *rc) /* get rid of pinned extents */ trans = btrfs_join_transaction(rc->extent_root, 1); - btrfs_commit_transaction(trans, rc->extent_root); + if (IS_ERR(trans)) + err = PTR_ERR(trans); + else + btrfs_commit_transaction(trans, rc->extent_root); out_free: btrfs_free_block_rsv(rc->extent_root, rc->block_rsv); btrfs_free_path(path); @@ -4022,6 +4047,7 @@ static noinline_for_stack int mark_garbage_root(struct btrfs_root *root) int ret; trans = btrfs_start_transaction(root->fs_info->tree_root, 0); + BUG_ON(IS_ERR(trans)); memset(&root->root_item.drop_progress, 0, sizeof(root->root_item.drop_progress)); @@ -4125,6 +4151,11 @@ int btrfs_recover_relocation(struct btrfs_root *root) set_reloc_control(rc); trans = btrfs_join_transaction(rc->extent_root, 1); + if (IS_ERR(trans)) { + unset_reloc_control(rc); + err = PTR_ERR(trans); + goto out_free; + } rc->merge_reloc_tree = 1; @@ -4154,9 +4185,13 @@ int btrfs_recover_relocation(struct btrfs_root *root) unset_reloc_control(rc); trans = btrfs_join_transaction(rc->extent_root, 1); - btrfs_commit_transaction(trans, rc->extent_root); -out: + if (IS_ERR(trans)) + err = PTR_ERR(trans); + else + btrfs_commit_transaction(trans, rc->extent_root); +out_free: kfree(rc); +out: while (!list_empty(&reloc_roots)) { reloc_root = list_entry(reloc_roots.next, struct btrfs_root, root_list); diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index b2130c4..d39a989 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -155,7 +155,8 @@ enum { Opt_nossd, Opt_ssd_spread, Opt_thread_pool, Opt_noacl, Opt_compress, Opt_compress_type, Opt_compress_force, Opt_compress_force_type, Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard, - Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed, Opt_err, + Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed, + Opt_enospc_debug, Opt_err, }; static match_table_t tokens = { @@ -184,6 +185,7 @@ static match_table_t tokens = { {Opt_space_cache, "space_cache"}, {Opt_clear_cache, "clear_cache"}, {Opt_user_subvol_rm_allowed, "user_subvol_rm_allowed"}, + {Opt_enospc_debug, "enospc_debug"}, {Opt_err, NULL}, }; @@ -358,6 +360,9 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) case Opt_user_subvol_rm_allowed: btrfs_set_opt(info->mount_opt, USER_SUBVOL_RM_ALLOWED); break; + case Opt_enospc_debug: + btrfs_set_opt(info->mount_opt, ENOSPC_DEBUG); + break; case Opt_err: printk(KERN_INFO "btrfs: unrecognized mount option " "'%s'\n", p); @@ -383,7 +388,7 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags, struct btrfs_fs_devices **fs_devices) { substring_t args[MAX_OPT_ARGS]; - char *opts, *p; + char *opts, *orig, *p; int error = 0; int intarg; @@ -397,6 +402,7 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags, opts = kstrdup(options, GFP_KERNEL); if (!opts) return -ENOMEM; + orig = opts; while ((p = strsep(&opts, ",")) != NULL) { int token; @@ -432,7 +438,7 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags, } out_free_opts: - kfree(opts); + kfree(orig); out: /* * If no subvolume name is specified we use the default one. Allocate @@ -623,6 +629,8 @@ int btrfs_sync_fs(struct super_block *sb, int wait) btrfs_wait_ordered_extents(root, 0, 0); trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) + return PTR_ERR(trans); ret = btrfs_commit_transaction(trans, root); return ret; } @@ -761,6 +769,8 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, } btrfs_close_devices(fs_devices); + kfree(fs_info); + kfree(tree_root); } else { char b[BDEVNAME_SIZE]; diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index bae5c7b..3d73c8d 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -1161,6 +1161,11 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans, INIT_DELAYED_WORK(&ac->work, do_async_commit); ac->root = root; ac->newtrans = btrfs_join_transaction(root, 0); + if (IS_ERR(ac->newtrans)) { + int err = PTR_ERR(ac->newtrans); + kfree(ac); + return err; + } /* take transaction reference */ mutex_lock(&root->fs_info->trans_mutex); diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index 054744a..a4bbb85 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -338,6 +338,12 @@ static noinline int overwrite_item(struct btrfs_trans_handle *trans, } dst_copy = kmalloc(item_size, GFP_NOFS); src_copy = kmalloc(item_size, GFP_NOFS); + if (!dst_copy || !src_copy) { + btrfs_release_path(root, path); + kfree(dst_copy); + kfree(src_copy); + return -ENOMEM; + } read_extent_buffer(eb, src_copy, src_ptr, item_size); @@ -665,6 +671,9 @@ static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans, btrfs_dir_item_key_to_cpu(leaf, di, &location); name_len = btrfs_dir_name_len(leaf, di); name = kmalloc(name_len, GFP_NOFS); + if (!name) + return -ENOMEM; + read_extent_buffer(leaf, name, (unsigned long)(di + 1), name_len); btrfs_release_path(root, path); @@ -744,6 +753,9 @@ static noinline int backref_in_log(struct btrfs_root *log, int match = 0; path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + ret = btrfs_search_slot(NULL, log, key, path, 0, 0); if (ret != 0) goto out; @@ -967,6 +979,8 @@ static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans, key.offset = (u64)-1; path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; while (1) { ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); @@ -1178,6 +1192,9 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans, name_len = btrfs_dir_name_len(eb, di); name = kmalloc(name_len, GFP_NOFS); + if (!name) + return -ENOMEM; + log_type = btrfs_dir_type(eb, di); read_extent_buffer(eb, name, (unsigned long)(di + 1), name_len); @@ -1692,6 +1709,8 @@ static noinline int walk_down_log_tree(struct btrfs_trans_handle *trans, root_owner = btrfs_header_owner(parent); next = btrfs_find_create_tree_block(root, bytenr, blocksize); + if (!next) + return -ENOMEM; if (*level == 1) { wc->process_func(root, next, wc, ptr_gen); @@ -2032,6 +2051,7 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, wait_log_commit(trans, log_root_tree, log_root_tree->log_transid); mutex_unlock(&log_root_tree->log_mutex); + ret = 0; goto out; } atomic_set(&log_root_tree->log_commit[index2], 1); @@ -2096,7 +2116,7 @@ out: smp_mb(); if (waitqueue_active(&root->log_commit_wait[index1])) wake_up(&root->log_commit_wait[index1]); - return 0; + return ret; } static void free_log_tree(struct btrfs_trans_handle *trans, @@ -2194,6 +2214,9 @@ int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans, log = root->log_root; path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + di = btrfs_lookup_dir_item(trans, log, path, dir->i_ino, name, name_len, -1); if (IS_ERR(di)) { @@ -2594,6 +2617,9 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, ins_data = kmalloc(nr * sizeof(struct btrfs_key) + nr * sizeof(u32), GFP_NOFS); + if (!ins_data) + return -ENOMEM; + ins_sizes = (u32 *)ins_data; ins_keys = (struct btrfs_key *)(ins_data + nr * sizeof(u32)); @@ -2725,7 +2751,13 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, log = root->log_root; path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; dst_path = btrfs_alloc_path(); + if (!dst_path) { + btrfs_free_path(path); + return -ENOMEM; + } min_key.objectid = inode->i_ino; min_key.type = BTRFS_INODE_ITEM_KEY; @@ -3080,6 +3112,7 @@ int btrfs_recover_log_trees(struct btrfs_root *log_root_tree) BUG_ON(!path); trans = btrfs_start_transaction(fs_info->tree_root, 0); + BUG_ON(IS_ERR(trans)); wc.trans = trans; wc.pin = 1; diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index d158530..dd13eb8 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -1213,6 +1213,10 @@ static int btrfs_rm_dev_item(struct btrfs_root *root, return -ENOMEM; trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) { + btrfs_free_path(path); + return PTR_ERR(trans); + } key.objectid = BTRFS_DEV_ITEMS_OBJECTID; key.type = BTRFS_DEV_ITEM_KEY; key.offset = device->devid; @@ -1334,11 +1338,11 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) ret = btrfs_shrink_device(device, 0); if (ret) - goto error_brelse; + goto error_undo; ret = btrfs_rm_dev_item(root->fs_info->chunk_root, device); if (ret) - goto error_brelse; + goto error_undo; device->in_fs_metadata = 0; @@ -1412,6 +1416,13 @@ out: mutex_unlock(&root->fs_info->volume_mutex); mutex_unlock(&uuid_mutex); return ret; +error_undo: + if (device->writeable) { + list_add(&device->dev_alloc_list, + &root->fs_info->fs_devices->alloc_list); + root->fs_info->fs_devices->rw_devices++; + } + goto error_brelse; } /* @@ -1601,11 +1612,19 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path) ret = find_next_devid(root, &device->devid); if (ret) { + kfree(device->name); kfree(device); goto error; } trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) { + kfree(device->name); + kfree(device); + ret = PTR_ERR(trans); + goto error; + } + lock_chunks(root); device->writeable = 1; @@ -1621,7 +1640,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path) device->dev_root = root->fs_info->dev_root; device->bdev = bdev; device->in_fs_metadata = 1; - device->mode = 0; + device->mode = FMODE_EXCL; set_blocksize(device->bdev, 4096); if (seeding_dev) { @@ -1873,7 +1892,7 @@ static int btrfs_relocate_chunk(struct btrfs_root *root, return ret; trans = btrfs_start_transaction(root, 0); - BUG_ON(!trans); + BUG_ON(IS_ERR(trans)); lock_chunks(root); @@ -2047,7 +2066,7 @@ int btrfs_balance(struct btrfs_root *dev_root) BUG_ON(ret); trans = btrfs_start_transaction(dev_root, 0); - BUG_ON(!trans); + BUG_ON(IS_ERR(trans)); ret = btrfs_grow_device(trans, device, old_size); BUG_ON(ret); @@ -2213,6 +2232,11 @@ again: /* Shrinking succeeded, else we would be at "done". */ trans = btrfs_start_transaction(root, 0); + if (IS_ERR(trans)) { + ret = PTR_ERR(trans); + goto done; + } + lock_chunks(root); device->disk_total_bytes = new_size; diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c index a577653..d779cef 100644 --- a/fs/btrfs/xattr.c +++ b/fs/btrfs/xattr.c @@ -370,7 +370,8 @@ int btrfs_removexattr(struct dentry *dentry, const char *name) } int btrfs_xattr_security_init(struct btrfs_trans_handle *trans, - struct inode *inode, struct inode *dir) + struct inode *inode, struct inode *dir, + const struct qstr *qstr) { int err; size_t len; @@ -378,7 +379,8 @@ int btrfs_xattr_security_init(struct btrfs_trans_handle *trans, char *suffix; char *name; - err = security_inode_init_security(inode, dir, &suffix, &value, &len); + err = security_inode_init_security(inode, dir, qstr, &suffix, &value, + &len); if (err) { if (err == -EOPNOTSUPP) return 0; diff --git a/fs/btrfs/xattr.h b/fs/btrfs/xattr.h index 7a43fd6..b3cc803 100644 --- a/fs/btrfs/xattr.h +++ b/fs/btrfs/xattr.h @@ -37,6 +37,7 @@ extern int btrfs_setxattr(struct dentry *dentry, const char *name, extern int btrfs_removexattr(struct dentry *dentry, const char *name); extern int btrfs_xattr_security_init(struct btrfs_trans_handle *trans, - struct inode *inode, struct inode *dir); + struct inode *inode, struct inode *dir, + const struct qstr *qstr); #endif /* __XATTR__ */ diff --git a/fs/cachefiles/namei.c b/fs/cachefiles/namei.c index 42c7faf..a0358c2 100644 --- a/fs/cachefiles/namei.c +++ b/fs/cachefiles/namei.c @@ -275,6 +275,7 @@ static int cachefiles_bury_object(struct cachefiles_cache *cache, bool preemptive) { struct dentry *grave, *trap; + struct path path, path_to_graveyard; char nbuffer[8 + 8 + 1]; int ret; @@ -287,10 +288,18 @@ static int cachefiles_bury_object(struct cachefiles_cache *cache, /* non-directories can just be unlinked */ if (!S_ISDIR(rep->d_inode->i_mode)) { _debug("unlink stale object"); - ret = vfs_unlink(dir->d_inode, rep); - if (preemptive) - cachefiles_mark_object_buried(cache, rep); + path.mnt = cache->mnt; + path.dentry = dir; + ret = security_path_unlink(&path, rep); + if (ret < 0) { + cachefiles_io_error(cache, "Unlink security error"); + } else { + ret = vfs_unlink(dir->d_inode, rep); + + if (preemptive) + cachefiles_mark_object_buried(cache, rep); + } mutex_unlock(&dir->d_inode->i_mutex); @@ -379,12 +388,23 @@ try_again: } /* attempt the rename */ - ret = vfs_rename(dir->d_inode, rep, cache->graveyard->d_inode, grave); - if (ret != 0 && ret != -ENOMEM) - cachefiles_io_error(cache, "Rename failed with error %d", ret); + path.mnt = cache->mnt; + path.dentry = dir; + path_to_graveyard.mnt = cache->mnt; + path_to_graveyard.dentry = cache->graveyard; + ret = security_path_rename(&path, rep, &path_to_graveyard, grave); + if (ret < 0) { + cachefiles_io_error(cache, "Rename security error %d", ret); + } else { + ret = vfs_rename(dir->d_inode, rep, + cache->graveyard->d_inode, grave); + if (ret != 0 && ret != -ENOMEM) + cachefiles_io_error(cache, + "Rename failed with error %d", ret); - if (preemptive) - cachefiles_mark_object_buried(cache, rep); + if (preemptive) + cachefiles_mark_object_buried(cache, rep); + } unlock_rename(cache->graveyard, dir); dput(grave); @@ -448,6 +468,7 @@ int cachefiles_walk_to_object(struct cachefiles_object *parent, { struct cachefiles_cache *cache; struct dentry *dir, *next = NULL; + struct path path; unsigned long start; const char *name; int ret, nlen; @@ -458,6 +479,7 @@ int cachefiles_walk_to_object(struct cachefiles_object *parent, cache = container_of(parent->fscache.cache, struct cachefiles_cache, cache); + path.mnt = cache->mnt; ASSERT(parent->dentry); ASSERT(parent->dentry->d_inode); @@ -511,6 +533,10 @@ lookup_again: if (ret < 0) goto create_error; + path.dentry = dir; + ret = security_path_mkdir(&path, next, 0); + if (ret < 0) + goto create_error; start = jiffies; ret = vfs_mkdir(dir->d_inode, next, 0); cachefiles_hist(cachefiles_mkdir_histogram, start); @@ -536,6 +562,10 @@ lookup_again: if (ret < 0) goto create_error; + path.dentry = dir; + ret = security_path_mknod(&path, next, S_IFREG, 0); + if (ret < 0) + goto create_error; start = jiffies; ret = vfs_create(dir->d_inode, next, S_IFREG, NULL); cachefiles_hist(cachefiles_create_histogram, start); @@ -692,6 +722,7 @@ struct dentry *cachefiles_get_directory(struct cachefiles_cache *cache, { struct dentry *subdir; unsigned long start; + struct path path; int ret; _enter(",,%s", dirname); @@ -719,6 +750,11 @@ struct dentry *cachefiles_get_directory(struct cachefiles_cache *cache, _debug("attempt mkdir"); + path.mnt = cache->mnt; + path.dentry = dir; + ret = security_path_mkdir(&path, subdir, 0700); + if (ret < 0) + goto mkdir_error; ret = vfs_mkdir(dir->d_inode, subdir, 0700); if (ret < 0) goto mkdir_error; diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 60d27bc..6b61ded 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -1560,9 +1560,10 @@ retry_locked: /* NOTE: no side-effects allowed, until we take s_mutex */ revoking = cap->implemented & ~cap->issued; - if (revoking) - dout(" mds%d revoking %s\n", cap->mds, - ceph_cap_string(revoking)); + dout(" mds%d cap %p issued %s implemented %s revoking %s\n", + cap->mds, cap, ceph_cap_string(cap->issued), + ceph_cap_string(cap->implemented), + ceph_cap_string(revoking)); if (cap == ci->i_auth_cap && (cap->issued & CEPH_CAP_FILE_WR)) { @@ -1658,6 +1659,8 @@ ack: if (cap == ci->i_auth_cap && ci->i_dirty_caps) flushing = __mark_caps_flushing(inode, session); + else + flushing = 0; mds = cap->mds; /* remember mds, so we don't repeat */ sent++; @@ -1940,6 +1943,35 @@ void ceph_kick_flushing_caps(struct ceph_mds_client *mdsc, } } +static void kick_flushing_inode_caps(struct ceph_mds_client *mdsc, + struct ceph_mds_session *session, + struct inode *inode) +{ + struct ceph_inode_info *ci = ceph_inode(inode); + struct ceph_cap *cap; + int delayed = 0; + + spin_lock(&inode->i_lock); + cap = ci->i_auth_cap; + dout("kick_flushing_inode_caps %p flushing %s flush_seq %lld\n", inode, + ceph_cap_string(ci->i_flushing_caps), ci->i_cap_flush_seq); + __ceph_flush_snaps(ci, &session, 1); + if (ci->i_flushing_caps) { + delayed = __send_cap(mdsc, cap, CEPH_CAP_OP_FLUSH, + __ceph_caps_used(ci), + __ceph_caps_wanted(ci), + cap->issued | cap->implemented, + ci->i_flushing_caps, NULL); + if (delayed) { + spin_lock(&inode->i_lock); + __cap_delay_requeue(mdsc, ci); + spin_unlock(&inode->i_lock); + } + } else { + spin_unlock(&inode->i_lock); + } +} + /* * Take references to capabilities we hold, so that we don't release @@ -2687,7 +2719,7 @@ static void handle_cap_import(struct ceph_mds_client *mdsc, ceph_add_cap(inode, session, cap_id, -1, issued, wanted, seq, mseq, realmino, CEPH_CAP_FLAG_AUTH, NULL /* no caps context */); - try_flush_caps(inode, session, NULL); + kick_flushing_inode_caps(mdsc, session, inode); up_read(&mdsc->snap_rwsem); /* make sure we re-request max_size, if necessary */ @@ -2785,8 +2817,7 @@ void ceph_handle_caps(struct ceph_mds_session *session, case CEPH_CAP_OP_IMPORT: handle_cap_import(mdsc, inode, h, session, snaptrace, snaptrace_len); - ceph_check_caps(ceph_inode(inode), CHECK_CAPS_NODELAY, - session); + ceph_check_caps(ceph_inode(inode), 0, session); goto done_unlocked; } diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c index 0bc68de..ebafa65 100644 --- a/fs/ceph/dir.c +++ b/fs/ceph/dir.c @@ -409,7 +409,7 @@ more: spin_lock(&inode->i_lock); if (ci->i_release_count == fi->dir_release_count) { dout(" marking %p complete\n", inode); - ci->i_ceph_flags |= CEPH_I_COMPLETE; + /* ci->i_ceph_flags |= CEPH_I_COMPLETE; */ ci->i_max_offset = filp->f_pos; } spin_unlock(&inode->i_lock); @@ -496,6 +496,7 @@ struct dentry *ceph_finish_lookup(struct ceph_mds_request *req, /* .snap dir? */ if (err == -ENOENT && + ceph_snap(parent) == CEPH_NOSNAP && strcmp(dentry->d_name.name, fsc->mount_options->snapdir_name) == 0) { struct inode *inode = ceph_get_snapdir(parent); @@ -992,7 +993,7 @@ static int ceph_d_revalidate(struct dentry *dentry, struct nameidata *nd) { struct inode *dir; - if (nd->flags & LOOKUP_RCU) + if (nd && nd->flags & LOOKUP_RCU) return -ECHILD; dir = dentry->d_parent->d_inode; @@ -1029,28 +1030,8 @@ out_touch: static void ceph_dentry_release(struct dentry *dentry) { struct ceph_dentry_info *di = ceph_dentry(dentry); - struct inode *parent_inode = NULL; - u64 snapid = CEPH_NOSNAP; - if (!IS_ROOT(dentry)) { - parent_inode = dentry->d_parent->d_inode; - if (parent_inode) - snapid = ceph_snap(parent_inode); - } - dout("dentry_release %p parent %p\n", dentry, parent_inode); - if (parent_inode && snapid != CEPH_SNAPDIR) { - struct ceph_inode_info *ci = ceph_inode(parent_inode); - - spin_lock(&parent_inode->i_lock); - if (ci->i_shared_gen == di->lease_shared_gen || - snapid <= CEPH_MAXSNAP) { - dout(" clearing %p complete (d_release)\n", - parent_inode); - ci->i_ceph_flags &= ~CEPH_I_COMPLETE; - ci->i_release_count++; - } - spin_unlock(&parent_inode->i_lock); - } + dout("dentry_release %p\n", dentry); if (di) { ceph_dentry_lru_del(dentry); if (di->lease_session) diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index e835eff..193bfa5 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -707,13 +707,9 @@ static int fill_inode(struct inode *inode, (issued & CEPH_CAP_FILE_EXCL) == 0 && (ci->i_ceph_flags & CEPH_I_COMPLETE) == 0) { dout(" marking %p complete (empty)\n", inode); - ci->i_ceph_flags |= CEPH_I_COMPLETE; + /* ci->i_ceph_flags |= CEPH_I_COMPLETE; */ ci->i_max_offset = 2; } - - /* it may be better to set st_size in getattr instead? */ - if (ceph_test_mount_opt(ceph_sb_to_client(inode->i_sb), RBYTES)) - inode->i_size = ci->i_rbytes; break; default: pr_err("fill_inode %llx.%llx BAD mode 0%o\n", @@ -1819,7 +1815,11 @@ int ceph_getattr(struct vfsmount *mnt, struct dentry *dentry, else stat->dev = 0; if (S_ISDIR(inode->i_mode)) { - stat->size = ci->i_rbytes; + if (ceph_test_mount_opt(ceph_sb_to_client(inode->i_sb), + RBYTES)) + stat->size = ci->i_rbytes; + else + stat->size = ci->i_files + ci->i_subdirs; stat->blocks = 0; stat->blksize = 65536; } diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 1e30d19..a1ee8fa 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -693,9 +693,11 @@ static int __choose_mds(struct ceph_mds_client *mdsc, dout("choose_mds %p %llx.%llx " "frag %u mds%d (%d/%d)\n", inode, ceph_vinop(inode), - frag.frag, frag.mds, + frag.frag, mds, (int)r, frag.ndist); - return mds; + if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >= + CEPH_MDS_STATE_ACTIVE) + return mds; } /* since this file/dir wasn't known to be @@ -708,7 +710,9 @@ static int __choose_mds(struct ceph_mds_client *mdsc, dout("choose_mds %p %llx.%llx " "frag %u mds%d (auth)\n", inode, ceph_vinop(inode), frag.frag, mds); - return mds; + if (ceph_mdsmap_get_state(mdsc->mdsmap, mds) >= + CEPH_MDS_STATE_ACTIVE) + return mds; } } } diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index 39c243a..f40b913 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -584,10 +584,14 @@ static void queue_realm_cap_snaps(struct ceph_snap_realm *realm) if (lastinode) iput(lastinode); - dout("queue_realm_cap_snaps %p %llx children\n", realm, realm->ino); - list_for_each_entry(child, &realm->children, child_item) - queue_realm_cap_snaps(child); + list_for_each_entry(child, &realm->children, child_item) { + dout("queue_realm_cap_snaps %p %llx queue child %p %llx\n", + realm, realm->ino, child, child->ino); + list_del_init(&child->dirty_item); + list_add(&child->dirty_item, &realm->dirty_item); + } + list_del_init(&realm->dirty_item); dout("queue_realm_cap_snaps %p %llx done\n", realm, realm->ino); } @@ -683,7 +687,9 @@ more: * queue cap snaps _after_ we've built the new snap contexts, * so that i_head_snapc can be set appropriately. */ - list_for_each_entry(realm, &dirty_realms, dirty_item) { + while (!list_empty(&dirty_realms)) { + realm = list_first_entry(&dirty_realms, struct ceph_snap_realm, + dirty_item); queue_realm_cap_snaps(realm); } diff --git a/fs/ceph/super.c b/fs/ceph/super.c index bf6f0f3..9c50854 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -290,6 +290,8 @@ static int parse_mount_options(struct ceph_mount_options **pfsopt, fsopt->rsize = CEPH_MOUNT_RSIZE_DEFAULT; fsopt->snapdir_name = kstrdup(CEPH_SNAPDIRNAME_DEFAULT, GFP_KERNEL); + fsopt->caps_wanted_delay_min = CEPH_CAPS_WANTED_DELAY_MIN_DEFAULT; + fsopt->caps_wanted_delay_max = CEPH_CAPS_WANTED_DELAY_MAX_DEFAULT; fsopt->cap_release_safety = CEPH_CAP_RELEASE_SAFETY_DEFAULT; fsopt->max_readdir = CEPH_MAX_READDIR_DEFAULT; fsopt->max_readdir_bytes = CEPH_MAX_READDIR_BYTES_DEFAULT; diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index 6e12a6b..8c9eba6 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -219,6 +219,7 @@ static struct ceph_inode_xattr *__get_xattr(struct ceph_inode_info *ci, struct rb_node **p; struct rb_node *parent = NULL; struct ceph_inode_xattr *xattr = NULL; + int name_len = strlen(name); int c; p = &ci->i_xattrs.index.rb_node; @@ -226,6 +227,8 @@ static struct ceph_inode_xattr *__get_xattr(struct ceph_inode_info *ci, parent = *p; xattr = rb_entry(parent, struct ceph_inode_xattr, node); c = strncmp(name, xattr->name, xattr->name_len); + if (c == 0 && name_len > xattr->name_len) + c = 1; if (c < 0) p = &(*p)->rb_left; else if (c > 0) diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig index ee45648..7cb0f7f 100644 --- a/fs/cifs/Kconfig +++ b/fs/cifs/Kconfig @@ -3,6 +3,7 @@ config CIFS depends on INET select NLS select CRYPTO + select CRYPTO_MD4 select CRYPTO_MD5 select CRYPTO_HMAC select CRYPTO_ARC4 diff --git a/fs/cifs/Makefile b/fs/cifs/Makefile index 43b19dd..d875584 100644 --- a/fs/cifs/Makefile +++ b/fs/cifs/Makefile @@ -5,7 +5,7 @@ obj-$(CONFIG_CIFS) += cifs.o cifs-y := cifsfs.o cifssmb.o cifs_debug.o connect.o dir.o file.o inode.o \ link.o misc.o netmisc.o smbdes.o smbencrypt.o transport.o asn1.o \ - md4.o md5.o cifs_unicode.o nterr.o xattr.o cifsencrypt.o \ + cifs_unicode.o nterr.o xattr.o cifsencrypt.o \ readdir.o ioctl.o sess.o export.o cifs-$(CONFIG_CIFS_ACL) += cifsacl.o diff --git a/fs/cifs/README b/fs/cifs/README index 46af99a..fe16835 100644 --- a/fs/cifs/README +++ b/fs/cifs/README @@ -452,6 +452,11 @@ A partial list of the supported mount options follows: if oplock (caching token) is granted and held. Note that direct allows write operations larger than page size to be sent to the server. + strictcache Use for switching on strict cache mode. In this mode the + client read from the cache all the time it has Oplock Level II, + otherwise - read from the server. All written data are stored + in the cache, but if the client doesn't have Exclusive Oplock, + it writes the data to the server. acl Allow setfacl and getfacl to manage posix ACLs if server supports them. (default) noacl Do not allow setfacl and getfacl calls on this mount diff --git a/fs/cifs/cifs_debug.c b/fs/cifs/cifs_debug.c index ede9830..65829d3 100644 --- a/fs/cifs/cifs_debug.c +++ b/fs/cifs/cifs_debug.c @@ -79,11 +79,11 @@ void cifs_dump_mids(struct TCP_Server_Info *server) spin_lock(&GlobalMid_Lock); list_for_each(tmp, &server->pending_mid_q) { mid_entry = list_entry(tmp, struct mid_q_entry, qhead); - cERROR(1, "State: %d Cmd: %d Pid: %d Tsk: %p Mid %d", + cERROR(1, "State: %d Cmd: %d Pid: %d Cbdata: %p Mid %d", mid_entry->midState, (int)mid_entry->command, mid_entry->pid, - mid_entry->tsk, + mid_entry->callback_data, mid_entry->mid); #ifdef CONFIG_CIFS_STATS2 cERROR(1, "IsLarge: %d buf: %p time rcv: %ld now: %ld", @@ -218,11 +218,11 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v) mid_entry = list_entry(tmp3, struct mid_q_entry, qhead); seq_printf(m, "\tState: %d com: %d pid:" - " %d tsk: %p mid %d\n", + " %d cbdata: %p mid %d\n", mid_entry->midState, (int)mid_entry->command, mid_entry->pid, - mid_entry->tsk, + mid_entry->callback_data, mid_entry->mid); } spin_unlock(&GlobalMid_Lock); @@ -331,7 +331,7 @@ static int cifs_stats_proc_show(struct seq_file *m, void *v) atomic_read(&totSmBufAllocCount)); #endif /* CONFIG_CIFS_STATS2 */ - seq_printf(m, "Operations (MIDs): %d\n", midCount.counter); + seq_printf(m, "Operations (MIDs): %d\n", atomic_read(&midCount)); seq_printf(m, "\n%d session %d share reconnects\n", tcpSesReconnectCount.counter, tconInfoReconnectCount.counter); diff --git a/fs/cifs/cifs_dfs_ref.c b/fs/cifs/cifs_dfs_ref.c index 7ed3653..0a265ad 100644 --- a/fs/cifs/cifs_dfs_ref.c +++ b/fs/cifs/cifs_dfs_ref.c @@ -282,8 +282,6 @@ static struct vfsmount *cifs_dfs_do_automount(struct dentry *mntpt) cFYI(1, "in %s", __func__); BUG_ON(IS_ROOT(mntpt)); - xid = GetXid(); - /* * The MSDFS spec states that paths in DFS referral requests and * responses must be prefixed by a single '\' character instead of @@ -293,20 +291,21 @@ static struct vfsmount *cifs_dfs_do_automount(struct dentry *mntpt) mnt = ERR_PTR(-ENOMEM); full_path = build_path_from_dentry(mntpt); if (full_path == NULL) - goto free_xid; + goto cdda_exit; cifs_sb = CIFS_SB(mntpt->d_inode->i_sb); tlink = cifs_sb_tlink(cifs_sb); - mnt = ERR_PTR(-EINVAL); if (IS_ERR(tlink)) { mnt = ERR_CAST(tlink); goto free_full_path; } ses = tlink_tcon(tlink)->ses; + xid = GetXid(); rc = get_dfs_path(xid, ses, full_path + 1, cifs_sb->local_nls, &num_referrals, &referrals, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + FreeXid(xid); cifs_put_tlink(tlink); @@ -339,8 +338,7 @@ success: free_dfs_info_array(referrals, num_referrals); free_full_path: kfree(full_path); -free_xid: - FreeXid(xid); +cdda_exit: cFYI(1, "leaving %s" , __func__); return mnt; } diff --git a/fs/cifs/cifs_fs_sb.h b/fs/cifs/cifs_fs_sb.h index 7852cd6..ac51cd2 100644 --- a/fs/cifs/cifs_fs_sb.h +++ b/fs/cifs/cifs_fs_sb.h @@ -40,6 +40,7 @@ #define CIFS_MOUNT_FSCACHE 0x8000 /* local caching enabled */ #define CIFS_MOUNT_MF_SYMLINKS 0x10000 /* Minshall+French Symlinks enabled */ #define CIFS_MOUNT_MULTIUSER 0x20000 /* multiuser mount */ +#define CIFS_MOUNT_STRICT_IO 0x40000 /* strict cache mode */ struct cifs_sb_info { struct rb_root tlink_tree; diff --git a/fs/cifs/cifs_unicode.c b/fs/cifs/cifs_unicode.c index 430f510..fc0fd4f 100644 --- a/fs/cifs/cifs_unicode.c +++ b/fs/cifs/cifs_unicode.c @@ -44,10 +44,14 @@ cifs_ucs2_bytes(const __le16 *from, int maxbytes, int charlen, outlen = 0; int maxwords = maxbytes / 2; char tmp[NLS_MAX_CHARSET_SIZE]; + __u16 ftmp; - for (i = 0; i < maxwords && from[i]; i++) { - charlen = codepage->uni2char(le16_to_cpu(from[i]), tmp, - NLS_MAX_CHARSET_SIZE); + for (i = 0; i < maxwords; i++) { + ftmp = get_unaligned_le16(&from[i]); + if (ftmp == 0) + break; + + charlen = codepage->uni2char(ftmp, tmp, NLS_MAX_CHARSET_SIZE); if (charlen > 0) outlen += charlen; else @@ -58,9 +62,9 @@ cifs_ucs2_bytes(const __le16 *from, int maxbytes, } /* - * cifs_mapchar - convert a little-endian char to proper char in codepage + * cifs_mapchar - convert a host-endian char to proper char in codepage * @target - where converted character should be copied - * @src_char - 2 byte little-endian source character + * @src_char - 2 byte host-endian source character * @cp - codepage to which character should be converted * @mapchar - should character be mapped according to mapchars mount option? * @@ -69,7 +73,7 @@ cifs_ucs2_bytes(const __le16 *from, int maxbytes, * enough to hold the result of the conversion (at least NLS_MAX_CHARSET_SIZE). */ static int -cifs_mapchar(char *target, const __le16 src_char, const struct nls_table *cp, +cifs_mapchar(char *target, const __u16 src_char, const struct nls_table *cp, bool mapchar) { int len = 1; @@ -82,7 +86,7 @@ cifs_mapchar(char *target, const __le16 src_char, const struct nls_table *cp, * build_path_from_dentry are modified, as they use slash as * separator. */ - switch (le16_to_cpu(src_char)) { + switch (src_char) { case UNI_COLON: *target = ':'; break; @@ -109,8 +113,7 @@ out: return len; cp_convert: - len = cp->uni2char(le16_to_cpu(src_char), target, - NLS_MAX_CHARSET_SIZE); + len = cp->uni2char(src_char, target, NLS_MAX_CHARSET_SIZE); if (len <= 0) { *target = '?'; len = 1; @@ -149,6 +152,7 @@ cifs_from_ucs2(char *to, const __le16 *from, int tolen, int fromlen, int nullsize = nls_nullsize(codepage); int fromwords = fromlen / 2; char tmp[NLS_MAX_CHARSET_SIZE]; + __u16 ftmp; /* * because the chars can be of varying widths, we need to take care @@ -158,19 +162,23 @@ cifs_from_ucs2(char *to, const __le16 *from, int tolen, int fromlen, */ safelen = tolen - (NLS_MAX_CHARSET_SIZE + nullsize); - for (i = 0; i < fromwords && from[i]; i++) { + for (i = 0; i < fromwords; i++) { + ftmp = get_unaligned_le16(&from[i]); + if (ftmp == 0) + break; + /* * check to see if converting this character might make the * conversion bleed into the null terminator */ if (outlen >= safelen) { - charlen = cifs_mapchar(tmp, from[i], codepage, mapchar); + charlen = cifs_mapchar(tmp, ftmp, codepage, mapchar); if ((outlen + charlen) > (tolen - nullsize)) break; } /* put converted char into 'to' buffer */ - charlen = cifs_mapchar(&to[outlen], from[i], codepage, mapchar); + charlen = cifs_mapchar(&to[outlen], ftmp, codepage, mapchar); outlen += charlen; } @@ -193,24 +201,21 @@ cifs_strtoUCS(__le16 *to, const char *from, int len, { int charlen; int i; - wchar_t *wchar_to = (wchar_t *)to; /* needed to quiet sparse */ + wchar_t wchar_to; /* needed to quiet sparse */ for (i = 0; len && *from; i++, from += charlen, len -= charlen) { - - /* works for 2.4.0 kernel or later */ - charlen = codepage->char2uni(from, len, &wchar_to[i]); + charlen = codepage->char2uni(from, len, &wchar_to); if (charlen < 1) { - cERROR(1, "strtoUCS: char2uni of %d returned %d", - (int)*from, charlen); + cERROR(1, "strtoUCS: char2uni of 0x%x returned %d", + *from, charlen); /* A question mark */ - to[i] = cpu_to_le16(0x003f); + wchar_to = 0x003f; charlen = 1; - } else - to[i] = cpu_to_le16(wchar_to[i]); - + } + put_unaligned_le16(wchar_to, &to[i]); } - to[i] = 0; + put_unaligned_le16(0, &to[i]); return i; } @@ -252,3 +257,79 @@ cifs_strndup_from_ucs(const char *src, const int maxlen, const bool is_unicode, return dst; } +/* + * Convert 16 bit Unicode pathname to wire format from string in current code + * page. Conversion may involve remapping up the six characters that are + * only legal in POSIX-like OS (if they are present in the string). Path + * names are little endian 16 bit Unicode on the wire + */ +int +cifsConvertToUCS(__le16 *target, const char *source, int maxlen, + const struct nls_table *cp, int mapChars) +{ + int i, j, charlen; + int len_remaining = maxlen; + char src_char; + __u16 temp; + + if (!mapChars) + return cifs_strtoUCS(target, source, PATH_MAX, cp); + + for (i = 0, j = 0; i < maxlen; j++) { + src_char = source[i]; + switch (src_char) { + case 0: + put_unaligned_le16(0, &target[j]); + goto ctoUCS_out; + case ':': + temp = UNI_COLON; + break; + case '*': + temp = UNI_ASTERIK; + break; + case '?': + temp = UNI_QUESTION; + break; + case '<': + temp = UNI_LESSTHAN; + break; + case '>': + temp = UNI_GRTRTHAN; + break; + case '|': + temp = UNI_PIPE; + break; + /* + * FIXME: We can not handle remapping backslash (UNI_SLASH) + * until all the calls to build_path_from_dentry are modified, + * as they use backslash as separator. + */ + default: + charlen = cp->char2uni(source+i, len_remaining, + &temp); + /* + * if no match, use question mark, which at least in + * some cases serves as wild card + */ + if (charlen < 1) { + temp = 0x003f; + charlen = 1; + } + len_remaining -= charlen; + /* + * character may take more than one byte in the source + * string, but will take exactly two bytes in the + * target string + */ + i += charlen; + continue; + } + put_unaligned_le16(temp, &target[j]); + i++; /* move to next char in source string */ + len_remaining--; + } + +ctoUCS_out: + return i; +} + diff --git a/fs/cifs/cifsacl.c b/fs/cifs/cifsacl.c index a437ec3..beeebf1 100644 --- a/fs/cifs/cifsacl.c +++ b/fs/cifs/cifsacl.c @@ -41,9 +41,12 @@ static struct cifs_wksid wksidarr[NUM_WK_SIDS] = { ; -/* security id for everyone */ +/* security id for everyone/world system group */ static const struct cifs_sid sid_everyone = { 1, 1, {0, 0, 0, 0, 0, 1}, {0} }; +/* security id for Authenticated Users system group */ +static const struct cifs_sid sid_authusers = { + 1, 1, {0, 0, 0, 0, 0, 5}, {11} }; /* group users */ static const struct cifs_sid sid_user = {1, 2 , {0, 0, 0, 0, 0, 5}, {} }; @@ -365,10 +368,14 @@ static void parse_dacl(struct cifs_acl *pdacl, char *end_of_acl, if (num_aces > 0) { umode_t user_mask = S_IRWXU; umode_t group_mask = S_IRWXG; - umode_t other_mask = S_IRWXO; + umode_t other_mask = S_IRWXU | S_IRWXG | S_IRWXO; ppace = kmalloc(num_aces * sizeof(struct cifs_ace *), GFP_KERNEL); + if (!ppace) { + cERROR(1, "DACL memory allocation error"); + return; + } for (i = 0; i < num_aces; ++i) { ppace[i] = (struct cifs_ace *) (acl_base + acl_size); @@ -390,6 +397,12 @@ static void parse_dacl(struct cifs_acl *pdacl, char *end_of_acl, ppace[i]->type, &fattr->cf_mode, &other_mask); + if (compare_sids(&(ppace[i]->sid), &sid_authusers)) + access_flags_to_mode(ppace[i]->access_req, + ppace[i]->type, + &fattr->cf_mode, + &other_mask); + /* memcpy((void *)(&(cifscred->aces[i])), (void *)ppace[i], diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index 66f3d50..a51585f 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -24,7 +24,6 @@ #include "cifspdu.h" #include "cifsglob.h" #include "cifs_debug.h" -#include "md5.h" #include "cifs_unicode.h" #include "cifsproto.h" #include "ntlmssp.h" @@ -37,11 +36,6 @@ /* Note that the smb header signature field on input contains the sequence number before this function is called */ -extern void mdfour(unsigned char *out, unsigned char *in, int n); -extern void E_md4hash(const unsigned char *passwd, unsigned char *p16); -extern void SMBencrypt(unsigned char *passwd, const unsigned char *c8, - unsigned char *p24); - static int cifs_calculate_signature(const struct smb_hdr *cifs_pdu, struct TCP_Server_Info *server, char *signature) { @@ -234,6 +228,7 @@ int cifs_verify_signature(struct smb_hdr *cifs_pdu, /* first calculate 24 bytes ntlm response and then 16 byte session key */ int setup_ntlm_response(struct cifsSesInfo *ses) { + int rc = 0; unsigned int temp_len = CIFS_SESS_KEY_SIZE + CIFS_AUTH_RESP_SIZE; char temp_key[CIFS_SESS_KEY_SIZE]; @@ -247,13 +242,26 @@ int setup_ntlm_response(struct cifsSesInfo *ses) } ses->auth_key.len = temp_len; - SMBNTencrypt(ses->password, ses->server->cryptkey, + rc = SMBNTencrypt(ses->password, ses->server->cryptkey, ses->auth_key.response + CIFS_SESS_KEY_SIZE); + if (rc) { + cFYI(1, "%s Can't generate NTLM response, error: %d", + __func__, rc); + return rc; + } - E_md4hash(ses->password, temp_key); - mdfour(ses->auth_key.response, temp_key, CIFS_SESS_KEY_SIZE); + rc = E_md4hash(ses->password, temp_key); + if (rc) { + cFYI(1, "%s Can't generate NT hash, error: %d", __func__, rc); + return rc; + } - return 0; + rc = mdfour(ses->auth_key.response, temp_key, CIFS_SESS_KEY_SIZE); + if (rc) + cFYI(1, "%s Can't generate NTLM session key, error: %d", + __func__, rc); + + return rc; } #ifdef CONFIG_CIFS_WEAK_PW_HASH @@ -649,9 +657,10 @@ calc_seckey(struct cifsSesInfo *ses) get_random_bytes(sec_key, CIFS_SESS_KEY_SIZE); tfm_arc4 = crypto_alloc_blkcipher("ecb(arc4)", 0, CRYPTO_ALG_ASYNC); - if (!tfm_arc4 || IS_ERR(tfm_arc4)) { + if (IS_ERR(tfm_arc4)) { + rc = PTR_ERR(tfm_arc4); cERROR(1, "could not allocate crypto API arc4\n"); - return PTR_ERR(tfm_arc4); + return rc; } desc.tfm = tfm_arc4; @@ -700,14 +709,13 @@ cifs_crypto_shash_allocate(struct TCP_Server_Info *server) unsigned int size; server->secmech.hmacmd5 = crypto_alloc_shash("hmac(md5)", 0, 0); - if (!server->secmech.hmacmd5 || - IS_ERR(server->secmech.hmacmd5)) { + if (IS_ERR(server->secmech.hmacmd5)) { cERROR(1, "could not allocate crypto hmacmd5\n"); return PTR_ERR(server->secmech.hmacmd5); } server->secmech.md5 = crypto_alloc_shash("md5", 0, 0); - if (!server->secmech.md5 || IS_ERR(server->secmech.md5)) { + if (IS_ERR(server->secmech.md5)) { cERROR(1, "could not allocate crypto md5\n"); rc = PTR_ERR(server->secmech.md5); goto crypto_allocate_md5_fail; diff --git a/fs/cifs/cifsencrypt.h b/fs/cifs/cifsencrypt.h deleted file mode 100644 index 15d2ec0..0000000 --- a/fs/cifs/cifsencrypt.h +++ /dev/null @@ -1,33 +0,0 @@ -/* - * fs/cifs/cifsencrypt.h - * - * Copyright (c) International Business Machines Corp., 2005 - * Author(s): Steve French (sfrench@us.ibm.com) - * - * Externs for misc. small encryption routines - * so we do not have to put them in cifsproto.h - * - * This library is free software; you can redistribute it and/or modify - * it under the terms of the GNU Lesser General Public License as published - * by the Free Software Foundation; either version 2.1 of the License, or - * (at your option) any later version. - * - * This library is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See - * the GNU Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public License - * along with this library; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ - -/* md4.c */ -extern void mdfour(unsigned char *out, unsigned char *in, int n); -/* smbdes.c */ -extern void E_P16(unsigned char *p14, unsigned char *p16); -extern void E_P24(unsigned char *p21, const unsigned char *c8, - unsigned char *p24); - - - diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index d9f652a..f297013 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -77,7 +77,11 @@ unsigned int cifs_max_pending = CIFS_MAX_REQ; module_param(cifs_max_pending, int, 0); MODULE_PARM_DESC(cifs_max_pending, "Simultaneous requests to server. " "Default: 50 Range: 2 to 256"); - +unsigned short echo_retries = 5; +module_param(echo_retries, ushort, 0644); +MODULE_PARM_DESC(echo_retries, "Number of echo attempts before giving up and " + "reconnecting server. Default: 5. 0 means " + "never reconnect."); extern mempool_t *cifs_sm_req_poolp; extern mempool_t *cifs_req_poolp; extern mempool_t *cifs_mid_poolp; @@ -596,10 +600,17 @@ static ssize_t cifs_file_aio_write(struct kiocb *iocb, const struct iovec *iov, { struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode; ssize_t written; + int rc; written = generic_file_aio_write(iocb, iov, nr_segs, pos); - if (!CIFS_I(inode)->clientCanCacheAll) - filemap_fdatawrite(inode->i_mapping); + + if (CIFS_I(inode)->clientCanCacheAll) + return written; + + rc = filemap_fdatawrite(inode->i_mapping); + if (rc) + cFYI(1, "cifs_file_aio_write: %d rc on %p inode", rc, inode); + return written; } @@ -729,6 +740,25 @@ const struct file_operations cifs_file_ops = { .setlease = cifs_setlease, }; +const struct file_operations cifs_file_strict_ops = { + .read = do_sync_read, + .write = do_sync_write, + .aio_read = cifs_strict_readv, + .aio_write = cifs_strict_writev, + .open = cifs_open, + .release = cifs_close, + .lock = cifs_lock, + .fsync = cifs_strict_fsync, + .flush = cifs_flush, + .mmap = cifs_file_strict_mmap, + .splice_read = generic_file_splice_read, + .llseek = cifs_llseek, +#ifdef CONFIG_CIFS_POSIX + .unlocked_ioctl = cifs_ioctl, +#endif /* CONFIG_CIFS_POSIX */ + .setlease = cifs_setlease, +}; + const struct file_operations cifs_file_direct_ops = { /* no aio, no readv - BB reevaluate whether they can be done with directio, no cache */ @@ -747,6 +777,7 @@ const struct file_operations cifs_file_direct_ops = { .llseek = cifs_llseek, .setlease = cifs_setlease, }; + const struct file_operations cifs_file_nobrl_ops = { .read = do_sync_read, .write = do_sync_write, @@ -765,6 +796,24 @@ const struct file_operations cifs_file_nobrl_ops = { .setlease = cifs_setlease, }; +const struct file_operations cifs_file_strict_nobrl_ops = { + .read = do_sync_read, + .write = do_sync_write, + .aio_read = cifs_strict_readv, + .aio_write = cifs_strict_writev, + .open = cifs_open, + .release = cifs_close, + .fsync = cifs_strict_fsync, + .flush = cifs_flush, + .mmap = cifs_file_strict_mmap, + .splice_read = generic_file_splice_read, + .llseek = cifs_llseek, +#ifdef CONFIG_CIFS_POSIX + .unlocked_ioctl = cifs_ioctl, +#endif /* CONFIG_CIFS_POSIX */ + .setlease = cifs_setlease, +}; + const struct file_operations cifs_file_direct_nobrl_ops = { /* no mmap, no aio, no readv - BB reevaluate whether they can be done with directio, no cache */ diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index 851030f..a9371b6 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -61,6 +61,7 @@ extern int cifs_rename(struct inode *, struct dentry *, struct inode *, struct dentry *); extern int cifs_revalidate_file(struct file *filp); extern int cifs_revalidate_dentry(struct dentry *); +extern void cifs_invalidate_mapping(struct inode *inode); extern int cifs_getattr(struct vfsmount *, struct dentry *, struct kstat *); extern int cifs_setattr(struct dentry *, struct iattr *); @@ -72,19 +73,27 @@ extern const struct inode_operations cifs_dfs_referral_inode_operations; /* Functions related to files and directories */ extern const struct file_operations cifs_file_ops; extern const struct file_operations cifs_file_direct_ops; /* if directio mnt */ -extern const struct file_operations cifs_file_nobrl_ops; -extern const struct file_operations cifs_file_direct_nobrl_ops; /* no brlocks */ +extern const struct file_operations cifs_file_strict_ops; /* if strictio mnt */ +extern const struct file_operations cifs_file_nobrl_ops; /* no brlocks */ +extern const struct file_operations cifs_file_direct_nobrl_ops; +extern const struct file_operations cifs_file_strict_nobrl_ops; extern int cifs_open(struct inode *inode, struct file *file); extern int cifs_close(struct inode *inode, struct file *file); extern int cifs_closedir(struct inode *inode, struct file *file); extern ssize_t cifs_user_read(struct file *file, char __user *read_data, - size_t read_size, loff_t *poffset); + size_t read_size, loff_t *poffset); +extern ssize_t cifs_strict_readv(struct kiocb *iocb, const struct iovec *iov, + unsigned long nr_segs, loff_t pos); extern ssize_t cifs_user_write(struct file *file, const char __user *write_data, - size_t write_size, loff_t *poffset); + size_t write_size, loff_t *poffset); +extern ssize_t cifs_strict_writev(struct kiocb *iocb, const struct iovec *iov, + unsigned long nr_segs, loff_t pos); extern int cifs_lock(struct file *, int, struct file_lock *); extern int cifs_fsync(struct file *, int); +extern int cifs_strict_fsync(struct file *, int); extern int cifs_flush(struct file *, fl_owner_t id); extern int cifs_file_mmap(struct file * , struct vm_area_struct *); +extern int cifs_file_strict_mmap(struct file * , struct vm_area_struct *); extern const struct file_operations cifs_dir_ops; extern int cifs_dir_open(struct inode *inode, struct file *file); extern int cifs_readdir(struct file *file, void *direntry, filldir_t filldir); @@ -118,5 +127,5 @@ extern long cifs_ioctl(struct file *filep, unsigned int cmd, unsigned long arg); extern const struct export_operations cifs_export_ops; #endif /* EXPERIMENTAL */ -#define CIFS_VERSION "1.68" +#define CIFS_VERSION "1.71" #endif /* _CIFSFS_H */ diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 606ca8b..17afb0f 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -161,46 +161,41 @@ struct TCP_Server_Info { int srv_count; /* reference counter */ /* 15 character server name + 0x20 16th byte indicating type = srv */ char server_RFC1001_name[RFC1001_NAME_LEN_WITH_NULL]; + enum statusEnum tcpStatus; /* what we think the status is */ char *hostname; /* hostname portion of UNC string */ struct socket *ssocket; struct sockaddr_storage dstaddr; struct sockaddr_storage srcaddr; /* locally bind to this IP */ +#ifdef CONFIG_NET_NS + struct net *net; +#endif wait_queue_head_t response_q; wait_queue_head_t request_q; /* if more than maxmpx to srvr must block*/ struct list_head pending_mid_q; - void *Server_NlsInfo; /* BB - placeholder for future NLS info */ - unsigned short server_codepage; /* codepage for the server */ - enum protocolEnum protocolType; - char versionMajor; - char versionMinor; - bool svlocal:1; /* local server or remote */ bool noblocksnd; /* use blocking sendmsg */ bool noautotune; /* do not autotune send buf sizes */ bool tcp_nodelay; atomic_t inFlight; /* number of requests on the wire to server */ -#ifdef CONFIG_CIFS_STATS2 - atomic_t inSend; /* requests trying to send */ - atomic_t num_waiters; /* blocked waiting to get in sendrecv */ -#endif - enum statusEnum tcpStatus; /* what we think the status is */ struct mutex srv_mutex; struct task_struct *tsk; char server_GUID[16]; char secMode; + bool session_estab; /* mark when very first sess is established */ + u16 dialect; /* dialect index that server chose */ enum securityEnum secType; unsigned int maxReq; /* Clients should submit no more */ /* than maxReq distinct unanswered SMBs to the server when using */ /* multiplexed reads or writes */ unsigned int maxBuf; /* maxBuf specifies the maximum */ /* message size the server can send or receive for non-raw SMBs */ + /* maxBuf is returned by SMB NegotiateProtocol so maxBuf is only 0 */ + /* when socket is setup (and during reconnect) before NegProt sent */ unsigned int max_rw; /* maxRw specifies the maximum */ /* message size the server can send or receive for */ /* SMB_COM_WRITE_RAW or SMB_COM_READ_RAW. */ unsigned int max_vcs; /* maximum number of smb sessions, at least those that can be specified uniquely with vcnumbers */ - char sessid[4]; /* unique token id for this session */ - /* (returned on Negotiate */ int capabilities; /* allow selective disabling of caps by smb sess */ int timeAdj; /* Adjust for difference in server time zone in sec */ __u16 CurrentMid; /* multiplex id - rotating counter */ @@ -210,20 +205,53 @@ struct TCP_Server_Info { __u32 sequence_number; /* for signing, protected by srv_mutex */ struct session_key session_key; unsigned long lstrp; /* when we got last response from this server */ - u16 dialect; /* dialect index that server chose */ struct cifs_secmech secmech; /* crypto sec mech functs, descriptors */ /* extended security flavors that server supports */ + bool sec_ntlmssp; /* supports NTLMSSP */ + bool sec_kerberosu2u; /* supports U2U Kerberos */ bool sec_kerberos; /* supports plain Kerberos */ bool sec_mskerberos; /* supports legacy MS Kerberos */ - bool sec_kerberosu2u; /* supports U2U Kerberos */ - bool sec_ntlmssp; /* supports NTLMSSP */ - bool session_estab; /* mark when very first sess is established */ + struct delayed_work echo; /* echo ping workqueue job */ #ifdef CONFIG_CIFS_FSCACHE struct fscache_cookie *fscache; /* client index cache cookie */ #endif +#ifdef CONFIG_CIFS_STATS2 + atomic_t inSend; /* requests trying to send */ + atomic_t num_waiters; /* blocked waiting to get in sendrecv */ +#endif }; /* + * Macros to allow the TCP_Server_Info->net field and related code to drop out + * when CONFIG_NET_NS isn't set. + */ + +#ifdef CONFIG_NET_NS + +static inline struct net *cifs_net_ns(struct TCP_Server_Info *srv) +{ + return srv->net; +} + +static inline void cifs_set_net_ns(struct TCP_Server_Info *srv, struct net *net) +{ + srv->net = net; +} + +#else + +static inline struct net *cifs_net_ns(struct TCP_Server_Info *srv) +{ + return &init_net; +} + +static inline void cifs_set_net_ns(struct TCP_Server_Info *srv, struct net *net) +{ +} + +#endif + +/* * Session structure. One of these for each uid session with a particular host */ struct cifsSesInfo { @@ -446,11 +474,11 @@ struct cifsInodeInfo { /* BB add in lists for dirty pages i.e. write caching info for oplock */ struct list_head openFileList; __u32 cifsAttrs; /* e.g. DOS archive bit, sparse, compressed, system */ - unsigned long time; /* jiffies of last update/check of inode */ - bool clientCanCacheRead:1; /* read oplock */ - bool clientCanCacheAll:1; /* read and writebehind oplock */ - bool delete_pending:1; /* DELETE_ON_CLOSE is set */ - bool invalid_mapping:1; /* pagecache is invalid */ + bool clientCanCacheRead; /* read oplock */ + bool clientCanCacheAll; /* read and writebehind oplock */ + bool delete_pending; /* DELETE_ON_CLOSE is set */ + bool invalid_mapping; /* pagecache is invalid */ + unsigned long time; /* jiffies of last update of inode */ u64 server_eof; /* current file size on server */ u64 uniqueid; /* server inode number */ u64 createtime; /* creation time on server */ @@ -508,6 +536,18 @@ static inline void cifs_stats_bytes_read(struct cifsTconInfo *tcon, #endif +struct mid_q_entry; + +/* + * This is the prototype for the mid callback function. When creating one, + * take special care to avoid deadlocks. Things to bear in mind: + * + * - it will be called by cifsd + * - the GlobalMid_Lock will be held + * - the mid will be removed from the pending_mid_q list + */ +typedef void (mid_callback_t)(struct mid_q_entry *mid); + /* one of these for every pending CIFS request to the server */ struct mid_q_entry { struct list_head qhead; /* mids waiting on reply from this server */ @@ -519,7 +559,8 @@ struct mid_q_entry { unsigned long when_sent; /* time when smb send finished */ unsigned long when_received; /* when demux complete (taken off wire) */ #endif - struct task_struct *tsk; /* task waiting for response */ + mid_callback_t *callback; /* call completion callback */ + void *callback_data; /* general purpose pointer for callback */ struct smb_hdr *resp_buf; /* response buffer */ int midState; /* wish this were enum but can not pass to wait_event */ __u8 command; /* smb command code */ @@ -613,7 +654,7 @@ static inline void free_dfs_info_array(struct dfs_info3_param *param, #define MID_REQUEST_SUBMITTED 2 #define MID_RESPONSE_RECEIVED 4 #define MID_RETRY_NEEDED 8 /* session closed while this request out */ -#define MID_NO_RESP_NEEDED 0x10 +#define MID_RESPONSE_MALFORMED 0x10 /* Types of response buffer returned from SendReceive2 */ #define CIFS_NO_BUFFER 0 /* Response buffer not returned */ @@ -622,12 +663,9 @@ static inline void free_dfs_info_array(struct dfs_info3_param *param, #define CIFS_IOVEC 4 /* array of response buffers */ /* Type of Request to SendReceive2 */ -#define CIFS_STD_OP 0 /* normal request timeout */ -#define CIFS_LONG_OP 1 /* long op (up to 45 sec, oplock time) */ -#define CIFS_VLONG_OP 2 /* sloow op - can take up to 180 seconds */ -#define CIFS_BLOCKING_OP 4 /* operation can block */ -#define CIFS_ASYNC_OP 8 /* do not wait for response */ -#define CIFS_TIMEOUT_MASK 0x00F /* only one of 5 above set in req */ +#define CIFS_BLOCKING_OP 1 /* operation can block */ +#define CIFS_ASYNC_OP 2 /* do not wait for response */ +#define CIFS_TIMEOUT_MASK 0x003 /* only one of above set in req */ #define CIFS_LOG_ERROR 0x010 /* log NT STATUS if non-zero */ #define CIFS_LARGE_BUF_OP 0x020 /* large request buffer */ #define CIFS_NO_RESP 0x040 /* no response buffer required */ @@ -790,6 +828,9 @@ GLOBAL_EXTERN unsigned int cifs_min_rcv; /* min size of big ntwrk buf pool */ GLOBAL_EXTERN unsigned int cifs_min_small; /* min size of small buf pool */ GLOBAL_EXTERN unsigned int cifs_max_pending; /* MAX requests at once to server*/ +/* reconnect after this many failed echo attempts */ +GLOBAL_EXTERN unsigned short echo_retries; + void cifs_oplock_break(struct work_struct *work); void cifs_oplock_break_get(struct cifsFileInfo *cfile); void cifs_oplock_break_put(struct cifsFileInfo *cfile); diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h index de36b09..b5c8cc5 100644 --- a/fs/cifs/cifspdu.h +++ b/fs/cifs/cifspdu.h @@ -23,6 +23,7 @@ #define _CIFSPDU_H #include <net/sock.h> +#include <asm/unaligned.h> #include "smbfsctl.h" #ifdef CONFIG_CIFS_WEAK_PW_HASH @@ -50,6 +51,7 @@ #define SMB_COM_SETATTR 0x09 /* trivial response */ #define SMB_COM_LOCKING_ANDX 0x24 /* trivial response */ #define SMB_COM_COPY 0x29 /* trivial rsp, fail filename ignrd*/ +#define SMB_COM_ECHO 0x2B /* echo request */ #define SMB_COM_OPEN_ANDX 0x2D /* Legacy open for old servers */ #define SMB_COM_READ_ANDX 0x2E #define SMB_COM_WRITE_ANDX 0x2F @@ -425,11 +427,49 @@ struct smb_hdr { __u16 Mid; __u8 WordCount; } __attribute__((packed)); -/* given a pointer to an smb_hdr retrieve the value of byte count */ -#define BCC(smb_var) (*(__u16 *)((char *)(smb_var) + sizeof(struct smb_hdr) + (2 * (smb_var)->WordCount))) -#define BCC_LE(smb_var) (*(__le16 *)((char *)(smb_var) + sizeof(struct smb_hdr) + (2 * (smb_var)->WordCount))) + +/* given a pointer to an smb_hdr retrieve a char pointer to the byte count */ +#define BCC(smb_var) ((unsigned char *)(smb_var) + sizeof(struct smb_hdr) + \ + (2 * (smb_var)->WordCount)) + /* given a pointer to an smb_hdr retrieve the pointer to the byte area */ -#define pByteArea(smb_var) ((unsigned char *)(smb_var) + sizeof(struct smb_hdr) + (2 * (smb_var)->WordCount) + 2) +#define pByteArea(smb_var) (BCC(smb_var) + 2) + +/* get the converted ByteCount for a SMB packet and return it */ +static inline __u16 +get_bcc(struct smb_hdr *hdr) +{ + __u16 *bc_ptr = (__u16 *)BCC(hdr); + + return get_unaligned(bc_ptr); +} + +/* get the unconverted ByteCount for a SMB packet and return it */ +static inline __u16 +get_bcc_le(struct smb_hdr *hdr) +{ + __le16 *bc_ptr = (__le16 *)BCC(hdr); + + return get_unaligned_le16(bc_ptr); +} + +/* set the ByteCount for a SMB packet in host-byte order */ +static inline void +put_bcc(__u16 count, struct smb_hdr *hdr) +{ + __u16 *bc_ptr = (__u16 *)BCC(hdr); + + put_unaligned(count, bc_ptr); +} + +/* set the ByteCount for a SMB packet in little-endian */ +static inline void +put_bcc_le(__u16 count, struct smb_hdr *hdr) +{ + __le16 *bc_ptr = (__le16 *)BCC(hdr); + + put_unaligned_le16(count, bc_ptr); +} /* * Computer Name Length (since Netbios name was length 16 with last byte 0x20) @@ -760,6 +800,20 @@ typedef struct smb_com_tconx_rsp_ext { * */ +typedef struct smb_com_echo_req { + struct smb_hdr hdr; + __le16 EchoCount; + __le16 ByteCount; + char Data[1]; +} __attribute__((packed)) ECHO_REQ; + +typedef struct smb_com_echo_rsp { + struct smb_hdr hdr; + __le16 SequenceNumber; + __le16 ByteCount; + char Data[1]; +} __attribute__((packed)) ECHO_RSP; + typedef struct smb_com_logoff_andx_req { struct smb_hdr hdr; /* wct = 2 */ __u8 AndXCommand; diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index e6d1481..8096f27 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -61,6 +61,12 @@ extern char *cifs_compose_mount_options(const char *sb_mountdata, const char *fullpath, const struct dfs_info3_param *ref, char **devname); /* extern void renew_parental_timestamps(struct dentry *direntry);*/ +extern struct mid_q_entry *AllocMidQEntry(const struct smb_hdr *smb_buffer, + struct TCP_Server_Info *server); +extern void DeleteMidQEntry(struct mid_q_entry *midEntry); +extern int cifs_call_async(struct TCP_Server_Info *server, + struct smb_hdr *in_buf, mid_callback_t *callback, + void *cbdata); extern int SendReceive(const unsigned int /* xid */ , struct cifsSesInfo *, struct smb_hdr * /* input */ , struct smb_hdr * /* out */ , @@ -79,6 +85,8 @@ extern int checkSMB(struct smb_hdr *smb, __u16 mid, unsigned int length); extern bool is_valid_oplock_break(struct smb_hdr *smb, struct TCP_Server_Info *); extern bool is_size_safe_to_change(struct cifsInodeInfo *, __u64 eof); +extern void cifs_update_eof(struct cifsInodeInfo *cifsi, loff_t offset, + unsigned int bytes_written); extern struct cifsFileInfo *find_writable_file(struct cifsInodeInfo *, bool); extern struct cifsFileInfo *find_readable_file(struct cifsInodeInfo *, bool); extern unsigned int smbCalcSize(struct smb_hdr *ptr); @@ -347,12 +355,13 @@ extern int CIFSSMBLock(const int xid, struct cifsTconInfo *tcon, const __u16 netfid, const __u64 len, const __u64 offset, const __u32 numUnlock, const __u32 numLock, const __u8 lockType, - const bool waitFlag); + const bool waitFlag, const __u8 oplock_level); extern int CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, const __u16 smb_file_id, const int get_flag, const __u64 len, struct file_lock *, const __u16 lock_type, const bool waitFlag); extern int CIFSSMBTDis(const int xid, struct cifsTconInfo *tcon); +extern int CIFSSMBEcho(struct TCP_Server_Info *server); extern int CIFSSMBLogoff(const int xid, struct cifsSesInfo *ses); extern struct cifsSesInfo *sesInfoAlloc(void); @@ -366,7 +375,7 @@ extern int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *, extern int cifs_verify_signature(struct smb_hdr *, struct TCP_Server_Info *server, __u32 expected_sequence_number); -extern void SMBNTencrypt(unsigned char *, unsigned char *, unsigned char *); +extern int SMBNTencrypt(unsigned char *, unsigned char *, unsigned char *); extern int setup_ntlm_response(struct cifsSesInfo *); extern int setup_ntlmv2_rsp(struct cifsSesInfo *, const struct nls_table *); extern int cifs_crypto_shash_allocate(struct TCP_Server_Info *); @@ -416,4 +425,11 @@ extern bool CIFSCouldBeMFSymlink(const struct cifs_fattr *fattr); extern int CIFSCheckMFSymlink(struct cifs_fattr *fattr, const unsigned char *path, struct cifs_sb_info *cifs_sb, int xid); +extern int mdfour(unsigned char *, unsigned char *, int); +extern int E_md4hash(const unsigned char *passwd, unsigned char *p16); +extern void SMBencrypt(unsigned char *passwd, const unsigned char *c8, + unsigned char *p24); +extern void E_P16(unsigned char *p14, unsigned char *p16); +extern void E_P24(unsigned char *p21, const unsigned char *c8, + unsigned char *p24); #endif /* _CIFSPROTO_H */ diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 2f6795e..904aa47 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -136,9 +136,6 @@ cifs_reconnect_tcon(struct cifsTconInfo *tcon, int smb_command) } } - if (ses->status == CifsExiting) - return -EIO; - /* * Give demultiplex thread up to 10 seconds to reconnect, should be * greater than cifs socket timeout which is 7 seconds @@ -156,7 +153,7 @@ cifs_reconnect_tcon(struct cifsTconInfo *tcon, int smb_command) * retrying until process is killed or server comes * back on-line */ - if (!tcon->retry || ses->status == CifsExiting) { + if (!tcon->retry) { cFYI(1, "gave up waiting on reconnect in smb_init"); return -EHOSTDOWN; } @@ -331,37 +328,35 @@ smb_init_no_reconnect(int smb_command, int wct, struct cifsTconInfo *tcon, static int validate_t2(struct smb_t2_rsp *pSMB) { - int rc = -EINVAL; - int total_size; - char *pBCC; + unsigned int total_size; + + /* check for plausible wct */ + if (pSMB->hdr.WordCount < 10) + goto vt2_err; - /* check for plausible wct, bcc and t2 data and parm sizes */ /* check for parm and data offset going beyond end of smb */ - if (pSMB->hdr.WordCount >= 10) { - if ((le16_to_cpu(pSMB->t2_rsp.ParameterOffset) <= 1024) && - (le16_to_cpu(pSMB->t2_rsp.DataOffset) <= 1024)) { - /* check that bcc is at least as big as parms + data */ - /* check that bcc is less than negotiated smb buffer */ - total_size = le16_to_cpu(pSMB->t2_rsp.ParameterCount); - if (total_size < 512) { - total_size += - le16_to_cpu(pSMB->t2_rsp.DataCount); - /* BCC le converted in SendReceive */ - pBCC = (pSMB->hdr.WordCount * 2) + - sizeof(struct smb_hdr) + - (char *)pSMB; - if ((total_size <= (*(u16 *)pBCC)) && - (total_size < - CIFSMaxBufSize+MAX_CIFS_HDR_SIZE)) { - return 0; - } - } - } - } + if (get_unaligned_le16(&pSMB->t2_rsp.ParameterOffset) > 1024 || + get_unaligned_le16(&pSMB->t2_rsp.DataOffset) > 1024) + goto vt2_err; + + /* check that bcc is at least as big as parms + data */ + /* check that bcc is less than negotiated smb buffer */ + total_size = get_unaligned_le16(&pSMB->t2_rsp.ParameterCount); + if (total_size >= 512) + goto vt2_err; + + total_size += get_unaligned_le16(&pSMB->t2_rsp.DataCount); + if (total_size > get_bcc(&pSMB->hdr) || + total_size >= CIFSMaxBufSize + MAX_CIFS_HDR_SIZE) + goto vt2_err; + + return 0; +vt2_err: cifs_dump_mem("Invalid transact2 SMB: ", (char *)pSMB, sizeof(struct smb_t2_rsp) + 16); - return rc; + return -EINVAL; } + int CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) { @@ -452,7 +447,6 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) server->maxBuf = min((__u32)le16_to_cpu(rsp->MaxBufSize), (__u32)CIFSMaxBufSize + MAX_CIFS_HDR_SIZE); server->max_vcs = le16_to_cpu(rsp->MaxNumberVcs); - GETU32(server->sessid) = le32_to_cpu(rsp->SessionKey); /* even though we do not use raw we might as well set this accurately, in case we ever find a need for it */ if ((le16_to_cpu(rsp->RawMode) & RAW_ENABLE) == RAW_ENABLE) { @@ -566,7 +560,6 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) (__u32) CIFSMaxBufSize + MAX_CIFS_HDR_SIZE); server->max_rw = le32_to_cpu(pSMBr->MaxRawSize); cFYI(DBG2, "Max buf = %d", ses->server->maxBuf); - GETU32(ses->server->sessid) = le32_to_cpu(pSMBr->SessionKey); server->capabilities = le32_to_cpu(pSMBr->Capabilities); server->timeAdj = (int)(__s16)le16_to_cpu(pSMBr->ServerTimeZone); server->timeAdj *= 60; @@ -706,6 +699,53 @@ CIFSSMBTDis(const int xid, struct cifsTconInfo *tcon) return rc; } +/* + * This is a no-op for now. We're not really interested in the reply, but + * rather in the fact that the server sent one and that server->lstrp + * gets updated. + * + * FIXME: maybe we should consider checking that the reply matches request? + */ +static void +cifs_echo_callback(struct mid_q_entry *mid) +{ + struct TCP_Server_Info *server = mid->callback_data; + + DeleteMidQEntry(mid); + atomic_dec(&server->inFlight); + wake_up(&server->request_q); +} + +int +CIFSSMBEcho(struct TCP_Server_Info *server) +{ + ECHO_REQ *smb; + int rc = 0; + + cFYI(1, "In echo request"); + + rc = small_smb_init(SMB_COM_ECHO, 0, NULL, (void **)&smb); + if (rc) + return rc; + + /* set up echo request */ + smb->hdr.Tid = cpu_to_le16(0xffff); + smb->hdr.WordCount = 1; + put_unaligned_le16(1, &smb->EchoCount); + put_bcc_le(1, &smb->hdr); + smb->Data[0] = 'a'; + smb->hdr.smb_buf_length += 3; + + rc = cifs_call_async(server, (struct smb_hdr *)smb, + cifs_echo_callback, server); + if (rc) + cFYI(1, "Echo request failed: %d", rc); + + cifs_small_buf_release(smb); + + return rc; +} + int CIFSSMBLogoff(const int xid, struct cifsSesInfo *ses) { @@ -1193,7 +1233,7 @@ OldOpenRetry: pSMB->ByteCount = cpu_to_le16(count); /* long_op set to 1 to allow for oplock break timeouts */ rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, - (struct smb_hdr *)pSMBr, &bytes_returned, CIFS_LONG_OP); + (struct smb_hdr *)pSMBr, &bytes_returned, 0); cifs_stats_inc(&tcon->num_opens); if (rc) { cFYI(1, "Error in Open = %d", rc); @@ -1306,7 +1346,7 @@ openRetry: pSMB->ByteCount = cpu_to_le16(count); /* long_op set to 1 to allow for oplock break timeouts */ rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, - (struct smb_hdr *)pSMBr, &bytes_returned, CIFS_LONG_OP); + (struct smb_hdr *)pSMBr, &bytes_returned, 0); cifs_stats_inc(&tcon->num_opens); if (rc) { cFYI(1, "Error in Open = %d", rc); @@ -1388,7 +1428,7 @@ CIFSSMBRead(const int xid, struct cifsTconInfo *tcon, const int netfid, iov[0].iov_base = (char *)pSMB; iov[0].iov_len = pSMB->hdr.smb_buf_length + 4; rc = SendReceive2(xid, tcon->ses, iov, 1 /* num iovecs */, - &resp_buf_type, CIFS_STD_OP | CIFS_LOG_ERROR); + &resp_buf_type, CIFS_LOG_ERROR); cifs_stats_inc(&tcon->num_reads); pSMBr = (READ_RSP *)iov[0].iov_base; if (rc) { @@ -1663,7 +1703,8 @@ int CIFSSMBLock(const int xid, struct cifsTconInfo *tcon, const __u16 smb_file_id, const __u64 len, const __u64 offset, const __u32 numUnlock, - const __u32 numLock, const __u8 lockType, const bool waitFlag) + const __u32 numLock, const __u8 lockType, + const bool waitFlag, const __u8 oplock_level) { int rc = 0; LOCK_REQ *pSMB = NULL; @@ -1691,6 +1732,7 @@ CIFSSMBLock(const int xid, struct cifsTconInfo *tcon, pSMB->NumberOfLocks = cpu_to_le16(numLock); pSMB->NumberOfUnlocks = cpu_to_le16(numUnlock); pSMB->LockType = lockType; + pSMB->OplockLevel = oplock_level; pSMB->AndXCommand = 0xFF; /* none */ pSMB->Fid = smb_file_id; /* netfid stays le */ @@ -3087,7 +3129,7 @@ CIFSSMBGetCIFSACL(const int xid, struct cifsTconInfo *tcon, __u16 fid, iov[0].iov_len = pSMB->hdr.smb_buf_length + 4; rc = SendReceive2(xid, tcon->ses, iov, 1 /* num iovec */, &buf_type, - CIFS_STD_OP); + 0); cifs_stats_inc(&tcon->num_acl_get); if (rc) { cFYI(1, "Send error in QuerySecDesc = %d", rc); @@ -4869,7 +4911,6 @@ CIFSSMBSetFileSize(const int xid, struct cifsTconInfo *tcon, __u64 size, __u16 fid, __u32 pid_of_opener, bool SetAllocation) { struct smb_com_transaction2_sfi_req *pSMB = NULL; - char *data_offset; struct file_end_of_file_info *parm_data; int rc = 0; __u16 params, param_offset, offset, byte_count, count; @@ -4893,8 +4934,6 @@ CIFSSMBSetFileSize(const int xid, struct cifsTconInfo *tcon, __u64 size, param_offset = offsetof(struct smb_com_transaction2_sfi_req, Fid) - 4; offset = param_offset + params; - data_offset = (char *) (&pSMB->hdr.Protocol) + offset; - count = sizeof(struct file_end_of_file_info); pSMB->MaxParameterCount = cpu_to_le16(2); /* BB find exact max SMB PDU from sess structure BB */ @@ -5562,7 +5601,7 @@ QAllEAsRetry: } /* make sure list_len doesn't go past end of SMB */ - end_of_smb = (char *)pByteArea(&pSMBr->hdr) + BCC(&pSMBr->hdr); + end_of_smb = (char *)pByteArea(&pSMBr->hdr) + get_bcc(&pSMBr->hdr); if ((char *)ea_response_data + list_len > end_of_smb) { cFYI(1, "EA list appears to go beyond SMB"); rc = -EIO; diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 9f59887..8d6c17a 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -52,8 +52,8 @@ #define CIFS_PORT 445 #define RFC1001_PORT 139 -extern void SMBNTencrypt(unsigned char *passwd, unsigned char *c8, - unsigned char *p24); +/* SMB echo "timeout" -- FIXME: tunable? */ +#define SMB_ECHO_INTERVAL (60 * HZ) extern mempool_t *cifs_req_poolp; @@ -84,6 +84,7 @@ struct smb_vol { bool no_xattr:1; /* set if xattr (EA) support should be disabled*/ bool server_ino:1; /* use inode numbers from server ie UniqueId */ bool direct_io:1; + bool strict_io:1; /* strict cache behavior */ bool remap:1; /* set to remap seven reserved chars in filenames */ bool posix_paths:1; /* unset to not ask for posix pathnames. */ bool no_linux_ext:1; @@ -152,6 +153,7 @@ cifs_reconnect(struct TCP_Server_Info *server) /* before reconnecting the tcp session, mark the smb session (uid) and the tid bad so they are not used until reconnected */ + cFYI(1, "%s: marking sessions and tcons for reconnect", __func__); spin_lock(&cifs_tcp_ses_lock); list_for_each(tmp, &server->smb_ses_list) { ses = list_entry(tmp, struct cifsSesInfo, smb_ses_list); @@ -163,7 +165,9 @@ cifs_reconnect(struct TCP_Server_Info *server) } } spin_unlock(&cifs_tcp_ses_lock); + /* do not want to be sending data on a socket we are freeing */ + cFYI(1, "%s: tearing down socket", __func__); mutex_lock(&server->srv_mutex); if (server->ssocket) { cFYI(1, "State: 0x%x Flags: 0x%lx", server->ssocket->state, @@ -180,22 +184,20 @@ cifs_reconnect(struct TCP_Server_Info *server) kfree(server->session_key.response); server->session_key.response = NULL; server->session_key.len = 0; + server->lstrp = jiffies; + mutex_unlock(&server->srv_mutex); + /* mark submitted MIDs for retry and issue callback */ + cFYI(1, "%s: issuing mid callbacks", __func__); spin_lock(&GlobalMid_Lock); - list_for_each(tmp, &server->pending_mid_q) { - mid_entry = list_entry(tmp, struct - mid_q_entry, - qhead); - if (mid_entry->midState == MID_REQUEST_SUBMITTED) { - /* Mark other intransit requests as needing - retry so we do not immediately mark the - session bad again (ie after we reconnect - below) as they timeout too */ + list_for_each_safe(tmp, tmp2, &server->pending_mid_q) { + mid_entry = list_entry(tmp, struct mid_q_entry, qhead); + if (mid_entry->midState == MID_REQUEST_SUBMITTED) mid_entry->midState = MID_RETRY_NEEDED; - } + list_del_init(&mid_entry->qhead); + mid_entry->callback(mid_entry); } spin_unlock(&GlobalMid_Lock); - mutex_unlock(&server->srv_mutex); while ((server->tcpStatus != CifsExiting) && (server->tcpStatus != CifsGood)) { @@ -212,10 +214,9 @@ cifs_reconnect(struct TCP_Server_Info *server) if (server->tcpStatus != CifsExiting) server->tcpStatus = CifsGood; spin_unlock(&GlobalMid_Lock); - /* atomic_set(&server->inFlight,0);*/ - wake_up(&server->response_q); } } + return rc; } @@ -229,9 +230,8 @@ cifs_reconnect(struct TCP_Server_Info *server) static int check2ndT2(struct smb_hdr *pSMB, unsigned int maxBufSize) { struct smb_t2_rsp *pSMBt; - int total_data_size; - int data_in_this_rsp; int remaining; + __u16 total_data_size, data_in_this_rsp; if (pSMB->Command != SMB_COM_TRANSACTION2) return 0; @@ -245,8 +245,8 @@ static int check2ndT2(struct smb_hdr *pSMB, unsigned int maxBufSize) pSMBt = (struct smb_t2_rsp *)pSMB; - total_data_size = le16_to_cpu(pSMBt->t2_rsp.TotalDataCount); - data_in_this_rsp = le16_to_cpu(pSMBt->t2_rsp.DataCount); + total_data_size = get_unaligned_le16(&pSMBt->t2_rsp.TotalDataCount); + data_in_this_rsp = get_unaligned_le16(&pSMBt->t2_rsp.DataCount); remaining = total_data_size - data_in_this_rsp; @@ -272,21 +272,18 @@ static int coalesce_t2(struct smb_hdr *psecond, struct smb_hdr *pTargetSMB) { struct smb_t2_rsp *pSMB2 = (struct smb_t2_rsp *)psecond; struct smb_t2_rsp *pSMBt = (struct smb_t2_rsp *)pTargetSMB; - int total_data_size; - int total_in_buf; - int remaining; - int total_in_buf2; char *data_area_of_target; char *data_area_of_buf2; - __u16 byte_count; + int remaining; + __u16 byte_count, total_data_size, total_in_buf, total_in_buf2; - total_data_size = le16_to_cpu(pSMBt->t2_rsp.TotalDataCount); + total_data_size = get_unaligned_le16(&pSMBt->t2_rsp.TotalDataCount); - if (total_data_size != le16_to_cpu(pSMB2->t2_rsp.TotalDataCount)) { + if (total_data_size != + get_unaligned_le16(&pSMB2->t2_rsp.TotalDataCount)) cFYI(1, "total data size of primary and secondary t2 differ"); - } - total_in_buf = le16_to_cpu(pSMBt->t2_rsp.DataCount); + total_in_buf = get_unaligned_le16(&pSMBt->t2_rsp.DataCount); remaining = total_data_size - total_in_buf; @@ -296,28 +293,28 @@ static int coalesce_t2(struct smb_hdr *psecond, struct smb_hdr *pTargetSMB) if (remaining == 0) /* nothing to do, ignore */ return 0; - total_in_buf2 = le16_to_cpu(pSMB2->t2_rsp.DataCount); + total_in_buf2 = get_unaligned_le16(&pSMB2->t2_rsp.DataCount); if (remaining < total_in_buf2) { cFYI(1, "transact2 2nd response contains too much data"); } /* find end of first SMB data area */ data_area_of_target = (char *)&pSMBt->hdr.Protocol + - le16_to_cpu(pSMBt->t2_rsp.DataOffset); + get_unaligned_le16(&pSMBt->t2_rsp.DataOffset); /* validate target area */ - data_area_of_buf2 = (char *) &pSMB2->hdr.Protocol + - le16_to_cpu(pSMB2->t2_rsp.DataOffset); + data_area_of_buf2 = (char *)&pSMB2->hdr.Protocol + + get_unaligned_le16(&pSMB2->t2_rsp.DataOffset); data_area_of_target += total_in_buf; /* copy second buffer into end of first buffer */ memcpy(data_area_of_target, data_area_of_buf2, total_in_buf2); total_in_buf += total_in_buf2; - pSMBt->t2_rsp.DataCount = cpu_to_le16(total_in_buf); - byte_count = le16_to_cpu(BCC_LE(pTargetSMB)); + put_unaligned_le16(total_in_buf, &pSMBt->t2_rsp.DataCount); + byte_count = get_bcc_le(pTargetSMB); byte_count += total_in_buf2; - BCC_LE(pTargetSMB) = cpu_to_le16(byte_count); + put_bcc_le(byte_count, pTargetSMB); byte_count = pTargetSMB->smb_buf_length; byte_count += total_in_buf2; @@ -331,7 +328,31 @@ static int coalesce_t2(struct smb_hdr *psecond, struct smb_hdr *pTargetSMB) return 0; /* we are done */ } else /* more responses to go */ return 1; +} + +static void +cifs_echo_request(struct work_struct *work) +{ + int rc; + struct TCP_Server_Info *server = container_of(work, + struct TCP_Server_Info, echo.work); + + /* + * We cannot send an echo until the NEGOTIATE_PROTOCOL request is + * done, which is indicated by maxBuf != 0. Also, no need to ping if + * we got a response recently + */ + if (server->maxBuf == 0 || + time_before(jiffies, server->lstrp + SMB_ECHO_INTERVAL - HZ)) + goto requeue_echo; + + rc = CIFSSMBEcho(server); + if (rc) + cFYI(1, "Unable to send echo request to server: %s", + server->hostname); +requeue_echo: + queue_delayed_work(system_nrt_wq, &server->echo, SMB_ECHO_INTERVAL); } static int @@ -345,8 +366,7 @@ cifs_demultiplex_thread(struct TCP_Server_Info *server) struct msghdr smb_msg; struct kvec iov; struct socket *csocket = server->ssocket; - struct list_head *tmp; - struct cifsSesInfo *ses; + struct list_head *tmp, *tmp2; struct task_struct *task_to_wake = NULL; struct mid_q_entry *mid_entry; char temp; @@ -399,7 +419,20 @@ cifs_demultiplex_thread(struct TCP_Server_Info *server) smb_msg.msg_control = NULL; smb_msg.msg_controllen = 0; pdu_length = 4; /* enough to get RFC1001 header */ + incomplete_rcv: + if (echo_retries > 0 && + time_after(jiffies, server->lstrp + + (echo_retries * SMB_ECHO_INTERVAL))) { + cERROR(1, "Server %s has not responded in %d seconds. " + "Reconnecting...", server->hostname, + (echo_retries * SMB_ECHO_INTERVAL / HZ)); + cifs_reconnect(server); + csocket = server->ssocket; + wake_up(&server->response_q); + continue; + } + length = kernel_recvmsg(csocket, &smb_msg, &iov, 1, pdu_length, 0 /* BB other flags? */); @@ -550,25 +583,36 @@ incomplete_rcv: else if (reconnect == 1) continue; - length += 4; /* account for rfc1002 hdr */ + total_read += 4; /* account for rfc1002 hdr */ + dump_smb(smb_buffer, total_read); - dump_smb(smb_buffer, length); - if (checkSMB(smb_buffer, smb_buffer->Mid, total_read+4)) { - cifs_dump_mem("Bad SMB: ", smb_buffer, 48); - continue; - } + /* + * We know that we received enough to get to the MID as we + * checked the pdu_length earlier. Now check to see + * if the rest of the header is OK. We borrow the length + * var for the rest of the loop to avoid a new stack var. + * + * 48 bytes is enough to display the header and a little bit + * into the payload for debugging purposes. + */ + length = checkSMB(smb_buffer, smb_buffer->Mid, total_read); + if (length != 0) + cifs_dump_mem("Bad SMB: ", smb_buffer, + min_t(unsigned int, total_read, 48)); + mid_entry = NULL; + server->lstrp = jiffies; - task_to_wake = NULL; spin_lock(&GlobalMid_Lock); - list_for_each(tmp, &server->pending_mid_q) { + list_for_each_safe(tmp, tmp2, &server->pending_mid_q) { mid_entry = list_entry(tmp, struct mid_q_entry, qhead); if ((mid_entry->mid == smb_buffer->Mid) && (mid_entry->midState == MID_REQUEST_SUBMITTED) && (mid_entry->command == smb_buffer->Command)) { - if (check2ndT2(smb_buffer,server->maxBuf) > 0) { + if (length == 0 && + check2ndT2(smb_buffer, server->maxBuf) > 0) { /* We have a multipart transact2 resp */ isMultiRsp = true; if (mid_entry->resp_buf) { @@ -603,20 +647,24 @@ incomplete_rcv: mid_entry->resp_buf = smb_buffer; mid_entry->largeBuf = isLargeBuf; multi_t2_fnd: - task_to_wake = mid_entry->tsk; - mid_entry->midState = MID_RESPONSE_RECEIVED; + if (length == 0) + mid_entry->midState = + MID_RESPONSE_RECEIVED; + else + mid_entry->midState = + MID_RESPONSE_MALFORMED; #ifdef CONFIG_CIFS_STATS2 mid_entry->when_received = jiffies; #endif - /* so we do not time out requests to server - which is still responding (since server could - be busy but not dead) */ - server->lstrp = jiffies; + list_del_init(&mid_entry->qhead); + mid_entry->callback(mid_entry); break; } + mid_entry = NULL; } spin_unlock(&GlobalMid_Lock); - if (task_to_wake) { + + if (mid_entry != NULL) { /* Was previous buf put in mpx struct for multi-rsp? */ if (!isMultiRsp) { /* smb buffer will be freed by user thread */ @@ -625,11 +673,13 @@ multi_t2_fnd: else smallbuf = NULL; } - wake_up_process(task_to_wake); + } else if (length != 0) { + /* response sanity checks failed */ + continue; } else if (!is_valid_oplock_break(smb_buffer, server) && !isMultiRsp) { cERROR(1, "No task to wake, unknown frame received! " - "NumMids %d", midCount.counter); + "NumMids %d", atomic_read(&midCount)); cifs_dump_mem("Received Data is: ", (char *)smb_buffer, sizeof(struct smb_hdr)); #ifdef CONFIG_CIFS_DEBUG2 @@ -677,44 +727,16 @@ multi_t2_fnd: if (smallbuf) /* no sense logging a debug message if NULL */ cifs_small_buf_release(smallbuf); - /* - * BB: we shouldn't have to do any of this. It shouldn't be - * possible to exit from the thread with active SMB sessions - */ - spin_lock(&cifs_tcp_ses_lock); - if (list_empty(&server->pending_mid_q)) { - /* loop through server session structures attached to this and - mark them dead */ - list_for_each(tmp, &server->smb_ses_list) { - ses = list_entry(tmp, struct cifsSesInfo, - smb_ses_list); - ses->status = CifsExiting; - ses->server = NULL; - } - spin_unlock(&cifs_tcp_ses_lock); - } else { - /* although we can not zero the server struct pointer yet, - since there are active requests which may depnd on them, - mark the corresponding SMB sessions as exiting too */ - list_for_each(tmp, &server->smb_ses_list) { - ses = list_entry(tmp, struct cifsSesInfo, - smb_ses_list); - ses->status = CifsExiting; - } - + if (!list_empty(&server->pending_mid_q)) { spin_lock(&GlobalMid_Lock); - list_for_each(tmp, &server->pending_mid_q) { - mid_entry = list_entry(tmp, struct mid_q_entry, qhead); - if (mid_entry->midState == MID_REQUEST_SUBMITTED) { - cFYI(1, "Clearing Mid 0x%x - waking up ", + list_for_each_safe(tmp, tmp2, &server->pending_mid_q) { + mid_entry = list_entry(tmp, struct mid_q_entry, qhead); + cFYI(1, "Clearing Mid 0x%x - issuing callback", mid_entry->mid); - task_to_wake = mid_entry->tsk; - if (task_to_wake) - wake_up_process(task_to_wake); - } + list_del_init(&mid_entry->qhead); + mid_entry->callback(mid_entry); } spin_unlock(&GlobalMid_Lock); - spin_unlock(&cifs_tcp_ses_lock); /* 1/8th of sec is more than enough time for them to exit */ msleep(125); } @@ -732,18 +754,6 @@ multi_t2_fnd: coming home not much else we can do but free the memory */ } - /* last chance to mark ses pointers invalid - if there are any pointing to this (e.g - if a crazy root user tried to kill cifsd - kernel thread explicitly this might happen) */ - /* BB: This shouldn't be necessary, see above */ - spin_lock(&cifs_tcp_ses_lock); - list_for_each(tmp, &server->smb_ses_list) { - ses = list_entry(tmp, struct cifsSesInfo, smb_ses_list); - ses->server = NULL; - } - spin_unlock(&cifs_tcp_ses_lock); - kfree(server->hostname); task_to_wake = xchg(&server->tsk, NULL); kfree(server); @@ -1355,6 +1365,8 @@ cifs_parse_mount_options(char *options, const char *devname, vol->direct_io = 1; } else if (strnicmp(data, "forcedirectio", 13) == 0) { vol->direct_io = 1; + } else if (strnicmp(data, "strictcache", 11) == 0) { + vol->strict_io = 1; } else if (strnicmp(data, "noac", 4) == 0) { printk(KERN_WARNING "CIFS: Mount option noac not " "supported. Instead set " @@ -1579,6 +1591,9 @@ cifs_find_tcp_session(struct sockaddr *addr, struct smb_vol *vol) spin_lock(&cifs_tcp_ses_lock); list_for_each_entry(server, &cifs_tcp_ses_list, tcp_ses_list) { + if (!net_eq(cifs_net_ns(server), current->nsproxy->net_ns)) + continue; + if (!match_address(server, addr, (struct sockaddr *)&vol->srcaddr)) continue; @@ -1609,9 +1624,13 @@ cifs_put_tcp_session(struct TCP_Server_Info *server) return; } + put_net(cifs_net_ns(server)); + list_del_init(&server->tcp_ses_list); spin_unlock(&cifs_tcp_ses_lock); + cancel_delayed_work_sync(&server->echo); + spin_lock(&GlobalMid_Lock); server->tcpStatus = CifsExiting; spin_unlock(&GlobalMid_Lock); @@ -1681,6 +1700,7 @@ cifs_get_tcp_session(struct smb_vol *volume_info) goto out_err; } + cifs_set_net_ns(tcp_ses, get_net(current->nsproxy->net_ns)); tcp_ses->hostname = extract_hostname(volume_info->UNC); if (IS_ERR(tcp_ses->hostname)) { rc = PTR_ERR(tcp_ses->hostname); @@ -1701,8 +1721,10 @@ cifs_get_tcp_session(struct smb_vol *volume_info) volume_info->target_rfc1001_name, RFC1001_NAME_LEN_WITH_NULL); tcp_ses->session_estab = false; tcp_ses->sequence_number = 0; + tcp_ses->lstrp = jiffies; INIT_LIST_HEAD(&tcp_ses->tcp_ses_list); INIT_LIST_HEAD(&tcp_ses->smb_ses_list); + INIT_DELAYED_WORK(&tcp_ses->echo, cifs_echo_request); /* * at this point we are the only ones with the pointer @@ -1751,11 +1773,16 @@ cifs_get_tcp_session(struct smb_vol *volume_info) cifs_fscache_get_client_cookie(tcp_ses); + /* queue echo request delayed work */ + queue_delayed_work(system_nrt_wq, &tcp_ses->echo, SMB_ECHO_INTERVAL); + return tcp_ses; out_err_crypto_release: cifs_crypto_shash_release(tcp_ses); + put_net(cifs_net_ns(tcp_ses)); + out_err: if (tcp_ses) { if (!IS_ERR(tcp_ses->hostname)) @@ -2267,8 +2294,8 @@ generic_ip_connect(struct TCP_Server_Info *server) } if (socket == NULL) { - rc = sock_create_kern(sfamily, SOCK_STREAM, - IPPROTO_TCP, &socket); + rc = __sock_create(cifs_net_ns(server), sfamily, SOCK_STREAM, + IPPROTO_TCP, &socket, 1); if (rc < 0) { cERROR(1, "Error %d creating socket", rc); server->ssocket = NULL; @@ -2580,6 +2607,8 @@ static void setup_cifs_sb(struct smb_vol *pvolume_info, if (pvolume_info->multiuser) cifs_sb->mnt_cifs_flags |= (CIFS_MOUNT_MULTIUSER | CIFS_MOUNT_NO_PERM); + if (pvolume_info->strict_io) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_STRICT_IO; if (pvolume_info->direct_io) { cFYI(1, "mounting share using direct i/o"); cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_DIRECT_IO; @@ -2936,8 +2965,8 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, TCONX_RSP *pSMBr; unsigned char *bcc_ptr; int rc = 0; - int length, bytes_left; - __u16 count; + int length; + __u16 bytes_left, count; if (ses == NULL) return -EIO; @@ -2965,7 +2994,7 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, bcc_ptr++; /* skip password */ /* already aligned so no need to do it below */ } else { - pSMB->PasswordLength = cpu_to_le16(CIFS_SESS_KEY_SIZE); + pSMB->PasswordLength = cpu_to_le16(CIFS_AUTH_RESP_SIZE); /* BB FIXME add code to fail this if NTLMv2 or Kerberos specified as required (when that support is added to the vfs in the future) as only NTLM or the much @@ -2981,9 +3010,10 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, bcc_ptr); else #endif /* CIFS_WEAK_PW_HASH */ - SMBNTencrypt(tcon->password, ses->server->cryptkey, bcc_ptr); + rc = SMBNTencrypt(tcon->password, ses->server->cryptkey, + bcc_ptr); - bcc_ptr += CIFS_SESS_KEY_SIZE; + bcc_ptr += CIFS_AUTH_RESP_SIZE; if (ses->capabilities & CAP_UNICODE) { /* must align unicode strings */ *bcc_ptr = 0; /* null byte password */ @@ -3021,7 +3051,7 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, pSMB->ByteCount = cpu_to_le16(count); rc = SendReceive(xid, ses, smb_buffer, smb_buffer_response, &length, - CIFS_STD_OP); + 0); /* above now done in SendReceive */ if ((rc == 0) && (tcon != NULL)) { @@ -3031,7 +3061,7 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, tcon->need_reconnect = false; tcon->tid = smb_buffer_response->Tid; bcc_ptr = pByteArea(smb_buffer_response); - bytes_left = BCC(smb_buffer_response); + bytes_left = get_bcc(smb_buffer_response); length = strnlen(bcc_ptr, bytes_left - 2); if (smb_buffer->Flags2 & SMBFLG2_UNICODE) is_unicode = true; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index d843631..e964b1c 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -287,6 +287,7 @@ void cifsFileInfo_put(struct cifsFileInfo *cifs_file) struct inode *inode = cifs_file->dentry->d_inode; struct cifsTconInfo *tcon = tlink_tcon(cifs_file->tlink); struct cifsInodeInfo *cifsi = CIFS_I(inode); + struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsLockInfo *li, *tmp; spin_lock(&cifs_file_list_lock); @@ -302,6 +303,13 @@ void cifsFileInfo_put(struct cifsFileInfo *cifs_file) if (list_empty(&cifsi->openFileList)) { cFYI(1, "closing last open instance for inode %p", cifs_file->dentry->d_inode); + + /* in strict cache mode we need invalidate mapping on the last + close because it may cause a error when we open this file + again and get at least level II oplock */ + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_STRICT_IO) + CIFS_I(inode)->invalid_mapping = true; + cifs_set_oplock_level(cifsi, 0); } spin_unlock(&cifs_file_list_lock); @@ -338,7 +346,6 @@ int cifs_open(struct inode *inode, struct file *file) struct cifsTconInfo *tcon; struct tcon_link *tlink; struct cifsFileInfo *pCifsFile = NULL; - struct cifsInodeInfo *pCifsInode; char *full_path = NULL; bool posix_open_ok = false; __u16 netfid; @@ -353,8 +360,6 @@ int cifs_open(struct inode *inode, struct file *file) } tcon = tlink_tcon(tlink); - pCifsInode = CIFS_I(file->f_path.dentry->d_inode); - full_path = build_path_from_dentry(file->f_path.dentry); if (full_path == NULL) { rc = -ENOMEM; @@ -726,12 +731,12 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) /* BB we could chain these into one lock request BB */ rc = CIFSSMBLock(xid, tcon, netfid, length, pfLock->fl_start, - 0, 1, lockType, 0 /* wait flag */ ); + 0, 1, lockType, 0 /* wait flag */, 0); if (rc == 0) { rc = CIFSSMBLock(xid, tcon, netfid, length, pfLock->fl_start, 1 /* numUnlock */ , 0 /* numLock */ , lockType, - 0 /* wait flag */ ); + 0 /* wait flag */, 0); pfLock->fl_type = F_UNLCK; if (rc != 0) cERROR(1, "Error unlocking previously locked " @@ -748,13 +753,13 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) rc = CIFSSMBLock(xid, tcon, netfid, length, pfLock->fl_start, 0, 1, lockType | LOCKING_ANDX_SHARED_LOCK, - 0 /* wait flag */); + 0 /* wait flag */, 0); if (rc == 0) { rc = CIFSSMBLock(xid, tcon, netfid, length, pfLock->fl_start, 1, 0, lockType | LOCKING_ANDX_SHARED_LOCK, - 0 /* wait flag */); + 0 /* wait flag */, 0); pfLock->fl_type = F_RDLCK; if (rc != 0) cERROR(1, "Error unlocking " @@ -797,8 +802,8 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) if (numLock) { rc = CIFSSMBLock(xid, tcon, netfid, length, - pfLock->fl_start, - 0, numLock, lockType, wait_flag); + pfLock->fl_start, 0, numLock, lockType, + wait_flag, 0); if (rc == 0) { /* For Windows locks we must store them. */ @@ -818,9 +823,9 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) (pfLock->fl_start + length) >= (li->offset + li->length)) { stored_rc = CIFSSMBLock(xid, tcon, - netfid, - li->length, li->offset, - 1, 0, li->type, false); + netfid, li->length, + li->offset, 1, 0, + li->type, false, 0); if (stored_rc) rc = stored_rc; else { @@ -839,31 +844,8 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) return rc; } -/* - * Set the timeout on write requests past EOF. For some servers (Windows) - * these calls can be very long. - * - * If we're writing >10M past the EOF we give a 180s timeout. Anything less - * than that gets a 45s timeout. Writes not past EOF get 15s timeouts. - * The 10M cutoff is totally arbitrary. A better scheme for this would be - * welcome if someone wants to suggest one. - * - * We may be able to do a better job with this if there were some way to - * declare that a file should be sparse. - */ -static int -cifs_write_timeout(struct cifsInodeInfo *cifsi, loff_t offset) -{ - if (offset <= cifsi->server_eof) - return CIFS_STD_OP; - else if (offset > (cifsi->server_eof + (10 * 1024 * 1024))) - return CIFS_VLONG_OP; - else - return CIFS_LONG_OP; -} - /* update the file size (if needed) after a write */ -static void +void cifs_update_eof(struct cifsInodeInfo *cifsi, loff_t offset, unsigned int bytes_written) { @@ -882,7 +864,7 @@ ssize_t cifs_user_write(struct file *file, const char __user *write_data, unsigned int total_written; struct cifs_sb_info *cifs_sb; struct cifsTconInfo *pTcon; - int xid, long_op; + int xid; struct cifsFileInfo *open_file; struct cifsInodeInfo *cifsi = CIFS_I(inode); @@ -903,7 +885,6 @@ ssize_t cifs_user_write(struct file *file, const char __user *write_data, xid = GetXid(); - long_op = cifs_write_timeout(cifsi, *poffset); for (total_written = 0; write_size > total_written; total_written += bytes_written) { rc = -EAGAIN; @@ -931,7 +912,7 @@ ssize_t cifs_user_write(struct file *file, const char __user *write_data, min_t(const int, cifs_sb->wsize, write_size - total_written), *poffset, &bytes_written, - NULL, write_data + total_written, long_op); + NULL, write_data + total_written, 0); } if (rc || (bytes_written == 0)) { if (total_written) @@ -944,8 +925,6 @@ ssize_t cifs_user_write(struct file *file, const char __user *write_data, cifs_update_eof(cifsi, *poffset, bytes_written); *poffset += bytes_written; } - long_op = CIFS_STD_OP; /* subsequent writes fast - - 15 seconds is plenty */ } cifs_stats_bytes_written(pTcon, total_written); @@ -974,7 +953,7 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file, unsigned int total_written; struct cifs_sb_info *cifs_sb; struct cifsTconInfo *pTcon; - int xid, long_op; + int xid; struct dentry *dentry = open_file->dentry; struct cifsInodeInfo *cifsi = CIFS_I(dentry->d_inode); @@ -987,7 +966,6 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file, xid = GetXid(); - long_op = cifs_write_timeout(cifsi, *poffset); for (total_written = 0; write_size > total_written; total_written += bytes_written) { rc = -EAGAIN; @@ -1017,7 +995,7 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file, rc = CIFSSMBWrite2(xid, pTcon, open_file->netfid, len, *poffset, &bytes_written, - iov, 1, long_op); + iov, 1, 0); } else rc = CIFSSMBWrite(xid, pTcon, open_file->netfid, @@ -1025,7 +1003,7 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file, write_size - total_written), *poffset, &bytes_written, write_data + total_written, - NULL, long_op); + NULL, 0); } if (rc || (bytes_written == 0)) { if (total_written) @@ -1038,8 +1016,6 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file, cifs_update_eof(cifsi, *poffset, bytes_written); *poffset += bytes_written; } - long_op = CIFS_STD_OP; /* subsequent writes fast - - 15 seconds is plenty */ } cifs_stats_bytes_written(pTcon, total_written); @@ -1167,7 +1143,6 @@ static int cifs_partialpagewrite(struct page *page, unsigned from, unsigned to) char *write_data; int rc = -EFAULT; int bytes_written = 0; - struct cifs_sb_info *cifs_sb; struct inode *inode; struct cifsFileInfo *open_file; @@ -1175,7 +1150,6 @@ static int cifs_partialpagewrite(struct page *page, unsigned from, unsigned to) return -EFAULT; inode = page->mapping->host; - cifs_sb = CIFS_SB(inode->i_sb); offset += (loff_t)from; write_data = kmap(page); @@ -1239,7 +1213,7 @@ static int cifs_writepages(struct address_space *mapping, struct pagevec pvec; int rc = 0; int scanned = 0; - int xid, long_op; + int xid; cifs_sb = CIFS_SB(mapping->host->i_sb); @@ -1377,43 +1351,67 @@ retry: break; } if (n_iov) { +retry_write: open_file = find_writable_file(CIFS_I(mapping->host), false); if (!open_file) { cERROR(1, "No writable handles for inode"); rc = -EBADF; } else { - long_op = cifs_write_timeout(cifsi, offset); rc = CIFSSMBWrite2(xid, tcon, open_file->netfid, bytes_to_write, offset, &bytes_written, iov, n_iov, - long_op); + 0); cifsFileInfo_put(open_file); - cifs_update_eof(cifsi, offset, bytes_written); } - if (rc || bytes_written < bytes_to_write) { - cERROR(1, "Write2 ret %d, wrote %d", - rc, bytes_written); - mapping_set_error(mapping, rc); - } else { + cFYI(1, "Write2 rc=%d, wrote=%u", rc, bytes_written); + + /* + * For now, treat a short write as if nothing got + * written. A zero length write however indicates + * ENOSPC or EFBIG. We have no way to know which + * though, so call it ENOSPC for now. EFBIG would + * get translated to AS_EIO anyway. + * + * FIXME: make it take into account the data that did + * get written + */ + if (rc == 0) { + if (bytes_written == 0) + rc = -ENOSPC; + else if (bytes_written < bytes_to_write) + rc = -EAGAIN; + } + + /* retry on data-integrity flush */ + if (wbc->sync_mode == WB_SYNC_ALL && rc == -EAGAIN) + goto retry_write; + + /* fix the stats and EOF */ + if (bytes_written > 0) { cifs_stats_bytes_written(tcon, bytes_written); + cifs_update_eof(cifsi, offset, bytes_written); } for (i = 0; i < n_iov; i++) { page = pvec.pages[first + i]; - /* Should we also set page error on - success rc but too little data written? */ - /* BB investigate retry logic on temporary - server crash cases and how recovery works - when page marked as error */ - if (rc) + /* on retryable write error, redirty page */ + if (rc == -EAGAIN) + redirty_page_for_writepage(wbc, page); + else if (rc != 0) SetPageError(page); kunmap(page); unlock_page(page); end_page_writeback(page); page_cache_release(page); } + + if (rc != -EAGAIN) + mapping_set_error(mapping, rc); + else + rc = 0; + if ((wbc->nr_to_write -= n_iov) <= 0) done = 1; index = next; @@ -1525,27 +1523,47 @@ static int cifs_write_end(struct file *file, struct address_space *mapping, return rc; } -int cifs_fsync(struct file *file, int datasync) +int cifs_strict_fsync(struct file *file, int datasync) { int xid; int rc = 0; struct cifsTconInfo *tcon; struct cifsFileInfo *smbfile = file->private_data; struct inode *inode = file->f_path.dentry->d_inode; + struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); xid = GetXid(); cFYI(1, "Sync file - name: %s datasync: 0x%x", file->f_path.dentry->d_name.name, datasync); - rc = filemap_write_and_wait(inode->i_mapping); - if (rc == 0) { - struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); + if (!CIFS_I(inode)->clientCanCacheRead) + cifs_invalidate_mapping(inode); - tcon = tlink_tcon(smbfile->tlink); - if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOSSYNC)) - rc = CIFSSMBFlush(xid, tcon, smbfile->netfid); - } + tcon = tlink_tcon(smbfile->tlink); + if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOSSYNC)) + rc = CIFSSMBFlush(xid, tcon, smbfile->netfid); + + FreeXid(xid); + return rc; +} + +int cifs_fsync(struct file *file, int datasync) +{ + int xid; + int rc = 0; + struct cifsTconInfo *tcon; + struct cifsFileInfo *smbfile = file->private_data; + struct cifs_sb_info *cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); + + xid = GetXid(); + + cFYI(1, "Sync file - name: %s datasync: 0x%x", + file->f_path.dentry->d_name.name, datasync); + + tcon = tlink_tcon(smbfile->tlink); + if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOSSYNC)) + rc = CIFSSMBFlush(xid, tcon, smbfile->netfid); FreeXid(xid); return rc; @@ -1596,42 +1614,244 @@ int cifs_flush(struct file *file, fl_owner_t id) return rc; } -ssize_t cifs_user_read(struct file *file, char __user *read_data, - size_t read_size, loff_t *poffset) +static int +cifs_write_allocate_pages(struct page **pages, unsigned long num_pages) { - int rc = -EACCES; + int rc = 0; + unsigned long i; + + for (i = 0; i < num_pages; i++) { + pages[i] = alloc_page(__GFP_HIGHMEM); + if (!pages[i]) { + /* + * save number of pages we have already allocated and + * return with ENOMEM error + */ + num_pages = i; + rc = -ENOMEM; + goto error; + } + } + + return rc; + +error: + for (i = 0; i < num_pages; i++) + put_page(pages[i]); + return rc; +} + +static inline +size_t get_numpages(const size_t wsize, const size_t len, size_t *cur_len) +{ + size_t num_pages; + size_t clen; + + clen = min_t(const size_t, len, wsize); + num_pages = clen / PAGE_CACHE_SIZE; + if (clen % PAGE_CACHE_SIZE) + num_pages++; + + if (cur_len) + *cur_len = clen; + + return num_pages; +} + +static ssize_t +cifs_iovec_write(struct file *file, const struct iovec *iov, + unsigned long nr_segs, loff_t *poffset) +{ + unsigned int written; + unsigned long num_pages, npages, i; + size_t copied, len, cur_len; + ssize_t total_written = 0; + struct kvec *to_send; + struct page **pages; + struct iov_iter it; + struct inode *inode; + struct cifsFileInfo *open_file; + struct cifsTconInfo *pTcon; + struct cifs_sb_info *cifs_sb; + int xid, rc; + + len = iov_length(iov, nr_segs); + if (!len) + return 0; + + rc = generic_write_checks(file, poffset, &len, 0); + if (rc) + return rc; + + cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); + num_pages = get_numpages(cifs_sb->wsize, len, &cur_len); + + pages = kmalloc(sizeof(struct pages *)*num_pages, GFP_KERNEL); + if (!pages) + return -ENOMEM; + + to_send = kmalloc(sizeof(struct kvec)*(num_pages + 1), GFP_KERNEL); + if (!to_send) { + kfree(pages); + return -ENOMEM; + } + + rc = cifs_write_allocate_pages(pages, num_pages); + if (rc) { + kfree(pages); + kfree(to_send); + return rc; + } + + xid = GetXid(); + open_file = file->private_data; + pTcon = tlink_tcon(open_file->tlink); + inode = file->f_path.dentry->d_inode; + + iov_iter_init(&it, iov, nr_segs, len, 0); + npages = num_pages; + + do { + size_t save_len = cur_len; + for (i = 0; i < npages; i++) { + copied = min_t(const size_t, cur_len, PAGE_CACHE_SIZE); + copied = iov_iter_copy_from_user(pages[i], &it, 0, + copied); + cur_len -= copied; + iov_iter_advance(&it, copied); + to_send[i+1].iov_base = kmap(pages[i]); + to_send[i+1].iov_len = copied; + } + + cur_len = save_len - cur_len; + + do { + if (open_file->invalidHandle) { + rc = cifs_reopen_file(open_file, false); + if (rc != 0) + break; + } + rc = CIFSSMBWrite2(xid, pTcon, open_file->netfid, + cur_len, *poffset, &written, + to_send, npages, 0); + } while (rc == -EAGAIN); + + for (i = 0; i < npages; i++) + kunmap(pages[i]); + + if (written) { + len -= written; + total_written += written; + cifs_update_eof(CIFS_I(inode), *poffset, written); + *poffset += written; + } else if (rc < 0) { + if (!total_written) + total_written = rc; + break; + } + + /* get length and number of kvecs of the next write */ + npages = get_numpages(cifs_sb->wsize, len, &cur_len); + } while (len > 0); + + if (total_written > 0) { + spin_lock(&inode->i_lock); + if (*poffset > inode->i_size) + i_size_write(inode, *poffset); + spin_unlock(&inode->i_lock); + } + + cifs_stats_bytes_written(pTcon, total_written); + mark_inode_dirty_sync(inode); + + for (i = 0; i < num_pages; i++) + put_page(pages[i]); + kfree(to_send); + kfree(pages); + FreeXid(xid); + return total_written; +} + +static ssize_t cifs_user_writev(struct kiocb *iocb, const struct iovec *iov, + unsigned long nr_segs, loff_t pos) +{ + ssize_t written; + struct inode *inode; + + inode = iocb->ki_filp->f_path.dentry->d_inode; + + /* + * BB - optimize the way when signing is disabled. We can drop this + * extra memory-to-memory copying and use iovec buffers for constructing + * write request. + */ + + written = cifs_iovec_write(iocb->ki_filp, iov, nr_segs, &pos); + if (written > 0) { + CIFS_I(inode)->invalid_mapping = true; + iocb->ki_pos = pos; + } + + return written; +} + +ssize_t cifs_strict_writev(struct kiocb *iocb, const struct iovec *iov, + unsigned long nr_segs, loff_t pos) +{ + struct inode *inode; + + inode = iocb->ki_filp->f_path.dentry->d_inode; + + if (CIFS_I(inode)->clientCanCacheAll) + return generic_file_aio_write(iocb, iov, nr_segs, pos); + + /* + * In strict cache mode we need to write the data to the server exactly + * from the pos to pos+len-1 rather than flush all affected pages + * because it may cause a error with mandatory locks on these pages but + * not on the region from pos to ppos+len-1. + */ + + return cifs_user_writev(iocb, iov, nr_segs, pos); +} + +static ssize_t +cifs_iovec_read(struct file *file, const struct iovec *iov, + unsigned long nr_segs, loff_t *poffset) +{ + int rc; + int xid; + ssize_t total_read; unsigned int bytes_read = 0; - unsigned int total_read = 0; - unsigned int current_read_size; + size_t len, cur_len; + int iov_offset = 0; struct cifs_sb_info *cifs_sb; struct cifsTconInfo *pTcon; - int xid; struct cifsFileInfo *open_file; - char *smb_read_data; - char __user *current_offset; struct smb_com_read_rsp *pSMBr; + char *read_data; + + if (!nr_segs) + return 0; + + len = iov_length(iov, nr_segs); + if (!len) + return 0; xid = GetXid(); cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); - if (file->private_data == NULL) { - rc = -EBADF; - FreeXid(xid); - return rc; - } open_file = file->private_data; pTcon = tlink_tcon(open_file->tlink); if ((file->f_flags & O_ACCMODE) == O_WRONLY) cFYI(1, "attempting read on write only file instance"); - for (total_read = 0, current_offset = read_data; - read_size > total_read; - total_read += bytes_read, current_offset += bytes_read) { - current_read_size = min_t(const int, read_size - total_read, - cifs_sb->rsize); + for (total_read = 0; total_read < len; total_read += bytes_read) { + cur_len = min_t(const size_t, len - total_read, cifs_sb->rsize); rc = -EAGAIN; - smb_read_data = NULL; + read_data = NULL; + while (rc == -EAGAIN) { int buf_type = CIFS_NO_BUFFER; if (open_file->invalidHandle) { @@ -1639,27 +1859,25 @@ ssize_t cifs_user_read(struct file *file, char __user *read_data, if (rc != 0) break; } - rc = CIFSSMBRead(xid, pTcon, - open_file->netfid, - current_read_size, *poffset, - &bytes_read, &smb_read_data, - &buf_type); - pSMBr = (struct smb_com_read_rsp *)smb_read_data; - if (smb_read_data) { - if (copy_to_user(current_offset, - smb_read_data + - 4 /* RFC1001 length field */ + - le16_to_cpu(pSMBr->DataOffset), - bytes_read)) + rc = CIFSSMBRead(xid, pTcon, open_file->netfid, + cur_len, *poffset, &bytes_read, + &read_data, &buf_type); + pSMBr = (struct smb_com_read_rsp *)read_data; + if (read_data) { + char *data_offset = read_data + 4 + + le16_to_cpu(pSMBr->DataOffset); + if (memcpy_toiovecend(iov, data_offset, + iov_offset, bytes_read)) rc = -EFAULT; - if (buf_type == CIFS_SMALL_BUFFER) - cifs_small_buf_release(smb_read_data); + cifs_small_buf_release(read_data); else if (buf_type == CIFS_LARGE_BUFFER) - cifs_buf_release(smb_read_data); - smb_read_data = NULL; + cifs_buf_release(read_data); + read_data = NULL; + iov_offset += bytes_read; } } + if (rc || (bytes_read == 0)) { if (total_read) { break; @@ -1672,13 +1890,57 @@ ssize_t cifs_user_read(struct file *file, char __user *read_data, *poffset += bytes_read; } } + FreeXid(xid); return total_read; } +ssize_t cifs_user_read(struct file *file, char __user *read_data, + size_t read_size, loff_t *poffset) +{ + struct iovec iov; + iov.iov_base = read_data; + iov.iov_len = read_size; + + return cifs_iovec_read(file, &iov, 1, poffset); +} + +static ssize_t cifs_user_readv(struct kiocb *iocb, const struct iovec *iov, + unsigned long nr_segs, loff_t pos) +{ + ssize_t read; + + read = cifs_iovec_read(iocb->ki_filp, iov, nr_segs, &pos); + if (read > 0) + iocb->ki_pos = pos; + + return read; +} + +ssize_t cifs_strict_readv(struct kiocb *iocb, const struct iovec *iov, + unsigned long nr_segs, loff_t pos) +{ + struct inode *inode; + + inode = iocb->ki_filp->f_path.dentry->d_inode; + + if (CIFS_I(inode)->clientCanCacheRead) + return generic_file_aio_read(iocb, iov, nr_segs, pos); + + /* + * In strict cache mode we need to read from the server all the time + * if we don't have level II oplock because the server can delay mtime + * change - so we can't make a decision about inode invalidating. + * And we can also fail with pagereading if there are mandatory locks + * on pages affected by this read but not on the region from pos to + * pos+len-1. + */ + + return cifs_user_readv(iocb, iov, nr_segs, pos); +} static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size, - loff_t *poffset) + loff_t *poffset) { int rc = -EACCES; unsigned int bytes_read = 0; @@ -1746,6 +2008,21 @@ static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size, return total_read; } +int cifs_file_strict_mmap(struct file *file, struct vm_area_struct *vma) +{ + int rc, xid; + struct inode *inode = file->f_path.dentry->d_inode; + + xid = GetXid(); + + if (!CIFS_I(inode)->clientCanCacheRead) + cifs_invalidate_mapping(inode); + + rc = generic_file_mmap(file, vma); + FreeXid(xid); + return rc; +} + int cifs_file_mmap(struct file *file, struct vm_area_struct *vma) { int rc, xid; @@ -2192,7 +2469,8 @@ void cifs_oplock_break(struct work_struct *work) */ if (!cfile->oplock_break_cancelled) { rc = CIFSSMBLock(0, tlink_tcon(cfile->tlink), cfile->netfid, 0, - 0, 0, 0, LOCKING_ANDX_OPLOCK_RELEASE, false); + 0, 0, 0, LOCKING_ANDX_OPLOCK_RELEASE, false, + cinode->clientCanCacheRead ? 1 : 0); cFYI(1, "Oplock release rc = %d", rc); } diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 6c9ee80..8852470 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -44,13 +44,17 @@ static void cifs_set_ops(struct inode *inode) inode->i_fop = &cifs_file_direct_nobrl_ops; else inode->i_fop = &cifs_file_direct_ops; + } else if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_STRICT_IO) { + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_BRL) + inode->i_fop = &cifs_file_strict_nobrl_ops; + else + inode->i_fop = &cifs_file_strict_ops; } else if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_BRL) inode->i_fop = &cifs_file_nobrl_ops; else { /* not direct, send byte range locks */ inode->i_fop = &cifs_file_ops; } - /* check if server can support readpages */ if (cifs_sb_master_tcon(cifs_sb)->ses->server->maxBuf < PAGE_CACHE_SIZE + MAX_CIFS_HDR_SIZE) @@ -1679,7 +1683,7 @@ cifs_inode_needs_reval(struct inode *inode) /* * Zap the cache. Called when invalid_mapping flag is set. */ -static void +void cifs_invalidate_mapping(struct inode *inode) { int rc; diff --git a/fs/cifs/link.c b/fs/cifs/link.c index 306769d..e8804d3 100644 --- a/fs/cifs/link.c +++ b/fs/cifs/link.c @@ -28,7 +28,6 @@ #include "cifsproto.h" #include "cifs_debug.h" #include "cifs_fs_sb.h" -#include "md5.h" #define CIFS_MF_SYMLINK_LEN_OFFSET (4+1) #define CIFS_MF_SYMLINK_MD5_OFFSET (CIFS_MF_SYMLINK_LEN_OFFSET+(4+1)) @@ -47,6 +46,45 @@ md5_hash[12], md5_hash[13], md5_hash[14], md5_hash[15] static int +symlink_hash(unsigned int link_len, const char *link_str, u8 *md5_hash) +{ + int rc; + unsigned int size; + struct crypto_shash *md5; + struct sdesc *sdescmd5; + + md5 = crypto_alloc_shash("md5", 0, 0); + if (IS_ERR(md5)) { + rc = PTR_ERR(md5); + cERROR(1, "%s: Crypto md5 allocation error %d\n", __func__, rc); + return rc; + } + size = sizeof(struct shash_desc) + crypto_shash_descsize(md5); + sdescmd5 = kmalloc(size, GFP_KERNEL); + if (!sdescmd5) { + rc = -ENOMEM; + cERROR(1, "%s: Memory allocation failure\n", __func__); + goto symlink_hash_err; + } + sdescmd5->shash.tfm = md5; + sdescmd5->shash.flags = 0x0; + + rc = crypto_shash_init(&sdescmd5->shash); + if (rc) { + cERROR(1, "%s: Could not init md5 shash\n", __func__); + goto symlink_hash_err; + } + crypto_shash_update(&sdescmd5->shash, link_str, link_len); + rc = crypto_shash_final(&sdescmd5->shash, md5_hash); + +symlink_hash_err: + crypto_free_shash(md5); + kfree(sdescmd5); + + return rc; +} + +static int CIFSParseMFSymlink(const u8 *buf, unsigned int buf_len, unsigned int *_link_len, @@ -56,7 +94,6 @@ CIFSParseMFSymlink(const u8 *buf, unsigned int link_len; const char *md5_str1; const char *link_str; - struct MD5Context md5_ctx; u8 md5_hash[16]; char md5_str2[34]; @@ -70,9 +107,11 @@ CIFSParseMFSymlink(const u8 *buf, if (rc != 1) return -EINVAL; - cifs_MD5_init(&md5_ctx); - cifs_MD5_update(&md5_ctx, (const u8 *)link_str, link_len); - cifs_MD5_final(md5_hash, &md5_ctx); + rc = symlink_hash(link_len, link_str, md5_hash); + if (rc) { + cFYI(1, "%s: MD5 hash failure: %d\n", __func__, rc); + return rc; + } snprintf(md5_str2, sizeof(md5_str2), CIFS_MF_SYMLINK_MD5_FORMAT, @@ -94,9 +133,9 @@ CIFSParseMFSymlink(const u8 *buf, static int CIFSFormatMFSymlink(u8 *buf, unsigned int buf_len, const char *link_str) { + int rc; unsigned int link_len; unsigned int ofs; - struct MD5Context md5_ctx; u8 md5_hash[16]; if (buf_len != CIFS_MF_SYMLINK_FILE_SIZE) @@ -107,9 +146,11 @@ CIFSFormatMFSymlink(u8 *buf, unsigned int buf_len, const char *link_str) if (link_len > CIFS_MF_SYMLINK_LINK_MAXLEN) return -ENAMETOOLONG; - cifs_MD5_init(&md5_ctx); - cifs_MD5_update(&md5_ctx, (const u8 *)link_str, link_len); - cifs_MD5_final(md5_hash, &md5_ctx); + rc = symlink_hash(link_len, link_str, md5_hash); + if (rc) { + cFYI(1, "%s: MD5 hash failure: %d\n", __func__, rc); + return rc; + } snprintf(buf, buf_len, CIFS_MF_SYMLINK_LEN_FORMAT CIFS_MF_SYMLINK_MD5_FORMAT, diff --git a/fs/cifs/md4.c b/fs/cifs/md4.c deleted file mode 100644 index a725c26..0000000 --- a/fs/cifs/md4.c +++ /dev/null @@ -1,205 +0,0 @@ -/* - Unix SMB/Netbios implementation. - Version 1.9. - a implementation of MD4 designed for use in the SMB authentication protocol - Copyright (C) Andrew Tridgell 1997-1998. - Modified by Steve French (sfrench@us.ibm.com) 2002-2003 - - This program is free software; you can redistribute it and/or modify - it under the terms of the GNU General Public License as published by - the Free Software Foundation; either version 2 of the License, or - (at your option) any later version. - - This program is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - GNU General Public License for more details. - - You should have received a copy of the GNU General Public License - along with this program; if not, write to the Free Software - Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. -*/ -#include <linux/module.h> -#include <linux/fs.h> -#include "cifsencrypt.h" - -/* NOTE: This code makes no attempt to be fast! */ - -static __u32 -F(__u32 X, __u32 Y, __u32 Z) -{ - return (X & Y) | ((~X) & Z); -} - -static __u32 -G(__u32 X, __u32 Y, __u32 Z) -{ - return (X & Y) | (X & Z) | (Y & Z); -} - -static __u32 -H(__u32 X, __u32 Y, __u32 Z) -{ - return X ^ Y ^ Z; -} - -static __u32 -lshift(__u32 x, int s) -{ - x &= 0xFFFFFFFF; - return ((x << s) & 0xFFFFFFFF) | (x >> (32 - s)); -} - -#define ROUND1(a,b,c,d,k,s) (*a) = lshift((*a) + F(*b,*c,*d) + X[k], s) -#define ROUND2(a,b,c,d,k,s) (*a) = lshift((*a) + G(*b,*c,*d) + X[k] + (__u32)0x5A827999,s) -#define ROUND3(a,b,c,d,k,s) (*a) = lshift((*a) + H(*b,*c,*d) + X[k] + (__u32)0x6ED9EBA1,s) - -/* this applies md4 to 64 byte chunks */ -static void -mdfour64(__u32 *M, __u32 *A, __u32 *B, __u32 *C, __u32 *D) -{ - int j; - __u32 AA, BB, CC, DD; - __u32 X[16]; - - - for (j = 0; j < 16; j++) - X[j] = M[j]; - - AA = *A; - BB = *B; - CC = *C; - DD = *D; - - ROUND1(A, B, C, D, 0, 3); - ROUND1(D, A, B, C, 1, 7); - ROUND1(C, D, A, B, 2, 11); - ROUND1(B, C, D, A, 3, 19); - ROUND1(A, B, C, D, 4, 3); - ROUND1(D, A, B, C, 5, 7); - ROUND1(C, D, A, B, 6, 11); - ROUND1(B, C, D, A, 7, 19); - ROUND1(A, B, C, D, 8, 3); - ROUND1(D, A, B, C, 9, 7); - ROUND1(C, D, A, B, 10, 11); - ROUND1(B, C, D, A, 11, 19); - ROUND1(A, B, C, D, 12, 3); - ROUND1(D, A, B, C, 13, 7); - ROUND1(C, D, A, B, 14, 11); - ROUND1(B, C, D, A, 15, 19); - - ROUND2(A, B, C, D, 0, 3); - ROUND2(D, A, B, C, 4, 5); - ROUND2(C, D, A, B, 8, 9); - ROUND2(B, C, D, A, 12, 13); - ROUND2(A, B, C, D, 1, 3); - ROUND2(D, A, B, C, 5, 5); - ROUND2(C, D, A, B, 9, 9); - ROUND2(B, C, D, A, 13, 13); - ROUND2(A, B, C, D, 2, 3); - ROUND2(D, A, B, C, 6, 5); - ROUND2(C, D, A, B, 10, 9); - ROUND2(B, C, D, A, 14, 13); - ROUND2(A, B, C, D, 3, 3); - ROUND2(D, A, B, C, 7, 5); - ROUND2(C, D, A, B, 11, 9); - ROUND2(B, C, D, A, 15, 13); - - ROUND3(A, B, C, D, 0, 3); - ROUND3(D, A, B, C, 8, 9); - ROUND3(C, D, A, B, 4, 11); - ROUND3(B, C, D, A, 12, 15); - ROUND3(A, B, C, D, 2, 3); - ROUND3(D, A, B, C, 10, 9); - ROUND3(C, D, A, B, 6, 11); - ROUND3(B, C, D, A, 14, 15); - ROUND3(A, B, C, D, 1, 3); - ROUND3(D, A, B, C, 9, 9); - ROUND3(C, D, A, B, 5, 11); - ROUND3(B, C, D, A, 13, 15); - ROUND3(A, B, C, D, 3, 3); - ROUND3(D, A, B, C, 11, 9); - ROUND3(C, D, A, B, 7, 11); - ROUND3(B, C, D, A, 15, 15); - - *A += AA; - *B += BB; - *C += CC; - *D += DD; - - *A &= 0xFFFFFFFF; - *B &= 0xFFFFFFFF; - *C &= 0xFFFFFFFF; - *D &= 0xFFFFFFFF; - - for (j = 0; j < 16; j++) - X[j] = 0; -} - -static void -copy64(__u32 *M, unsigned char *in) -{ - int i; - - for (i = 0; i < 16; i++) - M[i] = (in[i * 4 + 3] << 24) | (in[i * 4 + 2] << 16) | - (in[i * 4 + 1] << 8) | (in[i * 4 + 0] << 0); -} - -static void -copy4(unsigned char *out, __u32 x) -{ - out[0] = x & 0xFF; - out[1] = (x >> 8) & 0xFF; - out[2] = (x >> 16) & 0xFF; - out[3] = (x >> 24) & 0xFF; -} - -/* produce a md4 message digest from data of length n bytes */ -void -mdfour(unsigned char *out, unsigned char *in, int n) -{ - unsigned char buf[128]; - __u32 M[16]; - __u32 b = n * 8; - int i; - __u32 A = 0x67452301; - __u32 B = 0xefcdab89; - __u32 C = 0x98badcfe; - __u32 D = 0x10325476; - - while (n > 64) { - copy64(M, in); - mdfour64(M, &A, &B, &C, &D); - in += 64; - n -= 64; - } - - for (i = 0; i < 128; i++) - buf[i] = 0; - memcpy(buf, in, n); - buf[n] = 0x80; - - if (n <= 55) { - copy4(buf + 56, b); - copy64(M, buf); - mdfour64(M, &A, &B, &C, &D); - } else { - copy4(buf + 120, b); - copy64(M, buf); - mdfour64(M, &A, &B, &C, &D); - copy64(M, buf + 64); - mdfour64(M, &A, &B, &C, &D); - } - - for (i = 0; i < 128; i++) - buf[i] = 0; - copy64(M, buf); - - copy4(out, A); - copy4(out + 4, B); - copy4(out + 8, C); - copy4(out + 12, D); - - A = B = C = D = 0; -} diff --git a/fs/cifs/md5.c b/fs/cifs/md5.c deleted file mode 100644 index 98b66a5..0000000 --- a/fs/cifs/md5.c +++ /dev/null @@ -1,366 +0,0 @@ -/* - * This code implements the MD5 message-digest algorithm. - * The algorithm is due to Ron Rivest. This code was - * written by Colin Plumb in 1993, no copyright is claimed. - * This code is in the public domain; do with it what you wish. - * - * Equivalent code is available from RSA Data Security, Inc. - * This code has been tested against that, and is equivalent, - * except that you don't need to include two pages of legalese - * with every copy. - * - * To compute the message digest of a chunk of bytes, declare an - * MD5Context structure, pass it to cifs_MD5_init, call cifs_MD5_update as - * needed on buffers full of bytes, and then call cifs_MD5_final, which - * will fill a supplied 16-byte array with the digest. - */ - -/* This code slightly modified to fit into Samba by - abartlet@samba.org Jun 2001 - and to fit the cifs vfs by - Steve French sfrench@us.ibm.com */ - -#include <linux/string.h> -#include "md5.h" - -static void MD5Transform(__u32 buf[4], __u32 const in[16]); - -/* - * Note: this code is harmless on little-endian machines. - */ -static void -byteReverse(unsigned char *buf, unsigned longs) -{ - __u32 t; - do { - t = (__u32) ((unsigned) buf[3] << 8 | buf[2]) << 16 | - ((unsigned) buf[1] << 8 | buf[0]); - *(__u32 *) buf = t; - buf += 4; - } while (--longs); -} - -/* - * Start MD5 accumulation. Set bit count to 0 and buffer to mysterious - * initialization constants. - */ -void -cifs_MD5_init(struct MD5Context *ctx) -{ - ctx->buf[0] = 0x67452301; - ctx->buf[1] = 0xefcdab89; - ctx->buf[2] = 0x98badcfe; - ctx->buf[3] = 0x10325476; - - ctx->bits[0] = 0; - ctx->bits[1] = 0; -} - -/* - * Update context to reflect the concatenation of another buffer full - * of bytes. - */ -void -cifs_MD5_update(struct MD5Context *ctx, unsigned char const *buf, unsigned len) -{ - register __u32 t; - - /* Update bitcount */ - - t = ctx->bits[0]; - if ((ctx->bits[0] = t + ((__u32) len << 3)) < t) - ctx->bits[1]++; /* Carry from low to high */ - ctx->bits[1] += len >> 29; - - t = (t >> 3) & 0x3f; /* Bytes already in shsInfo->data */ - - /* Handle any leading odd-sized chunks */ - - if (t) { - unsigned char *p = (unsigned char *) ctx->in + t; - - t = 64 - t; - if (len < t) { - memmove(p, buf, len); - return; - } - memmove(p, buf, t); - byteReverse(ctx->in, 16); - MD5Transform(ctx->buf, (__u32 *) ctx->in); - buf += t; - len -= t; - } - /* Process data in 64-byte chunks */ - - while (len >= 64) { - memmove(ctx->in, buf, 64); - byteReverse(ctx->in, 16); - MD5Transform(ctx->buf, (__u32 *) ctx->in); - buf += 64; - len -= 64; - } - - /* Handle any remaining bytes of data. */ - - memmove(ctx->in, buf, len); -} - -/* - * Final wrapup - pad to 64-byte boundary with the bit pattern - * 1 0* (64-bit count of bits processed, MSB-first) - */ -void -cifs_MD5_final(unsigned char digest[16], struct MD5Context *ctx) -{ - unsigned int count; - unsigned char *p; - - /* Compute number of bytes mod 64 */ - count = (ctx->bits[0] >> 3) & 0x3F; - - /* Set the first char of padding to 0x80. This is safe since there is - always at least one byte free */ - p = ctx->in + count; - *p++ = 0x80; - - /* Bytes of padding needed to make 64 bytes */ - count = 64 - 1 - count; - - /* Pad out to 56 mod 64 */ - if (count < 8) { - /* Two lots of padding: Pad the first block to 64 bytes */ - memset(p, 0, count); - byteReverse(ctx->in, 16); - MD5Transform(ctx->buf, (__u32 *) ctx->in); - - /* Now fill the next block with 56 bytes */ - memset(ctx->in, 0, 56); - } else { - /* Pad block to 56 bytes */ - memset(p, 0, count - 8); - } - byteReverse(ctx->in, 14); - - /* Append length in bits and transform */ - ((__u32 *) ctx->in)[14] = ctx->bits[0]; - ((__u32 *) ctx->in)[15] = ctx->bits[1]; - - MD5Transform(ctx->buf, (__u32 *) ctx->in); - byteReverse((unsigned char *) ctx->buf, 4); - memmove(digest, ctx->buf, 16); - memset(ctx, 0, sizeof(*ctx)); /* In case it's sensitive */ -} - -/* The four core functions - F1 is optimized somewhat */ - -/* #define F1(x, y, z) (x & y | ~x & z) */ -#define F1(x, y, z) (z ^ (x & (y ^ z))) -#define F2(x, y, z) F1(z, x, y) -#define F3(x, y, z) (x ^ y ^ z) -#define F4(x, y, z) (y ^ (x | ~z)) - -/* This is the central step in the MD5 algorithm. */ -#define MD5STEP(f, w, x, y, z, data, s) \ - (w += f(x, y, z) + data, w = w<<s | w>>(32-s), w += x) - -/* - * The core of the MD5 algorithm, this alters an existing MD5 hash to - * reflect the addition of 16 longwords of new data. cifs_MD5_update blocks - * the data and converts bytes into longwords for this routine. - */ -static void -MD5Transform(__u32 buf[4], __u32 const in[16]) -{ - register __u32 a, b, c, d; - - a = buf[0]; - b = buf[1]; - c = buf[2]; - d = buf[3]; - - MD5STEP(F1, a, b, c, d, in[0] + 0xd76aa478, 7); - MD5STEP(F1, d, a, b, c, in[1] + 0xe8c7b756, 12); - MD5STEP(F1, c, d, a, b, in[2] + 0x242070db, 17); - MD5STEP(F1, b, c, d, a, in[3] + 0xc1bdceee, 22); - MD5STEP(F1, a, b, c, d, in[4] + 0xf57c0faf, 7); - MD5STEP(F1, d, a, b, c, in[5] + 0x4787c62a, 12); - MD5STEP(F1, c, d, a, b, in[6] + 0xa8304613, 17); - MD5STEP(F1, b, c, d, a, in[7] + 0xfd469501, 22); - MD5STEP(F1, a, b, c, d, in[8] + 0x698098d8, 7); - MD5STEP(F1, d, a, b, c, in[9] + 0x8b44f7af, 12); - MD5STEP(F1, c, d, a, b, in[10] + 0xffff5bb1, 17); - MD5STEP(F1, b, c, d, a, in[11] + 0x895cd7be, 22); - MD5STEP(F1, a, b, c, d, in[12] + 0x6b901122, 7); - MD5STEP(F1, d, a, b, c, in[13] + 0xfd987193, 12); - MD5STEP(F1, c, d, a, b, in[14] + 0xa679438e, 17); - MD5STEP(F1, b, c, d, a, in[15] + 0x49b40821, 22); - - MD5STEP(F2, a, b, c, d, in[1] + 0xf61e2562, 5); - MD5STEP(F2, d, a, b, c, in[6] + 0xc040b340, 9); - MD5STEP(F2, c, d, a, b, in[11] + 0x265e5a51, 14); - MD5STEP(F2, b, c, d, a, in[0] + 0xe9b6c7aa, 20); - MD5STEP(F2, a, b, c, d, in[5] + 0xd62f105d, 5); - MD5STEP(F2, d, a, b, c, in[10] + 0x02441453, 9); - MD5STEP(F2, c, d, a, b, in[15] + 0xd8a1e681, 14); - MD5STEP(F2, b, c, d, a, in[4] + 0xe7d3fbc8, 20); - MD5STEP(F2, a, b, c, d, in[9] + 0x21e1cde6, 5); - MD5STEP(F2, d, a, b, c, in[14] + 0xc33707d6, 9); - MD5STEP(F2, c, d, a, b, in[3] + 0xf4d50d87, 14); - MD5STEP(F2, b, c, d, a, in[8] + 0x455a14ed, 20); - MD5STEP(F2, a, b, c, d, in[13] + 0xa9e3e905, 5); - MD5STEP(F2, d, a, b, c, in[2] + 0xfcefa3f8, 9); - MD5STEP(F2, c, d, a, b, in[7] + 0x676f02d9, 14); - MD5STEP(F2, b, c, d, a, in[12] + 0x8d2a4c8a, 20); - - MD5STEP(F3, a, b, c, d, in[5] + 0xfffa3942, 4); - MD5STEP(F3, d, a, b, c, in[8] + 0x8771f681, 11); - MD5STEP(F3, c, d, a, b, in[11] + 0x6d9d6122, 16); - MD5STEP(F3, b, c, d, a, in[14] + 0xfde5380c, 23); - MD5STEP(F3, a, b, c, d, in[1] + 0xa4beea44, 4); - MD5STEP(F3, d, a, b, c, in[4] + 0x4bdecfa9, 11); - MD5STEP(F3, c, d, a, b, in[7] + 0xf6bb4b60, 16); - MD5STEP(F3, b, c, d, a, in[10] + 0xbebfbc70, 23); - MD5STEP(F3, a, b, c, d, in[13] + 0x289b7ec6, 4); - MD5STEP(F3, d, a, b, c, in[0] + 0xeaa127fa, 11); - MD5STEP(F3, c, d, a, b, in[3] + 0xd4ef3085, 16); - MD5STEP(F3, b, c, d, a, in[6] + 0x04881d05, 23); - MD5STEP(F3, a, b, c, d, in[9] + 0xd9d4d039, 4); - MD5STEP(F3, d, a, b, c, in[12] + 0xe6db99e5, 11); - MD5STEP(F3, c, d, a, b, in[15] + 0x1fa27cf8, 16); - MD5STEP(F3, b, c, d, a, in[2] + 0xc4ac5665, 23); - - MD5STEP(F4, a, b, c, d, in[0] + 0xf4292244, 6); - MD5STEP(F4, d, a, b, c, in[7] + 0x432aff97, 10); - MD5STEP(F4, c, d, a, b, in[14] + 0xab9423a7, 15); - MD5STEP(F4, b, c, d, a, in[5] + 0xfc93a039, 21); - MD5STEP(F4, a, b, c, d, in[12] + 0x655b59c3, 6); - MD5STEP(F4, d, a, b, c, in[3] + 0x8f0ccc92, 10); - MD5STEP(F4, c, d, a, b, in[10] + 0xffeff47d, 15); - MD5STEP(F4, b, c, d, a, in[1] + 0x85845dd1, 21); - MD5STEP(F4, a, b, c, d, in[8] + 0x6fa87e4f, 6); - MD5STEP(F4, d, a, b, c, in[15] + 0xfe2ce6e0, 10); - MD5STEP(F4, c, d, a, b, in[6] + 0xa3014314, 15); - MD5STEP(F4, b, c, d, a, in[13] + 0x4e0811a1, 21); - MD5STEP(F4, a, b, c, d, in[4] + 0xf7537e82, 6); - MD5STEP(F4, d, a, b, c, in[11] + 0xbd3af235, 10); - MD5STEP(F4, c, d, a, b, in[2] + 0x2ad7d2bb, 15); - MD5STEP(F4, b, c, d, a, in[9] + 0xeb86d391, 21); - - buf[0] += a; - buf[1] += b; - buf[2] += c; - buf[3] += d; -} - -#if 0 /* currently unused */ -/*********************************************************************** - the rfc 2104 version of hmac_md5 initialisation. -***********************************************************************/ -static void -hmac_md5_init_rfc2104(unsigned char *key, int key_len, - struct HMACMD5Context *ctx) -{ - int i; - - /* if key is longer than 64 bytes reset it to key=MD5(key) */ - if (key_len > 64) { - unsigned char tk[16]; - struct MD5Context tctx; - - cifs_MD5_init(&tctx); - cifs_MD5_update(&tctx, key, key_len); - cifs_MD5_final(tk, &tctx); - - key = tk; - key_len = 16; - } - - /* start out by storing key in pads */ - memset(ctx->k_ipad, 0, sizeof(ctx->k_ipad)); - memset(ctx->k_opad, 0, sizeof(ctx->k_opad)); - memcpy(ctx->k_ipad, key, key_len); - memcpy(ctx->k_opad, key, key_len); - - /* XOR key with ipad and opad values */ - for (i = 0; i < 64; i++) { - ctx->k_ipad[i] ^= 0x36; - ctx->k_opad[i] ^= 0x5c; - } - - cifs_MD5_init(&ctx->ctx); - cifs_MD5_update(&ctx->ctx, ctx->k_ipad, 64); -} -#endif - -/*********************************************************************** - the microsoft version of hmac_md5 initialisation. -***********************************************************************/ -void -hmac_md5_init_limK_to_64(const unsigned char *key, int key_len, - struct HMACMD5Context *ctx) -{ - int i; - - /* if key is longer than 64 bytes truncate it */ - if (key_len > 64) - key_len = 64; - - /* start out by storing key in pads */ - memset(ctx->k_ipad, 0, sizeof(ctx->k_ipad)); - memset(ctx->k_opad, 0, sizeof(ctx->k_opad)); - memcpy(ctx->k_ipad, key, key_len); - memcpy(ctx->k_opad, key, key_len); - - /* XOR key with ipad and opad values */ - for (i = 0; i < 64; i++) { - ctx->k_ipad[i] ^= 0x36; - ctx->k_opad[i] ^= 0x5c; - } - - cifs_MD5_init(&ctx->ctx); - cifs_MD5_update(&ctx->ctx, ctx->k_ipad, 64); -} - -/*********************************************************************** - update hmac_md5 "inner" buffer -***********************************************************************/ -void -hmac_md5_update(const unsigned char *text, int text_len, - struct HMACMD5Context *ctx) -{ - cifs_MD5_update(&ctx->ctx, text, text_len); /* then text of datagram */ -} - -/*********************************************************************** - finish off hmac_md5 "inner" buffer and generate outer one. -***********************************************************************/ -void -hmac_md5_final(unsigned char *digest, struct HMACMD5Context *ctx) -{ - struct MD5Context ctx_o; - - cifs_MD5_final(digest, &ctx->ctx); - - cifs_MD5_init(&ctx_o); - cifs_MD5_update(&ctx_o, ctx->k_opad, 64); - cifs_MD5_update(&ctx_o, digest, 16); - cifs_MD5_final(digest, &ctx_o); -} - -/*********************************************************** - single function to calculate an HMAC MD5 digest from data. - use the microsoft hmacmd5 init method because the key is 16 bytes. -************************************************************/ -#if 0 /* currently unused */ -static void -hmac_md5(unsigned char key[16], unsigned char *data, int data_len, - unsigned char *digest) -{ - struct HMACMD5Context ctx; - hmac_md5_init_limK_to_64(key, 16, &ctx); - if (data_len != 0) - hmac_md5_update(data, data_len, &ctx); - - hmac_md5_final(digest, &ctx); -} -#endif diff --git a/fs/cifs/md5.h b/fs/cifs/md5.h deleted file mode 100644 index 6fba8cb..0000000 --- a/fs/cifs/md5.h +++ /dev/null @@ -1,38 +0,0 @@ -#ifndef MD5_H -#define MD5_H -#ifndef HEADER_MD5_H -/* Try to avoid clashes with OpenSSL */ -#define HEADER_MD5_H -#endif - -struct MD5Context { - __u32 buf[4]; - __u32 bits[2]; - unsigned char in[64]; -}; -#endif /* !MD5_H */ - -#ifndef _HMAC_MD5_H -struct HMACMD5Context { - struct MD5Context ctx; - unsigned char k_ipad[65]; - unsigned char k_opad[65]; -}; -#endif /* _HMAC_MD5_H */ - -void cifs_MD5_init(struct MD5Context *context); -void cifs_MD5_update(struct MD5Context *context, unsigned char const *buf, - unsigned len); -void cifs_MD5_final(unsigned char digest[16], struct MD5Context *context); - -/* The following definitions come from lib/hmacmd5.c */ - -/* void hmac_md5_init_rfc2104(unsigned char *key, int key_len, - struct HMACMD5Context *ctx);*/ -void hmac_md5_init_limK_to_64(const unsigned char *key, int key_len, - struct HMACMD5Context *ctx); -void hmac_md5_update(const unsigned char *text, int text_len, - struct HMACMD5Context *ctx); -void hmac_md5_final(unsigned char *digest, struct HMACMD5Context *ctx); -/* void hmac_md5(unsigned char key[16], unsigned char *data, int data_len, - unsigned char *digest);*/ diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index 43f1028..2a930a7 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -236,10 +236,7 @@ __u16 GetNextMid(struct TCP_Server_Info *server) { __u16 mid = 0; __u16 last_mid; - int collision; - - if (server == NULL) - return mid; + bool collision; spin_lock(&GlobalMid_Lock); last_mid = server->CurrentMid; /* we do not want to loop forever */ @@ -252,24 +249,38 @@ __u16 GetNextMid(struct TCP_Server_Info *server) (and it would also have to have been a request that did not time out) */ while (server->CurrentMid != last_mid) { - struct list_head *tmp; struct mid_q_entry *mid_entry; + unsigned int num_mids; - collision = 0; + collision = false; if (server->CurrentMid == 0) server->CurrentMid++; - list_for_each(tmp, &server->pending_mid_q) { - mid_entry = list_entry(tmp, struct mid_q_entry, qhead); - - if ((mid_entry->mid == server->CurrentMid) && - (mid_entry->midState == MID_REQUEST_SUBMITTED)) { + num_mids = 0; + list_for_each_entry(mid_entry, &server->pending_mid_q, qhead) { + ++num_mids; + if (mid_entry->mid == server->CurrentMid && + mid_entry->midState == MID_REQUEST_SUBMITTED) { /* This mid is in use, try a different one */ - collision = 1; + collision = true; break; } } - if (collision == 0) { + + /* + * if we have more than 32k mids in the list, then something + * is very wrong. Possibly a local user is trying to DoS the + * box by issuing long-running calls and SIGKILL'ing them. If + * we get to 2^16 mids then we're in big trouble as this + * function could loop forever. + * + * Go ahead and assign out the mid in this situation, but force + * an eventual reconnect to clean out the pending_mid_q. + */ + if (num_mids > 32768) + server->tcpStatus = CifsNeedReconnect; + + if (!collision) { mid = server->CurrentMid; break; } @@ -381,29 +392,31 @@ header_assemble(struct smb_hdr *buffer, char smb_command /* command */ , } static int -checkSMBhdr(struct smb_hdr *smb, __u16 mid) +check_smb_hdr(struct smb_hdr *smb, __u16 mid) { - /* Make sure that this really is an SMB, that it is a response, - and that the message ids match */ - if ((*(__le32 *) smb->Protocol == cpu_to_le32(0x424d53ff)) && - (mid == smb->Mid)) { - if (smb->Flags & SMBFLG_RESPONSE) - return 0; - else { - /* only one valid case where server sends us request */ - if (smb->Command == SMB_COM_LOCKING_ANDX) - return 0; - else - cERROR(1, "Received Request not response"); - } - } else { /* bad signature or mid */ - if (*(__le32 *) smb->Protocol != cpu_to_le32(0x424d53ff)) - cERROR(1, "Bad protocol string signature header %x", - *(unsigned int *) smb->Protocol); - if (mid != smb->Mid) - cERROR(1, "Mids do not match"); + /* does it have the right SMB "signature" ? */ + if (*(__le32 *) smb->Protocol != cpu_to_le32(0x424d53ff)) { + cERROR(1, "Bad protocol string signature header 0x%x", + *(unsigned int *)smb->Protocol); + return 1; } - cERROR(1, "bad smb detected. The Mid=%d", smb->Mid); + + /* Make sure that message ids match */ + if (mid != smb->Mid) { + cERROR(1, "Mids do not match. received=%u expected=%u", + smb->Mid, mid); + return 1; + } + + /* if it's a response then accept */ + if (smb->Flags & SMBFLG_RESPONSE) + return 0; + + /* only one valid case where server sends us request */ + if (smb->Command == SMB_COM_LOCKING_ANDX) + return 0; + + cERROR(1, "Server sent request, not response. mid=%u", smb->Mid); return 1; } @@ -448,7 +461,7 @@ checkSMB(struct smb_hdr *smb, __u16 mid, unsigned int length) return 1; } - if (checkSMBhdr(smb, mid)) + if (check_smb_hdr(smb, mid)) return 1; clc_len = smbCalcSize_LE(smb); @@ -465,25 +478,26 @@ checkSMB(struct smb_hdr *smb, __u16 mid, unsigned int length) if (((4 + len) & 0xFFFF) == (clc_len & 0xFFFF)) return 0; /* bcc wrapped */ } - cFYI(1, "Calculated size %d vs length %d mismatch for mid %d", + cFYI(1, "Calculated size %u vs length %u mismatch for mid=%u", clc_len, 4 + len, smb->Mid); - /* Windows XP can return a few bytes too much, presumably - an illegal pad, at the end of byte range lock responses - so we allow for that three byte pad, as long as actual - received length is as long or longer than calculated length */ - /* We have now had to extend this more, since there is a - case in which it needs to be bigger still to handle a - malformed response to transact2 findfirst from WinXP when - access denied is returned and thus bcc and wct are zero - but server says length is 0x21 bytes too long as if the server - forget to reset the smb rfc1001 length when it reset the - wct and bcc to minimum size and drop the t2 parms and data */ - if ((4+len > clc_len) && (len <= clc_len + 512)) - return 0; - else { - cERROR(1, "RFC1001 size %d bigger than SMB for Mid=%d", + + if (4 + len < clc_len) { + cERROR(1, "RFC1001 size %u smaller than SMB for mid=%u", len, smb->Mid); return 1; + } else if (len > clc_len + 512) { + /* + * Some servers (Windows XP in particular) send more + * data than the lengths in the SMB packet would + * indicate on certain calls (byte range locks and + * trans2 find first calls in particular). While the + * client can handle such a frame by ignoring the + * trailing data, we choose limit the amount of extra + * data to 512 bytes. + */ + cERROR(1, "RFC1001 size %u more than 512 bytes larger " + "than SMB for mid=%u", len, smb->Mid); + return 1; } } return 0; @@ -571,7 +585,7 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) pCifsInode = CIFS_I(netfile->dentry->d_inode); cifs_set_oplock_level(pCifsInode, - pSMB->OplockLevel); + pSMB->OplockLevel ? OPLOCK_READ : 0); /* * cifs_oplock_break_put() can't be called * from here. Get reference after queueing @@ -637,77 +651,6 @@ dump_smb(struct smb_hdr *smb_buf, int smb_buf_length) return; } -/* Convert 16 bit Unicode pathname to wire format from string in current code - page. Conversion may involve remapping up the seven characters that are - only legal in POSIX-like OS (if they are present in the string). Path - names are little endian 16 bit Unicode on the wire */ -int -cifsConvertToUCS(__le16 *target, const char *source, int maxlen, - const struct nls_table *cp, int mapChars) -{ - int i, j, charlen; - int len_remaining = maxlen; - char src_char; - __u16 temp; - - if (!mapChars) - return cifs_strtoUCS(target, source, PATH_MAX, cp); - - for (i = 0, j = 0; i < maxlen; j++) { - src_char = source[i]; - switch (src_char) { - case 0: - target[j] = 0; - goto ctoUCS_out; - case ':': - target[j] = cpu_to_le16(UNI_COLON); - break; - case '*': - target[j] = cpu_to_le16(UNI_ASTERIK); - break; - case '?': - target[j] = cpu_to_le16(UNI_QUESTION); - break; - case '<': - target[j] = cpu_to_le16(UNI_LESSTHAN); - break; - case '>': - target[j] = cpu_to_le16(UNI_GRTRTHAN); - break; - case '|': - target[j] = cpu_to_le16(UNI_PIPE); - break; - /* BB We can not handle remapping slash until - all the calls to build_path_from_dentry - are modified, as they use slash as separator BB */ - /* case '\\': - target[j] = cpu_to_le16(UNI_SLASH); - break;*/ - default: - charlen = cp->char2uni(source+i, - len_remaining, &temp); - /* if no match, use question mark, which - at least in some cases servers as wild card */ - if (charlen < 1) { - target[j] = cpu_to_le16(0x003f); - charlen = 1; - } else - target[j] = cpu_to_le16(temp); - len_remaining -= charlen; - /* character may take more than one byte in the - the source string, but will take exactly two - bytes in the target string */ - i += charlen; - continue; - } - i++; /* move to next char in source string */ - len_remaining--; - } - -ctoUCS_out: - return i; -} - void cifs_autodisable_serverino(struct cifs_sb_info *cifs_sb) { diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c index 6783ce6..79f641e 100644 --- a/fs/cifs/netmisc.c +++ b/fs/cifs/netmisc.c @@ -170,7 +170,7 @@ cifs_convert_address(struct sockaddr *dst, const char *src, int len) { int rc, alen, slen; const char *pct; - char *endp, scope_id[13]; + char scope_id[13]; struct sockaddr_in *s4 = (struct sockaddr_in *) dst; struct sockaddr_in6 *s6 = (struct sockaddr_in6 *) dst; @@ -197,9 +197,9 @@ cifs_convert_address(struct sockaddr *dst, const char *src, int len) memcpy(scope_id, pct + 1, slen); scope_id[slen] = '\0'; - s6->sin6_scope_id = (u32) simple_strtoul(pct, &endp, 0); - if (endp != scope_id + slen) - return 0; + rc = strict_strtoul(scope_id, 0, + (unsigned long *)&s6->sin6_scope_id); + rc = (rc == 0) ? 1 : 0; } return rc; @@ -916,14 +916,14 @@ unsigned int smbCalcSize(struct smb_hdr *ptr) { return (sizeof(struct smb_hdr) + (2 * ptr->WordCount) + - 2 /* size of the bcc field */ + BCC(ptr)); + 2 /* size of the bcc field */ + get_bcc(ptr)); } unsigned int smbCalcSize_LE(struct smb_hdr *ptr) { return (sizeof(struct smb_hdr) + (2 * ptr->WordCount) + - 2 /* size of the bcc field */ + le16_to_cpu(BCC_LE(ptr))); + 2 /* size of the bcc field */ + get_bcc_le(ptr)); } /* The following are taken from fs/ntfs/util.c */ diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index 7f25cc3..f8e4cd2 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -764,7 +764,6 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir) { int rc = 0; int xid, i; - struct cifs_sb_info *cifs_sb; struct cifsTconInfo *pTcon; struct cifsFileInfo *cifsFile = NULL; char *current_entry; @@ -775,8 +774,6 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir) xid = GetXid(); - cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); - /* * Ensure FindFirst doesn't fail before doing filldir() for '.' and * '..'. Otherwise we won't be able to notify VFS in case of failure. diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c index eb74648..1676570 100644 --- a/fs/cifs/sess.c +++ b/fs/cifs/sess.c @@ -277,7 +277,7 @@ static void ascii_ssetup_strings(char **pbcc_area, struct cifsSesInfo *ses, } static void -decode_unicode_ssetup(char **pbcc_area, int bleft, struct cifsSesInfo *ses, +decode_unicode_ssetup(char **pbcc_area, __u16 bleft, struct cifsSesInfo *ses, const struct nls_table *nls_cp) { int len; @@ -323,7 +323,7 @@ decode_unicode_ssetup(char **pbcc_area, int bleft, struct cifsSesInfo *ses, return; } -static int decode_ascii_ssetup(char **pbcc_area, int bleft, +static int decode_ascii_ssetup(char **pbcc_area, __u16 bleft, struct cifsSesInfo *ses, const struct nls_table *nls_cp) { @@ -575,12 +575,11 @@ CIFS_SessSetup(unsigned int xid, struct cifsSesInfo *ses, char *str_area; SESSION_SETUP_ANDX *pSMB; __u32 capabilities; - int count; + __u16 count; int resp_buf_type; struct kvec iov[3]; enum securityEnum type; - __u16 action; - int bytes_remaining; + __u16 action, bytes_remaining; struct key *spnego_key = NULL; __le32 phase = NtLmNegotiate; /* NTLMSSP, if needed, is multistage */ u16 blob_len; @@ -657,13 +656,13 @@ ssetup_ntlmssp_authenticate: if (type == LANMAN) { #ifdef CONFIG_CIFS_WEAK_PW_HASH - char lnm_session_key[CIFS_SESS_KEY_SIZE]; + char lnm_session_key[CIFS_AUTH_RESP_SIZE]; pSMB->req.hdr.Flags2 &= ~SMBFLG2_UNICODE; /* no capabilities flags in old lanman negotiation */ - pSMB->old_req.PasswordLength = cpu_to_le16(CIFS_SESS_KEY_SIZE); + pSMB->old_req.PasswordLength = cpu_to_le16(CIFS_AUTH_RESP_SIZE); /* Calculate hash with password and copy into bcc_ptr. * Encryption Key (stored as in cryptkey) gets used if the @@ -676,8 +675,8 @@ ssetup_ntlmssp_authenticate: true : false, lnm_session_key); ses->flags |= CIFS_SES_LANMAN; - memcpy(bcc_ptr, (char *)lnm_session_key, CIFS_SESS_KEY_SIZE); - bcc_ptr += CIFS_SESS_KEY_SIZE; + memcpy(bcc_ptr, (char *)lnm_session_key, CIFS_AUTH_RESP_SIZE); + bcc_ptr += CIFS_AUTH_RESP_SIZE; /* can not sign if LANMAN negotiated so no need to calculate signing key? but what if server @@ -876,10 +875,10 @@ ssetup_ntlmssp_authenticate: count = iov[1].iov_len + iov[2].iov_len; smb_buf->smb_buf_length += count; - BCC_LE(smb_buf) = cpu_to_le16(count); + put_bcc_le(count, smb_buf); rc = SendReceive2(xid, ses, iov, 3 /* num_iovecs */, &resp_buf_type, - CIFS_STD_OP /* not long */ | CIFS_LOG_ERROR); + CIFS_LOG_ERROR); /* SMB request buf freed in SendReceive2 */ pSMB = (SESSION_SETUP_ANDX *)iov[0].iov_base; @@ -910,7 +909,7 @@ ssetup_ntlmssp_authenticate: cFYI(1, "UID = %d ", ses->Suid); /* response can have either 3 or 4 word count - Samba sends 3 */ /* and lanman response is 3 */ - bytes_remaining = BCC(smb_buf); + bytes_remaining = get_bcc(smb_buf); bcc_ptr = pByteArea(smb_buf); if (smb_buf->WordCount == 4) { diff --git a/fs/cifs/smbdes.c b/fs/cifs/smbdes.c index b6b6dcb..0472148 100644 --- a/fs/cifs/smbdes.c +++ b/fs/cifs/smbdes.c @@ -45,7 +45,6 @@ up with a different answer to the one above) */ #include <linux/slab.h> -#include "cifsencrypt.h" #define uchar unsigned char static uchar perm1[56] = { 57, 49, 41, 33, 25, 17, 9, diff --git a/fs/cifs/smbencrypt.c b/fs/cifs/smbencrypt.c index 192ea51..b5041c8 100644 --- a/fs/cifs/smbencrypt.c +++ b/fs/cifs/smbencrypt.c @@ -32,9 +32,8 @@ #include "cifs_unicode.h" #include "cifspdu.h" #include "cifsglob.h" -#include "md5.h" #include "cifs_debug.h" -#include "cifsencrypt.h" +#include "cifsproto.h" #ifndef false #define false 0 @@ -48,14 +47,58 @@ #define SSVALX(buf,pos,val) (CVAL(buf,pos)=(val)&0xFF,CVAL(buf,pos+1)=(val)>>8) #define SSVAL(buf,pos,val) SSVALX((buf),(pos),((__u16)(val))) -/*The following definitions come from libsmb/smbencrypt.c */ +/* produce a md4 message digest from data of length n bytes */ +int +mdfour(unsigned char *md4_hash, unsigned char *link_str, int link_len) +{ + int rc; + unsigned int size; + struct crypto_shash *md4; + struct sdesc *sdescmd4; + + md4 = crypto_alloc_shash("md4", 0, 0); + if (IS_ERR(md4)) { + rc = PTR_ERR(md4); + cERROR(1, "%s: Crypto md4 allocation error %d\n", __func__, rc); + return rc; + } + size = sizeof(struct shash_desc) + crypto_shash_descsize(md4); + sdescmd4 = kmalloc(size, GFP_KERNEL); + if (!sdescmd4) { + rc = -ENOMEM; + cERROR(1, "%s: Memory allocation failure\n", __func__); + goto mdfour_err; + } + sdescmd4->shash.tfm = md4; + sdescmd4->shash.flags = 0x0; + + rc = crypto_shash_init(&sdescmd4->shash); + if (rc) { + cERROR(1, "%s: Could not init md4 shash\n", __func__); + goto mdfour_err; + } + crypto_shash_update(&sdescmd4->shash, link_str, link_len); + rc = crypto_shash_final(&sdescmd4->shash, md4_hash); -void SMBencrypt(unsigned char *passwd, const unsigned char *c8, - unsigned char *p24); -void E_md4hash(const unsigned char *passwd, unsigned char *p16); -static void SMBOWFencrypt(unsigned char passwd[16], const unsigned char *c8, - unsigned char p24[24]); -void SMBNTencrypt(unsigned char *passwd, unsigned char *c8, unsigned char *p24); +mdfour_err: + crypto_free_shash(md4); + kfree(sdescmd4); + + return rc; +} + +/* Does the des encryption from the NT or LM MD4 hash. */ +static void +SMBOWFencrypt(unsigned char passwd[16], const unsigned char *c8, + unsigned char p24[24]) +{ + unsigned char p21[21]; + + memset(p21, '\0', 21); + + memcpy(p21, passwd, 16); + E_P24(p21, c8, p24); +} /* This implements the X/Open SMB password encryption @@ -118,9 +161,10 @@ _my_mbstowcs(__u16 *dst, const unsigned char *src, int len) * Creates the MD4 Hash of the users password in NT UNICODE. */ -void +int E_md4hash(const unsigned char *passwd, unsigned char *p16) { + int rc; int len; __u16 wpwd[129]; @@ -139,8 +183,10 @@ E_md4hash(const unsigned char *passwd, unsigned char *p16) /* Calculate length in bytes */ len = _my_wcslen(wpwd) * sizeof(__u16); - mdfour(p16, (unsigned char *) wpwd, len); + rc = mdfour(p16, (unsigned char *) wpwd, len); memset(wpwd, 0, 129 * 2); + + return rc; } #if 0 /* currently unused */ @@ -212,19 +258,6 @@ ntv2_owf_gen(const unsigned char owf[16], const char *user_n, } #endif -/* Does the des encryption from the NT or LM MD4 hash. */ -static void -SMBOWFencrypt(unsigned char passwd[16], const unsigned char *c8, - unsigned char p24[24]) -{ - unsigned char p21[21]; - - memset(p21, '\0', 21); - - memcpy(p21, passwd, 16); - E_P24(p21, c8, p24); -} - /* Does the des encryption from the FIRST 8 BYTES of the NT or LM MD4 hash. */ #if 0 /* currently unused */ static void @@ -242,16 +275,21 @@ NTLMSSPOWFencrypt(unsigned char passwd[8], #endif /* Does the NT MD4 hash then des encryption. */ - -void +int SMBNTencrypt(unsigned char *passwd, unsigned char *c8, unsigned char *p24) { + int rc; unsigned char p21[21]; memset(p21, '\0', 21); - E_md4hash(passwd, p21); + rc = E_md4hash(passwd, p21); + if (rc) { + cFYI(1, "%s Can't generate NT hash, error: %d", __func__, rc); + return rc; + } SMBOWFencrypt(p21, c8, p24); + return rc; } diff --git a/fs/cifs/transport.c b/fs/cifs/transport.c index 59ca81b..46d8756f 100644 --- a/fs/cifs/transport.c +++ b/fs/cifs/transport.c @@ -36,7 +36,13 @@ extern mempool_t *cifs_mid_poolp; -static struct mid_q_entry * +static void +wake_up_task(struct mid_q_entry *mid) +{ + wake_up_process(mid->callback_data); +} + +struct mid_q_entry * AllocMidQEntry(const struct smb_hdr *smb_buffer, struct TCP_Server_Info *server) { struct mid_q_entry *temp; @@ -58,28 +64,28 @@ AllocMidQEntry(const struct smb_hdr *smb_buffer, struct TCP_Server_Info *server) /* do_gettimeofday(&temp->when_sent);*/ /* easier to use jiffies */ /* when mid allocated can be before when sent */ temp->when_alloc = jiffies; - temp->tsk = current; + + /* + * The default is for the mid to be synchronous, so the + * default callback just wakes up the current task. + */ + temp->callback = wake_up_task; + temp->callback_data = current; } - spin_lock(&GlobalMid_Lock); - list_add_tail(&temp->qhead, &server->pending_mid_q); atomic_inc(&midCount); temp->midState = MID_REQUEST_ALLOCATED; - spin_unlock(&GlobalMid_Lock); return temp; } -static void +void DeleteMidQEntry(struct mid_q_entry *midEntry) { #ifdef CONFIG_CIFS_STATS2 unsigned long now; #endif - spin_lock(&GlobalMid_Lock); midEntry->midState = MID_FREE; - list_del(&midEntry->qhead); atomic_dec(&midCount); - spin_unlock(&GlobalMid_Lock); if (midEntry->largeBuf) cifs_buf_release(midEntry->resp_buf); else @@ -103,6 +109,16 @@ DeleteMidQEntry(struct mid_q_entry *midEntry) mempool_free(midEntry, cifs_mid_poolp); } +static void +delete_mid(struct mid_q_entry *mid) +{ + spin_lock(&GlobalMid_Lock); + list_del(&mid->qhead); + spin_unlock(&GlobalMid_Lock); + + DeleteMidQEntry(mid); +} + static int smb_sendv(struct TCP_Server_Info *server, struct kvec *iov, int n_vec) { @@ -220,9 +236,9 @@ smb_sendv(struct TCP_Server_Info *server, struct kvec *iov, int n_vec) server->tcpStatus = CifsNeedReconnect; } - if (rc < 0) { + if (rc < 0 && rc != -EINTR) cERROR(1, "Error %d sending data on socket to server", rc); - } else + else rc = 0; /* Don't want to modify the buffer as a @@ -244,31 +260,31 @@ smb_send(struct TCP_Server_Info *server, struct smb_hdr *smb_buffer, return smb_sendv(server, &iov, 1); } -static int wait_for_free_request(struct cifsSesInfo *ses, const int long_op) +static int wait_for_free_request(struct TCP_Server_Info *server, + const int long_op) { if (long_op == CIFS_ASYNC_OP) { /* oplock breaks must not be held up */ - atomic_inc(&ses->server->inFlight); + atomic_inc(&server->inFlight); return 0; } spin_lock(&GlobalMid_Lock); while (1) { - if (atomic_read(&ses->server->inFlight) >= - cifs_max_pending){ + if (atomic_read(&server->inFlight) >= cifs_max_pending) { spin_unlock(&GlobalMid_Lock); #ifdef CONFIG_CIFS_STATS2 - atomic_inc(&ses->server->num_waiters); + atomic_inc(&server->num_waiters); #endif - wait_event(ses->server->request_q, - atomic_read(&ses->server->inFlight) + wait_event(server->request_q, + atomic_read(&server->inFlight) < cifs_max_pending); #ifdef CONFIG_CIFS_STATS2 - atomic_dec(&ses->server->num_waiters); + atomic_dec(&server->num_waiters); #endif spin_lock(&GlobalMid_Lock); } else { - if (ses->server->tcpStatus == CifsExiting) { + if (server->tcpStatus == CifsExiting) { spin_unlock(&GlobalMid_Lock); return -ENOENT; } @@ -278,7 +294,7 @@ static int wait_for_free_request(struct cifsSesInfo *ses, const int long_op) /* update # of requests on the wire to server */ if (long_op != CIFS_BLOCKING_OP) - atomic_inc(&ses->server->inFlight); + atomic_inc(&server->inFlight); spin_unlock(&GlobalMid_Lock); break; } @@ -308,53 +324,85 @@ static int allocate_mid(struct cifsSesInfo *ses, struct smb_hdr *in_buf, *ppmidQ = AllocMidQEntry(in_buf, ses->server); if (*ppmidQ == NULL) return -ENOMEM; + spin_lock(&GlobalMid_Lock); + list_add_tail(&(*ppmidQ)->qhead, &ses->server->pending_mid_q); + spin_unlock(&GlobalMid_Lock); return 0; } -static int wait_for_response(struct cifsSesInfo *ses, - struct mid_q_entry *midQ, - unsigned long timeout, - unsigned long time_to_wait) +static int +wait_for_response(struct TCP_Server_Info *server, struct mid_q_entry *midQ) { - unsigned long curr_timeout; + int error; - for (;;) { - curr_timeout = timeout + jiffies; - wait_event_timeout(ses->server->response_q, - midQ->midState != MID_REQUEST_SUBMITTED, timeout); + error = wait_event_killable(server->response_q, + midQ->midState != MID_REQUEST_SUBMITTED); + if (error < 0) + return -ERESTARTSYS; - if (time_after(jiffies, curr_timeout) && - (midQ->midState == MID_REQUEST_SUBMITTED) && - ((ses->server->tcpStatus == CifsGood) || - (ses->server->tcpStatus == CifsNew))) { + return 0; +} - unsigned long lrt; - /* We timed out. Is the server still - sending replies ? */ - spin_lock(&GlobalMid_Lock); - lrt = ses->server->lstrp; - spin_unlock(&GlobalMid_Lock); +/* + * Send a SMB request and set the callback function in the mid to handle + * the result. Caller is responsible for dealing with timeouts. + */ +int +cifs_call_async(struct TCP_Server_Info *server, struct smb_hdr *in_buf, + mid_callback_t *callback, void *cbdata) +{ + int rc; + struct mid_q_entry *mid; - /* Calculate time_to_wait past last receive time. - Although we prefer not to time out if the - server is still responding - we will time - out if the server takes more than 15 (or 45 - or 180) seconds to respond to this request - and has not responded to any request from - other threads on the client within 10 seconds */ - lrt += time_to_wait; - if (time_after(jiffies, lrt)) { - /* No replies for time_to_wait. */ - cERROR(1, "server not responding"); - return -1; - } - } else { - return 0; - } + rc = wait_for_free_request(server, CIFS_ASYNC_OP); + if (rc) + return rc; + + /* enable signing if server requires it */ + if (server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) + in_buf->Flags2 |= SMBFLG2_SECURITY_SIGNATURE; + + mutex_lock(&server->srv_mutex); + mid = AllocMidQEntry(in_buf, server); + if (mid == NULL) { + mutex_unlock(&server->srv_mutex); + return -ENOMEM; } -} + /* put it on the pending_mid_q */ + spin_lock(&GlobalMid_Lock); + list_add_tail(&mid->qhead, &server->pending_mid_q); + spin_unlock(&GlobalMid_Lock); + + rc = cifs_sign_smb(in_buf, server, &mid->sequence_number); + if (rc) { + mutex_unlock(&server->srv_mutex); + goto out_err; + } + + mid->callback = callback; + mid->callback_data = cbdata; + mid->midState = MID_REQUEST_SUBMITTED; +#ifdef CONFIG_CIFS_STATS2 + atomic_inc(&server->inSend); +#endif + rc = smb_send(server, in_buf, in_buf->smb_buf_length); +#ifdef CONFIG_CIFS_STATS2 + atomic_dec(&server->inSend); + mid->when_sent = jiffies; +#endif + mutex_unlock(&server->srv_mutex); + if (rc) + goto out_err; + + return rc; +out_err: + delete_mid(mid); + atomic_dec(&server->inFlight); + wake_up(&server->request_q); + return rc; +} /* * @@ -382,6 +430,84 @@ SendReceiveNoRsp(const unsigned int xid, struct cifsSesInfo *ses, return rc; } +static int +sync_mid_result(struct mid_q_entry *mid, struct TCP_Server_Info *server) +{ + int rc = 0; + + cFYI(1, "%s: cmd=%d mid=%d state=%d", __func__, mid->command, + mid->mid, mid->midState); + + spin_lock(&GlobalMid_Lock); + /* ensure that it's no longer on the pending_mid_q */ + list_del_init(&mid->qhead); + + switch (mid->midState) { + case MID_RESPONSE_RECEIVED: + spin_unlock(&GlobalMid_Lock); + return rc; + case MID_REQUEST_SUBMITTED: + /* socket is going down, reject all calls */ + if (server->tcpStatus == CifsExiting) { + cERROR(1, "%s: canceling mid=%d cmd=0x%x state=%d", + __func__, mid->mid, mid->command, mid->midState); + rc = -EHOSTDOWN; + break; + } + case MID_RETRY_NEEDED: + rc = -EAGAIN; + break; + case MID_RESPONSE_MALFORMED: + rc = -EIO; + break; + default: + cERROR(1, "%s: invalid mid state mid=%d state=%d", __func__, + mid->mid, mid->midState); + rc = -EIO; + } + spin_unlock(&GlobalMid_Lock); + + DeleteMidQEntry(mid); + return rc; +} + +/* + * An NT cancel request header looks just like the original request except: + * + * The Command is SMB_COM_NT_CANCEL + * The WordCount is zeroed out + * The ByteCount is zeroed out + * + * This function mangles an existing request buffer into a + * SMB_COM_NT_CANCEL request and then sends it. + */ +static int +send_nt_cancel(struct TCP_Server_Info *server, struct smb_hdr *in_buf, + struct mid_q_entry *mid) +{ + int rc = 0; + + /* -4 for RFC1001 length and +2 for BCC field */ + in_buf->smb_buf_length = sizeof(struct smb_hdr) - 4 + 2; + in_buf->Command = SMB_COM_NT_CANCEL; + in_buf->WordCount = 0; + put_bcc_le(0, in_buf); + + mutex_lock(&server->srv_mutex); + rc = cifs_sign_smb(in_buf, server, &mid->sequence_number); + if (rc) { + mutex_unlock(&server->srv_mutex); + return rc; + } + rc = smb_send(server, in_buf, in_buf->smb_buf_length); + mutex_unlock(&server->srv_mutex); + + cFYI(1, "issued NT_CANCEL for mid %u, rc = %d", + in_buf->Mid, rc); + + return rc; +} + int SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, struct kvec *iov, int n_vec, int *pRespBufType /* ret */, @@ -390,7 +516,6 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, int rc = 0; int long_op; unsigned int receive_len; - unsigned long timeout; struct mid_q_entry *midQ; struct smb_hdr *in_buf = iov[0].iov_base; @@ -413,7 +538,7 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, to the same server. We may make this configurable later or use ses->maxReq */ - rc = wait_for_free_request(ses, long_op); + rc = wait_for_free_request(ses->server, long_op); if (rc) { cifs_small_buf_release(in_buf); return rc; @@ -452,70 +577,41 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, #endif mutex_unlock(&ses->server->srv_mutex); - cifs_small_buf_release(in_buf); - if (rc < 0) - goto out; - - if (long_op == CIFS_STD_OP) - timeout = 15 * HZ; - else if (long_op == CIFS_VLONG_OP) /* e.g. slow writes past EOF */ - timeout = 180 * HZ; - else if (long_op == CIFS_LONG_OP) - timeout = 45 * HZ; /* should be greater than - servers oplock break timeout (about 43 seconds) */ - else if (long_op == CIFS_ASYNC_OP) - goto out; - else if (long_op == CIFS_BLOCKING_OP) - timeout = 0x7FFFFFFF; /* large, but not so large as to wrap */ - else { - cERROR(1, "unknown timeout flag %d", long_op); - rc = -EIO; + if (rc < 0) { + cifs_small_buf_release(in_buf); goto out; } - /* wait for 15 seconds or until woken up due to response arriving or - due to last connection to this server being unmounted */ - if (signal_pending(current)) { - /* if signal pending do not hold up user for full smb timeout - but we still give response a chance to complete */ - timeout = 2 * HZ; + if (long_op == CIFS_ASYNC_OP) { + cifs_small_buf_release(in_buf); + goto out; } - /* No user interrupts in wait - wreaks havoc with performance */ - wait_for_response(ses, midQ, timeout, 10 * HZ); - - spin_lock(&GlobalMid_Lock); - - if (midQ->resp_buf == NULL) { - cERROR(1, "No response to cmd %d mid %d", - midQ->command, midQ->mid); + rc = wait_for_response(ses->server, midQ); + if (rc != 0) { + send_nt_cancel(ses->server, in_buf, midQ); + spin_lock(&GlobalMid_Lock); if (midQ->midState == MID_REQUEST_SUBMITTED) { - if (ses->server->tcpStatus == CifsExiting) - rc = -EHOSTDOWN; - else { - ses->server->tcpStatus = CifsNeedReconnect; - midQ->midState = MID_RETRY_NEEDED; - } - } - - if (rc != -EHOSTDOWN) { - if (midQ->midState == MID_RETRY_NEEDED) { - rc = -EAGAIN; - cFYI(1, "marking request for retry"); - } else { - rc = -EIO; - } + midQ->callback = DeleteMidQEntry; + spin_unlock(&GlobalMid_Lock); + cifs_small_buf_release(in_buf); + atomic_dec(&ses->server->inFlight); + wake_up(&ses->server->request_q); + return rc; } spin_unlock(&GlobalMid_Lock); - DeleteMidQEntry(midQ); - /* Update # of requests on wire to server */ + } + + cifs_small_buf_release(in_buf); + + rc = sync_mid_result(midQ, ses->server); + if (rc != 0) { atomic_dec(&ses->server->inFlight); wake_up(&ses->server->request_q); return rc; } - spin_unlock(&GlobalMid_Lock); receive_len = midQ->resp_buf->smb_buf_length; if (receive_len > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE) { @@ -559,19 +655,18 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, if (receive_len >= sizeof(struct smb_hdr) - 4 /* do not count RFC1001 header */ + (2 * midQ->resp_buf->WordCount) + 2 /* bcc */ ) - BCC(midQ->resp_buf) = - le16_to_cpu(BCC_LE(midQ->resp_buf)); + put_bcc(get_bcc_le(midQ->resp_buf), midQ->resp_buf); if ((flags & CIFS_NO_RESP) == 0) midQ->resp_buf = NULL; /* mark it so buf will not be freed by - DeleteMidQEntry */ + delete_mid */ } else { rc = -EIO; cFYI(1, "Bad MID state?"); } out: - DeleteMidQEntry(midQ); + delete_mid(midQ); atomic_dec(&ses->server->inFlight); wake_up(&ses->server->request_q); @@ -585,7 +680,6 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, { int rc = 0; unsigned int receive_len; - unsigned long timeout; struct mid_q_entry *midQ; if (ses == NULL) { @@ -610,7 +704,7 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, return -EIO; } - rc = wait_for_free_request(ses, long_op); + rc = wait_for_free_request(ses->server, long_op); if (rc) return rc; @@ -649,64 +743,31 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, if (rc < 0) goto out; - if (long_op == CIFS_STD_OP) - timeout = 15 * HZ; - /* wait for 15 seconds or until woken up due to response arriving or - due to last connection to this server being unmounted */ - else if (long_op == CIFS_ASYNC_OP) - goto out; - else if (long_op == CIFS_VLONG_OP) /* writes past EOF can be slow */ - timeout = 180 * HZ; - else if (long_op == CIFS_LONG_OP) - timeout = 45 * HZ; /* should be greater than - servers oplock break timeout (about 43 seconds) */ - else if (long_op == CIFS_BLOCKING_OP) - timeout = 0x7FFFFFFF; /* large but no so large as to wrap */ - else { - cERROR(1, "unknown timeout flag %d", long_op); - rc = -EIO; + if (long_op == CIFS_ASYNC_OP) goto out; - } - - if (signal_pending(current)) { - /* if signal pending do not hold up user for full smb timeout - but we still give response a chance to complete */ - timeout = 2 * HZ; - } - - /* No user interrupts in wait - wreaks havoc with performance */ - wait_for_response(ses, midQ, timeout, 10 * HZ); - spin_lock(&GlobalMid_Lock); - if (midQ->resp_buf == NULL) { - cERROR(1, "No response for cmd %d mid %d", - midQ->command, midQ->mid); + rc = wait_for_response(ses->server, midQ); + if (rc != 0) { + send_nt_cancel(ses->server, in_buf, midQ); + spin_lock(&GlobalMid_Lock); if (midQ->midState == MID_REQUEST_SUBMITTED) { - if (ses->server->tcpStatus == CifsExiting) - rc = -EHOSTDOWN; - else { - ses->server->tcpStatus = CifsNeedReconnect; - midQ->midState = MID_RETRY_NEEDED; - } - } - - if (rc != -EHOSTDOWN) { - if (midQ->midState == MID_RETRY_NEEDED) { - rc = -EAGAIN; - cFYI(1, "marking request for retry"); - } else { - rc = -EIO; - } + /* no longer considered to be "in-flight" */ + midQ->callback = DeleteMidQEntry; + spin_unlock(&GlobalMid_Lock); + atomic_dec(&ses->server->inFlight); + wake_up(&ses->server->request_q); + return rc; } spin_unlock(&GlobalMid_Lock); - DeleteMidQEntry(midQ); - /* Update # of requests on wire to server */ + } + + rc = sync_mid_result(midQ, ses->server); + if (rc != 0) { atomic_dec(&ses->server->inFlight); wake_up(&ses->server->request_q); return rc; } - spin_unlock(&GlobalMid_Lock); receive_len = midQ->resp_buf->smb_buf_length; if (receive_len > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE) { @@ -748,43 +809,20 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, if (receive_len >= sizeof(struct smb_hdr) - 4 /* do not count RFC1001 header */ + (2 * out_buf->WordCount) + 2 /* bcc */ ) - BCC(out_buf) = le16_to_cpu(BCC_LE(out_buf)); + put_bcc(get_bcc_le(midQ->resp_buf), midQ->resp_buf); } else { rc = -EIO; cERROR(1, "Bad MID state?"); } out: - DeleteMidQEntry(midQ); + delete_mid(midQ); atomic_dec(&ses->server->inFlight); wake_up(&ses->server->request_q); return rc; } -/* Send an NT_CANCEL SMB to cause the POSIX blocking lock to return. */ - -static int -send_nt_cancel(struct cifsTconInfo *tcon, struct smb_hdr *in_buf, - struct mid_q_entry *midQ) -{ - int rc = 0; - struct cifsSesInfo *ses = tcon->ses; - __u16 mid = in_buf->Mid; - - header_assemble(in_buf, SMB_COM_NT_CANCEL, tcon, 0); - in_buf->Mid = mid; - mutex_lock(&ses->server->srv_mutex); - rc = cifs_sign_smb(in_buf, ses->server, &midQ->sequence_number); - if (rc) { - mutex_unlock(&ses->server->srv_mutex); - return rc; - } - rc = smb_send(ses->server, in_buf, in_buf->smb_buf_length); - mutex_unlock(&ses->server->srv_mutex); - return rc; -} - /* We send a LOCKINGX_CANCEL_LOCK to cause the Windows blocking lock to return. */ @@ -807,7 +845,7 @@ send_lock_cancel(const unsigned int xid, struct cifsTconInfo *tcon, pSMB->hdr.Mid = GetNextMid(ses->server); return SendReceive(xid, ses, in_buf, out_buf, - &bytes_returned, CIFS_STD_OP); + &bytes_returned, 0); } int @@ -845,7 +883,7 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, return -EIO; } - rc = wait_for_free_request(ses, CIFS_BLOCKING_OP); + rc = wait_for_free_request(ses->server, CIFS_BLOCKING_OP); if (rc) return rc; @@ -863,7 +901,7 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, rc = cifs_sign_smb(in_buf, ses->server, &midQ->sequence_number); if (rc) { - DeleteMidQEntry(midQ); + delete_mid(midQ); mutex_unlock(&ses->server->srv_mutex); return rc; } @@ -880,7 +918,7 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, mutex_unlock(&ses->server->srv_mutex); if (rc < 0) { - DeleteMidQEntry(midQ); + delete_mid(midQ); return rc; } @@ -899,10 +937,9 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, if (in_buf->Command == SMB_COM_TRANSACTION2) { /* POSIX lock. We send a NT_CANCEL SMB to cause the blocking lock to return. */ - - rc = send_nt_cancel(tcon, in_buf, midQ); + rc = send_nt_cancel(ses->server, in_buf, midQ); if (rc) { - DeleteMidQEntry(midQ); + delete_mid(midQ); return rc; } } else { @@ -914,47 +951,33 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, /* If we get -ENOLCK back the lock may have already been removed. Don't exit in this case. */ if (rc && rc != -ENOLCK) { - DeleteMidQEntry(midQ); + delete_mid(midQ); return rc; } } - /* Wait 5 seconds for the response. */ - if (wait_for_response(ses, midQ, 5 * HZ, 5 * HZ) == 0) { - /* We got the response - restart system call. */ - rstart = 1; - } - } - - spin_lock(&GlobalMid_Lock); - if (midQ->resp_buf) { - spin_unlock(&GlobalMid_Lock); - receive_len = midQ->resp_buf->smb_buf_length; - } else { - cERROR(1, "No response for cmd %d mid %d", - midQ->command, midQ->mid); - if (midQ->midState == MID_REQUEST_SUBMITTED) { - if (ses->server->tcpStatus == CifsExiting) - rc = -EHOSTDOWN; - else { - ses->server->tcpStatus = CifsNeedReconnect; - midQ->midState = MID_RETRY_NEEDED; + rc = wait_for_response(ses->server, midQ); + if (rc) { + send_nt_cancel(ses->server, in_buf, midQ); + spin_lock(&GlobalMid_Lock); + if (midQ->midState == MID_REQUEST_SUBMITTED) { + /* no longer considered to be "in-flight" */ + midQ->callback = DeleteMidQEntry; + spin_unlock(&GlobalMid_Lock); + return rc; } + spin_unlock(&GlobalMid_Lock); } - if (rc != -EHOSTDOWN) { - if (midQ->midState == MID_RETRY_NEEDED) { - rc = -EAGAIN; - cFYI(1, "marking request for retry"); - } else { - rc = -EIO; - } - } - spin_unlock(&GlobalMid_Lock); - DeleteMidQEntry(midQ); - return rc; + /* We got the response - restart system call. */ + rstart = 1; } + rc = sync_mid_result(midQ, ses->server); + if (rc != 0) + return rc; + + receive_len = midQ->resp_buf->smb_buf_length; if (receive_len > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE) { cERROR(1, "Frame too large received. Length: %d Xid: %d", receive_len, xid); @@ -998,10 +1021,10 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, if (receive_len >= sizeof(struct smb_hdr) - 4 /* do not count RFC1001 header */ + (2 * out_buf->WordCount) + 2 /* bcc */ ) - BCC(out_buf) = le16_to_cpu(BCC_LE(out_buf)); + put_bcc(get_bcc_le(out_buf), out_buf); out: - DeleteMidQEntry(midQ); + delete_mid(midQ); if (rstart && rc == -EACCES) return -ERESTARTSYS; return rc; diff --git a/fs/coda/Makefile b/fs/coda/Makefile index 6c22e61..1bab69a 100644 --- a/fs/coda/Makefile +++ b/fs/coda/Makefile @@ -9,4 +9,4 @@ coda-objs := psdev.o cache.o cnode.o inode.o dir.o file.o upcall.o \ # If you want debugging output, please uncomment the following line. -# EXTRA_CFLAGS += -DDEBUG -DDEBUG_SMB_MALLOC=1 +# ccflags-y := -DDEBUG -DDEBUG_SMB_MALLOC=1 diff --git a/fs/compat.c b/fs/compat.c index f6fd0a0..72fe6cd 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -262,35 +262,19 @@ static int put_compat_statfs(struct compat_statfs __user *ubuf, struct kstatfs * */ asmlinkage long compat_sys_statfs(const char __user *pathname, struct compat_statfs __user *buf) { - struct path path; - int error; - - error = user_path(pathname, &path); - if (!error) { - struct kstatfs tmp; - error = vfs_statfs(&path, &tmp); - if (!error) - error = put_compat_statfs(buf, &tmp); - path_put(&path); - } + struct kstatfs tmp; + int error = user_statfs(pathname, &tmp); + if (!error) + error = put_compat_statfs(buf, &tmp); return error; } asmlinkage long compat_sys_fstatfs(unsigned int fd, struct compat_statfs __user *buf) { - struct file * file; struct kstatfs tmp; - int error; - - error = -EBADF; - file = fget(fd); - if (!file) - goto out; - error = vfs_statfs(&file->f_path, &tmp); + int error = fd_statfs(fd, &tmp); if (!error) error = put_compat_statfs(buf, &tmp); - fput(file); -out: return error; } @@ -329,41 +313,29 @@ static int put_compat_statfs64(struct compat_statfs64 __user *ubuf, struct kstat asmlinkage long compat_sys_statfs64(const char __user *pathname, compat_size_t sz, struct compat_statfs64 __user *buf) { - struct path path; + struct kstatfs tmp; int error; if (sz != sizeof(*buf)) return -EINVAL; - error = user_path(pathname, &path); - if (!error) { - struct kstatfs tmp; - error = vfs_statfs(&path, &tmp); - if (!error) - error = put_compat_statfs64(buf, &tmp); - path_put(&path); - } + error = user_statfs(pathname, &tmp); + if (!error) + error = put_compat_statfs64(buf, &tmp); return error; } asmlinkage long compat_sys_fstatfs64(unsigned int fd, compat_size_t sz, struct compat_statfs64 __user *buf) { - struct file * file; struct kstatfs tmp; int error; if (sz != sizeof(*buf)) return -EINVAL; - error = -EBADF; - file = fget(fd); - if (!file) - goto out; - error = vfs_statfs(&file->f_path, &tmp); + error = fd_statfs(fd, &tmp); if (!error) error = put_compat_statfs64(buf, &tmp); - fput(file); -out: return error; } @@ -1228,7 +1200,9 @@ compat_sys_preadv(unsigned long fd, const struct compat_iovec __user *vec, file = fget_light(fd, &fput_needed); if (!file) return -EBADF; - ret = compat_readv(file, vec, vlen, &pos); + ret = -ESPIPE; + if (file->f_mode & FMODE_PREAD) + ret = compat_readv(file, vec, vlen, &pos); fput_light(file, fput_needed); return ret; } @@ -1285,7 +1259,9 @@ compat_sys_pwritev(unsigned long fd, const struct compat_iovec __user *vec, file = fget_light(fd, &fput_needed); if (!file) return -EBADF; - ret = compat_writev(file, vec, vlen, &pos); + ret = -ESPIPE; + if (file->f_mode & FMODE_PWRITE) + ret = compat_writev(file, vec, vlen, &pos); fput_light(file, fput_needed); return ret; } @@ -1695,9 +1671,6 @@ int compat_set_fd_set(unsigned long nr, compat_ulong_t __user *ufdset, * Update: ERESTARTSYS breaks at least the xview clock binary, so * I'm trying ERESTARTNOHAND which restart only when you want to. */ -#define MAX_SELECT_SECONDS \ - ((unsigned long) (MAX_SCHEDULE_TIMEOUT / HZ)-1) - int compat_core_sys_select(int n, compat_ulong_t __user *inp, compat_ulong_t __user *outp, compat_ulong_t __user *exp, struct timespec *end_time) @@ -2308,3 +2281,16 @@ asmlinkage long compat_sys_timerfd_gettime(int ufd, } #endif /* CONFIG_TIMERFD */ + +#ifdef CONFIG_FHANDLE +/* + * Exactly like fs/open.c:sys_open_by_handle_at(), except that it + * doesn't set the O_LARGEFILE flag. + */ +asmlinkage long +compat_sys_open_by_handle_at(int mountdirfd, + struct file_handle __user *handle, int flags) +{ + return do_handle_open(mountdirfd, handle, flags); +} +#endif diff --git a/fs/dcache.c b/fs/dcache.c index 9f493ee..ad25c4c 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -176,6 +176,7 @@ static void d_free(struct dentry *dentry) /** * dentry_rcuwalk_barrier - invalidate in-progress rcu-walk lookups + * @dentry: the target dentry * After this call, in-progress rcu-walk path lookup will fail. This * should be called after unhashing, and after changing d_inode (if * the dentry has not already been unhashed). @@ -281,6 +282,7 @@ static void dentry_lru_move_tail(struct dentry *dentry) /** * d_kill - kill dentry and return parent * @dentry: dentry to kill + * @parent: parent dentry * * The dentry must already be unhashed and removed from the LRU. * @@ -294,8 +296,12 @@ static struct dentry *d_kill(struct dentry *dentry, struct dentry *parent) __releases(parent->d_lock) __releases(dentry->d_inode->i_lock) { - dentry->d_parent = NULL; list_del(&dentry->d_u.d_child); + /* + * Inform try_to_ascend() that we are no longer attached to the + * dentry tree + */ + dentry->d_flags |= DCACHE_DISCONNECTED; if (parent) spin_unlock(&parent->d_lock); dentry_iput(dentry); @@ -1010,6 +1016,35 @@ void shrink_dcache_for_umount(struct super_block *sb) } /* + * This tries to ascend one level of parenthood, but + * we can race with renaming, so we need to re-check + * the parenthood after dropping the lock and check + * that the sequence number still matches. + */ +static struct dentry *try_to_ascend(struct dentry *old, int locked, unsigned seq) +{ + struct dentry *new = old->d_parent; + + rcu_read_lock(); + spin_unlock(&old->d_lock); + spin_lock(&new->d_lock); + + /* + * might go back up the wrong parent if we have had a rename + * or deletion + */ + if (new != old->d_parent || + (old->d_flags & DCACHE_DISCONNECTED) || + (!locked && read_seqretry(&rename_lock, seq))) { + spin_unlock(&new->d_lock); + new = NULL; + } + rcu_read_unlock(); + return new; +} + + +/* * Search for at least 1 mount point in the dentry's subdirs. * We descend to the next level whenever the d_subdirs * list is non-empty and continue searching. @@ -1064,24 +1099,10 @@ resume: * All done at this level ... ascend and resume the search. */ if (this_parent != parent) { - struct dentry *tmp; - struct dentry *child; - - tmp = this_parent->d_parent; - rcu_read_lock(); - spin_unlock(&this_parent->d_lock); - child = this_parent; - this_parent = tmp; - spin_lock(&this_parent->d_lock); - /* might go back up the wrong parent if we have had a rename - * or deletion */ - if (this_parent != child->d_parent || - (!locked && read_seqretry(&rename_lock, seq))) { - spin_unlock(&this_parent->d_lock); - rcu_read_unlock(); + struct dentry *child = this_parent; + this_parent = try_to_ascend(this_parent, locked, seq); + if (!this_parent) goto rename_retry; - } - rcu_read_unlock(); next = child->d_u.d_child.next; goto resume; } @@ -1179,24 +1200,10 @@ resume: * All done at this level ... ascend and resume the search. */ if (this_parent != parent) { - struct dentry *tmp; - struct dentry *child; - - tmp = this_parent->d_parent; - rcu_read_lock(); - spin_unlock(&this_parent->d_lock); - child = this_parent; - this_parent = tmp; - spin_lock(&this_parent->d_lock); - /* might go back up the wrong parent if we have had a rename - * or deletion */ - if (this_parent != child->d_parent || - (!locked && read_seqretry(&rename_lock, seq))) { - spin_unlock(&this_parent->d_lock); - rcu_read_unlock(); + struct dentry *child = this_parent; + this_parent = try_to_ascend(this_parent, locked, seq); + if (!this_parent) goto rename_retry; - } - rcu_read_unlock(); next = child->d_u.d_child.next; goto resume; } @@ -1521,6 +1528,28 @@ struct dentry * d_alloc_root(struct inode * root_inode) } EXPORT_SYMBOL(d_alloc_root); +static struct dentry * __d_find_any_alias(struct inode *inode) +{ + struct dentry *alias; + + if (list_empty(&inode->i_dentry)) + return NULL; + alias = list_first_entry(&inode->i_dentry, struct dentry, d_alias); + __dget(alias); + return alias; +} + +static struct dentry * d_find_any_alias(struct inode *inode) +{ + struct dentry *de; + + spin_lock(&inode->i_lock); + de = __d_find_any_alias(inode); + spin_unlock(&inode->i_lock); + return de; +} + + /** * d_obtain_alias - find or allocate a dentry for a given inode * @inode: inode to allocate the dentry for @@ -1550,7 +1579,7 @@ struct dentry *d_obtain_alias(struct inode *inode) if (IS_ERR(inode)) return ERR_CAST(inode); - res = d_find_alias(inode); + res = d_find_any_alias(inode); if (res) goto out_iput; @@ -1563,7 +1592,7 @@ struct dentry *d_obtain_alias(struct inode *inode) spin_lock(&inode->i_lock); - res = __d_find_alias(inode, 0); + res = __d_find_any_alias(inode); if (res) { spin_unlock(&inode->i_lock); dput(tmp); @@ -1583,10 +1612,13 @@ struct dentry *d_obtain_alias(struct inode *inode) __bit_spin_unlock(0, (unsigned long *)&tmp->d_sb->s_anon.first); spin_unlock(&tmp->d_lock); spin_unlock(&inode->i_lock); + security_d_instantiate(tmp, inode); return tmp; out_iput: + if (res && !IS_ERR(res)) + security_d_instantiate(res, inode); iput(inode); return res; } @@ -1779,7 +1811,7 @@ struct dentry *__d_lookup_rcu(struct dentry *parent, struct qstr *name, * false-negative result. d_lookup() protects against concurrent * renames using rename_lock seqlock. * - * See Documentation/vfs/dcache-locking.txt for more details. + * See Documentation/filesystems/path-lookup.txt for more details. */ hlist_bl_for_each_entry_rcu(dentry, node, &b->head, d_hash) { struct inode *i; @@ -1899,7 +1931,7 @@ struct dentry *__d_lookup(struct dentry *parent, struct qstr *name) * false-negative result. d_lookup() protects against concurrent * renames using rename_lock seqlock. * - * See Documentation/vfs/dcache-locking.txt for more details. + * See Documentation/filesystems/path-lookup.txt for more details. */ rcu_read_lock(); @@ -1973,7 +2005,7 @@ out: /** * d_validate - verify dentry provided from insecure source (deprecated) * @dentry: The dentry alleged to be valid child of @dparent - * @parent: The parent dentry (known to be valid) + * @dparent: The parent dentry (known to be valid) * * An insecure source has sent us a dentry, here we verify it and dget() it. * This is used by ncpfs in its readdir implementation. @@ -2918,28 +2950,14 @@ resume: spin_unlock(&dentry->d_lock); } if (this_parent != root) { - struct dentry *tmp; - struct dentry *child; - - tmp = this_parent->d_parent; + struct dentry *child = this_parent; if (!(this_parent->d_flags & DCACHE_GENOCIDE)) { this_parent->d_flags |= DCACHE_GENOCIDE; this_parent->d_count--; } - rcu_read_lock(); - spin_unlock(&this_parent->d_lock); - child = this_parent; - this_parent = tmp; - spin_lock(&this_parent->d_lock); - /* might go back up the wrong parent if we have had a rename - * or deletion */ - if (this_parent != child->d_parent || - (!locked && read_seqretry(&rename_lock, seq))) { - spin_unlock(&this_parent->d_lock); - rcu_read_unlock(); + this_parent = try_to_ascend(this_parent, locked, seq); + if (!this_parent) goto rename_retry; - } - rcu_read_unlock(); next = child->d_u.d_child.next; goto resume; } diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index 37a8ca7..e7a7a2f 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -13,9 +13,6 @@ * */ -/* uncomment to get debug messages from the debug filesystem, ah the irony. */ -/* #define DEBUG */ - #include <linux/module.h> #include <linux/fs.h> #include <linux/mount.h> @@ -310,7 +307,7 @@ struct dentry *debugfs_create_symlink(const char *name, struct dentry *parent, } EXPORT_SYMBOL_GPL(debugfs_create_symlink); -static void __debugfs_remove(struct dentry *dentry, struct dentry *parent) +static int __debugfs_remove(struct dentry *dentry, struct dentry *parent) { int ret = 0; @@ -333,6 +330,7 @@ static void __debugfs_remove(struct dentry *dentry, struct dentry *parent) dput(dentry); } } + return ret; } /** @@ -351,7 +349,8 @@ static void __debugfs_remove(struct dentry *dentry, struct dentry *parent) void debugfs_remove(struct dentry *dentry) { struct dentry *parent; - + int ret; + if (!dentry) return; @@ -360,9 +359,10 @@ void debugfs_remove(struct dentry *dentry) return; mutex_lock(&parent->d_inode->i_mutex); - __debugfs_remove(dentry, parent); + ret = __debugfs_remove(dentry, parent); mutex_unlock(&parent->d_inode->i_mutex); - simple_release_fs(&debugfs_mount, &debugfs_mount_count); + if (!ret) + simple_release_fs(&debugfs_mount, &debugfs_mount_count); } EXPORT_SYMBOL_GPL(debugfs_remove); @@ -540,17 +540,5 @@ static int __init debugfs_init(void) return retval; } - -static void __exit debugfs_exit(void) -{ - debugfs_registered = false; - - simple_release_fs(&debugfs_mount, &debugfs_mount_count); - unregister_filesystem(&debug_fs_type); - kobject_put(debug_kobj); -} - core_initcall(debugfs_init); -module_exit(debugfs_exit); -MODULE_LICENSE("GPL"); diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c index 1bb547c..c6bd815 100644 --- a/fs/devpts/inode.c +++ b/fs/devpts/inode.c @@ -479,6 +479,7 @@ int devpts_pty_new(struct inode *ptmx_inode, struct tty_struct *tty) struct dentry *root = sb->s_root; struct pts_fs_info *fsi = DEVPTS_SB(sb); struct pts_mount_opts *opts = &fsi->mount_opts; + int ret = 0; char s[12]; /* We're supposed to be given the slave end of a pty */ @@ -504,11 +505,14 @@ int devpts_pty_new(struct inode *ptmx_inode, struct tty_struct *tty) if (!IS_ERR(dentry)) { d_add(dentry, inode); fsnotify_create(root->d_inode, dentry); + } else { + iput(inode); + ret = -ENOMEM; } mutex_unlock(&root->d_inode->i_mutex); - return 0; + return ret; } struct tty_struct *devpts_get_tty(struct inode *pts_inode, int number) @@ -544,17 +548,12 @@ void devpts_pty_kill(struct tty_struct *tty) mutex_lock(&root->d_inode->i_mutex); dentry = d_find_alias(inode); - if (IS_ERR(dentry)) - goto out; - - if (dentry) { - inode->i_nlink--; - d_delete(dentry); - dput(dentry); /* d_alloc_name() in devpts_pty_new() */ - } + inode->i_nlink--; + d_delete(dentry); + dput(dentry); /* d_alloc_name() in devpts_pty_new() */ dput(dentry); /* d_find_alias above */ -out: + mutex_unlock(&root->d_inode->i_mutex); } diff --git a/fs/direct-io.c b/fs/direct-io.c index 85882f6..dcb5577 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -325,12 +325,16 @@ void dio_end_io(struct bio *bio, int error) } EXPORT_SYMBOL_GPL(dio_end_io); -static int +static void dio_bio_alloc(struct dio *dio, struct block_device *bdev, sector_t first_sector, int nr_vecs) { struct bio *bio; + /* + * bio_alloc() is guaranteed to return a bio when called with + * __GFP_WAIT and we request a valid number of vectors. + */ bio = bio_alloc(GFP_KERNEL, nr_vecs); bio->bi_bdev = bdev; @@ -342,7 +346,6 @@ dio_bio_alloc(struct dio *dio, struct block_device *bdev, dio->bio = bio; dio->logical_offset_in_bio = dio->cur_page_fs_offset; - return 0; } /* @@ -583,8 +586,9 @@ static int dio_new_bio(struct dio *dio, sector_t start_sector) goto out; sector = start_sector << (dio->blkbits - 9); nr_pages = min(dio->pages_in_io, bio_get_nr_vecs(dio->map_bh.b_bdev)); + nr_pages = min(nr_pages, BIO_MAX_PAGES); BUG_ON(nr_pages <= 0); - ret = dio_bio_alloc(dio, dio->map_bh.b_bdev, sector, nr_pages); + dio_bio_alloc(dio, dio->map_bh.b_bdev, sector, nr_pages); dio->boundary = 0; out: return ret; @@ -641,11 +645,11 @@ static int dio_send_cur_page(struct dio *dio) /* * See whether this new request is contiguous with the old. * - * Btrfs cannot handl having logically non-contiguous requests - * submitted. For exmple if you have + * Btrfs cannot handle having logically non-contiguous requests + * submitted. For example if you have * * Logical: [0-4095][HOLE][8192-12287] - * Phyiscal: [0-4095] [4096-8181] + * Physical: [0-4095] [4096-8191] * * We cannot submit those pages together as one BIO. So if our * current logical offset in the file does not equal what would diff --git a/fs/dlm/ast.c b/fs/dlm/ast.c index 4314f0d..abc49f2 100644 --- a/fs/dlm/ast.c +++ b/fs/dlm/ast.c @@ -18,6 +18,7 @@ #define WAKE_ASTS 0 +static uint64_t ast_seq_count; static struct list_head ast_queue; static spinlock_t ast_queue_lock; static struct task_struct * astd_task; @@ -25,40 +26,186 @@ static unsigned long astd_wakeflags; static struct mutex astd_running; +static void dlm_dump_lkb_callbacks(struct dlm_lkb *lkb) +{ + int i; + + log_print("last_bast %x %llu flags %x mode %d sb %d %x", + lkb->lkb_id, + (unsigned long long)lkb->lkb_last_bast.seq, + lkb->lkb_last_bast.flags, + lkb->lkb_last_bast.mode, + lkb->lkb_last_bast.sb_status, + lkb->lkb_last_bast.sb_flags); + + log_print("last_cast %x %llu flags %x mode %d sb %d %x", + lkb->lkb_id, + (unsigned long long)lkb->lkb_last_cast.seq, + lkb->lkb_last_cast.flags, + lkb->lkb_last_cast.mode, + lkb->lkb_last_cast.sb_status, + lkb->lkb_last_cast.sb_flags); + + for (i = 0; i < DLM_CALLBACKS_SIZE; i++) { + log_print("cb %x %llu flags %x mode %d sb %d %x", + lkb->lkb_id, + (unsigned long long)lkb->lkb_callbacks[i].seq, + lkb->lkb_callbacks[i].flags, + lkb->lkb_callbacks[i].mode, + lkb->lkb_callbacks[i].sb_status, + lkb->lkb_callbacks[i].sb_flags); + } +} + void dlm_del_ast(struct dlm_lkb *lkb) { spin_lock(&ast_queue_lock); - if (lkb->lkb_ast_type & (AST_COMP | AST_BAST)) - list_del(&lkb->lkb_astqueue); + if (!list_empty(&lkb->lkb_astqueue)) + list_del_init(&lkb->lkb_astqueue); spin_unlock(&ast_queue_lock); } -void dlm_add_ast(struct dlm_lkb *lkb, int type, int mode) +int dlm_add_lkb_callback(struct dlm_lkb *lkb, uint32_t flags, int mode, + int status, uint32_t sbflags, uint64_t seq) { + struct dlm_ls *ls = lkb->lkb_resource->res_ls; + uint64_t prev_seq; + int prev_mode; + int i; + + for (i = 0; i < DLM_CALLBACKS_SIZE; i++) { + if (lkb->lkb_callbacks[i].seq) + continue; + + /* + * Suppress some redundant basts here, do more on removal. + * Don't even add a bast if the callback just before it + * is a bast for the same mode or a more restrictive mode. + * (the addional > PR check is needed for PR/CW inversion) + */ + + if ((i > 0) && (flags & DLM_CB_BAST) && + (lkb->lkb_callbacks[i-1].flags & DLM_CB_BAST)) { + + prev_seq = lkb->lkb_callbacks[i-1].seq; + prev_mode = lkb->lkb_callbacks[i-1].mode; + + if ((prev_mode == mode) || + (prev_mode > mode && prev_mode > DLM_LOCK_PR)) { + + log_debug(ls, "skip %x add bast %llu mode %d " + "for bast %llu mode %d", + lkb->lkb_id, + (unsigned long long)seq, + mode, + (unsigned long long)prev_seq, + prev_mode); + return 0; + } + } + + lkb->lkb_callbacks[i].seq = seq; + lkb->lkb_callbacks[i].flags = flags; + lkb->lkb_callbacks[i].mode = mode; + lkb->lkb_callbacks[i].sb_status = status; + lkb->lkb_callbacks[i].sb_flags = (sbflags & 0x000000FF); + break; + } + + if (i == DLM_CALLBACKS_SIZE) { + log_error(ls, "no callbacks %x %llu flags %x mode %d sb %d %x", + lkb->lkb_id, (unsigned long long)seq, + flags, mode, status, sbflags); + dlm_dump_lkb_callbacks(lkb); + return -1; + } + + return 0; +} + +int dlm_rem_lkb_callback(struct dlm_ls *ls, struct dlm_lkb *lkb, + struct dlm_callback *cb, int *resid) +{ + int i; + + *resid = 0; + + if (!lkb->lkb_callbacks[0].seq) + return -ENOENT; + + /* oldest undelivered cb is callbacks[0] */ + + memcpy(cb, &lkb->lkb_callbacks[0], sizeof(struct dlm_callback)); + memset(&lkb->lkb_callbacks[0], 0, sizeof(struct dlm_callback)); + + /* shift others down */ + + for (i = 1; i < DLM_CALLBACKS_SIZE; i++) { + if (!lkb->lkb_callbacks[i].seq) + break; + memcpy(&lkb->lkb_callbacks[i-1], &lkb->lkb_callbacks[i], + sizeof(struct dlm_callback)); + memset(&lkb->lkb_callbacks[i], 0, sizeof(struct dlm_callback)); + (*resid)++; + } + + /* if cb is a bast, it should be skipped if the blocking mode is + compatible with the last granted mode */ + + if ((cb->flags & DLM_CB_BAST) && lkb->lkb_last_cast.seq) { + if (dlm_modes_compat(cb->mode, lkb->lkb_last_cast.mode)) { + cb->flags |= DLM_CB_SKIP; + + log_debug(ls, "skip %x bast %llu mode %d " + "for cast %llu mode %d", + lkb->lkb_id, + (unsigned long long)cb->seq, + cb->mode, + (unsigned long long)lkb->lkb_last_cast.seq, + lkb->lkb_last_cast.mode); + return 0; + } + } + + if (cb->flags & DLM_CB_CAST) { + memcpy(&lkb->lkb_last_cast, cb, sizeof(struct dlm_callback)); + lkb->lkb_last_cast_time = ktime_get(); + } + + if (cb->flags & DLM_CB_BAST) { + memcpy(&lkb->lkb_last_bast, cb, sizeof(struct dlm_callback)); + lkb->lkb_last_bast_time = ktime_get(); + } + + return 0; +} + +void dlm_add_ast(struct dlm_lkb *lkb, uint32_t flags, int mode, int status, + uint32_t sbflags) +{ + uint64_t seq; + int rv; + + spin_lock(&ast_queue_lock); + + seq = ++ast_seq_count; + if (lkb->lkb_flags & DLM_IFL_USER) { - dlm_user_add_ast(lkb, type, mode); + spin_unlock(&ast_queue_lock); + dlm_user_add_ast(lkb, flags, mode, status, sbflags, seq); return; } - spin_lock(&ast_queue_lock); - if (!(lkb->lkb_ast_type & (AST_COMP | AST_BAST))) { + rv = dlm_add_lkb_callback(lkb, flags, mode, status, sbflags, seq); + if (rv < 0) { + spin_unlock(&ast_queue_lock); + return; + } + + if (list_empty(&lkb->lkb_astqueue)) { kref_get(&lkb->lkb_ref); list_add_tail(&lkb->lkb_astqueue, &ast_queue); - lkb->lkb_ast_first = type; } - - /* sanity check, this should not happen */ - - if ((type == AST_COMP) && (lkb->lkb_ast_type & AST_COMP)) - log_print("repeat cast %d castmode %d lock %x %s", - mode, lkb->lkb_castmode, - lkb->lkb_id, lkb->lkb_resource->res_name); - - lkb->lkb_ast_type |= type; - if (type == AST_BAST) - lkb->lkb_bastmode = mode; - else - lkb->lkb_castmode = mode; spin_unlock(&ast_queue_lock); set_bit(WAKE_ASTS, &astd_wakeflags); @@ -72,7 +219,8 @@ static void process_asts(void) struct dlm_lkb *lkb; void (*castfn) (void *astparam); void (*bastfn) (void *astparam, int mode); - int type, first, bastmode, castmode, do_bast, do_cast, last_castmode; + struct dlm_callback callbacks[DLM_CALLBACKS_SIZE]; + int i, rv, resid; repeat: spin_lock(&ast_queue_lock); @@ -83,54 +231,45 @@ repeat: if (dlm_locking_stopped(ls)) continue; - list_del(&lkb->lkb_astqueue); - type = lkb->lkb_ast_type; - lkb->lkb_ast_type = 0; - first = lkb->lkb_ast_first; - lkb->lkb_ast_first = 0; - bastmode = lkb->lkb_bastmode; - castmode = lkb->lkb_castmode; + /* we remove from astqueue list and remove everything in + lkb_callbacks before releasing the spinlock so empty + lkb_astqueue is always consistent with empty lkb_callbacks */ + + list_del_init(&lkb->lkb_astqueue); + castfn = lkb->lkb_astfn; bastfn = lkb->lkb_bastfn; - spin_unlock(&ast_queue_lock); - do_cast = (type & AST_COMP) && castfn; - do_bast = (type & AST_BAST) && bastfn; + memset(&callbacks, 0, sizeof(callbacks)); - /* Skip a bast if its blocking mode is compatible with the - granted mode of the preceding cast. */ + for (i = 0; i < DLM_CALLBACKS_SIZE; i++) { + rv = dlm_rem_lkb_callback(ls, lkb, &callbacks[i], &resid); + if (rv < 0) + break; + } + spin_unlock(&ast_queue_lock); - if (do_bast) { - if (first == AST_COMP) - last_castmode = castmode; - else - last_castmode = lkb->lkb_castmode_done; - if (dlm_modes_compat(bastmode, last_castmode)) - do_bast = 0; + if (resid) { + /* shouldn't happen, for loop should have removed all */ + log_error(ls, "callback resid %d lkb %x", + resid, lkb->lkb_id); } - if (first == AST_COMP) { - if (do_cast) - castfn(lkb->lkb_astparam); - if (do_bast) - bastfn(lkb->lkb_astparam, bastmode); - } else if (first == AST_BAST) { - if (do_bast) - bastfn(lkb->lkb_astparam, bastmode); - if (do_cast) + for (i = 0; i < DLM_CALLBACKS_SIZE; i++) { + if (!callbacks[i].seq) + break; + if (callbacks[i].flags & DLM_CB_SKIP) { + continue; + } else if (callbacks[i].flags & DLM_CB_BAST) { + bastfn(lkb->lkb_astparam, callbacks[i].mode); + } else if (callbacks[i].flags & DLM_CB_CAST) { + lkb->lkb_lksb->sb_status = callbacks[i].sb_status; + lkb->lkb_lksb->sb_flags = callbacks[i].sb_flags; castfn(lkb->lkb_astparam); - } else { - log_error(ls, "bad ast_first %d ast_type %d", - first, type); + } } - if (do_cast) - lkb->lkb_castmode_done = castmode; - if (do_bast) - lkb->lkb_bastmode_done = bastmode; - - /* this removes the reference added by dlm_add_ast - and may result in the lkb being freed */ + /* removes ref for ast_queue, may cause lkb to be freed */ dlm_put_lkb(lkb); cond_resched(); diff --git a/fs/dlm/ast.h b/fs/dlm/ast.h index bcb1aab..8aa89c9 100644 --- a/fs/dlm/ast.h +++ b/fs/dlm/ast.h @@ -13,8 +13,13 @@ #ifndef __ASTD_DOT_H__ #define __ASTD_DOT_H__ -void dlm_add_ast(struct dlm_lkb *lkb, int type, int mode); void dlm_del_ast(struct dlm_lkb *lkb); +int dlm_add_lkb_callback(struct dlm_lkb *lkb, uint32_t flags, int mode, + int status, uint32_t sbflags, uint64_t seq); +int dlm_rem_lkb_callback(struct dlm_ls *ls, struct dlm_lkb *lkb, + struct dlm_callback *cb, int *resid); +void dlm_add_ast(struct dlm_lkb *lkb, uint32_t flags, int mode, int status, + uint32_t sbflags); void dlm_astd_wake(void); int dlm_astd_start(void); diff --git a/fs/dlm/config.c b/fs/dlm/config.c index b54bca0..0d329ff 100644 --- a/fs/dlm/config.c +++ b/fs/dlm/config.c @@ -977,9 +977,9 @@ int dlm_our_addr(struct sockaddr_storage *addr, int num) /* Config file defaults */ #define DEFAULT_TCP_PORT 21064 #define DEFAULT_BUFFER_SIZE 4096 -#define DEFAULT_RSBTBL_SIZE 256 +#define DEFAULT_RSBTBL_SIZE 1024 #define DEFAULT_LKBTBL_SIZE 1024 -#define DEFAULT_DIRTBL_SIZE 512 +#define DEFAULT_DIRTBL_SIZE 1024 #define DEFAULT_RECOVER_TIMER 5 #define DEFAULT_TOSS_SECS 10 #define DEFAULT_SCAN_SECS 5 diff --git a/fs/dlm/debug_fs.c b/fs/dlm/debug_fs.c index 6b42ba8..5977923 100644 --- a/fs/dlm/debug_fs.c +++ b/fs/dlm/debug_fs.c @@ -257,12 +257,12 @@ static int print_format3_lock(struct seq_file *s, struct dlm_lkb *lkb, lkb->lkb_status, lkb->lkb_grmode, lkb->lkb_rqmode, - lkb->lkb_bastmode, + lkb->lkb_last_bast.mode, rsb_lookup, lkb->lkb_wait_type, lkb->lkb_lvbseq, (unsigned long long)ktime_to_ns(lkb->lkb_timestamp), - (unsigned long long)ktime_to_ns(lkb->lkb_time_bast)); + (unsigned long long)ktime_to_ns(lkb->lkb_last_bast_time)); return rv; } diff --git a/fs/dlm/dlm_internal.h b/fs/dlm/dlm_internal.h index f632b58..b942049 100644 --- a/fs/dlm/dlm_internal.h +++ b/fs/dlm/dlm_internal.h @@ -192,11 +192,6 @@ struct dlm_args { * lkb is a process copy, the nodeid specifies the lock master. */ -/* lkb_ast_type */ - -#define AST_COMP 1 -#define AST_BAST 2 - /* lkb_status */ #define DLM_LKSTS_WAITING 1 @@ -217,6 +212,20 @@ struct dlm_args { #define DLM_IFL_USER 0x00000001 #define DLM_IFL_ORPHAN 0x00000002 +#define DLM_CALLBACKS_SIZE 6 + +#define DLM_CB_CAST 0x00000001 +#define DLM_CB_BAST 0x00000002 +#define DLM_CB_SKIP 0x00000004 + +struct dlm_callback { + uint64_t seq; + uint32_t flags; /* DLM_CBF_ */ + int sb_status; /* copy to lksb status */ + uint8_t sb_flags; /* copy to lksb flags */ + int8_t mode; /* rq mode of bast, gr mode of cast */ +}; + struct dlm_lkb { struct dlm_rsb *lkb_resource; /* the rsb */ struct kref lkb_ref; @@ -236,13 +245,6 @@ struct dlm_lkb { int8_t lkb_wait_type; /* type of reply waiting for */ int8_t lkb_wait_count; - int8_t lkb_ast_type; /* type of ast queued for */ - int8_t lkb_ast_first; /* type of first ast queued */ - - int8_t lkb_bastmode; /* req mode of queued bast */ - int8_t lkb_castmode; /* gr mode of queued cast */ - int8_t lkb_bastmode_done; /* last delivered bastmode */ - int8_t lkb_castmode_done; /* last delivered castmode */ struct list_head lkb_idtbl_list; /* lockspace lkbtbl */ struct list_head lkb_statequeue; /* rsb g/c/w list */ @@ -251,10 +253,15 @@ struct dlm_lkb { struct list_head lkb_astqueue; /* need ast to be sent */ struct list_head lkb_ownqueue; /* list of locks for a process */ struct list_head lkb_time_list; - ktime_t lkb_time_bast; /* for debugging */ ktime_t lkb_timestamp; unsigned long lkb_timeout_cs; + struct dlm_callback lkb_callbacks[DLM_CALLBACKS_SIZE]; + struct dlm_callback lkb_last_cast; + struct dlm_callback lkb_last_bast; + ktime_t lkb_last_cast_time; /* for debugging */ + ktime_t lkb_last_bast_time; /* for debugging */ + char *lkb_lvbptr; struct dlm_lksb *lkb_lksb; /* caller's status block */ void (*lkb_astfn) (void *astparam); @@ -544,8 +551,6 @@ struct dlm_user_args { (dlm_user_proc) on the struct file, the process's locks point back to it*/ struct dlm_lksb lksb; - int old_mode; - int update_user_lvb; struct dlm_lksb __user *user_lksb; void __user *castparam; void __user *castaddr; diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index 64e5f3e..04b8c44 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -160,10 +160,10 @@ static const int __quecvt_compat_matrix[8][8] = { void dlm_print_lkb(struct dlm_lkb *lkb) { printk(KERN_ERR "lkb: nodeid %d id %x remid %x exflags %x flags %x\n" - " status %d rqmode %d grmode %d wait_type %d ast_type %d\n", + " status %d rqmode %d grmode %d wait_type %d\n", lkb->lkb_nodeid, lkb->lkb_id, lkb->lkb_remid, lkb->lkb_exflags, lkb->lkb_flags, lkb->lkb_status, lkb->lkb_rqmode, - lkb->lkb_grmode, lkb->lkb_wait_type, lkb->lkb_ast_type); + lkb->lkb_grmode, lkb->lkb_wait_type); } static void dlm_print_rsb(struct dlm_rsb *r) @@ -305,10 +305,7 @@ static void queue_cast(struct dlm_rsb *r, struct dlm_lkb *lkb, int rv) rv = -EDEADLK; } - lkb->lkb_lksb->sb_status = rv; - lkb->lkb_lksb->sb_flags = lkb->lkb_sbflags; - - dlm_add_ast(lkb, AST_COMP, lkb->lkb_grmode); + dlm_add_ast(lkb, DLM_CB_CAST, lkb->lkb_grmode, rv, lkb->lkb_sbflags); } static inline void queue_cast_overlap(struct dlm_rsb *r, struct dlm_lkb *lkb) @@ -319,13 +316,10 @@ static inline void queue_cast_overlap(struct dlm_rsb *r, struct dlm_lkb *lkb) static void queue_bast(struct dlm_rsb *r, struct dlm_lkb *lkb, int rqmode) { - lkb->lkb_time_bast = ktime_get(); - if (is_master_copy(lkb)) { - lkb->lkb_bastmode = rqmode; /* printed by debugfs */ send_bast(r, lkb, rqmode); } else { - dlm_add_ast(lkb, AST_BAST, rqmode); + dlm_add_ast(lkb, DLM_CB_BAST, rqmode, 0, 0); } } @@ -600,6 +594,7 @@ static int create_lkb(struct dlm_ls *ls, struct dlm_lkb **lkb_ret) INIT_LIST_HEAD(&lkb->lkb_ownqueue); INIT_LIST_HEAD(&lkb->lkb_rsb_lookup); INIT_LIST_HEAD(&lkb->lkb_time_list); + INIT_LIST_HEAD(&lkb->lkb_astqueue); get_random_bytes(&bucket, sizeof(bucket)); bucket &= (ls->ls_lkbtbl_size - 1); @@ -2819,9 +2814,9 @@ static void send_args(struct dlm_rsb *r, struct dlm_lkb *lkb, not from lkb fields */ if (lkb->lkb_bastfn) - ms->m_asts |= AST_BAST; + ms->m_asts |= DLM_CB_BAST; if (lkb->lkb_astfn) - ms->m_asts |= AST_COMP; + ms->m_asts |= DLM_CB_CAST; /* compare with switch in create_message; send_remove() doesn't use send_args() */ @@ -3122,8 +3117,8 @@ static int receive_request_args(struct dlm_ls *ls, struct dlm_lkb *lkb, lkb->lkb_grmode = DLM_LOCK_IV; lkb->lkb_rqmode = ms->m_rqmode; - lkb->lkb_bastfn = (ms->m_asts & AST_BAST) ? &fake_bastfn : NULL; - lkb->lkb_astfn = (ms->m_asts & AST_COMP) ? &fake_astfn : NULL; + lkb->lkb_bastfn = (ms->m_asts & DLM_CB_BAST) ? &fake_bastfn : NULL; + lkb->lkb_astfn = (ms->m_asts & DLM_CB_CAST) ? &fake_astfn : NULL; if (lkb->lkb_exflags & DLM_LKF_VALBLK) { /* lkb was just created so there won't be an lvb yet */ @@ -4412,8 +4407,8 @@ static int receive_rcom_lock_args(struct dlm_ls *ls, struct dlm_lkb *lkb, lkb->lkb_grmode = rl->rl_grmode; /* don't set lkb_status because add_lkb wants to itself */ - lkb->lkb_bastfn = (rl->rl_asts & AST_BAST) ? &fake_bastfn : NULL; - lkb->lkb_astfn = (rl->rl_asts & AST_COMP) ? &fake_astfn : NULL; + lkb->lkb_bastfn = (rl->rl_asts & DLM_CB_BAST) ? &fake_bastfn : NULL; + lkb->lkb_astfn = (rl->rl_asts & DLM_CB_CAST) ? &fake_astfn : NULL; if (lkb->lkb_exflags & DLM_LKF_VALBLK) { int lvblen = rc->rc_header.h_length - sizeof(struct dlm_rcom) - @@ -4589,7 +4584,6 @@ int dlm_user_request(struct dlm_ls *ls, struct dlm_user_args *ua, error = set_lock_args(mode, &ua->lksb, flags, namelen, timeout_cs, fake_astfn, ua, fake_bastfn, &args); lkb->lkb_flags |= DLM_IFL_USER; - ua->old_mode = DLM_LOCK_IV; if (error) { __put_lkb(ls, lkb); @@ -4658,7 +4652,6 @@ int dlm_user_convert(struct dlm_ls *ls, struct dlm_user_args *ua_tmp, ua->bastparam = ua_tmp->bastparam; ua->bastaddr = ua_tmp->bastaddr; ua->user_lksb = ua_tmp->user_lksb; - ua->old_mode = lkb->lkb_grmode; error = set_lock_args(mode, &ua->lksb, flags, 0, timeout_cs, fake_astfn, ua, fake_bastfn, &args); @@ -4917,8 +4910,9 @@ void dlm_clear_proc_locks(struct dlm_ls *ls, struct dlm_user_proc *proc) } list_for_each_entry_safe(lkb, safe, &proc->asts, lkb_astqueue) { - lkb->lkb_ast_type = 0; - list_del(&lkb->lkb_astqueue); + memset(&lkb->lkb_callbacks, 0, + sizeof(struct dlm_callback) * DLM_CALLBACKS_SIZE); + list_del_init(&lkb->lkb_astqueue); dlm_put_lkb(lkb); } @@ -4958,7 +4952,9 @@ static void purge_proc_locks(struct dlm_ls *ls, struct dlm_user_proc *proc) spin_lock(&proc->asts_spin); list_for_each_entry_safe(lkb, safe, &proc->asts, lkb_astqueue) { - list_del(&lkb->lkb_astqueue); + memset(&lkb->lkb_callbacks, 0, + sizeof(struct dlm_callback) * DLM_CALLBACKS_SIZE); + list_del_init(&lkb->lkb_astqueue); dlm_put_lkb(lkb); } spin_unlock(&proc->asts_spin); diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index 9c64ae9..bffa1e7 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -1468,15 +1468,15 @@ static void work_stop(void) static int work_start(void) { - recv_workqueue = alloc_workqueue("dlm_recv", WQ_MEM_RECLAIM | - WQ_HIGHPRI | WQ_FREEZEABLE, 0); + recv_workqueue = alloc_workqueue("dlm_recv", + WQ_UNBOUND | WQ_MEM_RECLAIM, 1); if (!recv_workqueue) { log_print("can't start dlm_recv"); return -ENOMEM; } - send_workqueue = alloc_workqueue("dlm_send", WQ_MEM_RECLAIM | - WQ_HIGHPRI | WQ_FREEZEABLE, 0); + send_workqueue = alloc_workqueue("dlm_send", + WQ_UNBOUND | WQ_MEM_RECLAIM, 1); if (!send_workqueue) { log_print("can't start dlm_send"); destroy_workqueue(recv_workqueue); diff --git a/fs/dlm/rcom.c b/fs/dlm/rcom.c index 3c83a49..f10a50f 100644 --- a/fs/dlm/rcom.c +++ b/fs/dlm/rcom.c @@ -321,9 +321,9 @@ static void pack_rcom_lock(struct dlm_rsb *r, struct dlm_lkb *lkb, rl->rl_wait_type = cpu_to_le16(lkb->lkb_wait_type); if (lkb->lkb_bastfn) - rl->rl_asts |= AST_BAST; + rl->rl_asts |= DLM_CB_BAST; if (lkb->lkb_astfn) - rl->rl_asts |= AST_COMP; + rl->rl_asts |= DLM_CB_CAST; rl->rl_namelen = cpu_to_le16(r->res_length); memcpy(rl->rl_name, r->res_name, r->res_length); diff --git a/fs/dlm/user.c b/fs/dlm/user.c index 66d6c16..d5ab3fe 100644 --- a/fs/dlm/user.c +++ b/fs/dlm/user.c @@ -24,6 +24,7 @@ #include "lock.h" #include "lvb_table.h" #include "user.h" +#include "ast.h" static const char name_prefix[] = "dlm"; static const struct file_operations device_fops; @@ -152,19 +153,16 @@ static void compat_output(struct dlm_lock_result *res, not related to the lifetime of the lkb struct which is managed entirely by refcount. */ -static int lkb_is_endoflife(struct dlm_lkb *lkb, int sb_status, int type) +static int lkb_is_endoflife(int mode, int status) { - switch (sb_status) { + switch (status) { case -DLM_EUNLOCK: return 1; case -DLM_ECANCEL: case -ETIMEDOUT: case -EDEADLK: - if (lkb->lkb_grmode == DLM_LOCK_IV) - return 1; - break; case -EAGAIN: - if (type == AST_COMP && lkb->lkb_grmode == DLM_LOCK_IV) + if (mode == DLM_LOCK_IV) return 1; break; } @@ -174,12 +172,13 @@ static int lkb_is_endoflife(struct dlm_lkb *lkb, int sb_status, int type) /* we could possibly check if the cancel of an orphan has resulted in the lkb being removed and then remove that lkb from the orphans list and free it */ -void dlm_user_add_ast(struct dlm_lkb *lkb, int type, int mode) +void dlm_user_add_ast(struct dlm_lkb *lkb, uint32_t flags, int mode, + int status, uint32_t sbflags, uint64_t seq) { struct dlm_ls *ls; struct dlm_user_args *ua; struct dlm_user_proc *proc; - int eol = 0, ast_type; + int rv; if (lkb->lkb_flags & (DLM_IFL_ORPHAN | DLM_IFL_DEAD)) return; @@ -200,49 +199,29 @@ void dlm_user_add_ast(struct dlm_lkb *lkb, int type, int mode) ua = lkb->lkb_ua; proc = ua->proc; - if (type == AST_BAST && ua->bastaddr == NULL) + if ((flags & DLM_CB_BAST) && ua->bastaddr == NULL) goto out; + if ((flags & DLM_CB_CAST) && lkb_is_endoflife(mode, status)) + lkb->lkb_flags |= DLM_IFL_ENDOFLIFE; + spin_lock(&proc->asts_spin); - ast_type = lkb->lkb_ast_type; - lkb->lkb_ast_type |= type; - if (type == AST_BAST) - lkb->lkb_bastmode = mode; - else - lkb->lkb_castmode = mode; + rv = dlm_add_lkb_callback(lkb, flags, mode, status, sbflags, seq); + if (rv < 0) { + spin_unlock(&proc->asts_spin); + goto out; + } - if (!ast_type) { + if (list_empty(&lkb->lkb_astqueue)) { kref_get(&lkb->lkb_ref); list_add_tail(&lkb->lkb_astqueue, &proc->asts); - lkb->lkb_ast_first = type; wake_up_interruptible(&proc->wait); } - if (type == AST_COMP && (ast_type & AST_COMP)) - log_debug(ls, "ast overlap %x status %x %x", - lkb->lkb_id, ua->lksb.sb_status, lkb->lkb_flags); - - eol = lkb_is_endoflife(lkb, ua->lksb.sb_status, type); - if (eol) { - lkb->lkb_flags |= DLM_IFL_ENDOFLIFE; - } - - /* We want to copy the lvb to userspace when the completion - ast is read if the status is 0, the lock has an lvb and - lvb_ops says we should. We could probably have set_lvb_lock() - set update_user_lvb instead and not need old_mode */ - - if ((lkb->lkb_ast_type & AST_COMP) && - (lkb->lkb_lksb->sb_status == 0) && - lkb->lkb_lksb->sb_lvbptr && - dlm_lvb_operations[ua->old_mode + 1][lkb->lkb_grmode + 1]) - ua->update_user_lvb = 1; - else - ua->update_user_lvb = 0; - spin_unlock(&proc->asts_spin); - if (eol) { + if (lkb->lkb_flags & DLM_IFL_ENDOFLIFE) { + /* N.B. spin_lock locks_spin, not asts_spin */ spin_lock(&proc->locks_spin); if (!list_empty(&lkb->lkb_ownqueue)) { list_del_init(&lkb->lkb_ownqueue); @@ -705,8 +684,9 @@ static int device_close(struct inode *inode, struct file *file) return 0; } -static int copy_result_to_user(struct dlm_user_args *ua, int compat, int type, - int mode, char __user *buf, size_t count) +static int copy_result_to_user(struct dlm_user_args *ua, int compat, + uint32_t flags, int mode, int copy_lvb, + char __user *buf, size_t count) { #ifdef CONFIG_COMPAT struct dlm_lock_result32 result32; @@ -730,7 +710,7 @@ static int copy_result_to_user(struct dlm_user_args *ua, int compat, int type, notes that a new blocking AST address and parameter are set even if the conversion fails, so maybe we should just do that. */ - if (type == AST_BAST) { + if (flags & DLM_CB_BAST) { result.user_astaddr = ua->bastaddr; result.user_astparam = ua->bastparam; result.bast_mode = mode; @@ -750,8 +730,7 @@ static int copy_result_to_user(struct dlm_user_args *ua, int compat, int type, /* copy lvb to userspace if there is one, it's been updated, and the user buffer has space for it */ - if (ua->update_user_lvb && ua->lksb.sb_lvbptr && - count >= len + DLM_USER_LVB_LEN) { + if (copy_lvb && ua->lksb.sb_lvbptr && count >= len + DLM_USER_LVB_LEN) { if (copy_to_user(buf+len, ua->lksb.sb_lvbptr, DLM_USER_LVB_LEN)) { error = -EFAULT; @@ -801,13 +780,12 @@ static ssize_t device_read(struct file *file, char __user *buf, size_t count, struct dlm_user_proc *proc = file->private_data; struct dlm_lkb *lkb; DECLARE_WAITQUEUE(wait, current); - int error = 0, removed; - int ret_type, ret_mode; - int bastmode, castmode, do_bast, do_cast; + struct dlm_callback cb; + int rv, resid, copy_lvb = 0; if (count == sizeof(struct dlm_device_version)) { - error = copy_version_to_user(buf, count); - return error; + rv = copy_version_to_user(buf, count); + return rv; } if (!proc) { @@ -854,92 +832,57 @@ static ssize_t device_read(struct file *file, char __user *buf, size_t count, } } - /* there may be both completion and blocking asts to return for - the lkb, don't remove lkb from asts list unless no asts remain */ + /* if we empty lkb_callbacks, we don't want to unlock the spinlock + without removing lkb_astqueue; so empty lkb_astqueue is always + consistent with empty lkb_callbacks */ lkb = list_entry(proc->asts.next, struct dlm_lkb, lkb_astqueue); - removed = 0; - ret_type = 0; - ret_mode = 0; - do_bast = lkb->lkb_ast_type & AST_BAST; - do_cast = lkb->lkb_ast_type & AST_COMP; - bastmode = lkb->lkb_bastmode; - castmode = lkb->lkb_castmode; - - /* when both are queued figure out which to do first and - switch first so the other goes in the next read */ - - if (do_cast && do_bast) { - if (lkb->lkb_ast_first == AST_COMP) { - ret_type = AST_COMP; - ret_mode = castmode; - lkb->lkb_ast_type &= ~AST_COMP; - lkb->lkb_ast_first = AST_BAST; - } else { - ret_type = AST_BAST; - ret_mode = bastmode; - lkb->lkb_ast_type &= ~AST_BAST; - lkb->lkb_ast_first = AST_COMP; - } - } else { - ret_type = lkb->lkb_ast_first; - ret_mode = (ret_type == AST_COMP) ? castmode : bastmode; - lkb->lkb_ast_type &= ~ret_type; - lkb->lkb_ast_first = 0; + rv = dlm_rem_lkb_callback(lkb->lkb_resource->res_ls, lkb, &cb, &resid); + if (rv < 0) { + /* this shouldn't happen; lkb should have been removed from + list when resid was zero */ + log_print("dlm_rem_lkb_callback empty %x", lkb->lkb_id); + list_del_init(&lkb->lkb_astqueue); + spin_unlock(&proc->asts_spin); + /* removes ref for proc->asts, may cause lkb to be freed */ + dlm_put_lkb(lkb); + goto try_another; } + if (!resid) + list_del_init(&lkb->lkb_astqueue); + spin_unlock(&proc->asts_spin); - /* if we're doing a bast but the bast is unnecessary, then - switch to do nothing or do a cast if that was needed next */ - - if ((ret_type == AST_BAST) && - dlm_modes_compat(bastmode, lkb->lkb_castmode_done)) { - ret_type = 0; - ret_mode = 0; - - if (do_cast) { - ret_type = AST_COMP; - ret_mode = castmode; - lkb->lkb_ast_type &= ~AST_COMP; - lkb->lkb_ast_first = 0; - } + if (cb.flags & DLM_CB_SKIP) { + /* removes ref for proc->asts, may cause lkb to be freed */ + if (!resid) + dlm_put_lkb(lkb); + goto try_another; } - if (lkb->lkb_ast_first != lkb->lkb_ast_type) { - log_print("device_read %x ast_first %x ast_type %x", - lkb->lkb_id, lkb->lkb_ast_first, lkb->lkb_ast_type); - } + if (cb.flags & DLM_CB_CAST) { + int old_mode, new_mode; - if (!lkb->lkb_ast_type) { - list_del(&lkb->lkb_astqueue); - removed = 1; - } - spin_unlock(&proc->asts_spin); + old_mode = lkb->lkb_last_cast.mode; + new_mode = cb.mode; - if (ret_type) { - error = copy_result_to_user(lkb->lkb_ua, - test_bit(DLM_PROC_FLAGS_COMPAT, &proc->flags), - ret_type, ret_mode, buf, count); + if (!cb.sb_status && lkb->lkb_lksb->sb_lvbptr && + dlm_lvb_operations[old_mode + 1][new_mode + 1]) + copy_lvb = 1; - if (ret_type == AST_COMP) - lkb->lkb_castmode_done = castmode; - if (ret_type == AST_BAST) - lkb->lkb_bastmode_done = bastmode; + lkb->lkb_lksb->sb_status = cb.sb_status; + lkb->lkb_lksb->sb_flags = cb.sb_flags; } - /* removes reference for the proc->asts lists added by - dlm_user_add_ast() and may result in the lkb being freed */ + rv = copy_result_to_user(lkb->lkb_ua, + test_bit(DLM_PROC_FLAGS_COMPAT, &proc->flags), + cb.flags, cb.mode, copy_lvb, buf, count); - if (removed) + /* removes ref for proc->asts, may cause lkb to be freed */ + if (!resid) dlm_put_lkb(lkb); - /* the bast that was queued was eliminated (see unnecessary above), - leaving nothing to return */ - - if (!ret_type) - goto try_another; - - return error; + return rv; } static unsigned int device_poll(struct file *file, poll_table *wait) diff --git a/fs/dlm/user.h b/fs/dlm/user.h index f196091..00499ab 100644 --- a/fs/dlm/user.h +++ b/fs/dlm/user.h @@ -9,7 +9,8 @@ #ifndef __USER_DOT_H__ #define __USER_DOT_H__ -void dlm_user_add_ast(struct dlm_lkb *lkb, int type, int mode); +void dlm_user_add_ast(struct dlm_lkb *lkb, uint32_t flags, int mode, + int status, uint32_t sbflags, uint64_t seq); int dlm_user_init(void); void dlm_user_exit(void); int dlm_device_deregister(struct dlm_ls *ls); diff --git a/fs/ecryptfs/dentry.c b/fs/ecryptfs/dentry.c index 6fc4f31..534c1d4 100644 --- a/fs/ecryptfs/dentry.c +++ b/fs/ecryptfs/dentry.c @@ -46,24 +46,28 @@ static int ecryptfs_d_revalidate(struct dentry *dentry, struct nameidata *nd) { struct dentry *lower_dentry; struct vfsmount *lower_mnt; - struct dentry *dentry_save; - struct vfsmount *vfsmount_save; + struct dentry *dentry_save = NULL; + struct vfsmount *vfsmount_save = NULL; int rc = 1; - if (nd->flags & LOOKUP_RCU) + if (nd && nd->flags & LOOKUP_RCU) return -ECHILD; lower_dentry = ecryptfs_dentry_to_lower(dentry); lower_mnt = ecryptfs_dentry_to_lower_mnt(dentry); if (!lower_dentry->d_op || !lower_dentry->d_op->d_revalidate) goto out; - dentry_save = nd->path.dentry; - vfsmount_save = nd->path.mnt; - nd->path.dentry = lower_dentry; - nd->path.mnt = lower_mnt; + if (nd) { + dentry_save = nd->path.dentry; + vfsmount_save = nd->path.mnt; + nd->path.dentry = lower_dentry; + nd->path.mnt = lower_mnt; + } rc = lower_dentry->d_op->d_revalidate(lower_dentry, nd); - nd->path.dentry = dentry_save; - nd->path.mnt = vfsmount_save; + if (nd) { + nd->path.dentry = dentry_save; + nd->path.mnt = vfsmount_save; + } if (dentry->d_inode) { struct inode *lower_inode = ecryptfs_inode_to_lower(dentry->d_inode); diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index dbc84ed..e007534 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -632,8 +632,7 @@ int ecryptfs_interpose(struct dentry *hidden_dentry, u32 flags); int ecryptfs_lookup_and_interpose_lower(struct dentry *ecryptfs_dentry, struct dentry *lower_dentry, - struct inode *ecryptfs_dir_inode, - struct nameidata *ecryptfs_nd); + struct inode *ecryptfs_dir_inode); int ecryptfs_decode_and_decrypt_filename(char **decrypted_name, size_t *decrypted_name_size, struct dentry *ecryptfs_dentry, diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c index 81e10e6..7d1050e 100644 --- a/fs/ecryptfs/file.c +++ b/fs/ecryptfs/file.c @@ -317,6 +317,7 @@ ecryptfs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg) const struct file_operations ecryptfs_dir_fops = { .readdir = ecryptfs_readdir, + .read = generic_read_dir, .unlocked_ioctl = ecryptfs_unlocked_ioctl, #ifdef CONFIG_COMPAT .compat_ioctl = ecryptfs_compat_ioctl, diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index bd33f87..b592938 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -74,16 +74,20 @@ ecryptfs_create_underlying_file(struct inode *lower_dir_inode, unsigned int flags_save; int rc; - dentry_save = nd->path.dentry; - vfsmount_save = nd->path.mnt; - flags_save = nd->flags; - nd->path.dentry = lower_dentry; - nd->path.mnt = lower_mnt; - nd->flags &= ~LOOKUP_OPEN; + if (nd) { + dentry_save = nd->path.dentry; + vfsmount_save = nd->path.mnt; + flags_save = nd->flags; + nd->path.dentry = lower_dentry; + nd->path.mnt = lower_mnt; + nd->flags &= ~LOOKUP_OPEN; + } rc = vfs_create(lower_dir_inode, lower_dentry, mode, nd); - nd->path.dentry = dentry_save; - nd->path.mnt = vfsmount_save; - nd->flags = flags_save; + if (nd) { + nd->path.dentry = dentry_save; + nd->path.mnt = vfsmount_save; + nd->flags = flags_save; + } return rc; } @@ -241,8 +245,7 @@ out: */ int ecryptfs_lookup_and_interpose_lower(struct dentry *ecryptfs_dentry, struct dentry *lower_dentry, - struct inode *ecryptfs_dir_inode, - struct nameidata *ecryptfs_nd) + struct inode *ecryptfs_dir_inode) { struct dentry *lower_dir_dentry; struct vfsmount *lower_mnt; @@ -290,8 +293,6 @@ int ecryptfs_lookup_and_interpose_lower(struct dentry *ecryptfs_dentry, goto out; if (special_file(lower_inode->i_mode)) goto out; - if (!ecryptfs_nd) - goto out; /* Released in this function */ page_virt = kmem_cache_zalloc(ecryptfs_header_cache_2, GFP_USER); if (!page_virt) { @@ -349,75 +350,6 @@ out: } /** - * ecryptfs_new_lower_dentry - * @name: The name of the new dentry. - * @lower_dir_dentry: Parent directory of the new dentry. - * @nd: nameidata from last lookup. - * - * Create a new dentry or get it from lower parent dir. - */ -static struct dentry * -ecryptfs_new_lower_dentry(struct qstr *name, struct dentry *lower_dir_dentry, - struct nameidata *nd) -{ - struct dentry *new_dentry; - struct dentry *tmp; - struct inode *lower_dir_inode; - - lower_dir_inode = lower_dir_dentry->d_inode; - - tmp = d_alloc(lower_dir_dentry, name); - if (!tmp) - return ERR_PTR(-ENOMEM); - - mutex_lock(&lower_dir_inode->i_mutex); - new_dentry = lower_dir_inode->i_op->lookup(lower_dir_inode, tmp, nd); - mutex_unlock(&lower_dir_inode->i_mutex); - - if (!new_dentry) - new_dentry = tmp; - else - dput(tmp); - - return new_dentry; -} - - -/** - * ecryptfs_lookup_one_lower - * @ecryptfs_dentry: The eCryptfs dentry that we are looking up - * @lower_dir_dentry: lower parent directory - * @name: lower file name - * - * Get the lower dentry from vfs. If lower dentry does not exist yet, - * create it. - */ -static struct dentry * -ecryptfs_lookup_one_lower(struct dentry *ecryptfs_dentry, - struct dentry *lower_dir_dentry, struct qstr *name) -{ - struct nameidata nd; - struct vfsmount *lower_mnt; - int err; - - lower_mnt = mntget(ecryptfs_dentry_to_lower_mnt( - ecryptfs_dentry->d_parent)); - err = vfs_path_lookup(lower_dir_dentry, lower_mnt, name->name , 0, &nd); - mntput(lower_mnt); - - if (!err) { - /* we dont need the mount */ - mntput(nd.path.mnt); - return nd.path.dentry; - } - if (err != -ENOENT) - return ERR_PTR(err); - - /* create a new lower dentry */ - return ecryptfs_new_lower_dentry(name, lower_dir_dentry, &nd); -} - -/** * ecryptfs_lookup * @ecryptfs_dir_inode: The eCryptfs directory inode * @ecryptfs_dentry: The eCryptfs dentry that we are looking up @@ -434,7 +366,6 @@ static struct dentry *ecryptfs_lookup(struct inode *ecryptfs_dir_inode, size_t encrypted_and_encoded_name_size; struct ecryptfs_mount_crypt_stat *mount_crypt_stat = NULL; struct dentry *lower_dir_dentry, *lower_dentry; - struct qstr lower_name; int rc = 0; if ((ecryptfs_dentry->d_name.len == 1 @@ -444,20 +375,14 @@ static struct dentry *ecryptfs_lookup(struct inode *ecryptfs_dir_inode, goto out_d_drop; } lower_dir_dentry = ecryptfs_dentry_to_lower(ecryptfs_dentry->d_parent); - lower_name.name = ecryptfs_dentry->d_name.name; - lower_name.len = ecryptfs_dentry->d_name.len; - lower_name.hash = ecryptfs_dentry->d_name.hash; - if (lower_dir_dentry->d_op && lower_dir_dentry->d_op->d_hash) { - rc = lower_dir_dentry->d_op->d_hash(lower_dir_dentry, - lower_dir_dentry->d_inode, &lower_name); - if (rc < 0) - goto out_d_drop; - } - lower_dentry = ecryptfs_lookup_one_lower(ecryptfs_dentry, - lower_dir_dentry, &lower_name); + mutex_lock(&lower_dir_dentry->d_inode->i_mutex); + lower_dentry = lookup_one_len(ecryptfs_dentry->d_name.name, + lower_dir_dentry, + ecryptfs_dentry->d_name.len); + mutex_unlock(&lower_dir_dentry->d_inode->i_mutex); if (IS_ERR(lower_dentry)) { rc = PTR_ERR(lower_dentry); - ecryptfs_printk(KERN_DEBUG, "%s: lookup_one_lower() returned " + ecryptfs_printk(KERN_DEBUG, "%s: lookup_one_len() returned " "[%d] on lower_dentry = [%s]\n", __func__, rc, encrypted_and_encoded_name); goto out_d_drop; @@ -479,28 +404,21 @@ static struct dentry *ecryptfs_lookup(struct inode *ecryptfs_dir_inode, "filename; rc = [%d]\n", __func__, rc); goto out_d_drop; } - lower_name.name = encrypted_and_encoded_name; - lower_name.len = encrypted_and_encoded_name_size; - lower_name.hash = full_name_hash(lower_name.name, lower_name.len); - if (lower_dir_dentry->d_op && lower_dir_dentry->d_op->d_hash) { - rc = lower_dir_dentry->d_op->d_hash(lower_dir_dentry, - lower_dir_dentry->d_inode, &lower_name); - if (rc < 0) - goto out_d_drop; - } - lower_dentry = ecryptfs_lookup_one_lower(ecryptfs_dentry, - lower_dir_dentry, &lower_name); + mutex_lock(&lower_dir_dentry->d_inode->i_mutex); + lower_dentry = lookup_one_len(encrypted_and_encoded_name, + lower_dir_dentry, + encrypted_and_encoded_name_size); + mutex_unlock(&lower_dir_dentry->d_inode->i_mutex); if (IS_ERR(lower_dentry)) { rc = PTR_ERR(lower_dentry); - ecryptfs_printk(KERN_DEBUG, "%s: lookup_one_lower() returned " + ecryptfs_printk(KERN_DEBUG, "%s: lookup_one_len() returned " "[%d] on lower_dentry = [%s]\n", __func__, rc, encrypted_and_encoded_name); goto out_d_drop; } lookup_and_interpose: rc = ecryptfs_lookup_and_interpose_lower(ecryptfs_dentry, lower_dentry, - ecryptfs_dir_inode, - ecryptfs_nd); + ecryptfs_dir_inode); goto out; out_d_drop: d_drop(ecryptfs_dentry); @@ -1092,6 +1010,8 @@ int ecryptfs_getattr(struct vfsmount *mnt, struct dentry *dentry, rc = vfs_getattr(ecryptfs_dentry_to_lower_mnt(dentry), ecryptfs_dentry_to_lower(dentry), &lower_stat); if (!rc) { + fsstack_copy_attr_all(dentry->d_inode, + ecryptfs_inode_to_lower(dentry->d_inode)); generic_fillattr(dentry->d_inode, stat); stat->blocks = lower_stat.blocks; } diff --git a/fs/eventfd.c b/fs/eventfd.c index e0194b3..d9a5917 100644 --- a/fs/eventfd.c +++ b/fs/eventfd.c @@ -99,7 +99,7 @@ EXPORT_SYMBOL_GPL(eventfd_ctx_get); * @ctx: [in] Pointer to eventfd context. * * The eventfd context reference must have been previously acquired either - * with eventfd_ctx_get() or eventfd_ctx_fdget()). + * with eventfd_ctx_get() or eventfd_ctx_fdget(). */ void eventfd_ctx_put(struct eventfd_ctx *ctx) { @@ -146,9 +146,9 @@ static void eventfd_ctx_do_read(struct eventfd_ctx *ctx, __u64 *cnt) * eventfd_ctx_remove_wait_queue - Read the current counter and removes wait queue. * @ctx: [in] Pointer to eventfd context. * @wait: [in] Wait queue to be removed. - * @cnt: [out] Pointer to the 64bit conter value. + * @cnt: [out] Pointer to the 64-bit counter value. * - * Returns zero if successful, or the following error codes: + * Returns %0 if successful, or the following error codes: * * -EAGAIN : The operation would have blocked. * @@ -175,11 +175,11 @@ EXPORT_SYMBOL_GPL(eventfd_ctx_remove_wait_queue); * eventfd_ctx_read - Reads the eventfd counter or wait if it is zero. * @ctx: [in] Pointer to eventfd context. * @no_wait: [in] Different from zero if the operation should not block. - * @cnt: [out] Pointer to the 64bit conter value. + * @cnt: [out] Pointer to the 64-bit counter value. * - * Returns zero if successful, or the following error codes: + * Returns %0 if successful, or the following error codes: * - * -EAGAIN : The operation would have blocked but @no_wait was nonzero. + * -EAGAIN : The operation would have blocked but @no_wait was non-zero. * -ERESTARTSYS : A signal interrupted the wait operation. * * If @no_wait is zero, the function might sleep until the eventfd internal diff --git a/fs/eventpoll.c b/fs/eventpoll.c index cc8a9b7..ff12f7a 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c @@ -62,7 +62,14 @@ * This mutex is acquired by ep_free() during the epoll file * cleanup path and it is also acquired by eventpoll_release_file() * if a file has been pushed inside an epoll set and it is then - * close()d without a previous call toepoll_ctl(EPOLL_CTL_DEL). + * close()d without a previous call to epoll_ctl(EPOLL_CTL_DEL). + * It is also acquired when inserting an epoll fd onto another epoll + * fd. We do this so that we walk the epoll tree and ensure that this + * insertion does not create a cycle of epoll file descriptors, which + * could lead to deadlock. We need a global mutex to prevent two + * simultaneous inserts (A into B and B into A) from racing and + * constructing a cycle without either insert observing that it is + * going to. * It is possible to drop the "ep->mtx" and to use the global * mutex "epmutex" (together with "ep->lock") to have it working, * but having "ep->mtx" will make the interface more scalable. @@ -145,11 +152,11 @@ struct epitem { /* * This structure is stored inside the "private_data" member of the file - * structure and rapresent the main data sructure for the eventpoll + * structure and represents the main data structure for the eventpoll * interface. */ struct eventpoll { - /* Protect the this structure access */ + /* Protect the access to this structure */ spinlock_t lock; /* @@ -224,6 +231,9 @@ static long max_user_watches __read_mostly; */ static DEFINE_MUTEX(epmutex); +/* Used to check for epoll file descriptor inclusion loops */ +static struct nested_calls poll_loop_ncalls; + /* Used for safe wake up implementation */ static struct nested_calls poll_safewake_ncalls; @@ -783,7 +793,7 @@ static struct epitem *ep_find(struct eventpoll *ep, struct file *file, int fd) /* * This is the callback that is passed to the wait queue wakeup - * machanism. It is called by the stored file descriptors when they + * mechanism. It is called by the stored file descriptors when they * have events to report. */ static int ep_poll_callback(wait_queue_t *wait, unsigned mode, int sync, void *key) @@ -814,9 +824,9 @@ static int ep_poll_callback(wait_queue_t *wait, unsigned mode, int sync, void *k goto out_unlock; /* - * If we are trasfering events to userspace, we can hold no locks + * If we are transferring events to userspace, we can hold no locks * (because we're accessing user memory, and because of linux f_op->poll() - * semantics). All the events that happens during that period of time are + * semantics). All the events that happen during that period of time are * chained in ep->ovflist and requeued later on. */ if (unlikely(ep->ovflist != EP_UNACTIVE_PTR)) { @@ -1114,6 +1124,17 @@ static int ep_send_events(struct eventpoll *ep, return ep_scan_ready_list(ep, ep_send_events_proc, &esed); } +static inline struct timespec ep_set_mstimeout(long ms) +{ + struct timespec now, ts = { + .tv_sec = ms / MSEC_PER_SEC, + .tv_nsec = NSEC_PER_MSEC * (ms % MSEC_PER_SEC), + }; + + ktime_get_ts(&now); + return timespec_add_safe(now, ts); +} + static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events, int maxevents, long timeout) { @@ -1121,12 +1142,11 @@ static int ep_poll(struct eventpoll *ep, struct epoll_event __user *events, unsigned long flags; long slack; wait_queue_t wait; - struct timespec end_time; ktime_t expires, *to = NULL; if (timeout > 0) { - ktime_get_ts(&end_time); - timespec_add_ns(&end_time, (u64)timeout * NSEC_PER_MSEC); + struct timespec end_time = ep_set_mstimeout(timeout); + slack = select_estimate_accuracy(&end_time); to = &expires; *to = timespec_to_ktime(end_time); @@ -1188,6 +1208,62 @@ retry: return res; } +/** + * ep_loop_check_proc - Callback function to be passed to the @ep_call_nested() + * API, to verify that adding an epoll file inside another + * epoll structure, does not violate the constraints, in + * terms of closed loops, or too deep chains (which can + * result in excessive stack usage). + * + * @priv: Pointer to the epoll file to be currently checked. + * @cookie: Original cookie for this call. This is the top-of-the-chain epoll + * data structure pointer. + * @call_nests: Current dept of the @ep_call_nested() call stack. + * + * Returns: Returns zero if adding the epoll @file inside current epoll + * structure @ep does not violate the constraints, or -1 otherwise. + */ +static int ep_loop_check_proc(void *priv, void *cookie, int call_nests) +{ + int error = 0; + struct file *file = priv; + struct eventpoll *ep = file->private_data; + struct rb_node *rbp; + struct epitem *epi; + + mutex_lock(&ep->mtx); + for (rbp = rb_first(&ep->rbr); rbp; rbp = rb_next(rbp)) { + epi = rb_entry(rbp, struct epitem, rbn); + if (unlikely(is_file_epoll(epi->ffd.file))) { + error = ep_call_nested(&poll_loop_ncalls, EP_MAX_NESTS, + ep_loop_check_proc, epi->ffd.file, + epi->ffd.file->private_data, current); + if (error != 0) + break; + } + } + mutex_unlock(&ep->mtx); + + return error; +} + +/** + * ep_loop_check - Performs a check to verify that adding an epoll file (@file) + * another epoll file (represented by @ep) does not create + * closed loops or too deep chains. + * + * @ep: Pointer to the epoll private data structure. + * @file: Pointer to the epoll file to be checked. + * + * Returns: Returns zero if adding the epoll @file inside current epoll + * structure @ep does not violate the constraints, or -1 otherwise. + */ +static int ep_loop_check(struct eventpoll *ep, struct file *file) +{ + return ep_call_nested(&poll_loop_ncalls, EP_MAX_NESTS, + ep_loop_check_proc, file, ep, current); +} + /* * Open an eventpoll file descriptor. */ @@ -1236,6 +1312,7 @@ SYSCALL_DEFINE4(epoll_ctl, int, epfd, int, op, int, fd, struct epoll_event __user *, event) { int error; + int did_lock_epmutex = 0; struct file *file, *tfile; struct eventpoll *ep; struct epitem *epi; @@ -1277,6 +1354,25 @@ SYSCALL_DEFINE4(epoll_ctl, int, epfd, int, op, int, fd, */ ep = file->private_data; + /* + * When we insert an epoll file descriptor, inside another epoll file + * descriptor, there is the change of creating closed loops, which are + * better be handled here, than in more critical paths. + * + * We hold epmutex across the loop check and the insert in this case, in + * order to prevent two separate inserts from racing and each doing the + * insert "at the same time" such that ep_loop_check passes on both + * before either one does the insert, thereby creating a cycle. + */ + if (unlikely(is_file_epoll(tfile) && op == EPOLL_CTL_ADD)) { + mutex_lock(&epmutex); + did_lock_epmutex = 1; + error = -ELOOP; + if (ep_loop_check(ep, tfile) != 0) + goto error_tgt_fput; + } + + mutex_lock(&ep->mtx); /* @@ -1312,6 +1408,9 @@ SYSCALL_DEFINE4(epoll_ctl, int, epfd, int, op, int, fd, mutex_unlock(&ep->mtx); error_tgt_fput: + if (unlikely(did_lock_epmutex)) + mutex_unlock(&epmutex); + fput(tfile); error_fput: fput(file); @@ -1431,6 +1530,12 @@ static int __init eventpoll_init(void) EP_ITEM_COST; BUG_ON(max_user_watches < 0); + /* + * Initialize the structure used to perform epoll file descriptor + * inclusion loops checks. + */ + ep_nested_calls_init(&poll_loop_ncalls); + /* Initialize the structure used to perform safe poll wait head wake ups */ ep_nested_calls_init(&poll_safewake_ncalls); @@ -115,13 +115,16 @@ SYSCALL_DEFINE1(uselib, const char __user *, library) struct file *file; char *tmp = getname(library); int error = PTR_ERR(tmp); + static const struct open_flags uselib_flags = { + .open_flag = O_LARGEFILE | O_RDONLY | __FMODE_EXEC, + .acc_mode = MAY_READ | MAY_EXEC | MAY_OPEN, + .intent = LOOKUP_OPEN + }; if (IS_ERR(tmp)) goto out; - file = do_filp_open(AT_FDCWD, tmp, - O_LARGEFILE | O_RDONLY | FMODE_EXEC, 0, - MAY_READ | MAY_EXEC | MAY_OPEN); + file = do_filp_open(AT_FDCWD, tmp, &uselib_flags, LOOKUP_FOLLOW); putname(tmp); error = PTR_ERR(file); if (IS_ERR(file)) @@ -721,10 +724,13 @@ struct file *open_exec(const char *name) { struct file *file; int err; + static const struct open_flags open_exec_flags = { + .open_flag = O_LARGEFILE | O_RDONLY | __FMODE_EXEC, + .acc_mode = MAY_EXEC | MAY_OPEN, + .intent = LOOKUP_OPEN + }; - file = do_filp_open(AT_FDCWD, name, - O_LARGEFILE | O_RDONLY | FMODE_EXEC, 0, - MAY_EXEC | MAY_OPEN); + file = do_filp_open(AT_FDCWD, name, &open_exec_flags, LOOKUP_FOLLOW); if (IS_ERR(file)) goto out; @@ -1869,7 +1875,7 @@ static void wait_for_dump_helpers(struct file *file) /* - * uhm_pipe_setup + * umh_pipe_setup * helper function to customize the process used * to collect the core in userspace. Specifically * it sets up a pipe and installs it as fd 0 (stdin) diff --git a/fs/exofs/inode.c b/fs/exofs/inode.c index 4268542..a755523 100644 --- a/fs/exofs/inode.c +++ b/fs/exofs/inode.c @@ -1030,7 +1030,6 @@ struct inode *exofs_iget(struct super_block *sb, unsigned long ino) memcpy(oi->i_data, fcb.i_data, sizeof(fcb.i_data)); } - inode->i_mapping->backing_dev_info = sb->s_bdi; if (S_ISREG(inode->i_mode)) { inode->i_op = &exofs_file_inode_operations; inode->i_fop = &exofs_file_operations; @@ -1131,7 +1130,6 @@ struct inode *exofs_new_inode(struct inode *dir, int mode) sbi = sb->s_fs_info; - inode->i_mapping->backing_dev_info = sb->s_bdi; sb->s_dirt = 1; inode_init_owner(inode, dir, mode); inode->i_ino = sbi->s_nextid++; diff --git a/fs/exofs/namei.c b/fs/exofs/namei.c index 264e95d..4d70db1 100644 --- a/fs/exofs/namei.c +++ b/fs/exofs/namei.c @@ -272,7 +272,6 @@ static int exofs_rename(struct inode *old_dir, struct dentry *old_dentry, new_de = exofs_find_entry(new_dir, new_dentry, &new_page); if (!new_de) goto out_dir; - inode_inc_link_count(old_inode); err = exofs_set_link(new_dir, new_de, new_page, old_inode); new_inode->i_ctime = CURRENT_TIME; if (dir_de) @@ -286,12 +285,9 @@ static int exofs_rename(struct inode *old_dir, struct dentry *old_dentry, if (new_dir->i_nlink >= EXOFS_LINK_MAX) goto out_dir; } - inode_inc_link_count(old_inode); err = exofs_add_link(new_dentry, old_inode); - if (err) { - inode_dec_link_count(old_inode); + if (err) goto out_dir; - } if (dir_de) inode_inc_link_count(new_dir); } @@ -299,7 +295,7 @@ static int exofs_rename(struct inode *old_dir, struct dentry *old_dentry, old_inode->i_ctime = CURRENT_TIME; exofs_delete_entry(old_de, old_page); - inode_dec_link_count(old_inode); + mark_inode_dirty(old_inode); if (dir_de) { err = exofs_set_link(old_inode, dir_de, dir_page, new_dir); diff --git a/fs/exportfs/expfs.c b/fs/exportfs/expfs.c index 4b68257..b05acb7 100644 --- a/fs/exportfs/expfs.c +++ b/fs/exportfs/expfs.c @@ -320,9 +320,14 @@ static int export_encode_fh(struct dentry *dentry, struct fid *fid, struct inode * inode = dentry->d_inode; int len = *max_len; int type = FILEID_INO32_GEN; - - if (len < 2 || (connectable && len < 4)) + + if (connectable && (len < 4)) { + *max_len = 4; + return 255; + } else if (len < 2) { + *max_len = 2; return 255; + } len = 2; fid->i32.ino = inode->i_ino; @@ -369,6 +374,8 @@ struct dentry *exportfs_decode_fh(struct vfsmount *mnt, struct fid *fid, /* * Try to get any dentry for the given file handle from the filesystem. */ + if (!nop || !nop->fh_to_dentry) + return ERR_PTR(-ESTALE); result = nop->fh_to_dentry(mnt->mnt_sb, fid, fh_len, fileid_type); if (!result) result = ERR_PTR(-ESTALE); diff --git a/fs/ext2/ext2.h b/fs/ext2/ext2.h index 6346a2a..1b48c33 100644 --- a/fs/ext2/ext2.h +++ b/fs/ext2/ext2.h @@ -110,7 +110,7 @@ extern struct ext2_dir_entry_2 * ext2_dotdot (struct inode *, struct page **); extern void ext2_set_link(struct inode *, struct ext2_dir_entry_2 *, struct page *, struct inode *, int); /* ialloc.c */ -extern struct inode * ext2_new_inode (struct inode *, int); +extern struct inode * ext2_new_inode (struct inode *, int, const struct qstr *); extern void ext2_free_inode (struct inode *); extern unsigned long ext2_count_free_inodes (struct super_block *); extern void ext2_check_inodes_bitmap (struct super_block *); diff --git a/fs/ext2/ialloc.c b/fs/ext2/ialloc.c index ad70479..ee9ed31 100644 --- a/fs/ext2/ialloc.c +++ b/fs/ext2/ialloc.c @@ -429,7 +429,8 @@ found: return group; } -struct inode *ext2_new_inode(struct inode *dir, int mode) +struct inode *ext2_new_inode(struct inode *dir, int mode, + const struct qstr *qstr) { struct super_block *sb; struct buffer_head *bitmap_bh = NULL; @@ -585,7 +586,7 @@ got: if (err) goto fail_free_drop; - err = ext2_init_security(inode,dir); + err = ext2_init_security(inode, dir, qstr); if (err) goto fail_free_drop; diff --git a/fs/ext2/namei.c b/fs/ext2/namei.c index 2e1d834..ed5c5d4 100644 --- a/fs/ext2/namei.c +++ b/fs/ext2/namei.c @@ -104,7 +104,7 @@ static int ext2_create (struct inode * dir, struct dentry * dentry, int mode, st dquot_initialize(dir); - inode = ext2_new_inode(dir, mode); + inode = ext2_new_inode(dir, mode, &dentry->d_name); if (IS_ERR(inode)) return PTR_ERR(inode); @@ -133,7 +133,7 @@ static int ext2_mknod (struct inode * dir, struct dentry *dentry, int mode, dev_ dquot_initialize(dir); - inode = ext2_new_inode (dir, mode); + inode = ext2_new_inode (dir, mode, &dentry->d_name); err = PTR_ERR(inode); if (!IS_ERR(inode)) { init_special_inode(inode, inode->i_mode, rdev); @@ -159,7 +159,7 @@ static int ext2_symlink (struct inode * dir, struct dentry * dentry, dquot_initialize(dir); - inode = ext2_new_inode (dir, S_IFLNK | S_IRWXUGO); + inode = ext2_new_inode (dir, S_IFLNK | S_IRWXUGO, &dentry->d_name); err = PTR_ERR(inode); if (IS_ERR(inode)) goto out; @@ -230,7 +230,7 @@ static int ext2_mkdir(struct inode * dir, struct dentry * dentry, int mode) inode_inc_link_count(dir); - inode = ext2_new_inode (dir, S_IFDIR | mode); + inode = ext2_new_inode(dir, S_IFDIR | mode, &dentry->d_name); err = PTR_ERR(inode); if (IS_ERR(inode)) goto out_dir; @@ -344,7 +344,6 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry, new_de = ext2_find_entry (new_dir, &new_dentry->d_name, &new_page); if (!new_de) goto out_dir; - inode_inc_link_count(old_inode); ext2_set_link(new_dir, new_de, new_page, old_inode, 1); new_inode->i_ctime = CURRENT_TIME_SEC; if (dir_de) @@ -356,12 +355,9 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry, if (new_dir->i_nlink >= EXT2_LINK_MAX) goto out_dir; } - inode_inc_link_count(old_inode); err = ext2_add_link(new_dentry, old_inode); - if (err) { - inode_dec_link_count(old_inode); + if (err) goto out_dir; - } if (dir_de) inode_inc_link_count(new_dir); } @@ -369,12 +365,11 @@ static int ext2_rename (struct inode * old_dir, struct dentry * old_dentry, /* * Like most other Unix systems, set the ctime for inodes on a * rename. - * inode_dec_link_count() will mark the inode dirty. */ old_inode->i_ctime = CURRENT_TIME_SEC; + mark_inode_dirty(old_inode); ext2_delete_entry (old_de, old_page); - inode_dec_link_count(old_inode); if (dir_de) { if (old_dir != new_dir) diff --git a/fs/ext2/xattr.h b/fs/ext2/xattr.h index a1a1c21..5e41ccc 100644 --- a/fs/ext2/xattr.h +++ b/fs/ext2/xattr.h @@ -116,9 +116,11 @@ exit_ext2_xattr(void) # endif /* CONFIG_EXT2_FS_XATTR */ #ifdef CONFIG_EXT2_FS_SECURITY -extern int ext2_init_security(struct inode *inode, struct inode *dir); +extern int ext2_init_security(struct inode *inode, struct inode *dir, + const struct qstr *qstr); #else -static inline int ext2_init_security(struct inode *inode, struct inode *dir) +static inline int ext2_init_security(struct inode *inode, struct inode *dir, + const struct qstr *qstr) { return 0; } diff --git a/fs/ext2/xattr_security.c b/fs/ext2/xattr_security.c index 3004e15..5d979b4 100644 --- a/fs/ext2/xattr_security.c +++ b/fs/ext2/xattr_security.c @@ -47,14 +47,15 @@ ext2_xattr_security_set(struct dentry *dentry, const char *name, } int -ext2_init_security(struct inode *inode, struct inode *dir) +ext2_init_security(struct inode *inode, struct inode *dir, + const struct qstr *qstr) { int err; size_t len; void *value; char *name; - err = security_inode_init_security(inode, dir, &name, &value, &len); + err = security_inode_init_security(inode, dir, qstr, &name, &value, &len); if (err) { if (err == -EOPNOTSUPP) return 0; diff --git a/fs/ext3/balloc.c b/fs/ext3/balloc.c index 045995c..15324218 100644 --- a/fs/ext3/balloc.c +++ b/fs/ext3/balloc.c @@ -1991,6 +1991,7 @@ ext3_grpblk_t ext3_trim_all_free(struct super_block *sb, unsigned int group, spin_unlock(sb_bgl_lock(sbi, group)); percpu_counter_sub(&sbi->s_freeblocks_counter, next - start); + free_blocks -= next - start; /* Do not issue a TRIM on extents smaller than minblocks */ if ((next - start) < minblocks) goto free_extent; @@ -2040,7 +2041,7 @@ free_extent: cond_resched(); /* No more suitable extents */ - if ((free_blocks - count) < minblocks) + if (free_blocks < minblocks) break; } @@ -2090,7 +2091,8 @@ int ext3_trim_fs(struct super_block *sb, struct fstrim_range *range) ext3_fsblk_t max_blks = le32_to_cpu(es->s_blocks_count); int ret = 0; - start = range->start >> sb->s_blocksize_bits; + start = (range->start >> sb->s_blocksize_bits) + + le32_to_cpu(es->s_first_data_block); len = range->len >> sb->s_blocksize_bits; minlen = range->minlen >> sb->s_blocksize_bits; trimmed = 0; @@ -2099,10 +2101,6 @@ int ext3_trim_fs(struct super_block *sb, struct fstrim_range *range) return -EINVAL; if (start >= max_blks) goto out; - if (start < le32_to_cpu(es->s_first_data_block)) { - len -= le32_to_cpu(es->s_first_data_block) - start; - start = le32_to_cpu(es->s_first_data_block); - } if (start + len > max_blks) len = max_blks - start; @@ -2129,10 +2127,15 @@ int ext3_trim_fs(struct super_block *sb, struct fstrim_range *range) if (free_blocks < minlen) continue; - if (len >= EXT3_BLOCKS_PER_GROUP(sb)) - len -= (EXT3_BLOCKS_PER_GROUP(sb) - first_block); - else + /* + * For all the groups except the last one, last block will + * always be EXT3_BLOCKS_PER_GROUP(sb), so we only need to + * change it for the last group in which case first_block + + * len < EXT3_BLOCKS_PER_GROUP(sb). + */ + if (first_block + len < EXT3_BLOCKS_PER_GROUP(sb)) last_block = first_block + len; + len -= last_block - first_block; ret = ext3_trim_all_free(sb, group, first_block, last_block, minlen); diff --git a/fs/ext3/ialloc.c b/fs/ext3/ialloc.c index 9724aef..bfc2dc4 100644 --- a/fs/ext3/ialloc.c +++ b/fs/ext3/ialloc.c @@ -404,7 +404,8 @@ static int find_group_other(struct super_block *sb, struct inode *parent) * For other inodes, search forward from the parent directory's block * group to find a free inode. */ -struct inode *ext3_new_inode(handle_t *handle, struct inode * dir, int mode) +struct inode *ext3_new_inode(handle_t *handle, struct inode * dir, + const struct qstr *qstr, int mode) { struct super_block *sb; struct buffer_head *bitmap_bh = NULL; @@ -589,7 +590,7 @@ got: if (err) goto fail_free_drop; - err = ext3_init_security(handle,inode, dir); + err = ext3_init_security(handle, inode, dir, qstr); if (err) goto fail_free_drop; diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index b27ba71..32f3b86 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c @@ -1540,8 +1540,8 @@ static int ext3_dx_add_entry(handle_t *handle, struct dentry *dentry, goto cleanup; node2 = (struct dx_node *)(bh2->b_data); entries2 = node2->entries; + memset(&node2->fake, 0, sizeof(struct fake_dirent)); node2->fake.rec_len = ext3_rec_len_to_disk(sb->s_blocksize); - node2->fake.inode = 0; BUFFER_TRACE(frame->bh, "get_write_access"); err = ext3_journal_get_write_access(handle, frame->bh); if (err) @@ -1710,7 +1710,7 @@ retry: if (IS_DIRSYNC(dir)) handle->h_sync = 1; - inode = ext3_new_inode (handle, dir, mode); + inode = ext3_new_inode (handle, dir, &dentry->d_name, mode); err = PTR_ERR(inode); if (!IS_ERR(inode)) { inode->i_op = &ext3_file_inode_operations; @@ -1746,7 +1746,7 @@ retry: if (IS_DIRSYNC(dir)) handle->h_sync = 1; - inode = ext3_new_inode (handle, dir, mode); + inode = ext3_new_inode (handle, dir, &dentry->d_name, mode); err = PTR_ERR(inode); if (!IS_ERR(inode)) { init_special_inode(inode, inode->i_mode, rdev); @@ -1784,7 +1784,7 @@ retry: if (IS_DIRSYNC(dir)) handle->h_sync = 1; - inode = ext3_new_inode (handle, dir, S_IFDIR | mode); + inode = ext3_new_inode (handle, dir, &dentry->d_name, S_IFDIR | mode); err = PTR_ERR(inode); if (IS_ERR(inode)) goto out_stop; @@ -2206,7 +2206,7 @@ retry: if (IS_DIRSYNC(dir)) handle->h_sync = 1; - inode = ext3_new_inode (handle, dir, S_IFLNK|S_IRWXUGO); + inode = ext3_new_inode (handle, dir, &dentry->d_name, S_IFLNK|S_IRWXUGO); err = PTR_ERR(inode); if (IS_ERR(inode)) goto out_stop; @@ -2253,13 +2253,6 @@ static int ext3_link (struct dentry * old_dentry, dquot_initialize(dir); - /* - * Return -ENOENT if we've raced with unlink and i_nlink is 0. Doing - * otherwise has the potential to corrupt the orphan inode list. - */ - if (inode->i_nlink == 0) - return -ENOENT; - retry: handle = ext3_journal_start(dir, EXT3_DATA_TRANS_BLOCKS(dir->i_sb) + EXT3_INDEX_EXTRA_TRANS_BLOCKS); diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 7aa767d..071689f 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -754,7 +754,7 @@ static int ext3_release_dquot(struct dquot *dquot); static int ext3_mark_dquot_dirty(struct dquot *dquot); static int ext3_write_info(struct super_block *sb, int type); static int ext3_quota_on(struct super_block *sb, int type, int format_id, - char *path); + struct path *path); static int ext3_quota_on_mount(struct super_block *sb, int type); static ssize_t ext3_quota_read(struct super_block *sb, int type, char *data, size_t len, loff_t off); @@ -1464,6 +1464,13 @@ static void ext3_orphan_cleanup (struct super_block * sb, return; } + /* Check if feature set allows readwrite operations */ + if (EXT3_HAS_RO_COMPAT_FEATURE(sb, ~EXT3_FEATURE_RO_COMPAT_SUPP)) { + ext3_msg(sb, KERN_INFO, "Skipping orphan cleanup due to " + "unknown ROCOMPAT features"); + return; + } + if (EXT3_SB(sb)->s_mount_state & EXT3_ERROR_FS) { if (es->s_last_orphan) jbd_debug(1, "Errors on filesystem, " @@ -1936,6 +1943,7 @@ static int ext3_fill_super (struct super_block *sb, void *data, int silent) sb->s_qcop = &ext3_qctl_operations; sb->dq_op = &ext3_quota_operations; #endif + memcpy(sb->s_uuid, es->s_uuid, sizeof(es->s_uuid)); INIT_LIST_HEAD(&sbi->s_orphan); /* unlinked but open files */ mutex_init(&sbi->s_orphan_lock); mutex_init(&sbi->s_resize_lock); @@ -2877,27 +2885,20 @@ static int ext3_quota_on_mount(struct super_block *sb, int type) * Standard function to be called on quota_on */ static int ext3_quota_on(struct super_block *sb, int type, int format_id, - char *name) + struct path *path) { int err; - struct path path; if (!test_opt(sb, QUOTA)) return -EINVAL; - err = kern_path(name, LOOKUP_FOLLOW, &path); - if (err) - return err; - /* Quotafile not on the same filesystem? */ - if (path.mnt->mnt_sb != sb) { - path_put(&path); + if (path->mnt->mnt_sb != sb) return -EXDEV; - } /* Journaling quota? */ if (EXT3_SB(sb)->s_qf_names[type]) { /* Quotafile not of fs root? */ - if (path.dentry->d_parent != sb->s_root) + if (path->dentry->d_parent != sb->s_root) ext3_msg(sb, KERN_WARNING, "warning: Quota file not on filesystem root. " "Journaled quota will not work."); @@ -2907,7 +2908,7 @@ static int ext3_quota_on(struct super_block *sb, int type, int format_id, * When we journal data on quota file, we have to flush journal to see * all updates to the file when we bypass pagecache... */ - if (ext3_should_journal_data(path.dentry->d_inode)) { + if (ext3_should_journal_data(path->dentry->d_inode)) { /* * We don't need to lock updates but journal_flush() could * otherwise be livelocked... @@ -2915,15 +2916,11 @@ static int ext3_quota_on(struct super_block *sb, int type, int format_id, journal_lock_updates(EXT3_SB(sb)->s_journal); err = journal_flush(EXT3_SB(sb)->s_journal); journal_unlock_updates(EXT3_SB(sb)->s_journal); - if (err) { - path_put(&path); + if (err) return err; - } } - err = dquot_quota_on_path(sb, type, format_id, &path); - path_put(&path); - return err; + return dquot_quota_on(sb, type, format_id, path); } /* Read data from quotafile - avoid pagecache and such because we cannot afford diff --git a/fs/ext3/xattr.h b/fs/ext3/xattr.h index 377fe72..2be4f69 100644 --- a/fs/ext3/xattr.h +++ b/fs/ext3/xattr.h @@ -128,10 +128,10 @@ exit_ext3_xattr(void) #ifdef CONFIG_EXT3_FS_SECURITY extern int ext3_init_security(handle_t *handle, struct inode *inode, - struct inode *dir); + struct inode *dir, const struct qstr *qstr); #else static inline int ext3_init_security(handle_t *handle, struct inode *inode, - struct inode *dir) + struct inode *dir, const struct qstr *qstr) { return 0; } diff --git a/fs/ext3/xattr_security.c b/fs/ext3/xattr_security.c index 03a99bf..b8d9f83 100644 --- a/fs/ext3/xattr_security.c +++ b/fs/ext3/xattr_security.c @@ -49,14 +49,15 @@ ext3_xattr_security_set(struct dentry *dentry, const char *name, } int -ext3_init_security(handle_t *handle, struct inode *inode, struct inode *dir) +ext3_init_security(handle_t *handle, struct inode *inode, struct inode *dir, + const struct qstr *qstr) { int err; size_t len; void *value; char *name; - err = security_inode_init_security(inode, dir, &name, &value, &len); + err = security_inode_init_security(inode, dir, qstr, &name, &value, &len); if (err) { if (err == -EOPNOTSUPP) return 0; diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 0c8d97b..3aa0b72 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -848,6 +848,7 @@ struct ext4_inode_info { atomic_t i_ioend_count; /* Number of outstanding io_end structs */ /* current io_end structure for async DIO write*/ ext4_io_end_t *cur_aio_dio; + atomic_t i_aiodio_unwritten; /* Nr. of inflight conversions pending */ spinlock_t i_block_reservation_lock; @@ -2119,6 +2120,15 @@ static inline void set_bitmap_uptodate(struct buffer_head *bh) #define in_range(b, first, len) ((b) >= (first) && (b) <= (first) + (len) - 1) +/* For ioend & aio unwritten conversion wait queues */ +#define EXT4_WQ_HASH_SZ 37 +#define ext4_ioend_wq(v) (&ext4__ioend_wq[((unsigned long)(v)) %\ + EXT4_WQ_HASH_SZ]) +#define ext4_aio_mutex(v) (&ext4__aio_mutex[((unsigned long)(v)) %\ + EXT4_WQ_HASH_SZ]) +extern wait_queue_head_t ext4__ioend_wq[EXT4_WQ_HASH_SZ]; +extern struct mutex ext4__aio_mutex[EXT4_WQ_HASH_SZ]; + #endif /* __KERNEL__ */ #endif /* _EXT4_H */ diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 63a7581..7516fb9 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -131,7 +131,7 @@ static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode, * fragmenting the file system's free space. Maybe we * should have some hueristics or some way to allow * userspace to pass a hint to file system, - * especiially if the latter case turns out to be + * especially if the latter case turns out to be * common. */ ex = path[depth].p_ext; @@ -2844,7 +2844,7 @@ fix_extent_len: * ext4_get_blocks_dio_write() when DIO to write * to an uninitialized extent. * - * Writing to an uninitized extent may result in splitting the uninitialized + * Writing to an uninitialized extent may result in splitting the uninitialized * extent into multiple /initialized uninitialized extents (up to three) * There are three possibilities: * a> There is no split required: Entire extent should be uninitialized @@ -3174,9 +3174,10 @@ ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode, * that this IO needs to convertion to written when IO is * completed */ - if (io) + if (io && !(io->flag & EXT4_IO_END_UNWRITTEN)) { io->flag = EXT4_IO_END_UNWRITTEN; - else + atomic_inc(&EXT4_I(inode)->i_aiodio_unwritten); + } else ext4_set_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN); if (ext4_should_dioread_nolock(inode)) map->m_flags |= EXT4_MAP_UNINIT; @@ -3463,9 +3464,10 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, * that we need to perform convertion when IO is done. */ if ((flags & EXT4_GET_BLOCKS_PRE_IO)) { - if (io) + if (io && !(io->flag & EXT4_IO_END_UNWRITTEN)) { io->flag = EXT4_IO_END_UNWRITTEN; - else + atomic_inc(&EXT4_I(inode)->i_aiodio_unwritten); + } else ext4_set_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN); } diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 2e8322c..7b80d54 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -55,11 +55,47 @@ static int ext4_release_file(struct inode *inode, struct file *filp) return 0; } +static void ext4_aiodio_wait(struct inode *inode) +{ + wait_queue_head_t *wq = ext4_ioend_wq(inode); + + wait_event(*wq, (atomic_read(&EXT4_I(inode)->i_aiodio_unwritten) == 0)); +} + +/* + * This tests whether the IO in question is block-aligned or not. + * Ext4 utilizes unwritten extents when hole-filling during direct IO, and they + * are converted to written only after the IO is complete. Until they are + * mapped, these blocks appear as holes, so dio_zero_block() will assume that + * it needs to zero out portions of the start and/or end block. If 2 AIO + * threads are at work on the same unwritten block, they must be synchronized + * or one thread will zero the other's data, causing corruption. + */ +static int +ext4_unaligned_aio(struct inode *inode, const struct iovec *iov, + unsigned long nr_segs, loff_t pos) +{ + struct super_block *sb = inode->i_sb; + int blockmask = sb->s_blocksize - 1; + size_t count = iov_length(iov, nr_segs); + loff_t final_size = pos + count; + + if (pos >= inode->i_size) + return 0; + + if ((pos & blockmask) || (final_size & blockmask)) + return 1; + + return 0; +} + static ssize_t ext4_file_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { struct inode *inode = iocb->ki_filp->f_path.dentry->d_inode; + int unaligned_aio = 0; + int ret; /* * If we have encountered a bitmap-format file, the size limit @@ -78,9 +114,31 @@ ext4_file_write(struct kiocb *iocb, const struct iovec *iov, nr_segs = iov_shorten((struct iovec *)iov, nr_segs, sbi->s_bitmap_maxbytes - pos); } + } else if (unlikely((iocb->ki_filp->f_flags & O_DIRECT) && + !is_sync_kiocb(iocb))) { + unaligned_aio = ext4_unaligned_aio(inode, iov, nr_segs, pos); } - return generic_file_aio_write(iocb, iov, nr_segs, pos); + /* Unaligned direct AIO must be serialized; see comment above */ + if (unaligned_aio) { + static unsigned long unaligned_warn_time; + + /* Warn about this once per day */ + if (printk_timed_ratelimit(&unaligned_warn_time, 60*60*24*HZ)) + ext4_msg(inode->i_sb, KERN_WARNING, + "Unaligned AIO/DIO on inode %ld by %s; " + "performance will be poor.", + inode->i_ino, current->comm); + mutex_lock(ext4_aio_mutex(inode)); + ext4_aiodio_wait(inode); + } + + ret = generic_file_aio_write(iocb, iov, nr_segs, pos); + + if (unaligned_aio) + mutex_unlock(ext4_aio_mutex(inode)); + + return ret; } static const struct vm_operations_struct ext4_file_vm_ops = { diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index eb9097a..78b79e1 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -1042,7 +1042,7 @@ got: if (err) goto fail_free_drop; - err = ext4_init_security(handle, inode, dir); + err = ext4_init_security(handle, inode, dir, qstr); if (err) goto fail_free_drop; diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 851f49b..d1fe09a 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -342,10 +342,15 @@ static struct kmem_cache *ext4_free_ext_cachep; /* We create slab caches for groupinfo data structures based on the * superblock block size. There will be one per mounted filesystem for * each unique s_blocksize_bits */ -#define NR_GRPINFO_CACHES \ - (EXT4_MAX_BLOCK_LOG_SIZE - EXT4_MIN_BLOCK_LOG_SIZE + 1) +#define NR_GRPINFO_CACHES 8 static struct kmem_cache *ext4_groupinfo_caches[NR_GRPINFO_CACHES]; +static const char *ext4_groupinfo_slab_names[NR_GRPINFO_CACHES] = { + "ext4_groupinfo_1k", "ext4_groupinfo_2k", "ext4_groupinfo_4k", + "ext4_groupinfo_8k", "ext4_groupinfo_16k", "ext4_groupinfo_32k", + "ext4_groupinfo_64k", "ext4_groupinfo_128k" +}; + static void ext4_mb_generate_from_pa(struct super_block *sb, void *bitmap, ext4_group_t group); static void ext4_mb_generate_from_freelist(struct super_block *sb, void *bitmap, @@ -2414,6 +2419,55 @@ err_freesgi: return -ENOMEM; } +static void ext4_groupinfo_destroy_slabs(void) +{ + int i; + + for (i = 0; i < NR_GRPINFO_CACHES; i++) { + if (ext4_groupinfo_caches[i]) + kmem_cache_destroy(ext4_groupinfo_caches[i]); + ext4_groupinfo_caches[i] = NULL; + } +} + +static int ext4_groupinfo_create_slab(size_t size) +{ + static DEFINE_MUTEX(ext4_grpinfo_slab_create_mutex); + int slab_size; + int blocksize_bits = order_base_2(size); + int cache_index = blocksize_bits - EXT4_MIN_BLOCK_LOG_SIZE; + struct kmem_cache *cachep; + + if (cache_index >= NR_GRPINFO_CACHES) + return -EINVAL; + + if (unlikely(cache_index < 0)) + cache_index = 0; + + mutex_lock(&ext4_grpinfo_slab_create_mutex); + if (ext4_groupinfo_caches[cache_index]) { + mutex_unlock(&ext4_grpinfo_slab_create_mutex); + return 0; /* Already created */ + } + + slab_size = offsetof(struct ext4_group_info, + bb_counters[blocksize_bits + 2]); + + cachep = kmem_cache_create(ext4_groupinfo_slab_names[cache_index], + slab_size, 0, SLAB_RECLAIM_ACCOUNT, + NULL); + + mutex_unlock(&ext4_grpinfo_slab_create_mutex); + if (!cachep) { + printk(KERN_EMERG "EXT4: no memory for groupinfo slab cache\n"); + return -ENOMEM; + } + + ext4_groupinfo_caches[cache_index] = cachep; + + return 0; +} + int ext4_mb_init(struct super_block *sb, int needs_recovery) { struct ext4_sb_info *sbi = EXT4_SB(sb); @@ -2421,9 +2475,6 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery) unsigned offset; unsigned max; int ret; - int cache_index; - struct kmem_cache *cachep; - char *namep = NULL; i = (sb->s_blocksize_bits + 2) * sizeof(*sbi->s_mb_offsets); @@ -2440,30 +2491,9 @@ int ext4_mb_init(struct super_block *sb, int needs_recovery) goto out; } - cache_index = sb->s_blocksize_bits - EXT4_MIN_BLOCK_LOG_SIZE; - cachep = ext4_groupinfo_caches[cache_index]; - if (!cachep) { - char name[32]; - int len = offsetof(struct ext4_group_info, - bb_counters[sb->s_blocksize_bits + 2]); - - sprintf(name, "ext4_groupinfo_%d", sb->s_blocksize_bits); - namep = kstrdup(name, GFP_KERNEL); - if (!namep) { - ret = -ENOMEM; - goto out; - } - - /* Need to free the kmem_cache_name() when we - * destroy the slab */ - cachep = kmem_cache_create(namep, len, 0, - SLAB_RECLAIM_ACCOUNT, NULL); - if (!cachep) { - ret = -ENOMEM; - goto out; - } - ext4_groupinfo_caches[cache_index] = cachep; - } + ret = ext4_groupinfo_create_slab(sb->s_blocksize); + if (ret < 0) + goto out; /* order 0 is regular bitmap */ sbi->s_mb_maxs[0] = sb->s_blocksize << 3; @@ -2520,7 +2550,6 @@ out: if (ret) { kfree(sbi->s_mb_offsets); kfree(sbi->s_mb_maxs); - kfree(namep); } return ret; } @@ -2734,7 +2763,6 @@ int __init ext4_init_mballoc(void) void ext4_exit_mballoc(void) { - int i; /* * Wait for completion of call_rcu()'s on ext4_pspace_cachep * before destroying the slab cache. @@ -2743,15 +2771,7 @@ void ext4_exit_mballoc(void) kmem_cache_destroy(ext4_pspace_cachep); kmem_cache_destroy(ext4_ac_cachep); kmem_cache_destroy(ext4_free_ext_cachep); - - for (i = 0; i < NR_GRPINFO_CACHES; i++) { - struct kmem_cache *cachep = ext4_groupinfo_caches[i]; - if (cachep) { - char *name = (char *)kmem_cache_name(cachep); - kmem_cache_destroy(cachep); - kfree(name); - } - } + ext4_groupinfo_destroy_slabs(); ext4_remove_debugfs_entry(); } diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 5485390..e781b7e 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -2304,13 +2304,6 @@ static int ext4_link(struct dentry *old_dentry, dquot_initialize(dir); - /* - * Return -ENOENT if we've raced with unlink and i_nlink is 0. Doing - * otherwise has the potential to corrupt the orphan inode list. - */ - if (inode->i_nlink == 0) - return -ENOENT; - retry: handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + EXT4_INDEX_EXTRA_TRANS_BLOCKS); diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 7270dcf..955cc30 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -32,14 +32,8 @@ static struct kmem_cache *io_page_cachep, *io_end_cachep; -#define WQ_HASH_SZ 37 -#define to_ioend_wq(v) (&ioend_wq[((unsigned long)v) % WQ_HASH_SZ]) -static wait_queue_head_t ioend_wq[WQ_HASH_SZ]; - int __init ext4_init_pageio(void) { - int i; - io_page_cachep = KMEM_CACHE(ext4_io_page, SLAB_RECLAIM_ACCOUNT); if (io_page_cachep == NULL) return -ENOMEM; @@ -48,9 +42,6 @@ int __init ext4_init_pageio(void) kmem_cache_destroy(io_page_cachep); return -ENOMEM; } - for (i = 0; i < WQ_HASH_SZ; i++) - init_waitqueue_head(&ioend_wq[i]); - return 0; } @@ -62,7 +53,7 @@ void ext4_exit_pageio(void) void ext4_ioend_wait(struct inode *inode) { - wait_queue_head_t *wq = to_ioend_wq(inode); + wait_queue_head_t *wq = ext4_ioend_wq(inode); wait_event(*wq, (atomic_read(&EXT4_I(inode)->i_ioend_count) == 0)); } @@ -87,7 +78,7 @@ void ext4_free_io_end(ext4_io_end_t *io) for (i = 0; i < io->num_io_pages; i++) put_io_page(io->pages[i]); io->num_io_pages = 0; - wq = to_ioend_wq(io->inode); + wq = ext4_ioend_wq(io->inode); if (atomic_dec_and_test(&EXT4_I(io->inode)->i_ioend_count) && waitqueue_active(wq)) wake_up_all(wq); @@ -102,6 +93,7 @@ int ext4_end_io_nolock(ext4_io_end_t *io) struct inode *inode = io->inode; loff_t offset = io->offset; ssize_t size = io->size; + wait_queue_head_t *wq; int ret = 0; ext4_debug("ext4_end_io_nolock: io 0x%p from inode %lu,list->next 0x%p," @@ -126,7 +118,16 @@ int ext4_end_io_nolock(ext4_io_end_t *io) if (io->iocb) aio_complete(io->iocb, io->result, 0); /* clear the DIO AIO unwritten flag */ - io->flag &= ~EXT4_IO_END_UNWRITTEN; + if (io->flag & EXT4_IO_END_UNWRITTEN) { + io->flag &= ~EXT4_IO_END_UNWRITTEN; + /* Wake up anyone waiting on unwritten extent conversion */ + wq = ext4_ioend_wq(io->inode); + if (atomic_dec_and_test(&EXT4_I(inode)->i_aiodio_unwritten) && + waitqueue_active(wq)) { + wake_up_all(wq); + } + } + return ret; } @@ -190,6 +191,7 @@ static void ext4_end_bio(struct bio *bio, int error) struct inode *inode; unsigned long flags; int i; + sector_t bi_sector = bio->bi_sector; BUG_ON(!io_end); bio->bi_private = NULL; @@ -207,9 +209,7 @@ static void ext4_end_bio(struct bio *bio, int error) if (error) SetPageError(page); BUG_ON(!head); - if (head->b_size == PAGE_CACHE_SIZE) - clear_buffer_dirty(head); - else { + if (head->b_size != PAGE_CACHE_SIZE) { loff_t offset; loff_t io_end_offset = io_end->offset + io_end->size; @@ -221,7 +221,6 @@ static void ext4_end_bio(struct bio *bio, int error) if (error) buffer_io_error(bh); - clear_buffer_dirty(bh); } if (buffer_delay(bh)) partial_write = 1; @@ -257,7 +256,7 @@ static void ext4_end_bio(struct bio *bio, int error) (unsigned long long) io_end->offset, (long) io_end->size, (unsigned long long) - bio->bi_sector >> (inode->i_blkbits - 9)); + bi_sector >> (inode->i_blkbits - 9)); } /* Add the io_end to per-inode completed io list*/ @@ -380,6 +379,7 @@ int ext4_bio_write_page(struct ext4_io_submit *io, blocksize = 1 << inode->i_blkbits; + BUG_ON(!PageLocked(page)); BUG_ON(PageWriteback(page)); set_page_writeback(page); ClearPageError(page); @@ -397,12 +397,14 @@ int ext4_bio_write_page(struct ext4_io_submit *io, for (bh = head = page_buffers(page), block_start = 0; bh != head || !block_start; block_start = block_end, bh = bh->b_this_page) { + block_end = block_start + blocksize; if (block_start >= len) { clear_buffer_dirty(bh); set_buffer_uptodate(bh); continue; } + clear_buffer_dirty(bh); ret = io_submit_add_bh(io, io_page, inode, wbc, bh); if (ret) { /* diff --git a/fs/ext4/super.c b/fs/ext4/super.c index cb10a06..203f9e4 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -77,6 +77,7 @@ static struct dentry *ext4_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *data); static void ext4_destroy_lazyinit_thread(void); static void ext4_unregister_li_request(struct super_block *sb); +static void ext4_clear_request_list(void); #if !defined(CONFIG_EXT3_FS) && !defined(CONFIG_EXT3_FS_MODULE) && defined(CONFIG_EXT4_USE_FOR_EXT23) static struct file_system_type ext3_fs_type = { @@ -832,6 +833,7 @@ static struct inode *ext4_alloc_inode(struct super_block *sb) ei->i_sync_tid = 0; ei->i_datasync_tid = 0; atomic_set(&ei->i_ioend_count, 0); + atomic_set(&ei->i_aiodio_unwritten, 0); return &ei->vfs_inode; } @@ -1161,7 +1163,7 @@ static int ext4_release_dquot(struct dquot *dquot); static int ext4_mark_dquot_dirty(struct dquot *dquot); static int ext4_write_info(struct super_block *sb, int type); static int ext4_quota_on(struct super_block *sb, int type, int format_id, - char *path); + struct path *path); static int ext4_quota_off(struct super_block *sb, int type); static int ext4_quota_on_mount(struct super_block *sb, int type); static ssize_t ext4_quota_read(struct super_block *sb, int type, char *data, @@ -2716,6 +2718,8 @@ static void ext4_unregister_li_request(struct super_block *sb) mutex_unlock(&ext4_li_info->li_list_mtx); } +static struct task_struct *ext4_lazyinit_task; + /* * This is the function where ext4lazyinit thread lives. It walks * through the request list searching for next scheduled filesystem. @@ -2784,6 +2788,10 @@ cont_thread: if (time_before(jiffies, next_wakeup)) schedule(); finish_wait(&eli->li_wait_daemon, &wait); + if (kthread_should_stop()) { + ext4_clear_request_list(); + goto exit_thread; + } } exit_thread: @@ -2808,6 +2816,7 @@ exit_thread: wake_up(&eli->li_wait_task); kfree(ext4_li_info); + ext4_lazyinit_task = NULL; ext4_li_info = NULL; mutex_unlock(&ext4_li_mtx); @@ -2830,11 +2839,10 @@ static void ext4_clear_request_list(void) static int ext4_run_lazyinit_thread(void) { - struct task_struct *t; - - t = kthread_run(ext4_lazyinit_thread, ext4_li_info, "ext4lazyinit"); - if (IS_ERR(t)) { - int err = PTR_ERR(t); + ext4_lazyinit_task = kthread_run(ext4_lazyinit_thread, + ext4_li_info, "ext4lazyinit"); + if (IS_ERR(ext4_lazyinit_task)) { + int err = PTR_ERR(ext4_lazyinit_task); ext4_clear_request_list(); del_timer_sync(&ext4_li_info->li_timer); kfree(ext4_li_info); @@ -2985,16 +2993,10 @@ static void ext4_destroy_lazyinit_thread(void) * If thread exited earlier * there's nothing to be done. */ - if (!ext4_li_info) + if (!ext4_li_info || !ext4_lazyinit_task) return; - ext4_clear_request_list(); - - while (ext4_li_info->li_task) { - wake_up(&ext4_li_info->li_wait_daemon); - wait_event(ext4_li_info->li_wait_task, - ext4_li_info->li_task == NULL); - } + kthread_stop(ext4_lazyinit_task); } static int ext4_fill_super(struct super_block *sb, void *data, int silent) @@ -3413,6 +3415,8 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) sb->s_qcop = &ext4_qctl_operations; sb->dq_op = &ext4_quota_operations; #endif + memcpy(sb->s_uuid, es->s_uuid, sizeof(es->s_uuid)); + INIT_LIST_HEAD(&sbi->s_orphan); /* unlinked but open files */ mutex_init(&sbi->s_orphan_lock); mutex_init(&sbi->s_resize_lock); @@ -3507,7 +3511,12 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) percpu_counter_set(&sbi->s_dirtyblocks_counter, 0); no_journal: - EXT4_SB(sb)->dio_unwritten_wq = create_workqueue("ext4-dio-unwritten"); + /* + * The maximum number of concurrent works can be high and + * concurrency isn't really necessary. Limit it to 1. + */ + EXT4_SB(sb)->dio_unwritten_wq = + alloc_workqueue("ext4-dio-unwritten", WQ_MEM_RECLAIM, 1); if (!EXT4_SB(sb)->dio_unwritten_wq) { printk(KERN_ERR "EXT4-fs: failed to create DIO workqueue\n"); goto failed_mount_wq; @@ -4558,27 +4567,20 @@ static int ext4_quota_on_mount(struct super_block *sb, int type) * Standard function to be called on quota_on */ static int ext4_quota_on(struct super_block *sb, int type, int format_id, - char *name) + struct path *path) { int err; - struct path path; if (!test_opt(sb, QUOTA)) return -EINVAL; - err = kern_path(name, LOOKUP_FOLLOW, &path); - if (err) - return err; - /* Quotafile not on the same filesystem? */ - if (path.mnt->mnt_sb != sb) { - path_put(&path); + if (path->mnt->mnt_sb != sb) return -EXDEV; - } /* Journaling quota? */ if (EXT4_SB(sb)->s_qf_names[type]) { /* Quotafile not in fs root? */ - if (path.dentry->d_parent != sb->s_root) + if (path->dentry->d_parent != sb->s_root) ext4_msg(sb, KERN_WARNING, "Quota file not on filesystem root. " "Journaled quota will not work"); @@ -4589,7 +4591,7 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id, * all updates to the file when we bypass pagecache... */ if (EXT4_SB(sb)->s_journal && - ext4_should_journal_data(path.dentry->d_inode)) { + ext4_should_journal_data(path->dentry->d_inode)) { /* * We don't need to lock updates but journal_flush() could * otherwise be livelocked... @@ -4597,15 +4599,11 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id, jbd2_journal_lock_updates(EXT4_SB(sb)->s_journal); err = jbd2_journal_flush(EXT4_SB(sb)->s_journal); jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal); - if (err) { - path_put(&path); + if (err) return err; - } } - err = dquot_quota_on_path(sb, type, format_id, &path); - path_put(&path); - return err; + return dquot_quota_on(sb, type, format_id, path); } static int ext4_quota_off(struct super_block *sb, int type) @@ -4779,7 +4777,7 @@ static struct file_system_type ext4_fs_type = { .fs_flags = FS_REQUIRES_DEV, }; -int __init ext4_init_feat_adverts(void) +static int __init ext4_init_feat_adverts(void) { struct ext4_features *ef; int ret = -ENOMEM; @@ -4803,23 +4801,44 @@ out: return ret; } +static void ext4_exit_feat_adverts(void) +{ + kobject_put(&ext4_feat->f_kobj); + wait_for_completion(&ext4_feat->f_kobj_unregister); + kfree(ext4_feat); +} + +/* Shared across all ext4 file systems */ +wait_queue_head_t ext4__ioend_wq[EXT4_WQ_HASH_SZ]; +struct mutex ext4__aio_mutex[EXT4_WQ_HASH_SZ]; + static int __init ext4_init_fs(void) { - int err; + int i, err; ext4_check_flag_values(); + + for (i = 0; i < EXT4_WQ_HASH_SZ; i++) { + mutex_init(&ext4__aio_mutex[i]); + init_waitqueue_head(&ext4__ioend_wq[i]); + } + err = ext4_init_pageio(); if (err) return err; err = ext4_init_system_zone(); if (err) - goto out5; + goto out7; ext4_kset = kset_create_and_add("ext4", NULL, fs_kobj); if (!ext4_kset) - goto out4; + goto out6; ext4_proc_root = proc_mkdir("fs/ext4", NULL); + if (!ext4_proc_root) + goto out5; err = ext4_init_feat_adverts(); + if (err) + goto out4; err = ext4_init_mballoc(); if (err) @@ -4849,12 +4868,14 @@ out1: out2: ext4_exit_mballoc(); out3: - kfree(ext4_feat); + ext4_exit_feat_adverts(); +out4: remove_proc_entry("fs/ext4", NULL); +out5: kset_unregister(ext4_kset); -out4: +out6: ext4_exit_system_zone(); -out5: +out7: ext4_exit_pageio(); return err; } @@ -4868,6 +4889,7 @@ static void __exit ext4_exit_fs(void) destroy_inodecache(); ext4_exit_xattr(); ext4_exit_mballoc(); + ext4_exit_feat_adverts(); remove_proc_entry("fs/ext4", NULL); kset_unregister(ext4_kset); ext4_exit_system_zone(); diff --git a/fs/ext4/xattr.h b/fs/ext4/xattr.h index 1ef1652..25b7387 100644 --- a/fs/ext4/xattr.h +++ b/fs/ext4/xattr.h @@ -145,10 +145,10 @@ ext4_expand_extra_isize_ea(struct inode *inode, int new_extra_isize, #ifdef CONFIG_EXT4_FS_SECURITY extern int ext4_init_security(handle_t *handle, struct inode *inode, - struct inode *dir); + struct inode *dir, const struct qstr *qstr); #else static inline int ext4_init_security(handle_t *handle, struct inode *inode, - struct inode *dir) + struct inode *dir, const struct qstr *qstr) { return 0; } diff --git a/fs/ext4/xattr_security.c b/fs/ext4/xattr_security.c index 9b21268..007c3bf 100644 --- a/fs/ext4/xattr_security.c +++ b/fs/ext4/xattr_security.c @@ -49,14 +49,15 @@ ext4_xattr_security_set(struct dentry *dentry, const char *name, } int -ext4_init_security(handle_t *handle, struct inode *inode, struct inode *dir) +ext4_init_security(handle_t *handle, struct inode *inode, struct inode *dir, + const struct qstr *qstr) { int err; size_t len; void *value; char *name; - err = security_inode_init_security(inode, dir, &name, &value, &len); + err = security_inode_init_security(inode, dir, qstr, &name, &value, &len); if (err) { if (err == -EOPNOTSUPP) return 0; diff --git a/fs/fat/inode.c b/fs/fat/inode.c index 86753fe..0e277ec 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -757,8 +757,10 @@ fat_encode_fh(struct dentry *de, __u32 *fh, int *lenp, int connectable) struct inode *inode = de->d_inode; u32 ipos_h, ipos_m, ipos_l; - if (len < 5) + if (len < 5) { + *lenp = 5; return 255; /* no room */ + } ipos_h = MSDOS_I(inode)->i_pos >> 8; ipos_m = (MSDOS_I(inode)->i_pos & 0xf0) << 24; diff --git a/fs/fat/namei_vfat.c b/fs/fat/namei_vfat.c index f88f752..adae3fb 100644 --- a/fs/fat/namei_vfat.c +++ b/fs/fat/namei_vfat.c @@ -43,7 +43,7 @@ static int vfat_revalidate_shortname(struct dentry *dentry) static int vfat_revalidate(struct dentry *dentry, struct nameidata *nd) { - if (nd->flags & LOOKUP_RCU) + if (nd && nd->flags & LOOKUP_RCU) return -ECHILD; /* This is not negative dentry. Always valid. */ @@ -54,7 +54,7 @@ static int vfat_revalidate(struct dentry *dentry, struct nameidata *nd) static int vfat_revalidate_ci(struct dentry *dentry, struct nameidata *nd) { - if (nd->flags & LOOKUP_RCU) + if (nd && nd->flags & LOOKUP_RCU) return -ECHILD; /* @@ -131,7 +131,7 @@ SYSCALL_DEFINE2(dup2, unsigned int, oldfd, unsigned int, newfd) SYSCALL_DEFINE1(dup, unsigned int, fildes) { int ret = -EBADF; - struct file *file = fget(fildes); + struct file *file = fget_raw(fildes); if (file) { ret = get_unused_fd(); @@ -426,15 +426,35 @@ static long do_fcntl(int fd, unsigned int cmd, unsigned long arg, return err; } +static int check_fcntl_cmd(unsigned cmd) +{ + switch (cmd) { + case F_DUPFD: + case F_DUPFD_CLOEXEC: + case F_GETFD: + case F_SETFD: + case F_GETFL: + return 1; + } + return 0; +} + SYSCALL_DEFINE3(fcntl, unsigned int, fd, unsigned int, cmd, unsigned long, arg) { struct file *filp; long err = -EBADF; - filp = fget(fd); + filp = fget_raw(fd); if (!filp) goto out; + if (unlikely(filp->f_mode & FMODE_PATH)) { + if (!check_fcntl_cmd(cmd)) { + fput(filp); + goto out; + } + } + err = security_file_fcntl(filp, cmd, arg); if (err) { fput(filp); @@ -456,10 +476,17 @@ SYSCALL_DEFINE3(fcntl64, unsigned int, fd, unsigned int, cmd, long err; err = -EBADF; - filp = fget(fd); + filp = fget_raw(fd); if (!filp) goto out; + if (unlikely(filp->f_mode & FMODE_PATH)) { + if (!check_fcntl_cmd(cmd)) { + fput(filp); + goto out; + } + } + err = security_file_fcntl(filp, cmd, arg); if (err) { fput(filp); @@ -808,14 +835,14 @@ static int __init fcntl_init(void) * Exceptions: O_NONBLOCK is a two bit define on parisc; O_NDELAY * is defined as O_NONBLOCK on some platforms and not on others. */ - BUILD_BUG_ON(18 - 1 /* for O_RDONLY being 0 */ != HWEIGHT32( + BUILD_BUG_ON(19 - 1 /* for O_RDONLY being 0 */ != HWEIGHT32( O_RDONLY | O_WRONLY | O_RDWR | O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC | O_APPEND | /* O_NONBLOCK | */ __O_SYNC | O_DSYNC | FASYNC | O_DIRECT | O_LARGEFILE | O_DIRECTORY | O_NOFOLLOW | O_NOATIME | O_CLOEXEC | - FMODE_EXEC + __FMODE_EXEC | O_PATH )); fasync_cache = kmem_cache_create("fasync_cache", diff --git a/fs/fhandle.c b/fs/fhandle.c new file mode 100644 index 0000000..bf93ad2 --- /dev/null +++ b/fs/fhandle.c @@ -0,0 +1,265 @@ +#include <linux/syscalls.h> +#include <linux/slab.h> +#include <linux/fs.h> +#include <linux/file.h> +#include <linux/mount.h> +#include <linux/namei.h> +#include <linux/exportfs.h> +#include <linux/fs_struct.h> +#include <linux/fsnotify.h> +#include <asm/uaccess.h> +#include "internal.h" + +static long do_sys_name_to_handle(struct path *path, + struct file_handle __user *ufh, + int __user *mnt_id) +{ + long retval; + struct file_handle f_handle; + int handle_dwords, handle_bytes; + struct file_handle *handle = NULL; + + /* + * We need t make sure wether the file system + * support decoding of the file handle + */ + if (!path->mnt->mnt_sb->s_export_op || + !path->mnt->mnt_sb->s_export_op->fh_to_dentry) + return -EOPNOTSUPP; + + if (copy_from_user(&f_handle, ufh, sizeof(struct file_handle))) + return -EFAULT; + + if (f_handle.handle_bytes > MAX_HANDLE_SZ) + return -EINVAL; + + handle = kmalloc(sizeof(struct file_handle) + f_handle.handle_bytes, + GFP_KERNEL); + if (!handle) + return -ENOMEM; + + /* convert handle size to multiple of sizeof(u32) */ + handle_dwords = f_handle.handle_bytes >> 2; + + /* we ask for a non connected handle */ + retval = exportfs_encode_fh(path->dentry, + (struct fid *)handle->f_handle, + &handle_dwords, 0); + handle->handle_type = retval; + /* convert handle size to bytes */ + handle_bytes = handle_dwords * sizeof(u32); + handle->handle_bytes = handle_bytes; + if ((handle->handle_bytes > f_handle.handle_bytes) || + (retval == 255) || (retval == -ENOSPC)) { + /* As per old exportfs_encode_fh documentation + * we could return ENOSPC to indicate overflow + * But file system returned 255 always. So handle + * both the values + */ + /* + * set the handle size to zero so we copy only + * non variable part of the file_handle + */ + handle_bytes = 0; + retval = -EOVERFLOW; + } else + retval = 0; + /* copy the mount id */ + if (copy_to_user(mnt_id, &path->mnt->mnt_id, sizeof(*mnt_id)) || + copy_to_user(ufh, handle, + sizeof(struct file_handle) + handle_bytes)) + retval = -EFAULT; + kfree(handle); + return retval; +} + +/** + * sys_name_to_handle_at: convert name to handle + * @dfd: directory relative to which name is interpreted if not absolute + * @name: name that should be converted to handle. + * @handle: resulting file handle + * @mnt_id: mount id of the file system containing the file + * @flag: flag value to indicate whether to follow symlink or not + * + * @handle->handle_size indicate the space available to store the + * variable part of the file handle in bytes. If there is not + * enough space, the field is updated to return the minimum + * value required. + */ +SYSCALL_DEFINE5(name_to_handle_at, int, dfd, const char __user *, name, + struct file_handle __user *, handle, int __user *, mnt_id, + int, flag) +{ + struct path path; + int lookup_flags; + int err; + + if ((flag & ~(AT_SYMLINK_FOLLOW | AT_EMPTY_PATH)) != 0) + return -EINVAL; + + lookup_flags = (flag & AT_SYMLINK_FOLLOW) ? LOOKUP_FOLLOW : 0; + if (flag & AT_EMPTY_PATH) + lookup_flags |= LOOKUP_EMPTY; + err = user_path_at(dfd, name, lookup_flags, &path); + if (!err) { + err = do_sys_name_to_handle(&path, handle, mnt_id); + path_put(&path); + } + return err; +} + +static struct vfsmount *get_vfsmount_from_fd(int fd) +{ + struct path path; + + if (fd == AT_FDCWD) { + struct fs_struct *fs = current->fs; + spin_lock(&fs->lock); + path = fs->pwd; + mntget(path.mnt); + spin_unlock(&fs->lock); + } else { + int fput_needed; + struct file *file = fget_light(fd, &fput_needed); + if (!file) + return ERR_PTR(-EBADF); + path = file->f_path; + mntget(path.mnt); + fput_light(file, fput_needed); + } + return path.mnt; +} + +static int vfs_dentry_acceptable(void *context, struct dentry *dentry) +{ + return 1; +} + +static int do_handle_to_path(int mountdirfd, struct file_handle *handle, + struct path *path) +{ + int retval = 0; + int handle_dwords; + + path->mnt = get_vfsmount_from_fd(mountdirfd); + if (IS_ERR(path->mnt)) { + retval = PTR_ERR(path->mnt); + goto out_err; + } + /* change the handle size to multiple of sizeof(u32) */ + handle_dwords = handle->handle_bytes >> 2; + path->dentry = exportfs_decode_fh(path->mnt, + (struct fid *)handle->f_handle, + handle_dwords, handle->handle_type, + vfs_dentry_acceptable, NULL); + if (IS_ERR(path->dentry)) { + retval = PTR_ERR(path->dentry); + goto out_mnt; + } + return 0; +out_mnt: + mntput(path->mnt); +out_err: + return retval; +} + +static int handle_to_path(int mountdirfd, struct file_handle __user *ufh, + struct path *path) +{ + int retval = 0; + struct file_handle f_handle; + struct file_handle *handle = NULL; + + /* + * With handle we don't look at the execute bit on the + * the directory. Ideally we would like CAP_DAC_SEARCH. + * But we don't have that + */ + if (!capable(CAP_DAC_READ_SEARCH)) { + retval = -EPERM; + goto out_err; + } + if (copy_from_user(&f_handle, ufh, sizeof(struct file_handle))) { + retval = -EFAULT; + goto out_err; + } + if ((f_handle.handle_bytes > MAX_HANDLE_SZ) || + (f_handle.handle_bytes == 0)) { + retval = -EINVAL; + goto out_err; + } + handle = kmalloc(sizeof(struct file_handle) + f_handle.handle_bytes, + GFP_KERNEL); + if (!handle) { + retval = -ENOMEM; + goto out_err; + } + /* copy the full handle */ + if (copy_from_user(handle, ufh, + sizeof(struct file_handle) + + f_handle.handle_bytes)) { + retval = -EFAULT; + goto out_handle; + } + + retval = do_handle_to_path(mountdirfd, handle, path); + +out_handle: + kfree(handle); +out_err: + return retval; +} + +long do_handle_open(int mountdirfd, + struct file_handle __user *ufh, int open_flag) +{ + long retval = 0; + struct path path; + struct file *file; + int fd; + + retval = handle_to_path(mountdirfd, ufh, &path); + if (retval) + return retval; + + fd = get_unused_fd_flags(open_flag); + if (fd < 0) { + path_put(&path); + return fd; + } + file = file_open_root(path.dentry, path.mnt, "", open_flag); + if (IS_ERR(file)) { + put_unused_fd(fd); + retval = PTR_ERR(file); + } else { + retval = fd; + fsnotify_open(file); + fd_install(fd, file); + } + path_put(&path); + return retval; +} + +/** + * sys_open_by_handle_at: Open the file handle + * @mountdirfd: directory file descriptor + * @handle: file handle to be opened + * @flag: open flags. + * + * @mountdirfd indicate the directory file descriptor + * of the mount point. file handle is decoded relative + * to the vfsmount pointed by the @mountdirfd. @flags + * value is same as the open(2) flags. + */ +SYSCALL_DEFINE3(open_by_handle_at, int, mountdirfd, + struct file_handle __user *, handle, + int, flags) +{ + long ret; + + if (force_o_largefile()) + flags |= O_LARGEFILE; + + ret = do_handle_open(mountdirfd, handle, flags); + return ret; +} @@ -66,8 +66,7 @@ static int fifo_open(struct inode *inode, struct file *filp) /* suppress POLLHUP until we have * seen a writer */ filp->f_version = pipe->w_counter; - } else - { + } else { wait_for_partner(inode, &pipe->w_counter); if(signal_pending(current)) goto err_rd; diff --git a/fs/file_table.c b/fs/file_table.c index c3e89ad..01e4c1e 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -125,13 +125,13 @@ struct file *get_empty_filp(void) goto fail; percpu_counter_inc(&nr_files); + f->f_cred = get_cred(cred); if (security_file_alloc(f)) goto fail_sec; INIT_LIST_HEAD(&f->f_u.fu_list); atomic_long_set(&f->f_count, 1); rwlock_init(&f->f_owner.lock); - f->f_cred = get_cred(cred); spin_lock_init(&f->f_lock); eventpoll_init_file(f); /* f->f_version: 0 */ @@ -190,7 +190,8 @@ struct file *alloc_file(struct path *path, fmode_t mode, file_take_write(file); WARN_ON(mnt_clone_write(path->mnt)); } - ima_counts_get(file); + if ((mode & (FMODE_READ | FMODE_WRITE)) == FMODE_READ) + i_readcount_inc(path->dentry->d_inode); return file; } EXPORT_SYMBOL(alloc_file); @@ -246,11 +247,15 @@ static void __fput(struct file *file) file->f_op->release(inode, file); security_file_free(file); ima_file_free(file); - if (unlikely(S_ISCHR(inode->i_mode) && inode->i_cdev != NULL)) + if (unlikely(S_ISCHR(inode->i_mode) && inode->i_cdev != NULL && + !(file->f_mode & FMODE_PATH))) { cdev_put(inode->i_cdev); + } fops_put(file->f_op); put_pid(file->f_owner.pid); file_sb_list_del(file); + if ((file->f_mode & (FMODE_READ | FMODE_WRITE)) == FMODE_READ) + i_readcount_dec(inode); if (file->f_mode & FMODE_WRITE) drop_file_write_access(file); file->f_path.dentry = NULL; @@ -276,11 +281,10 @@ struct file *fget(unsigned int fd) rcu_read_lock(); file = fcheck_files(files, fd); if (file) { - if (!atomic_long_inc_not_zero(&file->f_count)) { - /* File object ref couldn't be taken */ - rcu_read_unlock(); - return NULL; - } + /* File object ref couldn't be taken */ + if (file->f_mode & FMODE_PATH || + !atomic_long_inc_not_zero(&file->f_count)) + file = NULL; } rcu_read_unlock(); @@ -289,6 +293,25 @@ struct file *fget(unsigned int fd) EXPORT_SYMBOL(fget); +struct file *fget_raw(unsigned int fd) +{ + struct file *file; + struct files_struct *files = current->files; + + rcu_read_lock(); + file = fcheck_files(files, fd); + if (file) { + /* File object ref couldn't be taken */ + if (!atomic_long_inc_not_zero(&file->f_count)) + file = NULL; + } + rcu_read_unlock(); + + return file; +} + +EXPORT_SYMBOL(fget_raw); + /* * Lightweight file lookup - no refcnt increment if fd table isn't shared. * @@ -313,6 +336,33 @@ struct file *fget_light(unsigned int fd, int *fput_needed) *fput_needed = 0; if (atomic_read(&files->count) == 1) { file = fcheck_files(files, fd); + if (file && (file->f_mode & FMODE_PATH)) + file = NULL; + } else { + rcu_read_lock(); + file = fcheck_files(files, fd); + if (file) { + if (!(file->f_mode & FMODE_PATH) && + atomic_long_inc_not_zero(&file->f_count)) + *fput_needed = 1; + else + /* Didn't get the reference, someone's freed */ + file = NULL; + } + rcu_read_unlock(); + } + + return file; +} + +struct file *fget_raw_light(unsigned int fd, int *fput_needed) +{ + struct file *file; + struct files_struct *files = current->files; + + *fput_needed = 0; + if (atomic_read(&files->count) == 1) { + file = fcheck_files(files, fd); } else { rcu_read_lock(); file = fcheck_files(files, fd); diff --git a/fs/fuse/cuse.c b/fs/fuse/cuse.c index 3e87cce..7c39b885 100644 --- a/fs/fuse/cuse.c +++ b/fs/fuse/cuse.c @@ -458,7 +458,7 @@ static void cuse_fc_release(struct fuse_conn *fc) * @file: file struct being opened * * Userland CUSE server can create a CUSE device by opening /dev/cuse - * and replying to the initilaization request kernel sends. This + * and replying to the initialization request kernel sends. This * function is responsible for handling CUSE device initialization. * Because the fd opened by this function is used during * initialization, this function only creates cuse_conn and sends diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index bfed844..8bd0ef9 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -158,7 +158,7 @@ static int fuse_dentry_revalidate(struct dentry *entry, struct nameidata *nd) { struct inode *inode; - if (nd->flags & LOOKUP_RCU) + if (nd && nd->flags & LOOKUP_RCU) return -ECHILD; inode = entry->d_inode; @@ -1283,8 +1283,11 @@ static int fuse_do_setattr(struct dentry *entry, struct iattr *attr, if (err) return err; - if ((attr->ia_valid & ATTR_OPEN) && fc->atomic_o_trunc) - return 0; + if (attr->ia_valid & ATTR_OPEN) { + if (fc->atomic_o_trunc) + return 0; + file = NULL; + } if (attr->ia_valid & ATTR_SIZE) is_truncate = true; diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 95da1bc..9e0832d 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -86,18 +86,52 @@ struct fuse_file *fuse_file_get(struct fuse_file *ff) return ff; } +static void fuse_release_async(struct work_struct *work) +{ + struct fuse_req *req; + struct fuse_conn *fc; + struct path path; + + req = container_of(work, struct fuse_req, misc.release.work); + path = req->misc.release.path; + fc = get_fuse_conn(path.dentry->d_inode); + + fuse_put_request(fc, req); + path_put(&path); +} + static void fuse_release_end(struct fuse_conn *fc, struct fuse_req *req) { - path_put(&req->misc.release.path); + if (fc->destroy_req) { + /* + * If this is a fuseblk mount, then it's possible that + * releasing the path will result in releasing the + * super block and sending the DESTROY request. If + * the server is single threaded, this would hang. + * For this reason do the path_put() in a separate + * thread. + */ + atomic_inc(&req->count); + INIT_WORK(&req->misc.release.work, fuse_release_async); + schedule_work(&req->misc.release.work); + } else { + path_put(&req->misc.release.path); + } } -static void fuse_file_put(struct fuse_file *ff) +static void fuse_file_put(struct fuse_file *ff, bool sync) { if (atomic_dec_and_test(&ff->count)) { struct fuse_req *req = ff->reserved_req; - req->end = fuse_release_end; - fuse_request_send_background(ff->fc, req); + if (sync) { + fuse_request_send(ff->fc, req); + path_put(&req->misc.release.path); + fuse_put_request(ff->fc, req); + } else { + req->end = fuse_release_end; + fuse_request_send_background(ff->fc, req); + } kfree(ff); } } @@ -219,8 +253,12 @@ void fuse_release_common(struct file *file, int opcode) * Normally this will send the RELEASE request, however if * some asynchronous READ or WRITE requests are outstanding, * the sending will be delayed. + * + * Make the release synchronous if this is a fuseblk mount, + * synchronous RELEASE is allowed (and desirable) in this case + * because the server can be trusted not to screw up. */ - fuse_file_put(ff); + fuse_file_put(ff, ff->fc->destroy_req != NULL); } static int fuse_open(struct inode *inode, struct file *file) @@ -558,7 +596,7 @@ static void fuse_readpages_end(struct fuse_conn *fc, struct fuse_req *req) page_cache_release(page); } if (req->ff) - fuse_file_put(req->ff); + fuse_file_put(req->ff, false); } static void fuse_send_readpages(struct fuse_req *req, struct file *file) @@ -1137,7 +1175,7 @@ static ssize_t fuse_direct_write(struct file *file, const char __user *buf, static void fuse_writepage_free(struct fuse_conn *fc, struct fuse_req *req) { __free_page(req->pages[0]); - fuse_file_put(req->ff); + fuse_file_put(req->ff, false); } static void fuse_writepage_finish(struct fuse_conn *fc, struct fuse_req *req) diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index ae5744a..d428694 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -21,6 +21,7 @@ #include <linux/rwsem.h> #include <linux/rbtree.h> #include <linux/poll.h> +#include <linux/workqueue.h> /** Max number of pages that can be used in a single read request */ #define FUSE_MAX_PAGES_PER_REQ 32 @@ -262,7 +263,10 @@ struct fuse_req { /** Data for asynchronous requests */ union { struct { - struct fuse_release_in in; + union { + struct fuse_release_in in; + struct work_struct work; + }; struct path path; } release; struct fuse_init_in init_in; diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 9e3f68c..051b1a0 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -637,8 +637,10 @@ static int fuse_encode_fh(struct dentry *dentry, u32 *fh, int *max_len, u64 nodeid; u32 generation; - if (*max_len < len) + if (*max_len < len) { + *max_len = len; return 255; + } nodeid = get_fuse_inode(inode)->nodeid; generation = inode->i_generation; diff --git a/fs/gfs2/Makefile b/fs/gfs2/Makefile index 21f7e46..f3d23ef 100644 --- a/fs/gfs2/Makefile +++ b/fs/gfs2/Makefile @@ -1,4 +1,4 @@ -EXTRA_CFLAGS := -I$(src) +ccflags-y := -I$(src) obj-$(CONFIG_GFS2_FS) += gfs2.o gfs2-y := acl.o bmap.o dir.o xattr.o glock.o \ glops.o inode.o log.o lops.o main.o meta_io.o \ diff --git a/fs/gfs2/acl.c b/fs/gfs2/acl.c index 7118f1a..cbc0715 100644 --- a/fs/gfs2/acl.c +++ b/fs/gfs2/acl.c @@ -80,8 +80,11 @@ int gfs2_check_acl(struct inode *inode, int mask, unsigned int flags) struct posix_acl *acl; int error; - if (flags & IPERM_FLAG_RCU) - return -ECHILD; + if (flags & IPERM_FLAG_RCU) { + if (!negative_cached_acl(inode, ACL_TYPE_ACCESS)) + return -ECHILD; + return -EAGAIN; + } acl = gfs2_acl_get(GFS2_I(inode), ACL_TYPE_ACCESS); if (IS_ERR(acl)) diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 4f36f88..aad77e4 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -695,6 +695,7 @@ out: if (error == 0) return 0; + unlock_page(page); page_cache_release(page); gfs2_trans_end(sdp); diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index 3c4039d..ef3dc4b 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -21,6 +21,7 @@ #include "meta_io.h" #include "quota.h" #include "rgrp.h" +#include "super.h" #include "trans.h" #include "dir.h" #include "util.h" @@ -757,7 +758,7 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_rgrp_list rlist; u64 bn, bstart; - u32 blen; + u32 blen, btotal; __be64 *p; unsigned int rg_blocks = 0; int metadata; @@ -839,6 +840,7 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, bstart = 0; blen = 0; + btotal = 0; for (p = top; p < bottom; p++) { if (!*p) @@ -851,9 +853,11 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, else { if (bstart) { if (metadata) - gfs2_free_meta(ip, bstart, blen); + __gfs2_free_meta(ip, bstart, blen); else - gfs2_free_data(ip, bstart, blen); + __gfs2_free_data(ip, bstart, blen); + + btotal += blen; } bstart = bn; @@ -865,11 +869,17 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, } if (bstart) { if (metadata) - gfs2_free_meta(ip, bstart, blen); + __gfs2_free_meta(ip, bstart, blen); else - gfs2_free_data(ip, bstart, blen); + __gfs2_free_data(ip, bstart, blen); + + btotal += blen; } + gfs2_statfs_change(sdp, 0, +btotal, 0); + gfs2_quota_change(ip, -(s64)btotal, ip->i_inode.i_uid, + ip->i_inode.i_gid); + ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; gfs2_dinode_out(ip, dibh->b_data); diff --git a/fs/gfs2/dentry.c b/fs/gfs2/dentry.c index 4a45633..0da8da2 100644 --- a/fs/gfs2/dentry.c +++ b/fs/gfs2/dentry.c @@ -44,7 +44,7 @@ static int gfs2_drevalidate(struct dentry *dentry, struct nameidata *nd) int error; int had_lock = 0; - if (nd->flags & LOOKUP_RCU) + if (nd && nd->flags & LOOKUP_RCU) return -ECHILD; parent = dget_parent(dentry); diff --git a/fs/gfs2/export.c b/fs/gfs2/export.c index 9023db8..b5a5e60 100644 --- a/fs/gfs2/export.c +++ b/fs/gfs2/export.c @@ -36,9 +36,13 @@ static int gfs2_encode_fh(struct dentry *dentry, __u32 *p, int *len, struct super_block *sb = inode->i_sb; struct gfs2_inode *ip = GFS2_I(inode); - if (*len < GFS2_SMALL_FH_SIZE || - (connectable && *len < GFS2_LARGE_FH_SIZE)) + if (connectable && (*len < GFS2_LARGE_FH_SIZE)) { + *len = GFS2_LARGE_FH_SIZE; return 255; + } else if (*len < GFS2_SMALL_FH_SIZE) { + *len = GFS2_SMALL_FH_SIZE; + return 255; + } fh[0] = cpu_to_be32(ip->i_no_formal_ino >> 32); fh[1] = cpu_to_be32(ip->i_no_formal_ino & 0xFFFFFFFF); diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index 7cfdcb9..4074b95 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -448,15 +448,20 @@ static int gfs2_mmap(struct file *file, struct vm_area_struct *vma) { struct gfs2_inode *ip = GFS2_I(file->f_mapping->host); - if (!(file->f_flags & O_NOATIME)) { + if (!(file->f_flags & O_NOATIME) && + !IS_NOATIME(&ip->i_inode)) { struct gfs2_holder i_gh; int error; - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &i_gh); + gfs2_holder_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &i_gh); error = gfs2_glock_nq(&i_gh); - file_accessed(file); - if (error == 0) - gfs2_glock_dq_uninit(&i_gh); + if (error == 0) { + file_accessed(file); + gfs2_glock_dq(&i_gh); + } + gfs2_holder_uninit(&i_gh); + if (error) + return error; } vma->vm_ops = &gfs2_vm_ops; vma->vm_flags |= VM_CAN_NONLINEAR; @@ -617,8 +622,7 @@ static void empty_write_end(struct page *page, unsigned from, { struct gfs2_inode *ip = GFS2_I(page->mapping->host); - page_zero_new_buffers(page, from, to); - flush_dcache_page(page); + zero_user(page, from, to-from); mark_page_accessed(page); if (!gfs2_is_writeback(ip)) @@ -627,36 +631,43 @@ static void empty_write_end(struct page *page, unsigned from, block_commit_write(page, from, to); } -static int write_empty_blocks(struct page *page, unsigned from, unsigned to) +static int needs_empty_write(sector_t block, struct inode *inode) { - unsigned start, end, next; - struct buffer_head *bh, *head; int error; + struct buffer_head bh_map = { .b_state = 0, .b_blocknr = 0 }; - if (!page_has_buffers(page)) { - error = __block_write_begin(page, from, to - from, gfs2_block_map); - if (unlikely(error)) - return error; + bh_map.b_size = 1 << inode->i_blkbits; + error = gfs2_block_map(inode, block, &bh_map, 0); + if (unlikely(error)) + return error; + return !buffer_mapped(&bh_map); +} - empty_write_end(page, from, to); - return 0; - } +static int write_empty_blocks(struct page *page, unsigned from, unsigned to) +{ + struct inode *inode = page->mapping->host; + unsigned start, end, next, blksize; + sector_t block = page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits); + int ret; - bh = head = page_buffers(page); + blksize = 1 << inode->i_blkbits; next = end = 0; while (next < from) { - next += bh->b_size; - bh = bh->b_this_page; + next += blksize; + block++; } start = next; do { - next += bh->b_size; - if (buffer_mapped(bh)) { + next += blksize; + ret = needs_empty_write(block, inode); + if (unlikely(ret < 0)) + return ret; + if (ret == 0) { if (end) { - error = __block_write_begin(page, start, end - start, - gfs2_block_map); - if (unlikely(error)) - return error; + ret = __block_write_begin(page, start, end - start, + gfs2_block_map); + if (unlikely(ret)) + return ret; empty_write_end(page, start, end); end = 0; } @@ -664,13 +675,13 @@ static int write_empty_blocks(struct page *page, unsigned from, unsigned to) } else end = next; - bh = bh->b_this_page; + block++; } while (next < to); if (end) { - error = __block_write_begin(page, start, end - start, gfs2_block_map); - if (unlikely(error)) - return error; + ret = __block_write_begin(page, start, end - start, gfs2_block_map); + if (unlikely(ret)) + return ret; empty_write_end(page, start, end); } @@ -976,8 +987,10 @@ static void do_unflock(struct file *file, struct file_lock *fl) mutex_lock(&fp->f_fl_mutex); flock_lock_file_wait(file, fl); - if (fl_gh->gh_gl) - gfs2_glock_dq_uninit(fl_gh); + if (fl_gh->gh_gl) { + gfs2_glock_dq_wait(fl_gh); + gfs2_holder_uninit(fl_gh); + } mutex_unlock(&fp->f_fl_mutex); } diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index 08a8beb..e243131 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -26,6 +26,9 @@ #include <linux/freezer.h> #include <linux/workqueue.h> #include <linux/jiffies.h> +#include <linux/rcupdate.h> +#include <linux/rculist_bl.h> +#include <linux/bit_spinlock.h> #include "gfs2.h" #include "incore.h" @@ -41,10 +44,6 @@ #define CREATE_TRACE_POINTS #include "trace_gfs2.h" -struct gfs2_gl_hash_bucket { - struct hlist_head hb_list; -}; - struct gfs2_glock_iter { int hash; /* hash bucket index */ struct gfs2_sbd *sdp; /* incore superblock */ @@ -54,7 +53,6 @@ struct gfs2_glock_iter { typedef void (*glock_examiner) (struct gfs2_glock * gl); -static int gfs2_dump_lockstate(struct gfs2_sbd *sdp); static int __dump_glock(struct seq_file *seq, const struct gfs2_glock *gl); #define GLOCK_BUG_ON(gl,x) do { if (unlikely(x)) { __dump_glock(NULL, gl); BUG(); } } while(0) static void do_xmote(struct gfs2_glock *gl, struct gfs2_holder *gh, unsigned int target); @@ -70,57 +68,9 @@ static DEFINE_SPINLOCK(lru_lock); #define GFS2_GL_HASH_SIZE (1 << GFS2_GL_HASH_SHIFT) #define GFS2_GL_HASH_MASK (GFS2_GL_HASH_SIZE - 1) -static struct gfs2_gl_hash_bucket gl_hash_table[GFS2_GL_HASH_SIZE]; +static struct hlist_bl_head gl_hash_table[GFS2_GL_HASH_SIZE]; static struct dentry *gfs2_root; -/* - * Despite what you might think, the numbers below are not arbitrary :-) - * They are taken from the ipv4 routing hash code, which is well tested - * and thus should be nearly optimal. Later on we might tweek the numbers - * but for now this should be fine. - * - * The reason for putting the locks in a separate array from the list heads - * is that we can have fewer locks than list heads and save memory. We use - * the same hash function for both, but with a different hash mask. - */ -#if defined(CONFIG_SMP) || defined(CONFIG_DEBUG_SPINLOCK) || \ - defined(CONFIG_PROVE_LOCKING) - -#ifdef CONFIG_LOCKDEP -# define GL_HASH_LOCK_SZ 256 -#else -# if NR_CPUS >= 32 -# define GL_HASH_LOCK_SZ 4096 -# elif NR_CPUS >= 16 -# define GL_HASH_LOCK_SZ 2048 -# elif NR_CPUS >= 8 -# define GL_HASH_LOCK_SZ 1024 -# elif NR_CPUS >= 4 -# define GL_HASH_LOCK_SZ 512 -# else -# define GL_HASH_LOCK_SZ 256 -# endif -#endif - -/* We never want more locks than chains */ -#if GFS2_GL_HASH_SIZE < GL_HASH_LOCK_SZ -# undef GL_HASH_LOCK_SZ -# define GL_HASH_LOCK_SZ GFS2_GL_HASH_SIZE -#endif - -static rwlock_t gl_hash_locks[GL_HASH_LOCK_SZ]; - -static inline rwlock_t *gl_lock_addr(unsigned int x) -{ - return &gl_hash_locks[x & (GL_HASH_LOCK_SZ-1)]; -} -#else /* not SMP, so no spinlocks required */ -static inline rwlock_t *gl_lock_addr(unsigned int x) -{ - return NULL; -} -#endif - /** * gl_hash() - Turn glock number into hash bucket number * @lock: The glock number @@ -141,25 +91,35 @@ static unsigned int gl_hash(const struct gfs2_sbd *sdp, return h; } -/** - * glock_free() - Perform a few checks and then release struct gfs2_glock - * @gl: The glock to release - * - * Also calls lock module to release its internal structure for this glock. - * - */ +static inline void spin_lock_bucket(unsigned int hash) +{ + struct hlist_bl_head *bl = &gl_hash_table[hash]; + bit_spin_lock(0, (unsigned long *)bl); +} -static void glock_free(struct gfs2_glock *gl) +static inline void spin_unlock_bucket(unsigned int hash) +{ + struct hlist_bl_head *bl = &gl_hash_table[hash]; + __bit_spin_unlock(0, (unsigned long *)bl); +} + +static void gfs2_glock_dealloc(struct rcu_head *rcu) +{ + struct gfs2_glock *gl = container_of(rcu, struct gfs2_glock, gl_rcu); + + if (gl->gl_ops->go_flags & GLOF_ASPACE) + kmem_cache_free(gfs2_glock_aspace_cachep, gl); + else + kmem_cache_free(gfs2_glock_cachep, gl); +} + +void gfs2_glock_free(struct gfs2_glock *gl) { struct gfs2_sbd *sdp = gl->gl_sbd; - struct address_space *mapping = gfs2_glock2aspace(gl); - struct kmem_cache *cachep = gfs2_glock_cachep; - GLOCK_BUG_ON(gl, mapping && mapping->nrpages); - trace_gfs2_glock_put(gl); - if (mapping) - cachep = gfs2_glock_aspace_cachep; - sdp->sd_lockstruct.ls_ops->lm_put_lock(cachep, gl); + call_rcu(&gl->gl_rcu, gfs2_glock_dealloc); + if (atomic_dec_and_test(&sdp->sd_glock_disposal)) + wake_up(&sdp->sd_glock_wait); } /** @@ -185,34 +145,49 @@ static int demote_ok(const struct gfs2_glock *gl) { const struct gfs2_glock_operations *glops = gl->gl_ops; + /* assert_spin_locked(&gl->gl_spin); */ + if (gl->gl_state == LM_ST_UNLOCKED) return 0; - if (!list_empty(&gl->gl_holders)) + if (test_bit(GLF_LFLUSH, &gl->gl_flags)) + return 0; + if ((gl->gl_name.ln_type != LM_TYPE_INODE) && + !list_empty(&gl->gl_holders)) return 0; if (glops->go_demote_ok) return glops->go_demote_ok(gl); return 1; } + /** - * gfs2_glock_schedule_for_reclaim - Add a glock to the reclaim list + * __gfs2_glock_schedule_for_reclaim - Add a glock to the reclaim list * @gl: the glock * + * If the glock is demotable, then we add it (or move it) to the end + * of the glock LRU list. */ -static void gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl) +static void __gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl) { - int may_reclaim; - may_reclaim = (demote_ok(gl) && - (atomic_read(&gl->gl_ref) == 1 || - (gl->gl_name.ln_type == LM_TYPE_INODE && - atomic_read(&gl->gl_ref) <= 2))); - spin_lock(&lru_lock); - if (list_empty(&gl->gl_lru) && may_reclaim) { + if (demote_ok(gl)) { + spin_lock(&lru_lock); + + if (!list_empty(&gl->gl_lru)) + list_del_init(&gl->gl_lru); + else + atomic_inc(&lru_count); + list_add_tail(&gl->gl_lru, &lru_list); - atomic_inc(&lru_count); + spin_unlock(&lru_lock); } - spin_unlock(&lru_lock); +} + +void gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl) +{ + spin_lock(&gl->gl_spin); + __gfs2_glock_schedule_for_reclaim(gl); + spin_unlock(&gl->gl_spin); } /** @@ -227,7 +202,6 @@ void gfs2_glock_put_nolock(struct gfs2_glock *gl) { if (atomic_dec_and_test(&gl->gl_ref)) GLOCK_BUG_ON(gl, 1); - gfs2_glock_schedule_for_reclaim(gl); } /** @@ -236,30 +210,26 @@ void gfs2_glock_put_nolock(struct gfs2_glock *gl) * */ -int gfs2_glock_put(struct gfs2_glock *gl) +void gfs2_glock_put(struct gfs2_glock *gl) { - int rv = 0; + struct gfs2_sbd *sdp = gl->gl_sbd; + struct address_space *mapping = gfs2_glock2aspace(gl); - write_lock(gl_lock_addr(gl->gl_hash)); - if (atomic_dec_and_lock(&gl->gl_ref, &lru_lock)) { - hlist_del(&gl->gl_list); + if (atomic_dec_and_test(&gl->gl_ref)) { + spin_lock_bucket(gl->gl_hash); + hlist_bl_del_rcu(&gl->gl_list); + spin_unlock_bucket(gl->gl_hash); + spin_lock(&lru_lock); if (!list_empty(&gl->gl_lru)) { list_del_init(&gl->gl_lru); atomic_dec(&lru_count); } spin_unlock(&lru_lock); - write_unlock(gl_lock_addr(gl->gl_hash)); GLOCK_BUG_ON(gl, !list_empty(&gl->gl_holders)); - glock_free(gl); - rv = 1; - goto out; + GLOCK_BUG_ON(gl, mapping && mapping->nrpages); + trace_gfs2_glock_put(gl); + sdp->sd_lockstruct.ls_ops->lm_put_lock(gl); } - spin_lock(&gl->gl_spin); - gfs2_glock_schedule_for_reclaim(gl); - spin_unlock(&gl->gl_spin); - write_unlock(gl_lock_addr(gl->gl_hash)); -out: - return rv; } /** @@ -275,17 +245,15 @@ static struct gfs2_glock *search_bucket(unsigned int hash, const struct lm_lockname *name) { struct gfs2_glock *gl; - struct hlist_node *h; + struct hlist_bl_node *h; - hlist_for_each_entry(gl, h, &gl_hash_table[hash].hb_list, gl_list) { + hlist_bl_for_each_entry_rcu(gl, h, &gl_hash_table[hash], gl_list) { if (!lm_name_equal(&gl->gl_name, name)) continue; if (gl->gl_sbd != sdp) continue; - - atomic_inc(&gl->gl_ref); - - return gl; + if (atomic_inc_not_zero(&gl->gl_ref)) + return gl; } return NULL; @@ -743,10 +711,11 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, struct gfs2_glock *gl, *tmp; unsigned int hash = gl_hash(sdp, &name); struct address_space *mapping; + struct kmem_cache *cachep; - read_lock(gl_lock_addr(hash)); + rcu_read_lock(); gl = search_bucket(hash, sdp, &name); - read_unlock(gl_lock_addr(hash)); + rcu_read_unlock(); *glp = gl; if (gl) @@ -755,9 +724,10 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, return -ENOENT; if (glops->go_flags & GLOF_ASPACE) - gl = kmem_cache_alloc(gfs2_glock_aspace_cachep, GFP_KERNEL); + cachep = gfs2_glock_aspace_cachep; else - gl = kmem_cache_alloc(gfs2_glock_cachep, GFP_KERNEL); + cachep = gfs2_glock_cachep; + gl = kmem_cache_alloc(cachep, GFP_KERNEL); if (!gl) return -ENOMEM; @@ -790,15 +760,16 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, mapping->writeback_index = 0; } - write_lock(gl_lock_addr(hash)); + spin_lock_bucket(hash); tmp = search_bucket(hash, sdp, &name); if (tmp) { - write_unlock(gl_lock_addr(hash)); - glock_free(gl); + spin_unlock_bucket(hash); + kmem_cache_free(cachep, gl); + atomic_dec(&sdp->sd_glock_disposal); gl = tmp; } else { - hlist_add_head(&gl->gl_list, &gl_hash_table[hash].hb_list); - write_unlock(gl_lock_addr(hash)); + hlist_bl_add_head_rcu(&gl->gl_list, &gl_hash_table[hash]); + spin_unlock_bucket(hash); } *glp = gl; @@ -1007,13 +978,13 @@ fail: insert_pt = &gh2->gh_list; } set_bit(GLF_QUEUED, &gl->gl_flags); + trace_gfs2_glock_queue(gh, 1); if (likely(insert_pt == NULL)) { list_add_tail(&gh->gh_list, &gl->gl_holders); if (unlikely(gh->gh_flags & LM_FLAG_PRIORITY)) goto do_cancel; return; } - trace_gfs2_glock_queue(gh, 1); list_add_tail(&gh->gh_list, insert_pt); do_cancel: gh = list_entry(gl->gl_holders.next, struct gfs2_holder, gh_list); @@ -1113,6 +1084,7 @@ void gfs2_glock_dq(struct gfs2_holder *gh) !test_bit(GLF_DEMOTE, &gl->gl_flags)) fast_path = 1; } + __gfs2_glock_schedule_for_reclaim(gl); trace_gfs2_glock_queue(gh, 0); spin_unlock(&gl->gl_spin); if (likely(fast_path)) @@ -1276,10 +1248,8 @@ int gfs2_glock_nq_m(unsigned int num_gh, struct gfs2_holder *ghs) void gfs2_glock_dq_m(unsigned int num_gh, struct gfs2_holder *ghs) { - unsigned int x; - - for (x = 0; x < num_gh; x++) - gfs2_glock_dq(&ghs[x]); + while (num_gh--) + gfs2_glock_dq(&ghs[num_gh]); } /** @@ -1291,10 +1261,8 @@ void gfs2_glock_dq_m(unsigned int num_gh, struct gfs2_holder *ghs) void gfs2_glock_dq_uninit_m(unsigned int num_gh, struct gfs2_holder *ghs) { - unsigned int x; - - for (x = 0; x < num_gh; x++) - gfs2_glock_dq_uninit(&ghs[x]); + while (num_gh--) + gfs2_glock_dq_uninit(&ghs[num_gh]); } void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state) @@ -1440,42 +1408,30 @@ static struct shrinker glock_shrinker = { * @sdp: the filesystem * @bucket: the bucket * - * Returns: 1 if the bucket has entries */ -static int examine_bucket(glock_examiner examiner, struct gfs2_sbd *sdp, +static void examine_bucket(glock_examiner examiner, const struct gfs2_sbd *sdp, unsigned int hash) { - struct gfs2_glock *gl, *prev = NULL; - int has_entries = 0; - struct hlist_head *head = &gl_hash_table[hash].hb_list; + struct gfs2_glock *gl; + struct hlist_bl_head *head = &gl_hash_table[hash]; + struct hlist_bl_node *pos; - read_lock(gl_lock_addr(hash)); - /* Can't use hlist_for_each_entry - don't want prefetch here */ - if (hlist_empty(head)) - goto out; - gl = list_entry(head->first, struct gfs2_glock, gl_list); - while(1) { - if (!sdp || gl->gl_sbd == sdp) { - gfs2_glock_hold(gl); - read_unlock(gl_lock_addr(hash)); - if (prev) - gfs2_glock_put(prev); - prev = gl; + rcu_read_lock(); + hlist_bl_for_each_entry_rcu(gl, pos, head, gl_list) { + if ((gl->gl_sbd == sdp) && atomic_read(&gl->gl_ref)) examiner(gl); - has_entries = 1; - read_lock(gl_lock_addr(hash)); - } - if (gl->gl_list.next == NULL) - break; - gl = list_entry(gl->gl_list.next, struct gfs2_glock, gl_list); } -out: - read_unlock(gl_lock_addr(hash)); - if (prev) - gfs2_glock_put(prev); + rcu_read_unlock(); cond_resched(); - return has_entries; +} + +static void glock_hash_walk(glock_examiner examiner, const struct gfs2_sbd *sdp) +{ + unsigned x; + + for (x = 0; x < GFS2_GL_HASH_SIZE; x++) + examine_bucket(examiner, sdp, x); } @@ -1529,10 +1485,21 @@ static void clear_glock(struct gfs2_glock *gl) void gfs2_glock_thaw(struct gfs2_sbd *sdp) { - unsigned x; + glock_hash_walk(thaw_glock, sdp); +} - for (x = 0; x < GFS2_GL_HASH_SIZE; x++) - examine_bucket(thaw_glock, sdp, x); +static int dump_glock(struct seq_file *seq, struct gfs2_glock *gl) +{ + int ret; + spin_lock(&gl->gl_spin); + ret = __dump_glock(seq, gl); + spin_unlock(&gl->gl_spin); + return ret; +} + +static void dump_glock_func(struct gfs2_glock *gl) +{ + dump_glock(NULL, gl); } /** @@ -1545,13 +1512,10 @@ void gfs2_glock_thaw(struct gfs2_sbd *sdp) void gfs2_gl_hash_clear(struct gfs2_sbd *sdp) { - unsigned int x; - - for (x = 0; x < GFS2_GL_HASH_SIZE; x++) - examine_bucket(clear_glock, sdp, x); + glock_hash_walk(clear_glock, sdp); flush_workqueue(glock_workqueue); wait_event(sdp->sd_glock_wait, atomic_read(&sdp->sd_glock_disposal) == 0); - gfs2_dump_lockstate(sdp); + glock_hash_walk(dump_glock_func, sdp); } void gfs2_glock_finish_truncate(struct gfs2_inode *ip) @@ -1717,73 +1681,22 @@ out: return error; } -static int dump_glock(struct seq_file *seq, struct gfs2_glock *gl) -{ - int ret; - spin_lock(&gl->gl_spin); - ret = __dump_glock(seq, gl); - spin_unlock(&gl->gl_spin); - return ret; -} -/** - * gfs2_dump_lockstate - print out the current lockstate - * @sdp: the filesystem - * @ub: the buffer to copy the information into - * - * If @ub is NULL, dump the lockstate to the console. - * - */ - -static int gfs2_dump_lockstate(struct gfs2_sbd *sdp) -{ - struct gfs2_glock *gl; - struct hlist_node *h; - unsigned int x; - int error = 0; - - for (x = 0; x < GFS2_GL_HASH_SIZE; x++) { - - read_lock(gl_lock_addr(x)); - - hlist_for_each_entry(gl, h, &gl_hash_table[x].hb_list, gl_list) { - if (gl->gl_sbd != sdp) - continue; - - error = dump_glock(NULL, gl); - if (error) - break; - } - - read_unlock(gl_lock_addr(x)); - - if (error) - break; - } - - - return error; -} int __init gfs2_glock_init(void) { unsigned i; for(i = 0; i < GFS2_GL_HASH_SIZE; i++) { - INIT_HLIST_HEAD(&gl_hash_table[i].hb_list); - } -#ifdef GL_HASH_LOCK_SZ - for(i = 0; i < GL_HASH_LOCK_SZ; i++) { - rwlock_init(&gl_hash_locks[i]); + INIT_HLIST_BL_HEAD(&gl_hash_table[i]); } -#endif glock_workqueue = alloc_workqueue("glock_workqueue", WQ_MEM_RECLAIM | - WQ_HIGHPRI | WQ_FREEZEABLE, 0); + WQ_HIGHPRI | WQ_FREEZABLE, 0); if (IS_ERR(glock_workqueue)) return PTR_ERR(glock_workqueue); gfs2_delete_workqueue = alloc_workqueue("delete_workqueue", - WQ_MEM_RECLAIM | WQ_FREEZEABLE, + WQ_MEM_RECLAIM | WQ_FREEZABLE, 0); if (IS_ERR(gfs2_delete_workqueue)) { destroy_workqueue(glock_workqueue); @@ -1802,62 +1715,54 @@ void gfs2_glock_exit(void) destroy_workqueue(gfs2_delete_workqueue); } +static inline struct gfs2_glock *glock_hash_chain(unsigned hash) +{ + return hlist_bl_entry(hlist_bl_first_rcu(&gl_hash_table[hash]), + struct gfs2_glock, gl_list); +} + +static inline struct gfs2_glock *glock_hash_next(struct gfs2_glock *gl) +{ + return hlist_bl_entry(rcu_dereference(gl->gl_list.next), + struct gfs2_glock, gl_list); +} + static int gfs2_glock_iter_next(struct gfs2_glock_iter *gi) { struct gfs2_glock *gl; -restart: - read_lock(gl_lock_addr(gi->hash)); - gl = gi->gl; - if (gl) { - gi->gl = hlist_entry(gl->gl_list.next, - struct gfs2_glock, gl_list); - } else { - gi->gl = hlist_entry(gl_hash_table[gi->hash].hb_list.first, - struct gfs2_glock, gl_list); - } - if (gi->gl) - gfs2_glock_hold(gi->gl); - read_unlock(gl_lock_addr(gi->hash)); - if (gl) - gfs2_glock_put(gl); - while (gi->gl == NULL) { - gi->hash++; - if (gi->hash >= GFS2_GL_HASH_SIZE) - return 1; - read_lock(gl_lock_addr(gi->hash)); - gi->gl = hlist_entry(gl_hash_table[gi->hash].hb_list.first, - struct gfs2_glock, gl_list); - if (gi->gl) - gfs2_glock_hold(gi->gl); - read_unlock(gl_lock_addr(gi->hash)); - } - - if (gi->sdp != gi->gl->gl_sbd) - goto restart; + do { + gl = gi->gl; + if (gl) { + gi->gl = glock_hash_next(gl); + } else { + gi->gl = glock_hash_chain(gi->hash); + } + while (gi->gl == NULL) { + gi->hash++; + if (gi->hash >= GFS2_GL_HASH_SIZE) { + rcu_read_unlock(); + return 1; + } + gi->gl = glock_hash_chain(gi->hash); + } + /* Skip entries for other sb and dead entries */ + } while (gi->sdp != gi->gl->gl_sbd || atomic_read(&gi->gl->gl_ref) == 0); return 0; } -static void gfs2_glock_iter_free(struct gfs2_glock_iter *gi) -{ - if (gi->gl) - gfs2_glock_put(gi->gl); - gi->gl = NULL; -} - static void *gfs2_glock_seq_start(struct seq_file *seq, loff_t *pos) { struct gfs2_glock_iter *gi = seq->private; loff_t n = *pos; gi->hash = 0; + rcu_read_lock(); do { - if (gfs2_glock_iter_next(gi)) { - gfs2_glock_iter_free(gi); + if (gfs2_glock_iter_next(gi)) return NULL; - } } while (n--); return gi->gl; @@ -1870,10 +1775,8 @@ static void *gfs2_glock_seq_next(struct seq_file *seq, void *iter_ptr, (*pos)++; - if (gfs2_glock_iter_next(gi)) { - gfs2_glock_iter_free(gi); + if (gfs2_glock_iter_next(gi)) return NULL; - } return gi->gl; } @@ -1881,7 +1784,10 @@ static void *gfs2_glock_seq_next(struct seq_file *seq, void *iter_ptr, static void gfs2_glock_seq_stop(struct seq_file *seq, void *iter_ptr) { struct gfs2_glock_iter *gi = seq->private; - gfs2_glock_iter_free(gi); + + if (gi->gl) + rcu_read_unlock(); + gi->gl = NULL; } static int gfs2_glock_seq_show(struct seq_file *seq, void *iter_ptr) diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h index 691851c..aea1606 100644 --- a/fs/gfs2/glock.h +++ b/fs/gfs2/glock.h @@ -118,7 +118,7 @@ struct lm_lockops { int (*lm_mount) (struct gfs2_sbd *sdp, const char *fsname); void (*lm_unmount) (struct gfs2_sbd *sdp); void (*lm_withdraw) (struct gfs2_sbd *sdp); - void (*lm_put_lock) (struct kmem_cache *cachep, struct gfs2_glock *gl); + void (*lm_put_lock) (struct gfs2_glock *gl); int (*lm_lock) (struct gfs2_glock *gl, unsigned int req_state, unsigned int flags); void (*lm_cancel) (struct gfs2_glock *gl); @@ -174,7 +174,7 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, int create, struct gfs2_glock **glp); void gfs2_glock_hold(struct gfs2_glock *gl); void gfs2_glock_put_nolock(struct gfs2_glock *gl); -int gfs2_glock_put(struct gfs2_glock *gl); +void gfs2_glock_put(struct gfs2_glock *gl); void gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, unsigned flags, struct gfs2_holder *gh); void gfs2_holder_reinit(unsigned int state, unsigned flags, @@ -223,25 +223,22 @@ static inline int gfs2_glock_nq_init(struct gfs2_glock *gl, return error; } -/* Lock Value Block functions */ - -int gfs2_lvb_hold(struct gfs2_glock *gl); -void gfs2_lvb_unhold(struct gfs2_glock *gl); - -void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state); -void gfs2_glock_complete(struct gfs2_glock *gl, int ret); -void gfs2_reclaim_glock(struct gfs2_sbd *sdp); -void gfs2_gl_hash_clear(struct gfs2_sbd *sdp); -void gfs2_glock_finish_truncate(struct gfs2_inode *ip); -void gfs2_glock_thaw(struct gfs2_sbd *sdp); - -int __init gfs2_glock_init(void); -void gfs2_glock_exit(void); - -int gfs2_create_debugfs_file(struct gfs2_sbd *sdp); -void gfs2_delete_debugfs_file(struct gfs2_sbd *sdp); -int gfs2_register_debugfs(void); -void gfs2_unregister_debugfs(void); +extern void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state); +extern void gfs2_glock_complete(struct gfs2_glock *gl, int ret); +extern void gfs2_reclaim_glock(struct gfs2_sbd *sdp); +extern void gfs2_gl_hash_clear(struct gfs2_sbd *sdp); +extern void gfs2_glock_finish_truncate(struct gfs2_inode *ip); +extern void gfs2_glock_thaw(struct gfs2_sbd *sdp); +extern void gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl); +extern void gfs2_glock_free(struct gfs2_glock *gl); + +extern int __init gfs2_glock_init(void); +extern void gfs2_glock_exit(void); + +extern int gfs2_create_debugfs_file(struct gfs2_sbd *sdp); +extern void gfs2_delete_debugfs_file(struct gfs2_sbd *sdp); +extern int gfs2_register_debugfs(void); +extern void gfs2_unregister_debugfs(void); extern const struct lm_lockops gfs2_dlm_ops; diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c index 263561b..3754e3c 100644 --- a/fs/gfs2/glops.c +++ b/fs/gfs2/glops.c @@ -56,20 +56,26 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl) BUG_ON(current->journal_info); current->journal_info = &tr; - gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); while (!list_empty(head)) { bd = list_entry(head->next, struct gfs2_bufdata, bd_ail_gl_list); bh = bd->bd_bh; gfs2_remove_from_ail(bd); + spin_unlock(&sdp->sd_ail_lock); + bd->bd_bh = NULL; bh->b_private = NULL; bd->bd_blkno = bh->b_blocknr; + gfs2_log_lock(sdp); gfs2_assert_withdraw(sdp, !buffer_busy(bh)); gfs2_trans_add_revoke(sdp, bd); + gfs2_log_unlock(sdp); + + spin_lock(&sdp->sd_ail_lock); } gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count)); - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ail_lock); gfs2_trans_end(sdp); gfs2_log_flush(sdp, NULL); @@ -206,8 +212,17 @@ static void inode_go_inval(struct gfs2_glock *gl, int flags) static int inode_go_demote_ok(const struct gfs2_glock *gl) { struct gfs2_sbd *sdp = gl->gl_sbd; + struct gfs2_holder *gh; + if (sdp->sd_jindex == gl->gl_object || sdp->sd_rindex == gl->gl_object) return 0; + + if (!list_empty(&gl->gl_holders)) { + gh = list_entry(gl->gl_holders.next, struct gfs2_holder, gh_list); + if (gh->gh_list.next != &gl->gl_holders) + return 0; + } + return 1; } @@ -272,19 +287,6 @@ static int inode_go_dump(struct seq_file *seq, const struct gfs2_glock *gl) } /** - * rgrp_go_demote_ok - Check to see if it's ok to unlock a RG's glock - * @gl: the glock - * - * Returns: 1 if it's ok - */ - -static int rgrp_go_demote_ok(const struct gfs2_glock *gl) -{ - const struct address_space *mapping = (const struct address_space *)(gl + 1); - return !mapping->nrpages; -} - -/** * rgrp_go_lock - operation done after an rgrp lock is locked by * a first holder on this node. * @gl: the glock @@ -410,7 +412,6 @@ const struct gfs2_glock_operations gfs2_inode_glops = { const struct gfs2_glock_operations gfs2_rgrp_glops = { .go_xmote_th = rgrp_go_sync, .go_inval = rgrp_go_inval, - .go_demote_ok = rgrp_go_demote_ok, .go_lock = rgrp_go_lock, .go_unlock = rgrp_go_unlock, .go_dump = gfs2_rgrp_dump, diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index a79790c..870a89d 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -15,6 +15,8 @@ #include <linux/workqueue.h> #include <linux/dlm.h> #include <linux/buffer_head.h> +#include <linux/rcupdate.h> +#include <linux/rculist_bl.h> #define DIO_WAIT 0x00000010 #define DIO_METADATA 0x00000020 @@ -201,7 +203,7 @@ enum { }; struct gfs2_glock { - struct hlist_node gl_list; + struct hlist_bl_node gl_list; unsigned long gl_flags; /* GLF_... */ struct lm_lockname gl_name; atomic_t gl_ref; @@ -234,6 +236,7 @@ struct gfs2_glock { atomic_t gl_ail_count; struct delayed_work gl_work; struct work_struct gl_delete; + struct rcu_head gl_rcu; }; #define GFS2_MIN_LVB_SIZE 32 /* Min size of LVB that gfs2 supports */ @@ -314,6 +317,7 @@ enum { QDF_USER = 0, QDF_CHANGE = 1, QDF_LOCKED = 2, + QDF_REFRESH = 3, }; struct gfs2_quota_data { @@ -647,6 +651,7 @@ struct gfs2_sbd { unsigned int sd_log_flush_head; u64 sd_log_flush_wrapped; + spinlock_t sd_ail_lock; struct list_head sd_ail1_list; struct list_head sd_ail2_list; u64 sd_ail_sync_gen; diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 2232b3c..97d54a2 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -74,16 +74,14 @@ static struct inode *gfs2_iget(struct super_block *sb, u64 no_addr) } /** - * GFS2 lookup code fills in vfs inode contents based on info obtained - * from directory entry inside gfs2_inode_lookup(). This has caused issues - * with NFS code path since its get_dentry routine doesn't have the relevant - * directory entry when gfs2_inode_lookup() is invoked. Part of the code - * segment inside gfs2_inode_lookup code needs to get moved around. + * gfs2_set_iop - Sets inode operations + * @inode: The inode with correct i_mode filled in * - * Clears I_NEW as well. - **/ + * GFS2 lookup code fills in vfs inode contents based on info obtained + * from directory entry inside gfs2_inode_lookup(). + */ -void gfs2_set_iop(struct inode *inode) +static void gfs2_set_iop(struct inode *inode) { struct gfs2_sbd *sdp = GFS2_SB(inode); umode_t mode = inode->i_mode; @@ -106,8 +104,6 @@ void gfs2_set_iop(struct inode *inode) inode->i_op = &gfs2_file_iops; init_special_inode(inode, inode->i_mode, inode->i_rdev); } - - unlock_new_inode(inode); } /** @@ -119,10 +115,8 @@ void gfs2_set_iop(struct inode *inode) * Returns: A VFS inode, or an error */ -struct inode *gfs2_inode_lookup(struct super_block *sb, - unsigned int type, - u64 no_addr, - u64 no_formal_ino) +struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned int type, + u64 no_addr, u64 no_formal_ino) { struct inode *inode; struct gfs2_inode *ip; @@ -152,51 +146,37 @@ struct inode *gfs2_inode_lookup(struct super_block *sb, error = gfs2_glock_nq_init(io_gl, LM_ST_SHARED, GL_EXACT, &ip->i_iopen_gh); if (unlikely(error)) goto fail_iopen; - ip->i_iopen_gh.gh_gl->gl_object = ip; + ip->i_iopen_gh.gh_gl->gl_object = ip; gfs2_glock_put(io_gl); io_gl = NULL; - if ((type == DT_UNKNOWN) && (no_formal_ino == 0)) - goto gfs2_nfsbypass; - - inode->i_mode = DT2IF(type); - - /* - * We must read the inode in order to work out its type in - * this case. Note that this doesn't happen often as we normally - * know the type beforehand. This code path only occurs during - * unlinked inode recovery (where it is safe to do this glock, - * which is not true in the general case). - */ if (type == DT_UNKNOWN) { - struct gfs2_holder gh; - error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); - if (unlikely(error)) - goto fail_glock; - /* Inode is now uptodate */ - gfs2_glock_dq_uninit(&gh); + /* Inode glock must be locked already */ + error = gfs2_inode_refresh(GFS2_I(inode)); + if (error) + goto fail_refresh; + } else { + inode->i_mode = DT2IF(type); } gfs2_set_iop(inode); + unlock_new_inode(inode); } -gfs2_nfsbypass: return inode; -fail_glock: - gfs2_glock_dq(&ip->i_iopen_gh); + +fail_refresh: + ip->i_iopen_gh.gh_gl->gl_object = NULL; + gfs2_glock_dq_uninit(&ip->i_iopen_gh); fail_iopen: if (io_gl) gfs2_glock_put(io_gl); fail_put: - if (inode->i_state & I_NEW) - ip->i_gl->gl_object = NULL; + ip->i_gl->gl_object = NULL; gfs2_glock_put(ip->i_gl); fail: - if (inode->i_state & I_NEW) - iget_failed(inode); - else - iput(inode); + iget_failed(inode); return ERR_PTR(error); } @@ -221,14 +201,6 @@ struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr, if (IS_ERR(inode)) goto fail; - error = gfs2_inode_refresh(GFS2_I(inode)); - if (error) - goto fail_iput; - - /* Pick up the works we bypass in gfs2_inode_lookup */ - if (inode->i_state & I_NEW) - gfs2_set_iop(inode); - /* Two extra checks for NFS only */ if (no_formal_ino) { error = -ESTALE; @@ -791,14 +763,15 @@ fail: return error; } -static int gfs2_security_init(struct gfs2_inode *dip, struct gfs2_inode *ip) +static int gfs2_security_init(struct gfs2_inode *dip, struct gfs2_inode *ip, + const struct qstr *qstr) { int err; size_t len; void *value; char *name; - err = security_inode_init_security(&ip->i_inode, &dip->i_inode, + err = security_inode_init_security(&ip->i_inode, &dip->i_inode, qstr, &name, &value, &len); if (err) { @@ -882,7 +855,7 @@ struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name, if (error) goto fail_gunlock2; - error = gfs2_security_init(dip, GFS2_I(inode)); + error = gfs2_security_init(dip, GFS2_I(inode), name); if (error) goto fail_gunlock2; diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h index 732a183..3e00a66 100644 --- a/fs/gfs2/inode.h +++ b/fs/gfs2/inode.h @@ -96,7 +96,6 @@ err: return -EIO; } -extern void gfs2_set_iop(struct inode *inode); extern struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned type, u64 no_addr, u64 no_formal_ino); extern struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr, diff --git a/fs/gfs2/lock_dlm.c b/fs/gfs2/lock_dlm.c index 6e493ae..98c80d8 100644 --- a/fs/gfs2/lock_dlm.c +++ b/fs/gfs2/lock_dlm.c @@ -22,7 +22,6 @@ static void gdlm_ast(void *arg) { struct gfs2_glock *gl = arg; unsigned ret = gl->gl_state; - struct gfs2_sbd *sdp = gl->gl_sbd; BUG_ON(gl->gl_lksb.sb_flags & DLM_SBF_DEMOTED); @@ -31,12 +30,7 @@ static void gdlm_ast(void *arg) switch (gl->gl_lksb.sb_status) { case -DLM_EUNLOCK: /* Unlocked, so glock can be freed */ - if (gl->gl_ops->go_flags & GLOF_ASPACE) - kmem_cache_free(gfs2_glock_aspace_cachep, gl); - else - kmem_cache_free(gfs2_glock_cachep, gl); - if (atomic_dec_and_test(&sdp->sd_glock_disposal)) - wake_up(&sdp->sd_glock_wait); + gfs2_glock_free(gl); return; case -DLM_ECANCEL: /* Cancel while getting lock */ ret |= LM_OUT_CANCELED; @@ -164,16 +158,14 @@ static int gdlm_lock(struct gfs2_glock *gl, unsigned int req_state, GDLM_STRNAME_BYTES - 1, 0, gdlm_ast, gl, gdlm_bast); } -static void gdlm_put_lock(struct kmem_cache *cachep, struct gfs2_glock *gl) +static void gdlm_put_lock(struct gfs2_glock *gl) { struct gfs2_sbd *sdp = gl->gl_sbd; struct lm_lockstruct *ls = &sdp->sd_lockstruct; int error; if (gl->gl_lksb.sb_lkid == 0) { - kmem_cache_free(cachep, gl); - if (atomic_dec_and_test(&sdp->sd_glock_disposal)) - wake_up(&sdp->sd_glock_wait); + gfs2_glock_free(gl); return; } diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c index eb01f35..e7ed31f 100644 --- a/fs/gfs2/log.c +++ b/fs/gfs2/log.c @@ -67,7 +67,7 @@ unsigned int gfs2_struct2blk(struct gfs2_sbd *sdp, unsigned int nstruct, * @mapping: The associated mapping (maybe NULL) * @bd: The gfs2_bufdata to remove * - * The log lock _must_ be held when calling this function + * The ail lock _must_ be held when calling this function * */ @@ -88,8 +88,8 @@ void gfs2_remove_from_ail(struct gfs2_bufdata *bd) */ static void gfs2_ail1_start_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai) -__releases(&sdp->sd_log_lock) -__acquires(&sdp->sd_log_lock) +__releases(&sdp->sd_ail_lock) +__acquires(&sdp->sd_ail_lock) { struct gfs2_bufdata *bd, *s; struct buffer_head *bh; @@ -117,7 +117,7 @@ __acquires(&sdp->sd_log_lock) list_move(&bd->bd_ail_st_list, &ai->ai_ail1_list); get_bh(bh); - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ail_lock); lock_buffer(bh); if (test_clear_buffer_dirty(bh)) { bh->b_end_io = end_buffer_write_sync; @@ -126,7 +126,7 @@ __acquires(&sdp->sd_log_lock) unlock_buffer(bh); brelse(bh); } - gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); retry = 1; break; @@ -175,10 +175,10 @@ static void gfs2_ail1_start(struct gfs2_sbd *sdp) struct gfs2_ail *ai; int done = 0; - gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); head = &sdp->sd_ail1_list; if (list_empty(head)) { - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ail_lock); return; } sync_gen = sdp->sd_ail_sync_gen++; @@ -189,13 +189,13 @@ static void gfs2_ail1_start(struct gfs2_sbd *sdp) if (ai->ai_sync_gen >= sync_gen) continue; ai->ai_sync_gen = sync_gen; - gfs2_ail1_start_one(sdp, ai); /* This may drop log lock */ + gfs2_ail1_start_one(sdp, ai); /* This may drop ail lock */ done = 0; break; } } - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ail_lock); } static int gfs2_ail1_empty(struct gfs2_sbd *sdp, int flags) @@ -203,7 +203,7 @@ static int gfs2_ail1_empty(struct gfs2_sbd *sdp, int flags) struct gfs2_ail *ai, *s; int ret; - gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); list_for_each_entry_safe_reverse(ai, s, &sdp->sd_ail1_list, ai_list) { if (gfs2_ail1_empty_one(sdp, ai, flags)) @@ -214,7 +214,7 @@ static int gfs2_ail1_empty(struct gfs2_sbd *sdp, int flags) ret = list_empty(&sdp->sd_ail1_list); - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ail_lock); return ret; } @@ -247,7 +247,7 @@ static void ail2_empty(struct gfs2_sbd *sdp, unsigned int new_tail) int wrap = (new_tail < old_tail); int a, b, rm; - gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); list_for_each_entry_safe(ai, safe, &sdp->sd_ail2_list, ai_list) { a = (old_tail <= ai->ai_first); @@ -263,7 +263,7 @@ static void ail2_empty(struct gfs2_sbd *sdp, unsigned int new_tail) kfree(ai); } - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ail_lock); } /** @@ -421,7 +421,7 @@ static unsigned int current_tail(struct gfs2_sbd *sdp) struct gfs2_ail *ai; unsigned int tail; - gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); if (list_empty(&sdp->sd_ail1_list)) { tail = sdp->sd_log_head; @@ -430,7 +430,7 @@ static unsigned int current_tail(struct gfs2_sbd *sdp) tail = ai->ai_first; } - gfs2_log_unlock(sdp); + spin_unlock(&sdp->sd_ail_lock); return tail; } @@ -743,10 +743,12 @@ void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl) sdp->sd_log_commited_databuf = 0; sdp->sd_log_commited_revoke = 0; + spin_lock(&sdp->sd_ail_lock); if (!list_empty(&ai->ai_ail1_list)) { list_add(&ai->ai_list, &sdp->sd_ail1_list); ai = NULL; } + spin_unlock(&sdp->sd_ail_lock); gfs2_log_unlock(sdp); trace_gfs2_log_flush(sdp, 0); up_write(&sdp->sd_log_flush_lock); diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c index bf33f82..e919abf 100644 --- a/fs/gfs2/lops.c +++ b/fs/gfs2/lops.c @@ -51,8 +51,10 @@ static void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh) /* If this buffer is in the AIL and it has already been written * to in-place disk block, remove it from the AIL. */ + spin_lock(&sdp->sd_ail_lock); if (bd->bd_ail) list_move(&bd->bd_ail_st_list, &bd->bd_ail->ai_ail2_list); + spin_unlock(&sdp->sd_ail_lock); get_bh(bh); atomic_inc(&sdp->sd_log_pinned); trace_gfs2_pin(bd, 1); @@ -80,7 +82,7 @@ static void gfs2_unpin(struct gfs2_sbd *sdp, struct buffer_head *bh, mark_buffer_dirty(bh); clear_buffer_pinned(bh); - gfs2_log_lock(sdp); + spin_lock(&sdp->sd_ail_lock); if (bd->bd_ail) { list_del(&bd->bd_ail_st_list); brelse(bh); @@ -91,9 +93,11 @@ static void gfs2_unpin(struct gfs2_sbd *sdp, struct buffer_head *bh, } bd->bd_ail = ai; list_add(&bd->bd_ail_st_list, &ai->ai_ail1_list); - clear_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); + spin_unlock(&sdp->sd_ail_lock); + + if (test_and_clear_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags)) + gfs2_glock_schedule_for_reclaim(bd->bd_gl); trace_gfs2_pin(bd, 0); - gfs2_log_unlock(sdp); unlock_buffer(bh); atomic_dec(&sdp->sd_log_pinned); } diff --git a/fs/gfs2/main.c b/fs/gfs2/main.c index ebef7ab..888a5f5 100644 --- a/fs/gfs2/main.c +++ b/fs/gfs2/main.c @@ -14,6 +14,8 @@ #include <linux/module.h> #include <linux/init.h> #include <linux/gfs2_ondisk.h> +#include <linux/rcupdate.h> +#include <linux/rculist_bl.h> #include <asm/atomic.h> #include "gfs2.h" @@ -45,7 +47,7 @@ static void gfs2_init_glock_once(void *foo) { struct gfs2_glock *gl = foo; - INIT_HLIST_NODE(&gl->gl_list); + INIT_HLIST_BL_NODE(&gl->gl_list); spin_lock_init(&gl->gl_spin); INIT_LIST_HEAD(&gl->gl_holders); INIT_LIST_HEAD(&gl->gl_lru); @@ -59,14 +61,7 @@ static void gfs2_init_gl_aspace_once(void *foo) struct address_space *mapping = (struct address_space *)(gl + 1); gfs2_init_glock_once(gl); - memset(mapping, 0, sizeof(*mapping)); - INIT_RADIX_TREE(&mapping->page_tree, GFP_ATOMIC); - spin_lock_init(&mapping->tree_lock); - spin_lock_init(&mapping->i_mmap_lock); - INIT_LIST_HEAD(&mapping->private_list); - spin_lock_init(&mapping->private_lock); - INIT_RAW_PRIO_TREE_ROOT(&mapping->i_mmap); - INIT_LIST_HEAD(&mapping->i_mmap_nonlinear); + address_space_init_once(mapping); } /** @@ -144,7 +139,7 @@ static int __init init_gfs2_fs(void) error = -ENOMEM; gfs_recovery_wq = alloc_workqueue("gfs_recovery", - WQ_MEM_RECLAIM | WQ_FREEZEABLE, 0); + WQ_MEM_RECLAIM | WQ_FREEZABLE, 0); if (!gfs_recovery_wq) goto fail_wq; @@ -198,6 +193,8 @@ static void __exit exit_gfs2_fs(void) unregister_filesystem(&gfs2meta_fs_type); destroy_workqueue(gfs_recovery_wq); + rcu_barrier(); + kmem_cache_destroy(gfs2_quotad_cachep); kmem_cache_destroy(gfs2_rgrpd_cachep); kmem_cache_destroy(gfs2_bufdata_cachep); diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c index 939739c..01d97f4 100644 --- a/fs/gfs2/meta_io.c +++ b/fs/gfs2/meta_io.c @@ -326,6 +326,7 @@ void gfs2_remove_from_journal(struct buffer_head *bh, struct gfs2_trans *tr, int brelse(bh); } if (bd) { + spin_lock(&sdp->sd_ail_lock); if (bd->bd_ail) { gfs2_remove_from_ail(bd); bh->b_private = NULL; @@ -333,6 +334,7 @@ void gfs2_remove_from_journal(struct buffer_head *bh, struct gfs2_trans *tr, int bd->bd_blkno = bh->b_blocknr; gfs2_trans_add_revoke(sdp, bd); } + spin_unlock(&sdp->sd_ail_lock); } clear_buffer_dirty(bh); clear_buffer_uptodate(bh); diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 777927c..42ef243 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -99,6 +99,7 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb) init_waitqueue_head(&sdp->sd_log_waitq); init_waitqueue_head(&sdp->sd_logd_waitq); + spin_lock_init(&sdp->sd_ail_lock); INIT_LIST_HEAD(&sdp->sd_ail1_list); INIT_LIST_HEAD(&sdp->sd_ail2_list); @@ -928,17 +929,9 @@ static const match_table_t nolock_tokens = { { Opt_err, NULL }, }; -static void nolock_put_lock(struct kmem_cache *cachep, struct gfs2_glock *gl) -{ - struct gfs2_sbd *sdp = gl->gl_sbd; - kmem_cache_free(cachep, gl); - if (atomic_dec_and_test(&sdp->sd_glock_disposal)) - wake_up(&sdp->sd_glock_wait); -} - static const struct lm_lockops nolock_ops = { .lm_proto_name = "lock_nolock", - .lm_put_lock = nolock_put_lock, + .lm_put_lock = gfs2_glock_free, .lm_tokens = &nolock_tokens, }; diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c index d8b26ac..09e436a 100644 --- a/fs/gfs2/ops_inode.c +++ b/fs/gfs2/ops_inode.c @@ -1026,9 +1026,9 @@ static void gfs2_put_link(struct dentry *dentry, struct nameidata *nd, void *p) /** * gfs2_permission - - * @inode: - * @mask: - * @nd: passed from Linux VFS, ignored by us + * @inode: The inode + * @mask: The mask to be tested + * @flags: Indicates whether this is an RCU path walk or not * * This may be called from the VFS directly, or from within GFS2 with the * inode locked, so we look to see if the glock is already locked and only @@ -1044,11 +1044,11 @@ int gfs2_permission(struct inode *inode, int mask, unsigned int flags) int error; int unlock = 0; - if (flags & IPERM_FLAG_RCU) - return -ECHILD; ip = GFS2_I(inode); if (gfs2_glock_is_locked_by_me(ip->i_gl) == NULL) { + if (flags & IPERM_FLAG_RCU) + return -ECHILD; error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &i_gh); if (error) return error; diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index a689901..e23d986 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -834,6 +834,7 @@ static int do_sync(unsigned int num_qd, struct gfs2_quota_data **qda) goto out_end_trans; do_qc(qd, -qd->qd_change_sync); + set_bit(QDF_REFRESH, &qd->qd_flags); } error = 0; @@ -929,6 +930,7 @@ int gfs2_quota_lock(struct gfs2_inode *ip, u32 uid, u32 gid) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_alloc *al = ip->i_alloc; + struct gfs2_quota_data *qd; unsigned int x; int error = 0; @@ -942,7 +944,11 @@ int gfs2_quota_lock(struct gfs2_inode *ip, u32 uid, u32 gid) sort_qd, NULL); for (x = 0; x < al->al_qd_num; x++) { - error = do_glock(al->al_qd[x], NO_FORCE, &al->al_qd_ghs[x]); + int force = NO_FORCE; + qd = al->al_qd[x]; + if (test_and_clear_bit(QDF_REFRESH, &qd->qd_flags)) + force = FORCE; + error = do_glock(qd, force, &al->al_qd_ghs[x]); if (error) break; } @@ -1587,6 +1593,8 @@ static int gfs2_set_dqblk(struct super_block *sb, int type, qid_t id, offset = qd2offset(qd); alloc_required = gfs2_write_alloc_required(ip, offset, sizeof(struct gfs2_quota)); + if (gfs2_is_stuffed(ip)) + alloc_required = 1; if (alloc_required) { al = gfs2_alloc_get(ip); if (al == NULL) @@ -1600,7 +1608,9 @@ static int gfs2_set_dqblk(struct super_block *sb, int type, qid_t id, blocks += gfs2_rg_blocks(al); } - error = gfs2_trans_begin(sdp, blocks + RES_DINODE + 1, 0); + /* Some quotas span block boundaries and can update two blocks, + adding an extra block to the transaction to handle such quotas */ + error = gfs2_trans_begin(sdp, blocks + RES_DINODE + 2, 0); if (error) goto out_release; diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index 7293ea2..cf930cd 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -1602,7 +1602,7 @@ rgrp_error: * */ -void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen) +void __gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_rgrpd *rgd; @@ -1617,7 +1617,21 @@ void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen) gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); gfs2_trans_add_rg(rgd); +} +/** + * gfs2_free_data - free a contiguous run of data block(s) + * @ip: the inode these blocks are being freed from + * @bstart: first block of a run of contiguous blocks + * @blen: the length of the block run + * + */ + +void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen) +{ + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + + __gfs2_free_data(ip, bstart, blen); gfs2_statfs_change(sdp, 0, +blen, 0); gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid); } @@ -1630,7 +1644,7 @@ void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen) * */ -void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen) +void __gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_rgrpd *rgd; @@ -1645,10 +1659,24 @@ void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen) gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); gfs2_trans_add_rg(rgd); + gfs2_meta_wipe(ip, bstart, blen); +} +/** + * gfs2_free_meta - free a contiguous run of data block(s) + * @ip: the inode these blocks are being freed from + * @bstart: first block of a run of contiguous blocks + * @blen: the length of the block run + * + */ + +void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen) +{ + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + + __gfs2_free_meta(ip, bstart, blen); gfs2_statfs_change(sdp, 0, +blen, 0); gfs2_quota_change(ip, -(s64)blen, ip->i_inode.i_uid, ip->i_inode.i_gid); - gfs2_meta_wipe(ip, bstart, blen); } void gfs2_unlink_di(struct inode *inode) diff --git a/fs/gfs2/rgrp.h b/fs/gfs2/rgrp.h index 50c2bb0..a80e303 100644 --- a/fs/gfs2/rgrp.h +++ b/fs/gfs2/rgrp.h @@ -52,7 +52,9 @@ extern int gfs2_ri_update(struct gfs2_inode *ip); extern int gfs2_alloc_block(struct gfs2_inode *ip, u64 *bn, unsigned int *n); extern int gfs2_alloc_di(struct gfs2_inode *ip, u64 *bn, u64 *generation); +extern void __gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen); extern void gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen); +extern void __gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen); extern void gfs2_free_meta(struct gfs2_inode *ip, u64 bstart, u32 blen); extern void gfs2_free_di(struct gfs2_rgrpd *rgd, struct gfs2_inode *ip); extern void gfs2_unlink_di(struct inode *inode); diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index 16c2eca..ec73ed7 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -1336,6 +1336,7 @@ static void gfs2_evict_inode(struct inode *inode) if (error) goto out_truncate; + ip->i_iopen_gh.gh_flags |= GL_NOCACHE; gfs2_glock_dq_wait(&ip->i_iopen_gh); gfs2_holder_reinit(LM_ST_EXCLUSIVE, LM_FLAG_TRY_1CB | GL_NOCACHE, &ip->i_iopen_gh); error = gfs2_glock_nq(&ip->i_iopen_gh); diff --git a/fs/hfs/dir.c b/fs/hfs/dir.c index afa66aa..b4d70b1 100644 --- a/fs/hfs/dir.c +++ b/fs/hfs/dir.c @@ -238,46 +238,22 @@ static int hfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) } /* - * hfs_unlink() + * hfs_remove() * - * This is the unlink() entry in the inode_operations structure for - * regular HFS directories. The purpose is to delete an existing - * file, given the inode for the parent directory and the name - * (and its length) of the existing file. - */ -static int hfs_unlink(struct inode *dir, struct dentry *dentry) -{ - struct inode *inode; - int res; - - inode = dentry->d_inode; - res = hfs_cat_delete(inode->i_ino, dir, &dentry->d_name); - if (res) - return res; - - drop_nlink(inode); - hfs_delete_inode(inode); - inode->i_ctime = CURRENT_TIME_SEC; - mark_inode_dirty(inode); - - return res; -} - -/* - * hfs_rmdir() + * This serves as both unlink() and rmdir() in the inode_operations + * structure for regular HFS directories. The purpose is to delete + * an existing child, given the inode for the parent directory and + * the name (and its length) of the existing directory. * - * This is the rmdir() entry in the inode_operations structure for - * regular HFS directories. The purpose is to delete an existing - * directory, given the inode for the parent directory and the name - * (and its length) of the existing directory. + * HFS does not have hardlinks, so both rmdir and unlink set the + * link count to 0. The only difference is the emptiness check. */ -static int hfs_rmdir(struct inode *dir, struct dentry *dentry) +static int hfs_remove(struct inode *dir, struct dentry *dentry) { - struct inode *inode; + struct inode *inode = dentry->d_inode; int res; - inode = dentry->d_inode; - if (inode->i_size != 2) + if (S_ISDIR(inode->i_mode) && inode->i_size != 2) return -ENOTEMPTY; res = hfs_cat_delete(inode->i_ino, dir, &dentry->d_name); if (res) @@ -307,7 +283,7 @@ static int hfs_rename(struct inode *old_dir, struct dentry *old_dentry, /* Unlink destination if it already exists */ if (new_dentry->d_inode) { - res = hfs_unlink(new_dir, new_dentry); + res = hfs_remove(new_dir, new_dentry); if (res) return res; } @@ -332,9 +308,9 @@ const struct file_operations hfs_dir_operations = { const struct inode_operations hfs_dir_inode_operations = { .create = hfs_create, .lookup = hfs_lookup, - .unlink = hfs_unlink, + .unlink = hfs_remove, .mkdir = hfs_mkdir, - .rmdir = hfs_rmdir, + .rmdir = hfs_remove, .rename = hfs_rename, .setattr = hfs_inode_setattr, }; diff --git a/fs/hfsplus/extents.c b/fs/hfsplus/extents.c index 52a0bca..b1991a2 100644 --- a/fs/hfsplus/extents.c +++ b/fs/hfsplus/extents.c @@ -397,8 +397,8 @@ int hfsplus_file_extend(struct inode *inode) u32 start, len, goal; int res; - if (sbi->total_blocks - sbi->free_blocks + 8 > - sbi->alloc_file->i_size * 8) { + if (sbi->alloc_file->i_size * 8 < + sbi->total_blocks - sbi->free_blocks + 8) { /* extend alloc file */ printk(KERN_ERR "hfs: extend alloc file! " "(%llu,%u,%u)\n", diff --git a/fs/hfsplus/part_tbl.c b/fs/hfsplus/part_tbl.c index d66ad11..40ad88c 100644 --- a/fs/hfsplus/part_tbl.c +++ b/fs/hfsplus/part_tbl.c @@ -134,7 +134,7 @@ int hfs_part_find(struct super_block *sb, res = hfsplus_submit_bio(sb->s_bdev, *part_start + HFS_PMAP_BLK, data, READ); if (res) - return res; + goto out; switch (be16_to_cpu(*((__be16 *)data))) { case HFS_OLD_PMAP_MAGIC: @@ -147,7 +147,7 @@ int hfs_part_find(struct super_block *sb, res = -ENOENT; break; } - +out: kfree(data); return res; } diff --git a/fs/hfsplus/super.c b/fs/hfsplus/super.c index 9a3b479..b49b555 100644 --- a/fs/hfsplus/super.c +++ b/fs/hfsplus/super.c @@ -338,20 +338,22 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) struct inode *root, *inode; struct qstr str; struct nls_table *nls = NULL; - int err = -EINVAL; + int err; + err = -EINVAL; sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); if (!sbi) - return -ENOMEM; + goto out; sb->s_fs_info = sbi; mutex_init(&sbi->alloc_mutex); mutex_init(&sbi->vh_mutex); hfsplus_fill_defaults(sbi); + + err = -EINVAL; if (!hfsplus_parse_options(data, sbi)) { printk(KERN_ERR "hfs: unable to parse mount options\n"); - err = -EINVAL; - goto cleanup; + goto out_unload_nls; } /* temporarily use utf8 to correctly find the hidden dir below */ @@ -359,16 +361,14 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) sbi->nls = load_nls("utf8"); if (!sbi->nls) { printk(KERN_ERR "hfs: unable to load nls for utf8\n"); - err = -EINVAL; - goto cleanup; + goto out_unload_nls; } /* Grab the volume header */ if (hfsplus_read_wrapper(sb)) { if (!silent) printk(KERN_WARNING "hfs: unable to find HFS+ superblock\n"); - err = -EINVAL; - goto cleanup; + goto out_unload_nls; } vhdr = sbi->s_vhdr; @@ -377,7 +377,7 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) if (be16_to_cpu(vhdr->version) < HFSPLUS_MIN_VERSION || be16_to_cpu(vhdr->version) > HFSPLUS_CURRENT_VERSION) { printk(KERN_ERR "hfs: wrong filesystem version\n"); - goto cleanup; + goto out_free_vhdr; } sbi->total_blocks = be32_to_cpu(vhdr->total_blocks); sbi->free_blocks = be32_to_cpu(vhdr->free_blocks); @@ -421,19 +421,19 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) sbi->ext_tree = hfs_btree_open(sb, HFSPLUS_EXT_CNID); if (!sbi->ext_tree) { printk(KERN_ERR "hfs: failed to load extents file\n"); - goto cleanup; + goto out_free_vhdr; } sbi->cat_tree = hfs_btree_open(sb, HFSPLUS_CAT_CNID); if (!sbi->cat_tree) { printk(KERN_ERR "hfs: failed to load catalog file\n"); - goto cleanup; + goto out_close_ext_tree; } inode = hfsplus_iget(sb, HFSPLUS_ALLOC_CNID); if (IS_ERR(inode)) { printk(KERN_ERR "hfs: failed to load allocation file\n"); err = PTR_ERR(inode); - goto cleanup; + goto out_close_cat_tree; } sbi->alloc_file = inode; @@ -442,14 +442,7 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) if (IS_ERR(root)) { printk(KERN_ERR "hfs: failed to load root directory\n"); err = PTR_ERR(root); - goto cleanup; - } - sb->s_d_op = &hfsplus_dentry_operations; - sb->s_root = d_alloc_root(root); - if (!sb->s_root) { - iput(root); - err = -ENOMEM; - goto cleanup; + goto out_put_alloc_file; } str.len = sizeof(HFSP_HIDDENDIR_NAME) - 1; @@ -459,46 +452,69 @@ static int hfsplus_fill_super(struct super_block *sb, void *data, int silent) if (!hfs_brec_read(&fd, &entry, sizeof(entry))) { hfs_find_exit(&fd); if (entry.type != cpu_to_be16(HFSPLUS_FOLDER)) - goto cleanup; + goto out_put_root; inode = hfsplus_iget(sb, be32_to_cpu(entry.folder.id)); if (IS_ERR(inode)) { err = PTR_ERR(inode); - goto cleanup; + goto out_put_root; } sbi->hidden_dir = inode; } else hfs_find_exit(&fd); - if (sb->s_flags & MS_RDONLY) - goto out; + if (!(sb->s_flags & MS_RDONLY)) { + /* + * H+LX == hfsplusutils, H+Lx == this driver, H+lx is unused + * all three are registered with Apple for our use + */ + vhdr->last_mount_vers = cpu_to_be32(HFSP_MOUNT_VERSION); + vhdr->modify_date = hfsp_now2mt(); + be32_add_cpu(&vhdr->write_count, 1); + vhdr->attributes &= cpu_to_be32(~HFSPLUS_VOL_UNMNT); + vhdr->attributes |= cpu_to_be32(HFSPLUS_VOL_INCNSTNT); + hfsplus_sync_fs(sb, 1); - /* H+LX == hfsplusutils, H+Lx == this driver, H+lx is unused - * all three are registered with Apple for our use - */ - vhdr->last_mount_vers = cpu_to_be32(HFSP_MOUNT_VERSION); - vhdr->modify_date = hfsp_now2mt(); - be32_add_cpu(&vhdr->write_count, 1); - vhdr->attributes &= cpu_to_be32(~HFSPLUS_VOL_UNMNT); - vhdr->attributes |= cpu_to_be32(HFSPLUS_VOL_INCNSTNT); - hfsplus_sync_fs(sb, 1); - - if (!sbi->hidden_dir) { - mutex_lock(&sbi->vh_mutex); - sbi->hidden_dir = hfsplus_new_inode(sb, S_IFDIR); - hfsplus_create_cat(sbi->hidden_dir->i_ino, sb->s_root->d_inode, - &str, sbi->hidden_dir); - mutex_unlock(&sbi->vh_mutex); - - hfsplus_mark_inode_dirty(sbi->hidden_dir, HFSPLUS_I_CAT_DIRTY); + if (!sbi->hidden_dir) { + mutex_lock(&sbi->vh_mutex); + sbi->hidden_dir = hfsplus_new_inode(sb, S_IFDIR); + hfsplus_create_cat(sbi->hidden_dir->i_ino, root, &str, + sbi->hidden_dir); + mutex_unlock(&sbi->vh_mutex); + + hfsplus_mark_inode_dirty(sbi->hidden_dir, + HFSPLUS_I_CAT_DIRTY); + } } -out: + + sb->s_d_op = &hfsplus_dentry_operations; + sb->s_root = d_alloc_root(root); + if (!sb->s_root) { + err = -ENOMEM; + goto out_put_hidden_dir; + } + unload_nls(sbi->nls); sbi->nls = nls; return 0; -cleanup: - hfsplus_put_super(sb); +out_put_hidden_dir: + iput(sbi->hidden_dir); +out_put_root: + iput(sbi->alloc_file); +out_put_alloc_file: + iput(sbi->alloc_file); +out_close_cat_tree: + hfs_btree_close(sbi->cat_tree); +out_close_ext_tree: + hfs_btree_close(sbi->ext_tree); +out_free_vhdr: + kfree(sbi->s_vhdr); + kfree(sbi->s_backup_vhdr); +out_unload_nls: + unload_nls(sbi->nls); unload_nls(nls); + kfree(sbi); +out: return err; } diff --git a/fs/hfsplus/wrapper.c b/fs/hfsplus/wrapper.c index 1962317..3031d81 100644 --- a/fs/hfsplus/wrapper.c +++ b/fs/hfsplus/wrapper.c @@ -167,7 +167,7 @@ reread: break; case cpu_to_be16(HFSP_WRAP_MAGIC): if (!hfsplus_read_mdb(sbi->s_vhdr, &wd)) - goto out; + goto out_free_backup_vhdr; wd.ablk_size >>= HFSPLUS_SECTOR_SHIFT; part_start += wd.ablk_start + wd.embed_start * wd.ablk_size; part_size = wd.embed_count * wd.ablk_size; @@ -179,7 +179,7 @@ reread: * (should do this only for cdrom/loop though) */ if (hfs_part_find(sb, &part_start, &part_size)) - goto out; + goto out_free_backup_vhdr; goto reread; } diff --git a/fs/hpfs/Kconfig b/fs/hpfs/Kconfig index 63b6f56..0c39dc3 100644 --- a/fs/hpfs/Kconfig +++ b/fs/hpfs/Kconfig @@ -1,7 +1,7 @@ config HPFS_FS tristate "OS/2 HPFS file system support" depends on BLOCK - depends on BKL # nontrivial to fix + depends on BROKEN || !PREEMPT help OS/2 is IBM's operating system for PC's, the same as Warp, and HPFS is the file system used for organizing files on OS/2 hard disk diff --git a/fs/hpfs/dir.c b/fs/hpfs/dir.c index d32f63a..b3d7c0d 100644 --- a/fs/hpfs/dir.c +++ b/fs/hpfs/dir.c @@ -6,16 +6,15 @@ * directory VFS functions */ -#include <linux/smp_lock.h> #include <linux/slab.h> #include "hpfs_fn.h" static int hpfs_dir_release(struct inode *inode, struct file *filp) { - lock_kernel(); + hpfs_lock(inode->i_sb); hpfs_del_pos(inode, &filp->f_pos); /*hpfs_write_if_changed(inode);*/ - unlock_kernel(); + hpfs_unlock(inode->i_sb); return 0; } @@ -30,7 +29,7 @@ static loff_t hpfs_dir_lseek(struct file *filp, loff_t off, int whence) struct hpfs_inode_info *hpfs_inode = hpfs_i(i); struct super_block *s = i->i_sb; - lock_kernel(); + hpfs_lock(s); /*printk("dir lseek\n");*/ if (new_off == 0 || new_off == 1 || new_off == 11 || new_off == 12 || new_off == 13) goto ok; @@ -43,12 +42,12 @@ static loff_t hpfs_dir_lseek(struct file *filp, loff_t off, int whence) } mutex_unlock(&i->i_mutex); ok: - unlock_kernel(); + hpfs_unlock(s); return filp->f_pos = new_off; fail: mutex_unlock(&i->i_mutex); /*printk("illegal lseek: %016llx\n", new_off);*/ - unlock_kernel(); + hpfs_unlock(s); return -ESPIPE; } @@ -64,7 +63,7 @@ static int hpfs_readdir(struct file *filp, void *dirent, filldir_t filldir) int c1, c2 = 0; int ret = 0; - lock_kernel(); + hpfs_lock(inode->i_sb); if (hpfs_sb(inode->i_sb)->sb_chk) { if (hpfs_chk_sectors(inode->i_sb, inode->i_ino, 1, "dir_fnode")) { @@ -167,7 +166,7 @@ static int hpfs_readdir(struct file *filp, void *dirent, filldir_t filldir) hpfs_brelse4(&qbh); } out: - unlock_kernel(); + hpfs_unlock(inode->i_sb); return ret; } @@ -197,10 +196,10 @@ struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct name struct inode *result = NULL; struct hpfs_inode_info *hpfs_result; - lock_kernel(); + hpfs_lock(dir->i_sb); if ((err = hpfs_chk_name(name, &len))) { if (err == -ENAMETOOLONG) { - unlock_kernel(); + hpfs_unlock(dir->i_sb); return ERR_PTR(-ENAMETOOLONG); } goto end_add; @@ -298,7 +297,7 @@ struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct name end: end_add: - unlock_kernel(); + hpfs_unlock(dir->i_sb); d_add(dentry, result); return NULL; @@ -311,7 +310,7 @@ struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct name /*bail:*/ - unlock_kernel(); + hpfs_unlock(dir->i_sb); return ERR_PTR(-ENOENT); } diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c index c034088..2dbae20 100644 --- a/fs/hpfs/file.c +++ b/fs/hpfs/file.c @@ -6,16 +6,15 @@ * file VFS functions */ -#include <linux/smp_lock.h> #include "hpfs_fn.h" #define BLOCKS(size) (((size) + 511) >> 9) static int hpfs_file_release(struct inode *inode, struct file *file) { - lock_kernel(); + hpfs_lock(inode->i_sb); hpfs_write_if_changed(inode); - unlock_kernel(); + hpfs_unlock(inode->i_sb); return 0; } @@ -49,14 +48,14 @@ static secno hpfs_bmap(struct inode *inode, unsigned file_secno) static void hpfs_truncate(struct inode *i) { if (IS_IMMUTABLE(i)) return /*-EPERM*/; - lock_kernel(); + hpfs_lock(i->i_sb); hpfs_i(i)->i_n_secs = 0; i->i_blocks = 1 + ((i->i_size + 511) >> 9); hpfs_i(i)->mmu_private = i->i_size; hpfs_truncate_btree(i->i_sb, i->i_ino, 1, ((i->i_size + 511) >> 9)); hpfs_write_inode(i); hpfs_i(i)->i_n_secs = 0; - unlock_kernel(); + hpfs_unlock(i->i_sb); } static int hpfs_get_block(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create) diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h index 1c43dbe..c15adbc 100644 --- a/fs/hpfs/hpfs_fn.h +++ b/fs/hpfs/hpfs_fn.h @@ -342,3 +342,25 @@ static inline time32_t gmt_to_local(struct super_block *s, time_t t) extern struct timezone sys_tz; return t - sys_tz.tz_minuteswest * 60 - hpfs_sb(s)->sb_timeshift; } + +/* + * Locking: + * + * hpfs_lock() is a leftover from the big kernel lock. + * Right now, these functions are empty and only left + * for documentation purposes. The file system no longer + * works on SMP systems, so the lock is not needed + * any more. + * + * If someone is interested in making it work again, this + * would be the place to start by adding a per-superblock + * mutex and fixing all the bugs and performance issues + * caused by that. + */ +static inline void hpfs_lock(struct super_block *s) +{ +} + +static inline void hpfs_unlock(struct super_block *s) +{ +} diff --git a/fs/hpfs/inode.c b/fs/hpfs/inode.c index 1ae35ba..87f1f78 100644 --- a/fs/hpfs/inode.c +++ b/fs/hpfs/inode.c @@ -6,7 +6,6 @@ * inode VFS functions */ -#include <linux/smp_lock.h> #include <linux/slab.h> #include "hpfs_fn.h" @@ -267,7 +266,7 @@ int hpfs_setattr(struct dentry *dentry, struct iattr *attr) struct inode *inode = dentry->d_inode; int error = -EINVAL; - lock_kernel(); + hpfs_lock(inode->i_sb); if (inode->i_ino == hpfs_sb(inode->i_sb)->sb_root) goto out_unlock; if ((attr->ia_valid & ATTR_SIZE) && attr->ia_size > inode->i_size) @@ -290,7 +289,7 @@ int hpfs_setattr(struct dentry *dentry, struct iattr *attr) hpfs_write_inode(inode); out_unlock: - unlock_kernel(); + hpfs_unlock(inode->i_sb); return error; } @@ -307,8 +306,8 @@ void hpfs_evict_inode(struct inode *inode) truncate_inode_pages(&inode->i_data, 0); end_writeback(inode); if (!inode->i_nlink) { - lock_kernel(); + hpfs_lock(inode->i_sb); hpfs_remove_fnode(inode->i_sb, inode->i_ino); - unlock_kernel(); + hpfs_unlock(inode->i_sb); } } diff --git a/fs/hpfs/namei.c b/fs/hpfs/namei.c index f4ad9e3..d5f8c8a 100644 --- a/fs/hpfs/namei.c +++ b/fs/hpfs/namei.c @@ -6,7 +6,6 @@ * adding & removing files & directories */ #include <linux/sched.h> -#include <linux/smp_lock.h> #include "hpfs_fn.h" static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) @@ -25,7 +24,7 @@ static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) struct hpfs_dirent dee; int err; if ((err = hpfs_chk_name(name, &len))) return err==-ENOENT ? -EINVAL : err; - lock_kernel(); + hpfs_lock(dir->i_sb); err = -ENOSPC; fnode = hpfs_alloc_fnode(dir->i_sb, hpfs_i(dir)->i_dno, &fno, &bh); if (!fnode) @@ -103,7 +102,7 @@ static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) } d_instantiate(dentry, result); mutex_unlock(&hpfs_i(dir)->i_mutex); - unlock_kernel(); + hpfs_unlock(dir->i_sb); return 0; bail3: mutex_unlock(&hpfs_i(dir)->i_mutex); @@ -115,7 +114,7 @@ bail1: brelse(bh); hpfs_free_sectors(dir->i_sb, fno, 1); bail: - unlock_kernel(); + hpfs_unlock(dir->i_sb); return err; } @@ -132,7 +131,7 @@ static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struc int err; if ((err = hpfs_chk_name(name, &len))) return err==-ENOENT ? -EINVAL : err; - lock_kernel(); + hpfs_lock(dir->i_sb); err = -ENOSPC; fnode = hpfs_alloc_fnode(dir->i_sb, hpfs_i(dir)->i_dno, &fno, &bh); if (!fnode) @@ -195,7 +194,7 @@ static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struc } d_instantiate(dentry, result); mutex_unlock(&hpfs_i(dir)->i_mutex); - unlock_kernel(); + hpfs_unlock(dir->i_sb); return 0; bail2: @@ -205,7 +204,7 @@ bail1: brelse(bh); hpfs_free_sectors(dir->i_sb, fno, 1); bail: - unlock_kernel(); + hpfs_unlock(dir->i_sb); return err; } @@ -224,7 +223,7 @@ static int hpfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t if (hpfs_sb(dir->i_sb)->sb_eas < 2) return -EPERM; if (!new_valid_dev(rdev)) return -EINVAL; - lock_kernel(); + hpfs_lock(dir->i_sb); err = -ENOSPC; fnode = hpfs_alloc_fnode(dir->i_sb, hpfs_i(dir)->i_dno, &fno, &bh); if (!fnode) @@ -274,7 +273,7 @@ static int hpfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t d_instantiate(dentry, result); mutex_unlock(&hpfs_i(dir)->i_mutex); brelse(bh); - unlock_kernel(); + hpfs_unlock(dir->i_sb); return 0; bail2: mutex_unlock(&hpfs_i(dir)->i_mutex); @@ -283,7 +282,7 @@ bail1: brelse(bh); hpfs_free_sectors(dir->i_sb, fno, 1); bail: - unlock_kernel(); + hpfs_unlock(dir->i_sb); return err; } @@ -299,9 +298,9 @@ static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *sy struct inode *result; int err; if ((err = hpfs_chk_name(name, &len))) return err==-ENOENT ? -EINVAL : err; - lock_kernel(); + hpfs_lock(dir->i_sb); if (hpfs_sb(dir->i_sb)->sb_eas < 2) { - unlock_kernel(); + hpfs_unlock(dir->i_sb); return -EPERM; } err = -ENOSPC; @@ -354,7 +353,7 @@ static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *sy hpfs_write_inode_nolock(result); d_instantiate(dentry, result); mutex_unlock(&hpfs_i(dir)->i_mutex); - unlock_kernel(); + hpfs_unlock(dir->i_sb); return 0; bail2: mutex_unlock(&hpfs_i(dir)->i_mutex); @@ -363,7 +362,7 @@ bail1: brelse(bh); hpfs_free_sectors(dir->i_sb, fno, 1); bail: - unlock_kernel(); + hpfs_unlock(dir->i_sb); return err; } @@ -380,7 +379,7 @@ static int hpfs_unlink(struct inode *dir, struct dentry *dentry) int rep = 0; int err; - lock_kernel(); + hpfs_lock(dir->i_sb); hpfs_adjust_length(name, &len); again: mutex_lock(&hpfs_i(inode)->i_parent_mutex); @@ -416,7 +415,7 @@ again: dentry_unhash(dentry); if (!d_unhashed(dentry)) { dput(dentry); - unlock_kernel(); + hpfs_unlock(dir->i_sb); return -ENOSPC; } if (generic_permission(inode, MAY_WRITE, 0, NULL) || @@ -435,7 +434,7 @@ again: if (!err) goto again; } - unlock_kernel(); + hpfs_unlock(dir->i_sb); return -ENOSPC; default: drop_nlink(inode); @@ -448,7 +447,7 @@ out1: out: mutex_unlock(&hpfs_i(dir)->i_mutex); mutex_unlock(&hpfs_i(inode)->i_parent_mutex); - unlock_kernel(); + hpfs_unlock(dir->i_sb); return err; } @@ -466,7 +465,7 @@ static int hpfs_rmdir(struct inode *dir, struct dentry *dentry) int r; hpfs_adjust_length(name, &len); - lock_kernel(); + hpfs_lock(dir->i_sb); mutex_lock(&hpfs_i(inode)->i_parent_mutex); mutex_lock(&hpfs_i(dir)->i_mutex); err = -ENOENT; @@ -508,7 +507,7 @@ out1: out: mutex_unlock(&hpfs_i(dir)->i_mutex); mutex_unlock(&hpfs_i(inode)->i_parent_mutex); - unlock_kernel(); + hpfs_unlock(dir->i_sb); return err; } @@ -521,21 +520,21 @@ static int hpfs_symlink_readpage(struct file *file, struct page *page) int err; err = -EIO; - lock_kernel(); + hpfs_lock(i->i_sb); if (!(fnode = hpfs_map_fnode(i->i_sb, i->i_ino, &bh))) goto fail; err = hpfs_read_ea(i->i_sb, fnode, "SYMLINK", link, PAGE_SIZE); brelse(bh); if (err) goto fail; - unlock_kernel(); + hpfs_unlock(i->i_sb); SetPageUptodate(page); kunmap(page); unlock_page(page); return 0; fail: - unlock_kernel(); + hpfs_unlock(i->i_sb); SetPageError(page); kunmap(page); unlock_page(page); @@ -567,7 +566,7 @@ static int hpfs_rename(struct inode *old_dir, struct dentry *old_dentry, err = 0; hpfs_adjust_length(old_name, &old_len); - lock_kernel(); + hpfs_lock(i->i_sb); /* order doesn't matter, due to VFS exclusion */ mutex_lock(&hpfs_i(i)->i_parent_mutex); if (new_inode) @@ -659,7 +658,7 @@ end1: mutex_unlock(&hpfs_i(i)->i_parent_mutex); if (new_inode) mutex_unlock(&hpfs_i(new_inode)->i_parent_mutex); - unlock_kernel(); + hpfs_unlock(i->i_sb); return err; } diff --git a/fs/hpfs/super.c b/fs/hpfs/super.c index b30426b..c89b408 100644 --- a/fs/hpfs/super.c +++ b/fs/hpfs/super.c @@ -13,7 +13,6 @@ #include <linux/statfs.h> #include <linux/magic.h> #include <linux/sched.h> -#include <linux/smp_lock.h> #include <linux/bitmap.h> #include <linux/slab.h> @@ -103,15 +102,11 @@ static void hpfs_put_super(struct super_block *s) { struct hpfs_sb_info *sbi = hpfs_sb(s); - lock_kernel(); - kfree(sbi->sb_cp_table); kfree(sbi->sb_bmp_dir); unmark_dirty(s); s->s_fs_info = NULL; kfree(sbi); - - unlock_kernel(); } unsigned hpfs_count_one_bitmap(struct super_block *s, secno secno) @@ -143,7 +138,7 @@ static int hpfs_statfs(struct dentry *dentry, struct kstatfs *buf) struct super_block *s = dentry->d_sb; struct hpfs_sb_info *sbi = hpfs_sb(s); u64 id = huge_encode_dev(s->s_bdev->bd_dev); - lock_kernel(); + hpfs_lock(s); /*if (sbi->sb_n_free == -1) {*/ sbi->sb_n_free = count_bitmaps(s); @@ -160,7 +155,7 @@ static int hpfs_statfs(struct dentry *dentry, struct kstatfs *buf) buf->f_fsid.val[1] = (u32)(id >> 32); buf->f_namelen = 254; - unlock_kernel(); + hpfs_unlock(s); return 0; } @@ -406,7 +401,7 @@ static int hpfs_remount_fs(struct super_block *s, int *flags, char *data) *flags |= MS_NOATIME; - lock_kernel(); + hpfs_lock(s); lock_super(s); uid = sbi->sb_uid; gid = sbi->sb_gid; umask = 0777 & ~sbi->sb_mode; @@ -441,12 +436,12 @@ static int hpfs_remount_fs(struct super_block *s, int *flags, char *data) replace_mount_options(s, new_opts); unlock_super(s); - unlock_kernel(); + hpfs_unlock(s); return 0; out_err: unlock_super(s); - unlock_kernel(); + hpfs_unlock(s); kfree(new_opts); return -EINVAL; } @@ -484,13 +479,15 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) int o; - lock_kernel(); + if (num_possible_cpus() > 1) { + printk(KERN_ERR "HPFS is not SMP safe\n"); + return -EINVAL; + } save_mount_options(s, options); sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); if (!sbi) { - unlock_kernel(); return -ENOMEM; } s->s_fs_info = sbi; @@ -677,7 +674,6 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) root->i_blocks = 5; hpfs_brelse4(&qbh); } - unlock_kernel(); return 0; bail4: brelse(bh2); @@ -689,7 +685,6 @@ bail0: kfree(sbi->sb_cp_table); s->s_fs_info = NULL; kfree(sbi); - unlock_kernel(); return -EINVAL; } @@ -84,16 +84,13 @@ static struct hlist_head *inode_hashtable __read_mostly; DEFINE_SPINLOCK(inode_lock); /* - * iprune_sem provides exclusion between the kswapd or try_to_free_pages - * icache shrinking path, and the umount path. Without this exclusion, - * by the time prune_icache calls iput for the inode whose pages it has - * been invalidating, or by the time it calls clear_inode & destroy_inode - * from its final dispose_list, the struct super_block they refer to - * (for inode->i_sb->s_op) may already have been freed and reused. + * iprune_sem provides exclusion between the icache shrinking and the + * umount path. * - * We make this an rwsem because the fastpath is icache shrinking. In - * some cases a filesystem may be doing a significant amount of work in - * its inode reclaim code, so this should improve parallelism. + * We don't actually need it to protect anything in the umount path, + * but only need to cycle through it to make sure any inode that + * prune_icache took off the LRU list has been fully torn down by the + * time we are past evict_inodes. */ static DECLARE_RWSEM(iprune_sem); @@ -295,6 +292,20 @@ static void destroy_inode(struct inode *inode) call_rcu(&inode->i_rcu, i_callback); } +void address_space_init_once(struct address_space *mapping) +{ + memset(mapping, 0, sizeof(*mapping)); + INIT_RADIX_TREE(&mapping->page_tree, GFP_ATOMIC); + spin_lock_init(&mapping->tree_lock); + spin_lock_init(&mapping->i_mmap_lock); + INIT_LIST_HEAD(&mapping->private_list); + spin_lock_init(&mapping->private_lock); + INIT_RAW_PRIO_TREE_ROOT(&mapping->i_mmap); + INIT_LIST_HEAD(&mapping->i_mmap_nonlinear); + mutex_init(&mapping->unmap_mutex); +} +EXPORT_SYMBOL(address_space_init_once); + /* * These are initializations that only need to be done * once, because the fields are idempotent across use @@ -308,13 +319,7 @@ void inode_init_once(struct inode *inode) INIT_LIST_HEAD(&inode->i_devices); INIT_LIST_HEAD(&inode->i_wb_list); INIT_LIST_HEAD(&inode->i_lru); - INIT_RADIX_TREE(&inode->i_data.page_tree, GFP_ATOMIC); - spin_lock_init(&inode->i_data.tree_lock); - spin_lock_init(&inode->i_data.i_mmap_lock); - INIT_LIST_HEAD(&inode->i_data.private_list); - spin_lock_init(&inode->i_data.private_lock); - INIT_RAW_PRIO_TREE_ROOT(&inode->i_data.i_mmap); - INIT_LIST_HEAD(&inode->i_data.i_mmap_nonlinear); + address_space_init_once(&inode->i_data); i_size_ordered_init(inode); #ifdef CONFIG_FSNOTIFY INIT_HLIST_HEAD(&inode->i_fsnotify_marks); @@ -508,17 +513,12 @@ void evict_inodes(struct super_block *sb) struct inode *inode, *next; LIST_HEAD(dispose); - down_write(&iprune_sem); - spin_lock(&inode_lock); list_for_each_entry_safe(inode, next, &sb->s_inodes, i_sb_list) { if (atomic_read(&inode->i_count)) continue; - - if (inode->i_state & (I_NEW | I_FREEING | I_WILL_FREE)) { - WARN_ON(1); + if (inode->i_state & (I_NEW | I_FREEING | I_WILL_FREE)) continue; - } inode->i_state |= I_FREEING; @@ -534,28 +534,40 @@ void evict_inodes(struct super_block *sb) spin_unlock(&inode_lock); dispose_list(&dispose); + + /* + * Cycle through iprune_sem to make sure any inode that prune_icache + * moved off the list before we took the lock has been fully torn + * down. + */ + down_write(&iprune_sem); up_write(&iprune_sem); } /** * invalidate_inodes - attempt to free all inodes on a superblock * @sb: superblock to operate on + * @kill_dirty: flag to guide handling of dirty inodes * * Attempts to free all inodes for a given superblock. If there were any * busy inodes return a non-zero value, else zero. + * If @kill_dirty is set, discard dirty inodes too, otherwise treat + * them as busy. */ -int invalidate_inodes(struct super_block *sb) +int invalidate_inodes(struct super_block *sb, bool kill_dirty) { int busy = 0; struct inode *inode, *next; LIST_HEAD(dispose); - down_write(&iprune_sem); - spin_lock(&inode_lock); list_for_each_entry_safe(inode, next, &sb->s_inodes, i_sb_list) { if (inode->i_state & (I_NEW | I_FREEING | I_WILL_FREE)) continue; + if (inode->i_state & I_DIRTY && !kill_dirty) { + busy = 1; + continue; + } if (atomic_read(&inode->i_count)) { busy = 1; continue; @@ -575,7 +587,6 @@ int invalidate_inodes(struct super_block *sb) spin_unlock(&inode_lock); dispose_list(&dispose); - up_write(&iprune_sem); return busy; } @@ -1704,7 +1715,7 @@ void init_special_inode(struct inode *inode, umode_t mode, dev_t rdev) EXPORT_SYMBOL(init_special_inode); /** - * Init uid,gid,mode for new inode according to posix standards + * inode_init_owner - Init uid,gid,mode for new inode according to posix standards * @inode: New inode * @dir: Directory inode * @mode: mode of the new inode diff --git a/fs/internal.h b/fs/internal.h index 0663568..8318059 100644 --- a/fs/internal.h +++ b/fs/internal.h @@ -12,6 +12,7 @@ #include <linux/lglock.h> struct super_block; +struct file_system_type; struct linux_binprm; struct path; @@ -61,10 +62,9 @@ extern int check_unsafe_exec(struct linux_binprm *); extern int copy_mount_options(const void __user *, unsigned long *); extern int copy_mount_string(const void __user *, char **); -extern void free_vfsmnt(struct vfsmount *); -extern struct vfsmount *alloc_vfsmnt(const char *); extern unsigned int mnt_get_count(struct vfsmount *mnt); extern struct vfsmount *__lookup_mnt(struct vfsmount *, struct dentry *, int); +extern struct vfsmount *lookup_mnt(struct path *); extern void mnt_set_mountpoint(struct vfsmount *, struct dentry *, struct vfsmount *); extern void release_mounts(struct list_head *); @@ -99,6 +99,8 @@ extern struct file *get_empty_filp(void); extern int do_remount_sb(struct super_block *, int, void *, int); extern void __put_super(struct super_block *sb); extern void put_super(struct super_block *sb); +extern struct dentry *mount_fs(struct file_system_type *, + int, const char *, void *); /* * open.c @@ -106,10 +108,23 @@ extern void put_super(struct super_block *sb); struct nameidata; extern struct file *nameidata_to_filp(struct nameidata *); extern void release_open_intent(struct nameidata *); +struct open_flags { + int open_flag; + int mode; + int acc_mode; + int intent; +}; +extern struct file *do_filp_open(int dfd, const char *pathname, + const struct open_flags *op, int lookup_flags); +extern struct file *do_file_open_root(struct dentry *, struct vfsmount *, + const char *, const struct open_flags *, int lookup_flags); + +extern long do_handle_open(int mountdirfd, + struct file_handle __user *ufh, int open_flag); /* * inode.c */ extern int get_nr_dirty_inodes(void); extern void evict_inodes(struct super_block *); -extern int invalidate_inodes(struct super_block *); +extern int invalidate_inodes(struct super_block *, bool); @@ -273,6 +273,13 @@ int __generic_block_fiemap(struct inode *inode, len = isize; } + /* + * Some filesystems can't deal with being asked to map less than + * blocksize, so make sure our len is at least block length. + */ + if (logical_to_blk(inode, len) == 0) + len = blk_to_logical(inode, 1); + start_blk = logical_to_blk(inode, start); last_blk = logical_to_blk(inode, start + len - 1); @@ -541,6 +548,7 @@ int do_vfs_ioctl(struct file *filp, unsigned int fd, unsigned int cmd, { int error = 0; int __user *argp = (int __user *)arg; + struct inode *inode = filp->f_path.dentry->d_inode; switch (cmd) { case FIOCLEX: @@ -560,13 +568,11 @@ int do_vfs_ioctl(struct file *filp, unsigned int fd, unsigned int cmd, break; case FIOQSIZE: - if (S_ISDIR(filp->f_path.dentry->d_inode->i_mode) || - S_ISREG(filp->f_path.dentry->d_inode->i_mode) || - S_ISLNK(filp->f_path.dentry->d_inode->i_mode)) { - loff_t res = - inode_get_bytes(filp->f_path.dentry->d_inode); - error = copy_to_user((loff_t __user *)arg, &res, - sizeof(res)) ? -EFAULT : 0; + if (S_ISDIR(inode->i_mode) || S_ISREG(inode->i_mode) || + S_ISLNK(inode->i_mode)) { + loff_t res = inode_get_bytes(inode); + error = copy_to_user(argp, &res, sizeof(res)) ? + -EFAULT : 0; } else error = -ENOTTY; break; @@ -583,14 +589,10 @@ int do_vfs_ioctl(struct file *filp, unsigned int fd, unsigned int cmd, return ioctl_fiemap(filp, arg); case FIGETBSZ: - { - struct inode *inode = filp->f_path.dentry->d_inode; - int __user *p = (int __user *)arg; - return put_user(inode->i_sb->s_blocksize, p); - } + return put_user(inode->i_sb->s_blocksize, argp); default: - if (S_ISREG(filp->f_path.dentry->d_inode->i_mode)) + if (S_ISREG(inode->i_mode)) error = file_ioctl(filp, cmd, arg); else error = vfs_ioctl(filp, cmd, arg); diff --git a/fs/isofs/export.c b/fs/isofs/export.c index ed752cb..dd4687f 100644 --- a/fs/isofs/export.c +++ b/fs/isofs/export.c @@ -124,9 +124,13 @@ isofs_export_encode_fh(struct dentry *dentry, * offset of the inode and the upper 16 bits of fh32[1] to * hold the offset of the parent. */ - - if (len < 3 || (connectable && len < 5)) + if (connectable && (len < 5)) { + *max_len = 5; + return 255; + } else if (len < 3) { + *max_len = 3; return 255; + } len = 3; fh32[0] = ei->i_iget5_block; diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c index da1b5e4..eb11601 100644 --- a/fs/jbd/journal.c +++ b/fs/jbd/journal.c @@ -839,7 +839,7 @@ journal_t * journal_init_inode (struct inode *inode) err = journal_bmap(journal, 0, &blocknr); /* If that failed, give up */ if (err) { - printk(KERN_ERR "%s: Cannnot locate journal superblock\n", + printk(KERN_ERR "%s: Cannot locate journal superblock\n", __func__); goto out_err; } diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 9e46869..90407b8 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -473,7 +473,8 @@ int __jbd2_log_space_left(journal_t *journal) } /* - * Called under j_state_lock. Returns true if a transaction commit was started. + * Called with j_state_lock locked for writing. + * Returns true if a transaction commit was started. */ int __jbd2_log_start_commit(journal_t *journal, tid_t target) { @@ -520,11 +521,13 @@ int jbd2_journal_force_commit_nested(journal_t *journal) { transaction_t *transaction = NULL; tid_t tid; + int need_to_start = 0; read_lock(&journal->j_state_lock); if (journal->j_running_transaction && !current->journal_info) { transaction = journal->j_running_transaction; - __jbd2_log_start_commit(journal, transaction->t_tid); + if (!tid_geq(journal->j_commit_request, transaction->t_tid)) + need_to_start = 1; } else if (journal->j_committing_transaction) transaction = journal->j_committing_transaction; @@ -535,6 +538,8 @@ int jbd2_journal_force_commit_nested(journal_t *journal) tid = transaction->t_tid; read_unlock(&journal->j_state_lock); + if (need_to_start) + jbd2_log_start_commit(journal, tid); jbd2_log_wait_commit(journal, tid); return 1; } @@ -986,7 +991,7 @@ journal_t * jbd2_journal_init_inode (struct inode *inode) err = jbd2_journal_bmap(journal, 0, &blocknr); /* If that failed, give up */ if (err) { - printk(KERN_ERR "%s: Cannnot locate journal superblock\n", + printk(KERN_ERR "%s: Cannot locate journal superblock\n", __func__); goto out_err; } diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index faad2bd..1d11910 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -117,10 +117,10 @@ static inline void update_t_max_wait(transaction_t *transaction) static int start_this_handle(journal_t *journal, handle_t *handle, int gfp_mask) { - transaction_t *transaction; - int needed; - int nblocks = handle->h_buffer_credits; - transaction_t *new_transaction = NULL; + transaction_t *transaction, *new_transaction = NULL; + tid_t tid; + int needed, need_to_start; + int nblocks = handle->h_buffer_credits; if (nblocks > journal->j_max_transaction_buffers) { printk(KERN_ERR "JBD: %s wants too many credits (%d > %d)\n", @@ -222,8 +222,11 @@ repeat: atomic_sub(nblocks, &transaction->t_outstanding_credits); prepare_to_wait(&journal->j_wait_transaction_locked, &wait, TASK_UNINTERRUPTIBLE); - __jbd2_log_start_commit(journal, transaction->t_tid); + tid = transaction->t_tid; + need_to_start = !tid_geq(journal->j_commit_request, tid); read_unlock(&journal->j_state_lock); + if (need_to_start) + jbd2_log_start_commit(journal, tid); schedule(); finish_wait(&journal->j_wait_transaction_locked, &wait); goto repeat; @@ -442,7 +445,8 @@ int jbd2__journal_restart(handle_t *handle, int nblocks, int gfp_mask) { transaction_t *transaction = handle->h_transaction; journal_t *journal = transaction->t_journal; - int ret; + tid_t tid; + int need_to_start, ret; /* If we've had an abort of any type, don't even think about * actually doing the restart! */ @@ -465,8 +469,11 @@ int jbd2__journal_restart(handle_t *handle, int nblocks, int gfp_mask) spin_unlock(&transaction->t_handle_lock); jbd_debug(2, "restarting handle %p\n", handle); - __jbd2_log_start_commit(journal, transaction->t_tid); + tid = transaction->t_tid; + need_to_start = !tid_geq(journal->j_commit_request, tid); read_unlock(&journal->j_state_lock); + if (need_to_start) + jbd2_log_start_commit(journal, tid); lock_map_release(&handle->h_lockdep_map); handle->h_buffer_credits = nblocks; diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c index 9297865..82faddd 100644 --- a/fs/jffs2/dir.c +++ b/fs/jffs2/dir.c @@ -215,8 +215,7 @@ static int jffs2_create(struct inode *dir_i, struct dentry *dentry, int mode, no chance of AB-BA deadlock involving its f->sem). */ mutex_unlock(&f->sem); - ret = jffs2_do_create(c, dir_f, f, ri, - dentry->d_name.name, dentry->d_name.len); + ret = jffs2_do_create(c, dir_f, f, ri, &dentry->d_name); if (ret) goto fail; @@ -386,7 +385,7 @@ static int jffs2_symlink (struct inode *dir_i, struct dentry *dentry, const char jffs2_complete_reservation(c); - ret = jffs2_init_security(inode, dir_i); + ret = jffs2_init_security(inode, dir_i, &dentry->d_name); if (ret) goto fail; @@ -530,7 +529,7 @@ static int jffs2_mkdir (struct inode *dir_i, struct dentry *dentry, int mode) jffs2_complete_reservation(c); - ret = jffs2_init_security(inode, dir_i); + ret = jffs2_init_security(inode, dir_i, &dentry->d_name); if (ret) goto fail; @@ -703,7 +702,7 @@ static int jffs2_mknod (struct inode *dir_i, struct dentry *dentry, int mode, de jffs2_complete_reservation(c); - ret = jffs2_init_security(inode, dir_i); + ret = jffs2_init_security(inode, dir_i, &dentry->d_name); if (ret) goto fail; diff --git a/fs/jffs2/nodelist.h b/fs/jffs2/nodelist.h index 5a53d9b..e4619b0 100644 --- a/fs/jffs2/nodelist.h +++ b/fs/jffs2/nodelist.h @@ -401,7 +401,7 @@ int jffs2_write_inode_range(struct jffs2_sb_info *c, struct jffs2_inode_info *f, struct jffs2_raw_inode *ri, unsigned char *buf, uint32_t offset, uint32_t writelen, uint32_t *retlen); int jffs2_do_create(struct jffs2_sb_info *c, struct jffs2_inode_info *dir_f, struct jffs2_inode_info *f, - struct jffs2_raw_inode *ri, const char *name, int namelen); + struct jffs2_raw_inode *ri, const struct qstr *qstr); int jffs2_do_unlink(struct jffs2_sb_info *c, struct jffs2_inode_info *dir_f, const char *name, int namelen, struct jffs2_inode_info *dead_f, uint32_t time); int jffs2_do_link(struct jffs2_sb_info *c, struct jffs2_inode_info *dir_f, uint32_t ino, diff --git a/fs/jffs2/security.c b/fs/jffs2/security.c index 239f512..cfeb716 100644 --- a/fs/jffs2/security.c +++ b/fs/jffs2/security.c @@ -23,14 +23,15 @@ #include "nodelist.h" /* ---- Initial Security Label Attachment -------------- */ -int jffs2_init_security(struct inode *inode, struct inode *dir) +int jffs2_init_security(struct inode *inode, struct inode *dir, + const struct qstr *qstr) { int rc; size_t len; void *value; char *name; - rc = security_inode_init_security(inode, dir, &name, &value, &len); + rc = security_inode_init_security(inode, dir, qstr, &name, &value, &len); if (rc) { if (rc == -EOPNOTSUPP) return 0; diff --git a/fs/jffs2/write.c b/fs/jffs2/write.c index c819eb0..30d175b 100644 --- a/fs/jffs2/write.c +++ b/fs/jffs2/write.c @@ -424,7 +424,9 @@ int jffs2_write_inode_range(struct jffs2_sb_info *c, struct jffs2_inode_info *f, return ret; } -int jffs2_do_create(struct jffs2_sb_info *c, struct jffs2_inode_info *dir_f, struct jffs2_inode_info *f, struct jffs2_raw_inode *ri, const char *name, int namelen) +int jffs2_do_create(struct jffs2_sb_info *c, struct jffs2_inode_info *dir_f, + struct jffs2_inode_info *f, struct jffs2_raw_inode *ri, + const struct qstr *qstr) { struct jffs2_raw_dirent *rd; struct jffs2_full_dnode *fn; @@ -466,15 +468,15 @@ int jffs2_do_create(struct jffs2_sb_info *c, struct jffs2_inode_info *dir_f, str mutex_unlock(&f->sem); jffs2_complete_reservation(c); - ret = jffs2_init_security(&f->vfs_inode, &dir_f->vfs_inode); + ret = jffs2_init_security(&f->vfs_inode, &dir_f->vfs_inode, qstr); if (ret) return ret; ret = jffs2_init_acl_post(&f->vfs_inode); if (ret) return ret; - ret = jffs2_reserve_space(c, sizeof(*rd)+namelen, &alloclen, - ALLOC_NORMAL, JFFS2_SUMMARY_DIRENT_SIZE(namelen)); + ret = jffs2_reserve_space(c, sizeof(*rd)+qstr->len, &alloclen, + ALLOC_NORMAL, JFFS2_SUMMARY_DIRENT_SIZE(qstr->len)); if (ret) { /* Eep. */ @@ -493,19 +495,19 @@ int jffs2_do_create(struct jffs2_sb_info *c, struct jffs2_inode_info *dir_f, str rd->magic = cpu_to_je16(JFFS2_MAGIC_BITMASK); rd->nodetype = cpu_to_je16(JFFS2_NODETYPE_DIRENT); - rd->totlen = cpu_to_je32(sizeof(*rd) + namelen); + rd->totlen = cpu_to_je32(sizeof(*rd) + qstr->len); rd->hdr_crc = cpu_to_je32(crc32(0, rd, sizeof(struct jffs2_unknown_node)-4)); rd->pino = cpu_to_je32(dir_f->inocache->ino); rd->version = cpu_to_je32(++dir_f->highest_version); rd->ino = ri->ino; rd->mctime = ri->ctime; - rd->nsize = namelen; + rd->nsize = qstr->len; rd->type = DT_REG; rd->node_crc = cpu_to_je32(crc32(0, rd, sizeof(*rd)-8)); - rd->name_crc = cpu_to_je32(crc32(0, name, namelen)); + rd->name_crc = cpu_to_je32(crc32(0, qstr->name, qstr->len)); - fd = jffs2_write_dirent(c, dir_f, rd, name, namelen, ALLOC_NORMAL); + fd = jffs2_write_dirent(c, dir_f, rd, qstr->name, qstr->len, ALLOC_NORMAL); jffs2_free_raw_dirent(rd); diff --git a/fs/jffs2/xattr.h b/fs/jffs2/xattr.h index cf4f575..7be4beb 100644 --- a/fs/jffs2/xattr.h +++ b/fs/jffs2/xattr.h @@ -121,10 +121,11 @@ extern ssize_t jffs2_listxattr(struct dentry *, char *, size_t); #endif /* CONFIG_JFFS2_FS_XATTR */ #ifdef CONFIG_JFFS2_FS_SECURITY -extern int jffs2_init_security(struct inode *inode, struct inode *dir); +extern int jffs2_init_security(struct inode *inode, struct inode *dir, + const struct qstr *qstr); extern const struct xattr_handler jffs2_security_xattr_handler; #else -#define jffs2_init_security(inode,dir) (0) +#define jffs2_init_security(inode,dir,qstr) (0) #endif /* CONFIG_JFFS2_FS_SECURITY */ #endif /* _JFFS2_FS_XATTR_H_ */ diff --git a/fs/jfs/Makefile b/fs/jfs/Makefile index 3adb639..a58fa72 100644 --- a/fs/jfs/Makefile +++ b/fs/jfs/Makefile @@ -13,4 +13,4 @@ jfs-y := super.o file.o inode.o namei.o jfs_mount.o jfs_umount.o \ jfs-$(CONFIG_JFS_POSIX_ACL) += acl.o -EXTRA_CFLAGS += -D_JFS_4K +ccflags-y := -D_JFS_4K diff --git a/fs/jfs/jfs_xattr.h b/fs/jfs/jfs_xattr.h index 88b6cc5..e9e100f 100644 --- a/fs/jfs/jfs_xattr.h +++ b/fs/jfs/jfs_xattr.h @@ -62,10 +62,11 @@ extern ssize_t jfs_listxattr(struct dentry *, char *, size_t); extern int jfs_removexattr(struct dentry *, const char *); #ifdef CONFIG_JFS_SECURITY -extern int jfs_init_security(tid_t, struct inode *, struct inode *); +extern int jfs_init_security(tid_t, struct inode *, struct inode *, + const struct qstr *); #else static inline int jfs_init_security(tid_t tid, struct inode *inode, - struct inode *dir) + struct inode *dir, const struct qstr *qstr) { return 0; } diff --git a/fs/jfs/namei.c b/fs/jfs/namei.c index 81ead85..eaaf2b5 100644 --- a/fs/jfs/namei.c +++ b/fs/jfs/namei.c @@ -115,7 +115,7 @@ static int jfs_create(struct inode *dip, struct dentry *dentry, int mode, if (rc) goto out3; - rc = jfs_init_security(tid, ip, dip); + rc = jfs_init_security(tid, ip, dip, &dentry->d_name); if (rc) { txAbort(tid, 0); goto out3; @@ -253,7 +253,7 @@ static int jfs_mkdir(struct inode *dip, struct dentry *dentry, int mode) if (rc) goto out3; - rc = jfs_init_security(tid, ip, dip); + rc = jfs_init_security(tid, ip, dip, &dentry->d_name); if (rc) { txAbort(tid, 0); goto out3; @@ -809,9 +809,6 @@ static int jfs_link(struct dentry *old_dentry, if (ip->i_nlink == JFS_LINK_MAX) return -EMLINK; - if (ip->i_nlink == 0) - return -ENOENT; - dquot_initialize(dir); tid = txBegin(ip->i_sb, 0); @@ -932,7 +929,7 @@ static int jfs_symlink(struct inode *dip, struct dentry *dentry, mutex_lock_nested(&JFS_IP(dip)->commit_mutex, COMMIT_MUTEX_PARENT); mutex_lock_nested(&JFS_IP(ip)->commit_mutex, COMMIT_MUTEX_CHILD); - rc = jfs_init_security(tid, ip, dip); + rc = jfs_init_security(tid, ip, dip, &dentry->d_name); if (rc) goto out3; @@ -1395,7 +1392,7 @@ static int jfs_mknod(struct inode *dir, struct dentry *dentry, if (rc) goto out3; - rc = jfs_init_security(tid, ip, dir); + rc = jfs_init_security(tid, ip, dir, &dentry->d_name); if (rc) { txAbort(tid, 0); goto out3; @@ -1600,7 +1597,7 @@ out: static int jfs_ci_revalidate(struct dentry *dentry, struct nameidata *nd) { - if (nd->flags & LOOKUP_RCU) + if (nd && nd->flags & LOOKUP_RCU) return -ECHILD; /* * This is not negative dentry. Always valid. diff --git a/fs/jfs/xattr.c b/fs/jfs/xattr.c index 2d7f165..3fa4c322 100644 --- a/fs/jfs/xattr.c +++ b/fs/jfs/xattr.c @@ -1091,7 +1091,8 @@ int jfs_removexattr(struct dentry *dentry, const char *name) } #ifdef CONFIG_JFS_SECURITY -int jfs_init_security(tid_t tid, struct inode *inode, struct inode *dir) +int jfs_init_security(tid_t tid, struct inode *inode, struct inode *dir, + const struct qstr *qstr) { int rc; size_t len; @@ -1099,7 +1100,8 @@ int jfs_init_security(tid_t tid, struct inode *inode, struct inode *dir) char *suffix; char *name; - rc = security_inode_init_security(inode, dir, &suffix, &value, &len); + rc = security_inode_init_security(inode, dir, qstr, &suffix, &value, + &len); if (rc) { if (rc == -EOPNOTSUPP) return 0; diff --git a/fs/lockd/host.c b/fs/lockd/host.c index 5f1bcb2..b7c99bf 100644 --- a/fs/lockd/host.c +++ b/fs/lockd/host.c @@ -520,7 +520,7 @@ static struct nlm_host *next_host_state(struct hlist_head *cache, struct nsm_handle *nsm, const struct nlm_reboot *info) { - struct nlm_host *host = NULL; + struct nlm_host *host; struct hlist_head *chain; struct hlist_node *pos; @@ -532,12 +532,13 @@ static struct nlm_host *next_host_state(struct hlist_head *cache, host->h_state++; nlm_get_host(host); - goto out; + mutex_unlock(&nlm_host_mutex); + return host; } } -out: + mutex_unlock(&nlm_host_mutex); - return host; + return NULL; } /** @@ -145,7 +145,6 @@ static DEFINE_SPINLOCK(file_lock_lock); /* * Protects the two list heads above, plus the inode->i_flock list - * FIXME: should use a spinlock, once lockd and ceph are ready. */ void lock_flocks(void) { diff --git a/fs/minix/namei.c b/fs/minix/namei.c index ce7337d..6e6777f 100644 --- a/fs/minix/namei.c +++ b/fs/minix/namei.c @@ -213,7 +213,6 @@ static int minix_rename(struct inode * old_dir, struct dentry *old_dentry, new_de = minix_find_entry(new_dentry, &new_page); if (!new_de) goto out_dir; - inode_inc_link_count(old_inode); minix_set_link(new_de, new_page, old_inode); new_inode->i_ctime = CURRENT_TIME_SEC; if (dir_de) @@ -225,18 +224,15 @@ static int minix_rename(struct inode * old_dir, struct dentry *old_dentry, if (new_dir->i_nlink >= info->s_link_max) goto out_dir; } - inode_inc_link_count(old_inode); err = minix_add_link(new_dentry, old_inode); - if (err) { - inode_dec_link_count(old_inode); + if (err) goto out_dir; - } if (dir_de) inode_inc_link_count(new_dir); } minix_delete_entry(old_de, old_page); - inode_dec_link_count(old_inode); + mark_inode_dirty(old_inode); if (dir_de) { minix_set_link(dir_de, dir_page, new_dir); @@ -136,7 +136,7 @@ static int do_getname(const char __user *filename, char *page) return retval; } -char * getname(const char __user * filename) +static char *getname_flags(const char __user * filename, int flags) { char *tmp, *result; @@ -147,14 +147,21 @@ char * getname(const char __user * filename) result = tmp; if (retval < 0) { - __putname(tmp); - result = ERR_PTR(retval); + if (retval != -ENOENT || !(flags & LOOKUP_EMPTY)) { + __putname(tmp); + result = ERR_PTR(retval); + } } } audit_getname(result); return result; } +char *getname(const char __user * filename) +{ + return getname_flags(filename, 0); +} + #ifdef CONFIG_AUDITSYSCALL void putname(const char *name) { @@ -401,9 +408,11 @@ static int nameidata_drop_rcu(struct nameidata *nd) { struct fs_struct *fs = current->fs; struct dentry *dentry = nd->path.dentry; + int want_root = 0; BUG_ON(!(nd->flags & LOOKUP_RCU)); - if (nd->root.mnt) { + if (nd->root.mnt && !(nd->flags & LOOKUP_ROOT)) { + want_root = 1; spin_lock(&fs->lock); if (nd->root.mnt != fs->root.mnt || nd->root.dentry != fs->root.dentry) @@ -414,7 +423,7 @@ static int nameidata_drop_rcu(struct nameidata *nd) goto err; BUG_ON(nd->inode != dentry->d_inode); spin_unlock(&dentry->d_lock); - if (nd->root.mnt) { + if (want_root) { path_get(&nd->root); spin_unlock(&fs->lock); } @@ -427,7 +436,7 @@ static int nameidata_drop_rcu(struct nameidata *nd) err: spin_unlock(&dentry->d_lock); err_root: - if (nd->root.mnt) + if (want_root) spin_unlock(&fs->lock); return -ECHILD; } @@ -454,17 +463,11 @@ static int nameidata_dentry_drop_rcu(struct nameidata *nd, struct dentry *dentry { struct fs_struct *fs = current->fs; struct dentry *parent = nd->path.dentry; - - /* - * It can be possible to revalidate the dentry that we started - * the path walk with. force_reval_path may also revalidate the - * dentry already committed to the nameidata. - */ - if (unlikely(parent == dentry)) - return nameidata_drop_rcu(nd); + int want_root = 0; BUG_ON(!(nd->flags & LOOKUP_RCU)); - if (nd->root.mnt) { + if (nd->root.mnt && !(nd->flags & LOOKUP_ROOT)) { + want_root = 1; spin_lock(&fs->lock); if (nd->root.mnt != fs->root.mnt || nd->root.dentry != fs->root.dentry) @@ -484,7 +487,7 @@ static int nameidata_dentry_drop_rcu(struct nameidata *nd, struct dentry *dentry parent->d_count++; spin_unlock(&dentry->d_lock); spin_unlock(&parent->d_lock); - if (nd->root.mnt) { + if (want_root) { path_get(&nd->root); spin_unlock(&fs->lock); } @@ -498,7 +501,7 @@ err: spin_unlock(&dentry->d_lock); spin_unlock(&parent->d_lock); err_root: - if (nd->root.mnt) + if (want_root) spin_unlock(&fs->lock); return -ECHILD; } @@ -506,8 +509,16 @@ err_root: /* Try to drop out of rcu-walk mode if we were in it, otherwise do nothing. */ static inline int nameidata_dentry_drop_rcu_maybe(struct nameidata *nd, struct dentry *dentry) { - if (nd->flags & LOOKUP_RCU) - return nameidata_dentry_drop_rcu(nd, dentry); + if (nd->flags & LOOKUP_RCU) { + if (unlikely(nameidata_dentry_drop_rcu(nd, dentry))) { + nd->flags &= ~LOOKUP_RCU; + if (!(nd->flags & LOOKUP_ROOT)) + nd->root.mnt = NULL; + rcu_read_unlock(); + br_read_unlock(vfsmount_lock); + return -ECHILD; + } + } return 0; } @@ -526,7 +537,8 @@ static int nameidata_drop_rcu_last(struct nameidata *nd) BUG_ON(!(nd->flags & LOOKUP_RCU)); nd->flags &= ~LOOKUP_RCU; - nd->root.mnt = NULL; + if (!(nd->flags & LOOKUP_ROOT)) + nd->root.mnt = NULL; spin_lock(&dentry->d_lock); if (!__d_rcu_to_refcount(dentry, nd->seq)) goto err_unlock; @@ -547,53 +559,31 @@ err_unlock: return -ECHILD; } -/* Try to drop out of rcu-walk mode if we were in it, otherwise do nothing. */ -static inline int nameidata_drop_rcu_last_maybe(struct nameidata *nd) -{ - if (likely(nd->flags & LOOKUP_RCU)) - return nameidata_drop_rcu_last(nd); - return 0; -} - /** * release_open_intent - free up open intent resources * @nd: pointer to nameidata */ void release_open_intent(struct nameidata *nd) { - if (nd->intent.open.file->f_path.dentry == NULL) - put_filp(nd->intent.open.file); - else - fput(nd->intent.open.file); -} + struct file *file = nd->intent.open.file; -/* - * Call d_revalidate and handle filesystems that request rcu-walk - * to be dropped. This may be called and return in rcu-walk mode, - * regardless of success or error. If -ECHILD is returned, the caller - * must return -ECHILD back up the path walk stack so path walk may - * be restarted in ref-walk mode. - */ -static int d_revalidate(struct dentry *dentry, struct nameidata *nd) -{ - int status; - - status = dentry->d_op->d_revalidate(dentry, nd); - if (status == -ECHILD) { - if (nameidata_dentry_drop_rcu(nd, dentry)) - return status; - status = dentry->d_op->d_revalidate(dentry, nd); + if (file && !IS_ERR(file)) { + if (file->f_path.dentry == NULL) + put_filp(file); + else + fput(file); } +} - return status; +static inline int d_revalidate(struct dentry *dentry, struct nameidata *nd) +{ + return dentry->d_op->d_revalidate(dentry, nd); } -static inline struct dentry * +static struct dentry * do_revalidate(struct dentry *dentry, struct nameidata *nd) { - int status; - - status = d_revalidate(dentry, nd); + int status = d_revalidate(dentry, nd); if (unlikely(status <= 0)) { /* * The dentry failed validation. @@ -602,37 +592,18 @@ do_revalidate(struct dentry *dentry, struct nameidata *nd) * to return a fail status. */ if (status < 0) { - /* If we're in rcu-walk, we don't have a ref */ - if (!(nd->flags & LOOKUP_RCU)) - dput(dentry); + dput(dentry); dentry = ERR_PTR(status); - - } else { - /* Don't d_invalidate in rcu-walk mode */ - if (nameidata_dentry_drop_rcu_maybe(nd, dentry)) - return ERR_PTR(-ECHILD); - if (!d_invalidate(dentry)) { - dput(dentry); - dentry = NULL; - } + } else if (!d_invalidate(dentry)) { + dput(dentry); + dentry = NULL; } } return dentry; } -static inline int need_reval_dot(struct dentry *dentry) -{ - if (likely(!(dentry->d_flags & DCACHE_OP_REVALIDATE))) - return 0; - - if (likely(!(dentry->d_sb->s_type->fs_flags & FS_REVAL_DOT))) - return 0; - - return 1; -} - /* - * force_reval_path - force revalidation of a dentry + * handle_reval_path - force revalidation of a dentry * * In some situations the path walking code will trust dentries without * revalidating them. This causes problems for filesystems that depend on @@ -646,30 +617,28 @@ static inline int need_reval_dot(struct dentry *dentry) * invalidate the dentry. It's up to the caller to handle putting references * to the path if necessary. */ -static int -force_reval_path(struct path *path, struct nameidata *nd) +static inline int handle_reval_path(struct nameidata *nd) { + struct dentry *dentry = nd->path.dentry; int status; - struct dentry *dentry = path->dentry; - /* - * only check on filesystems where it's possible for the dentry to - * become stale. - */ - if (!need_reval_dot(dentry)) + if (likely(!(nd->flags & LOOKUP_JUMPED))) + return 0; + + if (likely(!(dentry->d_flags & DCACHE_OP_REVALIDATE))) + return 0; + + if (likely(!(dentry->d_sb->s_type->fs_flags & FS_REVAL_DOT))) return 0; + /* Note: we do not d_invalidate() */ status = d_revalidate(dentry, nd); if (status > 0) return 0; - if (!status) { - /* Don't d_invalidate in rcu-walk mode */ - if (nameidata_drop_rcu(nd)) - return -ECHILD; - d_invalidate(dentry); + if (!status) status = -ESTALE; - } + return status; } @@ -738,6 +707,7 @@ static __always_inline int __vfs_follow_link(struct nameidata *nd, const char *l path_put(&nd->path); nd->path = nd->root; path_get(&nd->root); + nd->flags |= LOOKUP_JUMPED; } nd->inode = nd->path.dentry->d_inode; @@ -767,18 +737,43 @@ static inline void path_to_nameidata(const struct path *path, nd->path.dentry = path->dentry; } +static inline void put_link(struct nameidata *nd, struct path *link, void *cookie) +{ + struct inode *inode = link->dentry->d_inode; + if (!IS_ERR(cookie) && inode->i_op->put_link) + inode->i_op->put_link(link->dentry, nd, cookie); + path_put(link); +} + static __always_inline int -__do_follow_link(const struct path *link, struct nameidata *nd, void **p) +follow_link(struct path *link, struct nameidata *nd, void **p) { int error; struct dentry *dentry = link->dentry; - touch_atime(link->mnt, dentry); - nd_set_link(nd, NULL); + BUG_ON(nd->flags & LOOKUP_RCU); if (link->mnt == nd->path.mnt) mntget(link->mnt); + if (unlikely(current->total_link_count >= 40)) { + *p = ERR_PTR(-ELOOP); /* no ->put_link(), please */ + path_put(&nd->path); + return -ELOOP; + } + cond_resched(); + current->total_link_count++; + + touch_atime(link->mnt, dentry); + nd_set_link(nd, NULL); + + error = security_inode_follow_link(link->dentry, nd); + if (error) { + *p = ERR_PTR(error); /* no ->put_link(), please */ + path_put(&nd->path); + return error; + } + nd->last_type = LAST_BIND; *p = dentry->d_inode->i_op->follow_link(dentry, nd); error = PTR_ERR(*p); @@ -788,50 +783,18 @@ __do_follow_link(const struct path *link, struct nameidata *nd, void **p) if (s) error = __vfs_follow_link(nd, s); else if (nd->last_type == LAST_BIND) { - error = force_reval_path(&nd->path, nd); - if (error) + nd->flags |= LOOKUP_JUMPED; + nd->inode = nd->path.dentry->d_inode; + if (nd->inode->i_op->follow_link) { + /* stepped on a _really_ weird one */ path_put(&nd->path); + error = -ELOOP; + } } } return error; } -/* - * This limits recursive symlink follows to 8, while - * limiting consecutive symlinks to 40. - * - * Without that kind of total limit, nasty chains of consecutive - * symlinks can cause almost arbitrarily long lookups. - */ -static inline int do_follow_link(struct path *path, struct nameidata *nd) -{ - void *cookie; - int err = -ELOOP; - if (current->link_count >= MAX_NESTED_LINKS) - goto loop; - if (current->total_link_count >= 40) - goto loop; - BUG_ON(nd->depth >= MAX_NESTED_LINKS); - cond_resched(); - err = security_inode_follow_link(path->dentry, nd); - if (err) - goto loop; - current->link_count++; - current->total_link_count++; - nd->depth++; - err = __do_follow_link(path, nd, &cookie); - if (!IS_ERR(cookie) && path->dentry->d_inode->i_op->put_link) - path->dentry->d_inode->i_op->put_link(path->dentry, nd, cookie); - path_put(path); - current->link_count--; - nd->depth--; - return err; -loop: - path_put_conditional(path, nd); - path_put(&nd->path); - return err; -} - static int follow_up_rcu(struct path *path) { struct vfsmount *parent; @@ -970,8 +933,7 @@ static int follow_managed(struct path *path, unsigned flags) if (managed & DCACHE_MANAGE_TRANSIT) { BUG_ON(!path->dentry->d_op); BUG_ON(!path->dentry->d_op->d_manage); - ret = path->dentry->d_op->d_manage(path->dentry, - false, false); + ret = path->dentry->d_op->d_manage(path->dentry, false); if (ret < 0) return ret == -EISDIR ? 0 : ret; } @@ -1036,7 +998,7 @@ static bool __follow_mount_rcu(struct nameidata *nd, struct path *path, struct vfsmount *mounted; if (unlikely(path->dentry->d_flags & DCACHE_MANAGE_TRANSIT) && !reverse_transit && - path->dentry->d_op->d_manage(path->dentry, false, true) < 0) + path->dentry->d_op->d_manage(path->dentry, true) < 0) return false; mounted = __lookup_mnt(path->mnt, path->dentry, 1); if (!mounted) @@ -1070,7 +1032,7 @@ static int follow_dotdot_rcu(struct nameidata *nd) seq = read_seqcount_begin(&parent->d_seq); if (read_seqcount_retry(&old->d_seq, nd->seq)) - return -ECHILD; + goto failed; inode = parent->d_inode; nd->path.dentry = parent; nd->seq = seq; @@ -1083,8 +1045,15 @@ static int follow_dotdot_rcu(struct nameidata *nd) } __follow_mount_rcu(nd, &nd->path, &inode, true); nd->inode = inode; - return 0; + +failed: + nd->flags &= ~LOOKUP_RCU; + if (!(nd->flags & LOOKUP_ROOT)) + nd->root.mnt = NULL; + rcu_read_unlock(); + br_read_unlock(vfsmount_lock); + return -ECHILD; } /* @@ -1095,7 +1064,7 @@ static int follow_dotdot_rcu(struct nameidata *nd) * Care must be taken as namespace_sem may be held (indicated by mounting_here * being true). */ -int follow_down(struct path *path, bool mounting_here) +int follow_down(struct path *path) { unsigned managed; int ret; @@ -1116,7 +1085,7 @@ int follow_down(struct path *path, bool mounting_here) BUG_ON(!path->dentry->d_op); BUG_ON(!path->dentry->d_op->d_manage); ret = path->dentry->d_op->d_manage( - path->dentry, mounting_here, false); + path->dentry, false); if (ret < 0) return ret == -EISDIR ? 0 : ret; } @@ -1218,57 +1187,85 @@ static int do_lookup(struct nameidata *nd, struct qstr *name, { struct vfsmount *mnt = nd->path.mnt; struct dentry *dentry, *parent = nd->path.dentry; - struct inode *dir; + int need_reval = 1; + int status = 1; int err; /* - * See if the low-level filesystem might want - * to use its own hash.. - */ - if (unlikely(parent->d_flags & DCACHE_OP_HASH)) { - err = parent->d_op->d_hash(parent, nd->inode, name); - if (err < 0) - return err; - } - - /* * Rename seqlock is not required here because in the off chance * of a false negative due to a concurrent rename, we're going to * do the non-racy lookup, below. */ if (nd->flags & LOOKUP_RCU) { unsigned seq; - *inode = nd->inode; dentry = __d_lookup_rcu(parent, name, &seq, inode); - if (!dentry) { - if (nameidata_drop_rcu(nd)) - return -ECHILD; - goto need_lookup; - } + if (!dentry) + goto unlazy; + /* Memory barrier in read_seqcount_begin of child is enough */ if (__read_seqcount_retry(&parent->d_seq, nd->seq)) return -ECHILD; - nd->seq = seq; - if (dentry->d_flags & DCACHE_OP_REVALIDATE) - goto need_revalidate; -done2: + + if (unlikely(dentry->d_flags & DCACHE_OP_REVALIDATE)) { + status = d_revalidate(dentry, nd); + if (unlikely(status <= 0)) { + if (status != -ECHILD) + need_reval = 0; + goto unlazy; + } + } path->mnt = mnt; path->dentry = dentry; if (likely(__follow_mount_rcu(nd, path, inode, false))) return 0; - if (nameidata_drop_rcu(nd)) - return -ECHILD; - /* fallthru */ +unlazy: + if (dentry) { + if (nameidata_dentry_drop_rcu(nd, dentry)) + return -ECHILD; + } else { + if (nameidata_drop_rcu(nd)) + return -ECHILD; + } + } else { + dentry = __d_lookup(parent, name); } - dentry = __d_lookup(parent, name); - if (!dentry) - goto need_lookup; -found: - if (dentry->d_flags & DCACHE_OP_REVALIDATE) - goto need_revalidate; -done: + +retry: + if (unlikely(!dentry)) { + struct inode *dir = parent->d_inode; + BUG_ON(nd->inode != dir); + + mutex_lock(&dir->i_mutex); + dentry = d_lookup(parent, name); + if (likely(!dentry)) { + dentry = d_alloc_and_lookup(parent, name, nd); + if (IS_ERR(dentry)) { + mutex_unlock(&dir->i_mutex); + return PTR_ERR(dentry); + } + /* known good */ + need_reval = 0; + status = 1; + } + mutex_unlock(&dir->i_mutex); + } + if (unlikely(dentry->d_flags & DCACHE_OP_REVALIDATE) && need_reval) + status = d_revalidate(dentry, nd); + if (unlikely(status <= 0)) { + if (status < 0) { + dput(dentry); + return status; + } + if (!d_invalidate(dentry)) { + dput(dentry); + dentry = NULL; + need_reval = 1; + goto retry; + } + } + path->mnt = mnt; path->dentry = dentry; err = follow_managed(path, nd->flags); @@ -1278,49 +1275,113 @@ done: } *inode = path->dentry->d_inode; return 0; +} -need_lookup: - dir = parent->d_inode; - BUG_ON(nd->inode != dir); +static inline int may_lookup(struct nameidata *nd) +{ + if (nd->flags & LOOKUP_RCU) { + int err = exec_permission(nd->inode, IPERM_FLAG_RCU); + if (err != -ECHILD) + return err; + if (nameidata_drop_rcu(nd)) + return -ECHILD; + } + return exec_permission(nd->inode, 0); +} - mutex_lock(&dir->i_mutex); - /* - * First re-do the cached lookup just in case it was created - * while we waited for the directory semaphore, or the first - * lookup failed due to an unrelated rename. - * - * This could use version numbering or similar to avoid unnecessary - * cache lookups, but then we'd have to do the first lookup in the - * non-racy way. However in the common case here, everything should - * be hot in cache, so would it be a big win? - */ - dentry = d_lookup(parent, name); - if (likely(!dentry)) { - dentry = d_alloc_and_lookup(parent, name, nd); - mutex_unlock(&dir->i_mutex); - if (IS_ERR(dentry)) - goto fail; - goto done; +static inline int handle_dots(struct nameidata *nd, int type) +{ + if (type == LAST_DOTDOT) { + if (nd->flags & LOOKUP_RCU) { + if (follow_dotdot_rcu(nd)) + return -ECHILD; + } else + follow_dotdot(nd); + } + return 0; +} + +static void terminate_walk(struct nameidata *nd) +{ + if (!(nd->flags & LOOKUP_RCU)) { + path_put(&nd->path); + } else { + nd->flags &= ~LOOKUP_RCU; + if (!(nd->flags & LOOKUP_ROOT)) + nd->root.mnt = NULL; + rcu_read_unlock(); + br_read_unlock(vfsmount_lock); } +} + +static inline int walk_component(struct nameidata *nd, struct path *path, + struct qstr *name, int type, int follow) +{ + struct inode *inode; + int err; /* - * Uhhuh! Nasty case: the cache was re-populated while - * we waited on the semaphore. Need to revalidate. + * "." and ".." are special - ".." especially so because it has + * to be able to know about the current root directory and + * parent relationships. */ - mutex_unlock(&dir->i_mutex); - goto found; + if (unlikely(type != LAST_NORM)) + return handle_dots(nd, type); + err = do_lookup(nd, name, path, &inode); + if (unlikely(err)) { + terminate_walk(nd); + return err; + } + if (!inode) { + path_to_nameidata(path, nd); + terminate_walk(nd); + return -ENOENT; + } + if (unlikely(inode->i_op->follow_link) && follow) { + if (nameidata_dentry_drop_rcu_maybe(nd, path->dentry)) + return -ECHILD; + BUG_ON(inode != path->dentry->d_inode); + return 1; + } + path_to_nameidata(path, nd); + nd->inode = inode; + return 0; +} -need_revalidate: - dentry = do_revalidate(dentry, nd); - if (!dentry) - goto need_lookup; - if (IS_ERR(dentry)) - goto fail; - if (nd->flags & LOOKUP_RCU) - goto done2; - goto done; +/* + * This limits recursive symlink follows to 8, while + * limiting consecutive symlinks to 40. + * + * Without that kind of total limit, nasty chains of consecutive + * symlinks can cause almost arbitrarily long lookups. + */ +static inline int nested_symlink(struct path *path, struct nameidata *nd) +{ + int res; -fail: - return PTR_ERR(dentry); + BUG_ON(nd->depth >= MAX_NESTED_LINKS); + if (unlikely(current->link_count >= MAX_NESTED_LINKS)) { + path_put_conditional(path, nd); + path_put(&nd->path); + return -ELOOP; + } + + nd->depth++; + current->link_count++; + + do { + struct path link = *path; + void *cookie; + + res = follow_link(&link, nd, &cookie); + if (!res) + res = walk_component(nd, path, &nd->last, + nd->last_type, LOOKUP_FOLLOW); + put_link(nd, &link, cookie); + } while (res > 0); + + current->link_count--; + nd->depth--; + return res; } /* @@ -1340,30 +1401,18 @@ static int link_path_walk(const char *name, struct nameidata *nd) while (*name=='/') name++; if (!*name) - goto return_reval; - - if (nd->depth) - lookup_flags = LOOKUP_FOLLOW | (nd->flags & LOOKUP_CONTINUE); + return 0; /* At this point we know we have a real path component. */ for(;;) { - struct inode *inode; unsigned long hash; struct qstr this; unsigned int c; + int type; nd->flags |= LOOKUP_CONTINUE; - if (nd->flags & LOOKUP_RCU) { - err = exec_permission(nd->inode, IPERM_FLAG_RCU); - if (err == -ECHILD) { - if (nameidata_drop_rcu(nd)) - return -ECHILD; - goto exec_again; - } - } else { -exec_again: - err = exec_permission(nd->inode, 0); - } + + err = may_lookup(nd); if (err) break; @@ -1379,56 +1428,43 @@ exec_again: this.len = name - (const char *) this.name; this.hash = end_name_hash(hash); + type = LAST_NORM; + if (this.name[0] == '.') switch (this.len) { + case 2: + if (this.name[1] == '.') { + type = LAST_DOTDOT; + nd->flags |= LOOKUP_JUMPED; + } + break; + case 1: + type = LAST_DOT; + } + if (likely(type == LAST_NORM)) { + struct dentry *parent = nd->path.dentry; + nd->flags &= ~LOOKUP_JUMPED; + if (unlikely(parent->d_flags & DCACHE_OP_HASH)) { + err = parent->d_op->d_hash(parent, nd->inode, + &this); + if (err < 0) + break; + } + } + /* remove trailing slashes? */ if (!c) goto last_component; while (*++name == '/'); if (!*name) - goto last_with_slashes; + goto last_component; - /* - * "." and ".." are special - ".." especially so because it has - * to be able to know about the current root directory and - * parent relationships. - */ - if (this.name[0] == '.') switch (this.len) { - default: - break; - case 2: - if (this.name[1] != '.') - break; - if (nd->flags & LOOKUP_RCU) { - if (follow_dotdot_rcu(nd)) - return -ECHILD; - } else - follow_dotdot(nd); - /* fallthrough */ - case 1: - continue; - } - /* This does the actual lookups.. */ - err = do_lookup(nd, &this, &next, &inode); - if (err) - break; - err = -ENOENT; - if (!inode) - goto out_dput; + err = walk_component(nd, &next, &this, type, LOOKUP_FOLLOW); + if (err < 0) + return err; - if (inode->i_op->follow_link) { - /* We commonly drop rcu-walk here */ - if (nameidata_dentry_drop_rcu_maybe(nd, next.dentry)) - return -ECHILD; - BUG_ON(inode != next.dentry->d_inode); - err = do_follow_link(&next, nd); + if (err) { + err = nested_symlink(&next, nd); if (err) - goto return_err; - nd->inode = nd->path.dentry->d_inode; - err = -ENOENT; - if (!nd->inode) - break; - } else { - path_to_nameidata(&next, nd); - nd->inode = inode; + return err; } err = -ENOTDIR; if (!nd->inode->i_op->lookup) @@ -1436,209 +1472,109 @@ exec_again: continue; /* here ends the main loop */ -last_with_slashes: - lookup_flags |= LOOKUP_FOLLOW | LOOKUP_DIRECTORY; last_component: /* Clear LOOKUP_CONTINUE iff it was previously unset */ nd->flags &= lookup_flags | ~LOOKUP_CONTINUE; - if (lookup_flags & LOOKUP_PARENT) - goto lookup_parent; - if (this.name[0] == '.') switch (this.len) { - default: - break; - case 2: - if (this.name[1] != '.') - break; - if (nd->flags & LOOKUP_RCU) { - if (follow_dotdot_rcu(nd)) - return -ECHILD; - } else - follow_dotdot(nd); - /* fallthrough */ - case 1: - goto return_reval; - } - err = do_lookup(nd, &this, &next, &inode); - if (err) - break; - if (inode && unlikely(inode->i_op->follow_link) && - (lookup_flags & LOOKUP_FOLLOW)) { - if (nameidata_dentry_drop_rcu_maybe(nd, next.dentry)) - return -ECHILD; - BUG_ON(inode != next.dentry->d_inode); - err = do_follow_link(&next, nd); - if (err) - goto return_err; - nd->inode = nd->path.dentry->d_inode; - } else { - path_to_nameidata(&next, nd); - nd->inode = inode; - } - err = -ENOENT; - if (!nd->inode) - break; - if (lookup_flags & LOOKUP_DIRECTORY) { - err = -ENOTDIR; - if (!nd->inode->i_op->lookup) - break; - } - goto return_base; -lookup_parent: nd->last = this; - nd->last_type = LAST_NORM; - if (this.name[0] != '.') - goto return_base; - if (this.len == 1) - nd->last_type = LAST_DOT; - else if (this.len == 2 && this.name[1] == '.') - nd->last_type = LAST_DOTDOT; - else - goto return_base; -return_reval: - /* - * We bypassed the ordinary revalidation routines. - * We may need to check the cached dentry for staleness. - */ - if (need_reval_dot(nd->path.dentry)) { - /* Note: we do not d_invalidate() */ - err = d_revalidate(nd->path.dentry, nd); - if (!err) - err = -ESTALE; - if (err < 0) - break; - } -return_base: - if (nameidata_drop_rcu_last_maybe(nd)) - return -ECHILD; + nd->last_type = type; return 0; -out_dput: - if (!(nd->flags & LOOKUP_RCU)) - path_put_conditional(&next, nd); - break; } - if (!(nd->flags & LOOKUP_RCU)) - path_put(&nd->path); -return_err: + terminate_walk(nd); return err; } -static inline int path_walk_rcu(const char *name, struct nameidata *nd) -{ - current->total_link_count = 0; - - return link_path_walk(name, nd); -} - -static inline int path_walk_simple(const char *name, struct nameidata *nd) -{ - current->total_link_count = 0; - - return link_path_walk(name, nd); -} - -static int path_walk(const char *name, struct nameidata *nd) -{ - struct path save = nd->path; - int result; - - current->total_link_count = 0; - - /* make sure the stuff we saved doesn't go away */ - path_get(&save); - - result = link_path_walk(name, nd); - if (result == -ESTALE) { - /* nd->path had been dropped */ - current->total_link_count = 0; - nd->path = save; - path_get(&nd->path); - nd->flags |= LOOKUP_REVAL; - result = link_path_walk(name, nd); - } - - path_put(&save); - - return result; -} - -static void path_finish_rcu(struct nameidata *nd) -{ - if (nd->flags & LOOKUP_RCU) { - /* RCU dangling. Cancel it. */ - nd->flags &= ~LOOKUP_RCU; - nd->root.mnt = NULL; - rcu_read_unlock(); - br_read_unlock(vfsmount_lock); - } - if (nd->file) - fput(nd->file); -} - -static int path_init_rcu(int dfd, const char *name, unsigned int flags, struct nameidata *nd) +static int path_init(int dfd, const char *name, unsigned int flags, + struct nameidata *nd, struct file **fp) { int retval = 0; int fput_needed; struct file *file; nd->last_type = LAST_ROOT; /* if there are only slashes... */ - nd->flags = flags | LOOKUP_RCU; + nd->flags = flags | LOOKUP_JUMPED; nd->depth = 0; + if (flags & LOOKUP_ROOT) { + struct inode *inode = nd->root.dentry->d_inode; + if (*name) { + if (!inode->i_op->lookup) + return -ENOTDIR; + retval = inode_permission(inode, MAY_EXEC); + if (retval) + return retval; + } + nd->path = nd->root; + nd->inode = inode; + if (flags & LOOKUP_RCU) { + br_read_lock(vfsmount_lock); + rcu_read_lock(); + nd->seq = __read_seqcount_begin(&nd->path.dentry->d_seq); + } else { + path_get(&nd->path); + } + return 0; + } + nd->root.mnt = NULL; - nd->file = NULL; if (*name=='/') { - struct fs_struct *fs = current->fs; - unsigned seq; - - br_read_lock(vfsmount_lock); - rcu_read_lock(); - - do { - seq = read_seqcount_begin(&fs->seq); - nd->root = fs->root; - nd->path = nd->root; - nd->seq = __read_seqcount_begin(&nd->path.dentry->d_seq); - } while (read_seqcount_retry(&fs->seq, seq)); - + if (flags & LOOKUP_RCU) { + br_read_lock(vfsmount_lock); + rcu_read_lock(); + set_root_rcu(nd); + } else { + set_root(nd); + path_get(&nd->root); + } + nd->path = nd->root; } else if (dfd == AT_FDCWD) { - struct fs_struct *fs = current->fs; - unsigned seq; - - br_read_lock(vfsmount_lock); - rcu_read_lock(); + if (flags & LOOKUP_RCU) { + struct fs_struct *fs = current->fs; + unsigned seq; - do { - seq = read_seqcount_begin(&fs->seq); - nd->path = fs->pwd; - nd->seq = __read_seqcount_begin(&nd->path.dentry->d_seq); - } while (read_seqcount_retry(&fs->seq, seq)); + br_read_lock(vfsmount_lock); + rcu_read_lock(); + do { + seq = read_seqcount_begin(&fs->seq); + nd->path = fs->pwd; + nd->seq = __read_seqcount_begin(&nd->path.dentry->d_seq); + } while (read_seqcount_retry(&fs->seq, seq)); + } else { + get_fs_pwd(current->fs, &nd->path); + } } else { struct dentry *dentry; - file = fget_light(dfd, &fput_needed); + file = fget_raw_light(dfd, &fput_needed); retval = -EBADF; if (!file) goto out_fail; dentry = file->f_path.dentry; - retval = -ENOTDIR; - if (!S_ISDIR(dentry->d_inode->i_mode)) - goto fput_fail; + if (*name) { + retval = -ENOTDIR; + if (!S_ISDIR(dentry->d_inode->i_mode)) + goto fput_fail; - retval = file_permission(file, MAY_EXEC); - if (retval) - goto fput_fail; + retval = file_permission(file, MAY_EXEC); + if (retval) + goto fput_fail; + } nd->path = file->f_path; - if (fput_needed) - nd->file = file; - - nd->seq = __read_seqcount_begin(&nd->path.dentry->d_seq); - br_read_lock(vfsmount_lock); - rcu_read_lock(); + if (flags & LOOKUP_RCU) { + if (fput_needed) + *fp = file; + nd->seq = __read_seqcount_begin(&nd->path.dentry->d_seq); + br_read_lock(vfsmount_lock); + rcu_read_lock(); + } else { + path_get(&file->f_path); + fput_light(file, fput_needed); + } } + nd->inode = nd->path.dentry->d_inode; return 0; @@ -1648,60 +1584,23 @@ out_fail: return retval; } -static int path_init(int dfd, const char *name, unsigned int flags, struct nameidata *nd) +static inline int lookup_last(struct nameidata *nd, struct path *path) { - int retval = 0; - int fput_needed; - struct file *file; - - nd->last_type = LAST_ROOT; /* if there are only slashes... */ - nd->flags = flags; - nd->depth = 0; - nd->root.mnt = NULL; + if (nd->last_type == LAST_NORM && nd->last.name[nd->last.len]) + nd->flags |= LOOKUP_FOLLOW | LOOKUP_DIRECTORY; - if (*name=='/') { - set_root(nd); - nd->path = nd->root; - path_get(&nd->root); - } else if (dfd == AT_FDCWD) { - get_fs_pwd(current->fs, &nd->path); - } else { - struct dentry *dentry; - - file = fget_light(dfd, &fput_needed); - retval = -EBADF; - if (!file) - goto out_fail; - - dentry = file->f_path.dentry; - - retval = -ENOTDIR; - if (!S_ISDIR(dentry->d_inode->i_mode)) - goto fput_fail; - - retval = file_permission(file, MAY_EXEC); - if (retval) - goto fput_fail; - - nd->path = file->f_path; - path_get(&file->f_path); - - fput_light(file, fput_needed); - } - nd->inode = nd->path.dentry->d_inode; - return 0; - -fput_fail: - fput_light(file, fput_needed); -out_fail: - return retval; + nd->flags &= ~LOOKUP_PARENT; + return walk_component(nd, path, &nd->last, nd->last_type, + nd->flags & LOOKUP_FOLLOW); } /* Returns 0 and nd will be valid on success; Retuns error, otherwise. */ -static int do_path_lookup(int dfd, const char *name, +static int path_lookupat(int dfd, const char *name, unsigned int flags, struct nameidata *nd) { - int retval; + struct file *base = NULL; + struct path path; + int err; /* * Path walking is largely split up into 2 different synchronisation @@ -1717,44 +1616,75 @@ static int do_path_lookup(int dfd, const char *name, * be handled by restarting a traditional ref-walk (which will always * be able to complete). */ - retval = path_init_rcu(dfd, name, flags, nd); - if (unlikely(retval)) - return retval; - retval = path_walk_rcu(name, nd); - path_finish_rcu(nd); - if (nd->root.mnt) { - path_put(&nd->root); - nd->root.mnt = NULL; + err = path_init(dfd, name, flags | LOOKUP_PARENT, nd, &base); + + if (unlikely(err)) + return err; + + current->total_link_count = 0; + err = link_path_walk(name, nd); + + if (!err && !(flags & LOOKUP_PARENT)) { + err = lookup_last(nd, &path); + while (err > 0) { + void *cookie; + struct path link = path; + nd->flags |= LOOKUP_PARENT; + err = follow_link(&link, nd, &cookie); + if (!err) + err = lookup_last(nd, &path); + put_link(nd, &link, cookie); + } } - if (unlikely(retval == -ECHILD || retval == -ESTALE)) { - /* slower, locked walk */ - if (retval == -ESTALE) - flags |= LOOKUP_REVAL; - retval = path_init(dfd, name, flags, nd); - if (unlikely(retval)) - return retval; - retval = path_walk(name, nd); - if (nd->root.mnt) { - path_put(&nd->root); - nd->root.mnt = NULL; + if (nd->flags & LOOKUP_RCU) { + /* went all way through without dropping RCU */ + BUG_ON(err); + if (nameidata_drop_rcu_last(nd)) + err = -ECHILD; + } + + if (!err) + err = handle_reval_path(nd); + + if (!err && nd->flags & LOOKUP_DIRECTORY) { + if (!nd->inode->i_op->lookup) { + path_put(&nd->path); + return -ENOTDIR; } } + if (base) + fput(base); + + if (nd->root.mnt && !(nd->flags & LOOKUP_ROOT)) { + path_put(&nd->root); + nd->root.mnt = NULL; + } + return err; +} + +static int do_path_lookup(int dfd, const char *name, + unsigned int flags, struct nameidata *nd) +{ + int retval = path_lookupat(dfd, name, flags | LOOKUP_RCU, nd); + if (unlikely(retval == -ECHILD)) + retval = path_lookupat(dfd, name, flags, nd); + if (unlikely(retval == -ESTALE)) + retval = path_lookupat(dfd, name, flags | LOOKUP_REVAL, nd); + if (likely(!retval)) { if (unlikely(!audit_dummy_context())) { if (nd->path.dentry && nd->inode) audit_inode(name, nd->path.dentry); } } - return retval; } -int path_lookup(const char *name, unsigned int flags, - struct nameidata *nd) +int kern_path_parent(const char *name, struct nameidata *nd) { - return do_path_lookup(AT_FDCWD, name, flags, nd); + return do_path_lookup(AT_FDCWD, name, LOOKUP_PARENT, nd); } int kern_path(const char *name, unsigned int flags, struct path *path) @@ -1778,29 +1708,10 @@ int vfs_path_lookup(struct dentry *dentry, struct vfsmount *mnt, const char *name, unsigned int flags, struct nameidata *nd) { - int retval; - - /* same as do_path_lookup */ - nd->last_type = LAST_ROOT; - nd->flags = flags; - nd->depth = 0; - - nd->path.dentry = dentry; - nd->path.mnt = mnt; - path_get(&nd->path); - nd->root = nd->path; - path_get(&nd->root); - nd->inode = nd->path.dentry->d_inode; - - retval = path_walk(name, nd); - if (unlikely(!retval && !audit_dummy_context() && nd->path.dentry && - nd->inode)) - audit_inode(name, nd->path.dentry); - - path_put(&nd->root); - nd->root.mnt = NULL; - - return retval; + nd->root.dentry = dentry; + nd->root.mnt = mnt; + /* the first argument of do_path_lookup() is ignored with LOOKUP_ROOT */ + return do_path_lookup(AT_FDCWD, name, flags | LOOKUP_ROOT, nd); } static struct dentry *__lookup_hash(struct qstr *name, @@ -1815,17 +1726,6 @@ static struct dentry *__lookup_hash(struct qstr *name, return ERR_PTR(err); /* - * See if the low-level filesystem might want - * to use its own hash.. - */ - if (base->d_flags & DCACHE_OP_HASH) { - err = base->d_op->d_hash(base, inode, name); - dentry = ERR_PTR(err); - if (err < 0) - goto out; - } - - /* * Don't bother with __d_lookup: callers are for creat as * well as unlink, so a lot of the time it would cost * a double lookup. @@ -1837,7 +1737,7 @@ static struct dentry *__lookup_hash(struct qstr *name, if (!dentry) dentry = d_alloc_and_lookup(base, name, nd); -out: + return dentry; } @@ -1851,28 +1751,6 @@ static struct dentry *lookup_hash(struct nameidata *nd) return __lookup_hash(&nd->last, nd->path.dentry, nd); } -static int __lookup_one_len(const char *name, struct qstr *this, - struct dentry *base, int len) -{ - unsigned long hash; - unsigned int c; - - this->name = name; - this->len = len; - if (!len) - return -EACCES; - - hash = init_name_hash(); - while (len--) { - c = *(const unsigned char *)name++; - if (c == '/' || c == '\0') - return -EACCES; - hash = partial_name_hash(c, hash); - } - this->hash = end_name_hash(hash); - return 0; -} - /** * lookup_one_len - filesystem helper to lookup single pathname component * @name: pathname component to lookup @@ -1886,14 +1764,34 @@ static int __lookup_one_len(const char *name, struct qstr *this, */ struct dentry *lookup_one_len(const char *name, struct dentry *base, int len) { - int err; struct qstr this; + unsigned long hash; + unsigned int c; WARN_ON_ONCE(!mutex_is_locked(&base->d_inode->i_mutex)); - err = __lookup_one_len(name, &this, base, len); - if (err) - return ERR_PTR(err); + this.name = name; + this.len = len; + if (!len) + return ERR_PTR(-EACCES); + + hash = init_name_hash(); + while (len--) { + c = *(const unsigned char *)name++; + if (c == '/' || c == '\0') + return ERR_PTR(-EACCES); + hash = partial_name_hash(c, hash); + } + this.hash = end_name_hash(hash); + /* + * See if the low-level filesystem might want + * to use its own hash.. + */ + if (base->d_flags & DCACHE_OP_HASH) { + int err = base->d_op->d_hash(base, base->d_inode, &this); + if (err < 0) + return ERR_PTR(err); + } return __lookup_hash(&this, base, NULL); } @@ -1902,7 +1800,7 @@ int user_path_at(int dfd, const char __user *name, unsigned flags, struct path *path) { struct nameidata nd; - char *tmp = getname(name); + char *tmp = getname_flags(name, flags); int err = PTR_ERR(tmp); if (!IS_ERR(tmp)) { @@ -2082,12 +1980,16 @@ int vfs_create(struct inode *dir, struct dentry *dentry, int mode, return error; } -int may_open(struct path *path, int acc_mode, int flag) +static int may_open(struct path *path, int acc_mode, int flag) { struct dentry *dentry = path->dentry; struct inode *inode = dentry->d_inode; int error; + /* O_PATH? */ + if (!acc_mode) + return 0; + if (!inode) return -ENOENT; @@ -2156,34 +2058,6 @@ static int handle_truncate(struct file *filp) } /* - * Be careful about ever adding any more callers of this - * function. Its flags must be in the namei format, not - * what get passed to sys_open(). - */ -static int __open_namei_create(struct nameidata *nd, struct path *path, - int open_flag, int mode) -{ - int error; - struct dentry *dir = nd->path.dentry; - - if (!IS_POSIXACL(dir->d_inode)) - mode &= ~current_umask(); - error = security_path_mknod(&nd->path, path->dentry, mode, 0); - if (error) - goto out_unlock; - error = vfs_create(dir->d_inode, path->dentry, mode, nd); -out_unlock: - mutex_unlock(&dir->d_inode->i_mutex); - dput(nd->path.dentry); - nd->path.dentry = path->dentry; - - if (error) - return error; - /* Don't check for write permission, don't truncate */ - return may_open(&nd->path, 0, open_flag & ~O_TRUNC); -} - -/* * Note that while the flag value (low two bits) for sys_open means: * 00 - read-only * 01 - write-only @@ -2207,128 +2081,115 @@ static inline int open_to_namei_flags(int flag) return flag; } -static int open_will_truncate(int flag, struct inode *inode) -{ - /* - * We'll never write to the fs underlying - * a device file. - */ - if (special_file(inode->i_mode)) - return 0; - return (flag & O_TRUNC); -} - -static struct file *finish_open(struct nameidata *nd, - int open_flag, int acc_mode) -{ - struct file *filp; - int will_truncate; - int error; - - will_truncate = open_will_truncate(open_flag, nd->path.dentry->d_inode); - if (will_truncate) { - error = mnt_want_write(nd->path.mnt); - if (error) - goto exit; - } - error = may_open(&nd->path, acc_mode, open_flag); - if (error) { - if (will_truncate) - mnt_drop_write(nd->path.mnt); - goto exit; - } - filp = nameidata_to_filp(nd); - if (!IS_ERR(filp)) { - error = ima_file_check(filp, acc_mode); - if (error) { - fput(filp); - filp = ERR_PTR(error); - } - } - if (!IS_ERR(filp)) { - if (will_truncate) { - error = handle_truncate(filp); - if (error) { - fput(filp); - filp = ERR_PTR(error); - } - } - } - /* - * It is now safe to drop the mnt write - * because the filp has had a write taken - * on its behalf. - */ - if (will_truncate) - mnt_drop_write(nd->path.mnt); - path_put(&nd->path); - return filp; - -exit: - if (!IS_ERR(nd->intent.open.file)) - release_open_intent(nd); - path_put(&nd->path); - return ERR_PTR(error); -} - /* - * Handle O_CREAT case for do_filp_open + * Handle the last step of open() */ static struct file *do_last(struct nameidata *nd, struct path *path, - int open_flag, int acc_mode, - int mode, const char *pathname) + const struct open_flags *op, const char *pathname) { struct dentry *dir = nd->path.dentry; + struct dentry *dentry; + int open_flag = op->open_flag; + int will_truncate = open_flag & O_TRUNC; + int want_write = 0; + int acc_mode = op->acc_mode; struct file *filp; - int error = -EISDIR; + int error; + + nd->flags &= ~LOOKUP_PARENT; + nd->flags |= op->intent; switch (nd->last_type) { case LAST_DOTDOT: - follow_dotdot(nd); - dir = nd->path.dentry; case LAST_DOT: - if (need_reval_dot(dir)) { - int status = d_revalidate(nd->path.dentry, nd); - if (!status) - status = -ESTALE; - if (status < 0) { - error = status; - goto exit; - } - } + error = handle_dots(nd, nd->last_type); + if (error) + return ERR_PTR(error); /* fallthrough */ case LAST_ROOT: - goto exit; + if (nd->flags & LOOKUP_RCU) { + if (nameidata_drop_rcu_last(nd)) + return ERR_PTR(-ECHILD); + } + error = handle_reval_path(nd); + if (error) + goto exit; + audit_inode(pathname, nd->path.dentry); + if (open_flag & O_CREAT) { + error = -EISDIR; + goto exit; + } + goto ok; case LAST_BIND: + /* can't be RCU mode here */ + error = handle_reval_path(nd); + if (error) + goto exit; audit_inode(pathname, dir); goto ok; } + if (!(open_flag & O_CREAT)) { + int symlink_ok = 0; + if (nd->last.name[nd->last.len]) + nd->flags |= LOOKUP_FOLLOW | LOOKUP_DIRECTORY; + if (open_flag & O_PATH && !(nd->flags & LOOKUP_FOLLOW)) + symlink_ok = 1; + /* we _can_ be in RCU mode here */ + error = walk_component(nd, path, &nd->last, LAST_NORM, + !symlink_ok); + if (error < 0) + return ERR_PTR(error); + if (error) /* symlink */ + return NULL; + /* sayonara */ + if (nd->flags & LOOKUP_RCU) { + if (nameidata_drop_rcu_last(nd)) + return ERR_PTR(-ECHILD); + } + + error = -ENOTDIR; + if (nd->flags & LOOKUP_DIRECTORY) { + if (!nd->inode->i_op->lookup) + goto exit; + } + audit_inode(pathname, nd->path.dentry); + goto ok; + } + + /* create side of things */ + + if (nd->flags & LOOKUP_RCU) { + if (nameidata_drop_rcu_last(nd)) + return ERR_PTR(-ECHILD); + } + + audit_inode(pathname, dir); + error = -EISDIR; /* trailing slashes? */ if (nd->last.name[nd->last.len]) goto exit; mutex_lock(&dir->d_inode->i_mutex); - path->dentry = lookup_hash(nd); - path->mnt = nd->path.mnt; - - error = PTR_ERR(path->dentry); - if (IS_ERR(path->dentry)) { + dentry = lookup_hash(nd); + error = PTR_ERR(dentry); + if (IS_ERR(dentry)) { mutex_unlock(&dir->d_inode->i_mutex); goto exit; } - if (IS_ERR(nd->intent.open.file)) { - error = PTR_ERR(nd->intent.open.file); - goto exit_mutex_unlock; - } + path->dentry = dentry; + path->mnt = nd->path.mnt; /* Negative dentry, just create the file */ - if (!path->dentry->d_inode) { + if (!dentry->d_inode) { + int mode = op->mode; + if (!IS_POSIXACL(dir->d_inode)) + mode &= ~current_umask(); /* * This write is needed to ensure that a - * ro->rw transition does not occur between + * rw->ro transition does not occur between * the time when the file is created and when * a permanent write count is taken through * the 'struct file' in nameidata_to_filp(). @@ -2336,22 +2197,21 @@ static struct file *do_last(struct nameidata *nd, struct path *path, error = mnt_want_write(nd->path.mnt); if (error) goto exit_mutex_unlock; - error = __open_namei_create(nd, path, open_flag, mode); - if (error) { - mnt_drop_write(nd->path.mnt); - goto exit; - } - filp = nameidata_to_filp(nd); - mnt_drop_write(nd->path.mnt); - path_put(&nd->path); - if (!IS_ERR(filp)) { - error = ima_file_check(filp, acc_mode); - if (error) { - fput(filp); - filp = ERR_PTR(error); - } - } - return filp; + want_write = 1; + /* Don't check for write permission, don't truncate */ + open_flag &= ~O_TRUNC; + will_truncate = 0; + acc_mode = MAY_OPEN; + error = security_path_mknod(&nd->path, dentry, mode, 0); + if (error) + goto exit_mutex_unlock; + error = vfs_create(dir->d_inode, dentry, mode, nd); + if (error) + goto exit_mutex_unlock; + mutex_unlock(&dir->d_inode->i_mutex); + dput(nd->path.dentry); + nd->path.dentry = dentry; + goto common; } /* @@ -2381,7 +2241,40 @@ static struct file *do_last(struct nameidata *nd, struct path *path, if (S_ISDIR(nd->inode->i_mode)) goto exit; ok: - filp = finish_open(nd, open_flag, acc_mode); + if (!S_ISREG(nd->inode->i_mode)) + will_truncate = 0; + + if (will_truncate) { + error = mnt_want_write(nd->path.mnt); + if (error) + goto exit; + want_write = 1; + } +common: + error = may_open(&nd->path, acc_mode, open_flag); + if (error) + goto exit; + filp = nameidata_to_filp(nd); + if (!IS_ERR(filp)) { + error = ima_file_check(filp, op->acc_mode); + if (error) { + fput(filp); + filp = ERR_PTR(error); + } + } + if (!IS_ERR(filp)) { + if (will_truncate) { + error = handle_truncate(filp); + if (error) { + fput(filp); + filp = ERR_PTR(error); + } + } + } +out: + if (want_write) + mnt_drop_write(nd->path.mnt); + path_put(&nd->path); return filp; exit_mutex_unlock: @@ -2389,199 +2282,103 @@ exit_mutex_unlock: exit_dput: path_put_conditional(path, nd); exit: - if (!IS_ERR(nd->intent.open.file)) - release_open_intent(nd); - path_put(&nd->path); - return ERR_PTR(error); + filp = ERR_PTR(error); + goto out; } -/* - * Note that the low bits of the passed in "open_flag" - * are not the same as in the local variable "flag". See - * open_to_namei_flags() for more details. - */ -struct file *do_filp_open(int dfd, const char *pathname, - int open_flag, int mode, int acc_mode) +static struct file *path_openat(int dfd, const char *pathname, + struct nameidata *nd, const struct open_flags *op, int flags) { + struct file *base = NULL; struct file *filp; - struct nameidata nd; - int error; struct path path; - int count = 0; - int flag = open_to_namei_flags(open_flag); - int flags; - - if (!(open_flag & O_CREAT)) - mode = 0; - - /* Must never be set by userspace */ - open_flag &= ~FMODE_NONOTIFY; - - /* - * O_SYNC is implemented as __O_SYNC|O_DSYNC. As many places only - * check for O_DSYNC if the need any syncing at all we enforce it's - * always set instead of having to deal with possibly weird behaviour - * for malicious applications setting only __O_SYNC. - */ - if (open_flag & __O_SYNC) - open_flag |= O_DSYNC; - - if (!acc_mode) - acc_mode = MAY_OPEN | ACC_MODE(open_flag); - - /* O_TRUNC implies we need access checks for write permissions */ - if (open_flag & O_TRUNC) - acc_mode |= MAY_WRITE; - - /* Allow the LSM permission hook to distinguish append - access from general write access. */ - if (open_flag & O_APPEND) - acc_mode |= MAY_APPEND; - - flags = LOOKUP_OPEN; - if (open_flag & O_CREAT) { - flags |= LOOKUP_CREATE; - if (open_flag & O_EXCL) - flags |= LOOKUP_EXCL; - } - if (open_flag & O_DIRECTORY) - flags |= LOOKUP_DIRECTORY; - if (!(open_flag & O_NOFOLLOW)) - flags |= LOOKUP_FOLLOW; + int error; filp = get_empty_filp(); if (!filp) return ERR_PTR(-ENFILE); - filp->f_flags = open_flag; - nd.intent.open.file = filp; - nd.intent.open.flags = flag; - nd.intent.open.create_mode = mode; + filp->f_flags = op->open_flag; + nd->intent.open.file = filp; + nd->intent.open.flags = open_to_namei_flags(op->open_flag); + nd->intent.open.create_mode = op->mode; - if (open_flag & O_CREAT) - goto creat; - - /* !O_CREAT, simple open */ - error = do_path_lookup(dfd, pathname, flags, &nd); + error = path_init(dfd, pathname, flags | LOOKUP_PARENT, nd, &base); if (unlikely(error)) goto out_filp; - error = -ELOOP; - if (!(nd.flags & LOOKUP_FOLLOW)) { - if (nd.inode->i_op->follow_link) - goto out_path; - } - error = -ENOTDIR; - if (nd.flags & LOOKUP_DIRECTORY) { - if (!nd.inode->i_op->lookup) - goto out_path; - } - audit_inode(pathname, nd.path.dentry); - filp = finish_open(&nd, open_flag, acc_mode); - return filp; -creat: - /* OK, have to create the file. Find the parent. */ - error = path_init_rcu(dfd, pathname, - LOOKUP_PARENT | (flags & LOOKUP_REVAL), &nd); - if (error) - goto out_filp; - error = path_walk_rcu(pathname, &nd); - path_finish_rcu(&nd); - if (unlikely(error == -ECHILD || error == -ESTALE)) { - /* slower, locked walk */ - if (error == -ESTALE) { -reval: - flags |= LOOKUP_REVAL; - } - error = path_init(dfd, pathname, - LOOKUP_PARENT | (flags & LOOKUP_REVAL), &nd); - if (error) - goto out_filp; - - error = path_walk_simple(pathname, &nd); - } + current->total_link_count = 0; + error = link_path_walk(pathname, nd); if (unlikely(error)) goto out_filp; - if (unlikely(!audit_dummy_context())) - audit_inode(pathname, nd.path.dentry); - /* - * We have the parent and last component. - */ - nd.flags = flags; - filp = do_last(&nd, &path, open_flag, acc_mode, mode, pathname); + filp = do_last(nd, &path, op, pathname); while (unlikely(!filp)) { /* trailing symlink */ struct path link = path; - struct inode *linki = link.dentry->d_inode; void *cookie; - error = -ELOOP; - if (!(nd.flags & LOOKUP_FOLLOW)) - goto exit_dput; - if (count++ == 32) - goto exit_dput; - /* - * This is subtle. Instead of calling do_follow_link() we do - * the thing by hands. The reason is that this way we have zero - * link_count and path_walk() (called from ->follow_link) - * honoring LOOKUP_PARENT. After that we have the parent and - * last component, i.e. we are in the same situation as after - * the first path_walk(). Well, almost - if the last component - * is normal we get its copy stored in nd->last.name and we will - * have to putname() it when we are done. Procfs-like symlinks - * just set LAST_BIND. - */ - nd.flags |= LOOKUP_PARENT; - error = security_inode_follow_link(link.dentry, &nd); - if (error) - goto exit_dput; - error = __do_follow_link(&link, &nd, &cookie); - if (unlikely(error)) { - if (!IS_ERR(cookie) && linki->i_op->put_link) - linki->i_op->put_link(link.dentry, &nd, cookie); - /* nd.path had been dropped */ - nd.path = link; - goto out_path; + if (!(nd->flags & LOOKUP_FOLLOW)) { + path_put_conditional(&path, nd); + path_put(&nd->path); + filp = ERR_PTR(-ELOOP); + break; } - nd.flags &= ~LOOKUP_PARENT; - filp = do_last(&nd, &path, open_flag, acc_mode, mode, pathname); - if (linki->i_op->put_link) - linki->i_op->put_link(link.dentry, &nd, cookie); - path_put(&link); + nd->flags |= LOOKUP_PARENT; + nd->flags &= ~(LOOKUP_OPEN|LOOKUP_CREATE|LOOKUP_EXCL); + error = follow_link(&link, nd, &cookie); + if (unlikely(error)) + filp = ERR_PTR(error); + else + filp = do_last(nd, &path, op, pathname); + put_link(nd, &link, cookie); } out: - if (nd.root.mnt) - path_put(&nd.root); - if (filp == ERR_PTR(-ESTALE) && !(flags & LOOKUP_REVAL)) - goto reval; + if (nd->root.mnt && !(nd->flags & LOOKUP_ROOT)) + path_put(&nd->root); + if (base) + fput(base); + release_open_intent(nd); return filp; -exit_dput: - path_put_conditional(&path, &nd); -out_path: - path_put(&nd.path); out_filp: - if (!IS_ERR(nd.intent.open.file)) - release_open_intent(&nd); filp = ERR_PTR(error); goto out; } -/** - * filp_open - open file and return file pointer - * - * @filename: path to open - * @flags: open flags as per the open(2) second argument - * @mode: mode for the new file if O_CREAT is set, else ignored - * - * This is the helper to open a file from kernelspace if you really - * have to. But in generally you should not do this, so please move - * along, nothing to see here.. - */ -struct file *filp_open(const char *filename, int flags, int mode) +struct file *do_filp_open(int dfd, const char *pathname, + const struct open_flags *op, int flags) { - return do_filp_open(AT_FDCWD, filename, flags, mode, 0); + struct nameidata nd; + struct file *filp; + + filp = path_openat(dfd, pathname, &nd, op, flags | LOOKUP_RCU); + if (unlikely(filp == ERR_PTR(-ECHILD))) + filp = path_openat(dfd, pathname, &nd, op, flags); + if (unlikely(filp == ERR_PTR(-ESTALE))) + filp = path_openat(dfd, pathname, &nd, op, flags | LOOKUP_REVAL); + return filp; +} + +struct file *do_file_open_root(struct dentry *dentry, struct vfsmount *mnt, + const char *name, const struct open_flags *op, int flags) +{ + struct nameidata nd; + struct file *file; + + nd.root.mnt = mnt; + nd.root.dentry = dentry; + + flags |= LOOKUP_ROOT; + + if (dentry->d_inode->i_op->follow_link && op->intent & LOOKUP_OPEN) + return ERR_PTR(-ELOOP); + + file = path_openat(-1, name, &nd, op, flags | LOOKUP_RCU); + if (unlikely(file == ERR_PTR(-ECHILD))) + file = path_openat(-1, name, &nd, op, flags); + if (unlikely(file == ERR_PTR(-ESTALE))) + file = path_openat(-1, name, &nd, op, flags | LOOKUP_REVAL); + return file; } -EXPORT_SYMBOL(filp_open); /** * lookup_create - lookup a dentry, creating it if it doesn't exist @@ -3120,7 +2917,11 @@ int vfs_link(struct dentry *old_dentry, struct inode *dir, struct dentry *new_de return error; mutex_lock(&inode->i_mutex); - error = dir->i_op->link(old_dentry, dir, new_dentry); + /* Make sure we don't allow creating hardlink to an unlinked file */ + if (inode->i_nlink == 0) + error = -ENOENT; + else + error = dir->i_op->link(old_dentry, dir, new_dentry); mutex_unlock(&inode->i_mutex); if (!error) fsnotify_link(dir, inode, new_dentry); @@ -3142,15 +2943,27 @@ SYSCALL_DEFINE5(linkat, int, olddfd, const char __user *, oldname, struct dentry *new_dentry; struct nameidata nd; struct path old_path; + int how = 0; int error; char *to; - if ((flags & ~AT_SYMLINK_FOLLOW) != 0) + if ((flags & ~(AT_SYMLINK_FOLLOW | AT_EMPTY_PATH)) != 0) return -EINVAL; + /* + * To use null names we require CAP_DAC_READ_SEARCH + * This ensures that not everyone will be able to create + * handlink using the passed filedescriptor. + */ + if (flags & AT_EMPTY_PATH) { + if (!capable(CAP_DAC_READ_SEARCH)) + return -ENOENT; + how = LOOKUP_EMPTY; + } + + if (flags & AT_SYMLINK_FOLLOW) + how |= LOOKUP_FOLLOW; - error = user_path_at(olddfd, oldname, - flags & AT_SYMLINK_FOLLOW ? LOOKUP_FOLLOW : 0, - &old_path); + error = user_path_at(olddfd, oldname, how, &old_path); if (error) return error; @@ -3587,7 +3400,7 @@ EXPORT_SYMBOL(page_readlink); EXPORT_SYMBOL(__page_symlink); EXPORT_SYMBOL(page_symlink); EXPORT_SYMBOL(page_symlink_inode_operations); -EXPORT_SYMBOL(path_lookup); +EXPORT_SYMBOL(kern_path_parent); EXPORT_SYMBOL(kern_path); EXPORT_SYMBOL(vfs_path_lookup); EXPORT_SYMBOL(inode_permission); diff --git a/fs/namespace.c b/fs/namespace.c index 7b0b953..9263995 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -196,7 +196,7 @@ unsigned int mnt_get_count(struct vfsmount *mnt) #endif } -struct vfsmount *alloc_vfsmnt(const char *name) +static struct vfsmount *alloc_vfsmnt(const char *name) { struct vfsmount *mnt = kmem_cache_zalloc(mnt_cache, GFP_KERNEL); if (mnt) { @@ -466,15 +466,7 @@ static void __mnt_unmake_readonly(struct vfsmount *mnt) br_write_unlock(vfsmount_lock); } -void simple_set_mnt(struct vfsmount *mnt, struct super_block *sb) -{ - mnt->mnt_sb = sb; - mnt->mnt_root = dget(sb->s_root); -} - -EXPORT_SYMBOL(simple_set_mnt); - -void free_vfsmnt(struct vfsmount *mnt) +static void free_vfsmnt(struct vfsmount *mnt) { kfree(mnt->mnt_devname); mnt_free_id(mnt); @@ -678,6 +670,36 @@ static struct vfsmount *skip_mnt_tree(struct vfsmount *p) return p; } +struct vfsmount * +vfs_kern_mount(struct file_system_type *type, int flags, const char *name, void *data) +{ + struct vfsmount *mnt; + struct dentry *root; + + if (!type) + return ERR_PTR(-ENODEV); + + mnt = alloc_vfsmnt(name); + if (!mnt) + return ERR_PTR(-ENOMEM); + + if (flags & MS_KERNMOUNT) + mnt->mnt_flags = MNT_INTERNAL; + + root = mount_fs(type, flags, name, data); + if (IS_ERR(root)) { + free_vfsmnt(mnt); + return ERR_CAST(root); + } + + mnt->mnt_root = root; + mnt->mnt_sb = root->d_sb; + mnt->mnt_mountpoint = mnt->mnt_root; + mnt->mnt_parent = mnt; + return mnt; +} +EXPORT_SYMBOL_GPL(vfs_kern_mount); + static struct vfsmount *clone_mnt(struct vfsmount *old, struct dentry *root, int flag) { @@ -978,7 +1000,13 @@ static int show_vfsmnt(struct seq_file *m, void *v) int err = 0; struct path mnt_path = { .dentry = mnt->mnt_root, .mnt = mnt }; - mangle(m, mnt->mnt_devname ? mnt->mnt_devname : "none"); + if (mnt->mnt_sb->s_op->show_devname) { + err = mnt->mnt_sb->s_op->show_devname(m, mnt); + if (err) + goto out; + } else { + mangle(m, mnt->mnt_devname ? mnt->mnt_devname : "none"); + } seq_putc(m, ' '); seq_path(m, &mnt_path, " \t\n\\"); seq_putc(m, ' '); @@ -1002,6 +1030,18 @@ const struct seq_operations mounts_op = { .show = show_vfsmnt }; +static int uuid_is_nil(u8 *uuid) +{ + int i; + u8 *cp = (u8 *)uuid; + + for (i = 0; i < 16; i++) { + if (*cp++) + return 0; + } + return 1; +} + static int show_mountinfo(struct seq_file *m, void *v) { struct proc_mounts *p = m->private; @@ -1013,7 +1053,12 @@ static int show_mountinfo(struct seq_file *m, void *v) seq_printf(m, "%i %i %u:%u ", mnt->mnt_id, mnt->mnt_parent->mnt_id, MAJOR(sb->s_dev), MINOR(sb->s_dev)); - seq_dentry(m, mnt->mnt_root, " \t\n\\"); + if (sb->s_op->show_path) + err = sb->s_op->show_path(m, mnt); + else + seq_dentry(m, mnt->mnt_root, " \t\n\\"); + if (err) + goto out; seq_putc(m, ' '); seq_path_root(m, &mnt_path, &root, " \t\n\\"); if (root.mnt != p->root.mnt || root.dentry != p->root.dentry) { @@ -1040,11 +1085,20 @@ static int show_mountinfo(struct seq_file *m, void *v) if (IS_MNT_UNBINDABLE(mnt)) seq_puts(m, " unbindable"); + if (!uuid_is_nil(mnt->mnt_sb->s_uuid)) + /* print the uuid */ + seq_printf(m, " uuid:%pU", mnt->mnt_sb->s_uuid); + /* Filesystem specific data */ seq_puts(m, " - "); show_type(m, sb); seq_putc(m, ' '); - mangle(m, mnt->mnt_devname ? mnt->mnt_devname : "none"); + if (sb->s_op->show_devname) + err = sb->s_op->show_devname(m, mnt); + else + mangle(m, mnt->mnt_devname ? mnt->mnt_devname : "none"); + if (err) + goto out; seq_puts(m, sb->s_flags & MS_RDONLY ? " ro" : " rw"); err = show_sb_opts(m, sb); if (err) @@ -1070,11 +1124,15 @@ static int show_vfsstat(struct seq_file *m, void *v) int err = 0; /* device */ - if (mnt->mnt_devname) { - seq_puts(m, "device "); - mangle(m, mnt->mnt_devname); - } else - seq_puts(m, "no device"); + if (mnt->mnt_sb->s_op->show_devname) { + err = mnt->mnt_sb->s_op->show_devname(m, mnt); + } else { + if (mnt->mnt_devname) { + seq_puts(m, "device "); + mangle(m, mnt->mnt_devname); + } else + seq_puts(m, "no device"); + } /* mount point */ seq_puts(m, " mounted on "); @@ -1088,7 +1146,8 @@ static int show_vfsstat(struct seq_file *m, void *v) /* optional statistics */ if (mnt->mnt_sb->s_op->show_stats) { seq_putc(m, ' '); - err = mnt->mnt_sb->s_op->show_stats(m, mnt); + if (!err) + err = mnt->mnt_sb->s_op->show_stats(m, mnt); } seq_putc(m, '\n'); @@ -1244,7 +1303,7 @@ static int do_umount(struct vfsmount *mnt, int flags) */ br_write_lock(vfsmount_lock); if (mnt_get_count(mnt) != 2) { - br_write_lock(vfsmount_lock); + br_write_unlock(vfsmount_lock); return -EBUSY; } br_write_unlock(vfsmount_lock); @@ -1604,9 +1663,35 @@ static int attach_recursive_mnt(struct vfsmount *source_mnt, return err; } +static int lock_mount(struct path *path) +{ + struct vfsmount *mnt; +retry: + mutex_lock(&path->dentry->d_inode->i_mutex); + if (unlikely(cant_mount(path->dentry))) { + mutex_unlock(&path->dentry->d_inode->i_mutex); + return -ENOENT; + } + down_write(&namespace_sem); + mnt = lookup_mnt(path); + if (likely(!mnt)) + return 0; + up_write(&namespace_sem); + mutex_unlock(&path->dentry->d_inode->i_mutex); + path_put(path); + path->mnt = mnt; + path->dentry = dget(mnt->mnt_root); + goto retry; +} + +static void unlock_mount(struct path *path) +{ + up_write(&namespace_sem); + mutex_unlock(&path->dentry->d_inode->i_mutex); +} + static int graft_tree(struct vfsmount *mnt, struct path *path) { - int err; if (mnt->mnt_sb->s_flags & MS_NOUSER) return -EINVAL; @@ -1614,16 +1699,10 @@ static int graft_tree(struct vfsmount *mnt, struct path *path) S_ISDIR(mnt->mnt_root->d_inode->i_mode)) return -ENOTDIR; - err = -ENOENT; - mutex_lock(&path->dentry->d_inode->i_mutex); - if (cant_mount(path->dentry)) - goto out_unlock; + if (d_unlinked(path->dentry)) + return -ENOENT; - if (!d_unlinked(path->dentry)) - err = attach_recursive_mnt(mnt, path, NULL); -out_unlock: - mutex_unlock(&path->dentry->d_inode->i_mutex); - return err; + return attach_recursive_mnt(mnt, path, NULL); } /* @@ -1686,6 +1765,7 @@ static int do_change_type(struct path *path, int flag) static int do_loopback(struct path *path, char *old_name, int recurse) { + LIST_HEAD(umount_list); struct path old_path; struct vfsmount *mnt = NULL; int err = mount_is_safe(path); @@ -1697,13 +1777,16 @@ static int do_loopback(struct path *path, char *old_name, if (err) return err; - down_write(&namespace_sem); + err = lock_mount(path); + if (err) + goto out; + err = -EINVAL; if (IS_MNT_UNBINDABLE(old_path.mnt)) - goto out; + goto out2; if (!check_mnt(path->mnt) || !check_mnt(old_path.mnt)) - goto out; + goto out2; err = -ENOMEM; if (recurse) @@ -1712,20 +1795,18 @@ static int do_loopback(struct path *path, char *old_name, mnt = clone_mnt(old_path.mnt, old_path.dentry, 0); if (!mnt) - goto out; + goto out2; err = graft_tree(mnt, path); if (err) { - LIST_HEAD(umount_list); - br_write_lock(vfsmount_lock); umount_tree(mnt, 0, &umount_list); br_write_unlock(vfsmount_lock); - release_mounts(&umount_list); } - +out2: + unlock_mount(path); + release_mounts(&umount_list); out: - up_write(&namespace_sem); path_put(&old_path); return err; } @@ -1767,6 +1848,10 @@ static int do_remount(struct path *path, int flags, int mnt_flags, if (path->dentry != path->mnt->mnt_root) return -EINVAL; + err = security_sb_remount(sb, data); + if (err) + return err; + down_write(&sb->s_umount); if (flags & MS_BIND) err = change_mount_flags(path->mnt, flags); @@ -1810,18 +1895,12 @@ static int do_move_mount(struct path *path, char *old_name) if (err) return err; - down_write(&namespace_sem); - err = follow_down(path, true); + err = lock_mount(path); if (err < 0) goto out; err = -EINVAL; if (!check_mnt(path->mnt) || !check_mnt(old_path.mnt)) - goto out; - - err = -ENOENT; - mutex_lock(&path->dentry->d_inode->i_mutex); - if (cant_mount(path->dentry)) goto out1; if (d_unlinked(path->dentry)) @@ -1863,16 +1942,87 @@ static int do_move_mount(struct path *path, char *old_name) * automatically */ list_del_init(&old_path.mnt->mnt_expire); out1: - mutex_unlock(&path->dentry->d_inode->i_mutex); + unlock_mount(path); out: - up_write(&namespace_sem); if (!err) path_put(&parent_path); path_put(&old_path); return err; } -static int do_add_mount(struct vfsmount *, struct path *, int); +static struct vfsmount *fs_set_subtype(struct vfsmount *mnt, const char *fstype) +{ + int err; + const char *subtype = strchr(fstype, '.'); + if (subtype) { + subtype++; + err = -EINVAL; + if (!subtype[0]) + goto err; + } else + subtype = ""; + + mnt->mnt_sb->s_subtype = kstrdup(subtype, GFP_KERNEL); + err = -ENOMEM; + if (!mnt->mnt_sb->s_subtype) + goto err; + return mnt; + + err: + mntput(mnt); + return ERR_PTR(err); +} + +struct vfsmount * +do_kern_mount(const char *fstype, int flags, const char *name, void *data) +{ + struct file_system_type *type = get_fs_type(fstype); + struct vfsmount *mnt; + if (!type) + return ERR_PTR(-ENODEV); + mnt = vfs_kern_mount(type, flags, name, data); + if (!IS_ERR(mnt) && (type->fs_flags & FS_HAS_SUBTYPE) && + !mnt->mnt_sb->s_subtype) + mnt = fs_set_subtype(mnt, fstype); + put_filesystem(type); + return mnt; +} +EXPORT_SYMBOL_GPL(do_kern_mount); + +/* + * add a mount into a namespace's mount tree + */ +static int do_add_mount(struct vfsmount *newmnt, struct path *path, int mnt_flags) +{ + int err; + + mnt_flags &= ~(MNT_SHARED | MNT_WRITE_HOLD | MNT_INTERNAL); + + err = lock_mount(path); + if (err) + return err; + + err = -EINVAL; + if (!(mnt_flags & MNT_SHRINKABLE) && !check_mnt(path->mnt)) + goto unlock; + + /* Refuse the same filesystem on the same mount point */ + err = -EBUSY; + if (path->mnt->mnt_sb == newmnt->mnt_sb && + path->mnt->mnt_root == path->dentry) + goto unlock; + + err = -EINVAL; + if (S_ISLNK(newmnt->mnt_root->d_inode->i_mode)) + goto unlock; + + newmnt->mnt_flags = mnt_flags; + err = graft_tree(newmnt, path); + +unlock: + unlock_mount(path); + return err; +} /* * create a new mount for userspace and request it to be added into the @@ -1932,43 +2082,6 @@ fail: return err; } -/* - * add a mount into a namespace's mount tree - */ -static int do_add_mount(struct vfsmount *newmnt, struct path *path, int mnt_flags) -{ - int err; - - mnt_flags &= ~(MNT_SHARED | MNT_WRITE_HOLD | MNT_INTERNAL); - - down_write(&namespace_sem); - /* Something was mounted here while we slept */ - err = follow_down(path, true); - if (err < 0) - goto unlock; - - err = -EINVAL; - if (!(mnt_flags & MNT_SHRINKABLE) && !check_mnt(path->mnt)) - goto unlock; - - /* Refuse the same filesystem on the same mount point */ - err = -EBUSY; - if (path->mnt->mnt_sb == newmnt->mnt_sb && - path->mnt->mnt_root == path->dentry) - goto unlock; - - err = -EINVAL; - if (S_ISLNK(newmnt->mnt_root->d_inode->i_mode)) - goto unlock; - - newmnt->mnt_flags = mnt_flags; - err = graft_tree(newmnt, path); - -unlock: - up_write(&namespace_sem); - return err; -} - /** * mnt_set_expiry - Put a mount on an expiration list * @mnt: The mount to list. @@ -2469,65 +2582,60 @@ SYSCALL_DEFINE2(pivot_root, const char __user *, new_root, error = user_path_dir(new_root, &new); if (error) goto out0; - error = -EINVAL; - if (!check_mnt(new.mnt)) - goto out1; error = user_path_dir(put_old, &old); if (error) goto out1; error = security_sb_pivotroot(&old, &new); - if (error) { - path_put(&old); - goto out1; - } + if (error) + goto out2; get_fs_root(current->fs, &root); - down_write(&namespace_sem); - mutex_lock(&old.dentry->d_inode->i_mutex); + error = lock_mount(&old); + if (error) + goto out3; + error = -EINVAL; if (IS_MNT_SHARED(old.mnt) || IS_MNT_SHARED(new.mnt->mnt_parent) || IS_MNT_SHARED(root.mnt->mnt_parent)) - goto out2; - if (!check_mnt(root.mnt)) - goto out2; + goto out4; + if (!check_mnt(root.mnt) || !check_mnt(new.mnt)) + goto out4; error = -ENOENT; - if (cant_mount(old.dentry)) - goto out2; if (d_unlinked(new.dentry)) - goto out2; + goto out4; if (d_unlinked(old.dentry)) - goto out2; + goto out4; error = -EBUSY; if (new.mnt == root.mnt || old.mnt == root.mnt) - goto out2; /* loop, on the same file system */ + goto out4; /* loop, on the same file system */ error = -EINVAL; if (root.mnt->mnt_root != root.dentry) - goto out2; /* not a mountpoint */ + goto out4; /* not a mountpoint */ if (root.mnt->mnt_parent == root.mnt) - goto out2; /* not attached */ + goto out4; /* not attached */ if (new.mnt->mnt_root != new.dentry) - goto out2; /* not a mountpoint */ + goto out4; /* not a mountpoint */ if (new.mnt->mnt_parent == new.mnt) - goto out2; /* not attached */ + goto out4; /* not attached */ /* make sure we can reach put_old from new_root */ tmp = old.mnt; - br_write_lock(vfsmount_lock); if (tmp != new.mnt) { for (;;) { if (tmp->mnt_parent == tmp) - goto out3; /* already mounted on put_old */ + goto out4; /* already mounted on put_old */ if (tmp->mnt_parent == new.mnt) break; tmp = tmp->mnt_parent; } if (!is_subdir(tmp->mnt_mountpoint, new.dentry)) - goto out3; + goto out4; } else if (!is_subdir(old.dentry, new.dentry)) - goto out3; + goto out4; + br_write_lock(vfsmount_lock); detach_mnt(new.mnt, &parent_path); detach_mnt(root.mnt, &root_parent); /* mount old root on put_old */ @@ -2537,22 +2645,21 @@ SYSCALL_DEFINE2(pivot_root, const char __user *, new_root, touch_mnt_namespace(current->nsproxy->mnt_ns); br_write_unlock(vfsmount_lock); chroot_fs_refs(&root, &new); - error = 0; - path_put(&root_parent); - path_put(&parent_path); -out2: - mutex_unlock(&old.dentry->d_inode->i_mutex); - up_write(&namespace_sem); +out4: + unlock_mount(&old); + if (!error) { + path_put(&root_parent); + path_put(&parent_path); + } +out3: path_put(&root); +out2: path_put(&old); out1: path_put(&new); out0: return error; -out3: - br_write_unlock(vfsmount_lock); - goto out2; } static void __init init_mount_tree(void) @@ -2627,3 +2734,9 @@ void put_mnt_ns(struct mnt_namespace *ns) kfree(ns); } EXPORT_SYMBOL(put_mnt_ns); + +struct vfsmount *kern_mount_data(struct file_system_type *type, void *data) +{ + return vfs_kern_mount(type, MS_KERNMOUNT, type->name, data); +} +EXPORT_SYMBOL_GPL(kern_mount_data); diff --git a/fs/ncpfs/Makefile b/fs/ncpfs/Makefile index 68ea095..c66af56 100644 --- a/fs/ncpfs/Makefile +++ b/fs/ncpfs/Makefile @@ -11,6 +11,6 @@ ncpfs-$(CONFIG_NCPFS_EXTRAS) += symlink.o ncpfs-$(CONFIG_NCPFS_NFS_NS) += symlink.o # If you want debugging output, please uncomment the following line -# EXTRA_CFLAGS += -DDEBUG_NCP=1 +# ccflags-y := -DDEBUG_NCP=1 CFLAGS_ncplib_kernel.o := -finline-functions diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c index 1990165..e3d2942 100644 --- a/fs/nfs/callback.c +++ b/fs/nfs/callback.c @@ -135,33 +135,6 @@ out_err: #if defined(CONFIG_NFS_V4_1) /* - * * CB_SEQUENCE operations will fail until the callback sessionid is set. - * */ -int nfs4_set_callback_sessionid(struct nfs_client *clp) -{ - struct svc_serv *serv = clp->cl_rpcclient->cl_xprt->bc_serv; - struct nfs4_sessionid *bc_sid; - - if (!serv->sv_bc_xprt) - return -EINVAL; - - /* on success freed in xprt_free */ - bc_sid = kmalloc(sizeof(struct nfs4_sessionid), GFP_KERNEL); - if (!bc_sid) - return -ENOMEM; - memcpy(bc_sid->data, &clp->cl_session->sess_id.data, - NFS4_MAX_SESSIONID_LEN); - spin_lock_bh(&serv->sv_cb_lock); - serv->sv_bc_xprt->xpt_bc_sid = bc_sid; - spin_unlock_bh(&serv->sv_cb_lock); - dprintk("%s set xpt_bc_sid=%u:%u:%u:%u for sv_bc_xprt %p\n", __func__, - ((u32 *)bc_sid->data)[0], ((u32 *)bc_sid->data)[1], - ((u32 *)bc_sid->data)[2], ((u32 *)bc_sid->data)[3], - serv->sv_bc_xprt); - return 0; -} - -/* * The callback service for NFSv4.1 callbacks */ static int @@ -266,10 +239,6 @@ static inline void nfs_callback_bc_serv(u32 minorversion, struct rpc_xprt *xprt, struct nfs_callback_data *cb_info) { } -int nfs4_set_callback_sessionid(struct nfs_client *clp) -{ - return 0; -} #endif /* CONFIG_NFS_V4_1 */ /* @@ -359,78 +328,58 @@ void nfs_callback_down(int minorversion) mutex_unlock(&nfs_callback_mutex); } -static int check_gss_callback_principal(struct nfs_client *clp, - struct svc_rqst *rqstp) +/* Boolean check of RPC_AUTH_GSS principal */ +int +check_gss_callback_principal(struct nfs_client *clp, struct svc_rqst *rqstp) { struct rpc_clnt *r = clp->cl_rpcclient; char *p = svc_gss_principal(rqstp); + if (rqstp->rq_authop->flavour != RPC_AUTH_GSS) + return 1; + /* No RPC_AUTH_GSS on NFSv4.1 back channel yet */ if (clp->cl_minorversion != 0) - return SVC_DROP; + return 0; /* * It might just be a normal user principal, in which case * userspace won't bother to tell us the name at all. */ if (p == NULL) - return SVC_DENIED; + return 0; /* Expect a GSS_C_NT_HOSTBASED_NAME like "nfs@serverhostname" */ if (memcmp(p, "nfs@", 4) != 0) - return SVC_DENIED; + return 0; p += 4; if (strcmp(p, r->cl_server) != 0) - return SVC_DENIED; - return SVC_OK; + return 0; + return 1; } -/* pg_authenticate method helper */ -static struct nfs_client *nfs_cb_find_client(struct svc_rqst *rqstp) -{ - struct nfs4_sessionid *sessionid = bc_xprt_sid(rqstp); - int is_cb_compound = rqstp->rq_proc == CB_COMPOUND ? 1 : 0; - - dprintk("--> %s rq_proc %d\n", __func__, rqstp->rq_proc); - if (svc_is_backchannel(rqstp)) - /* Sessionid (usually) set after CB_NULL ping */ - return nfs4_find_client_sessionid(svc_addr(rqstp), sessionid, - is_cb_compound); - else - /* No callback identifier in pg_authenticate */ - return nfs4_find_client_no_ident(svc_addr(rqstp)); -} - -/* pg_authenticate method for nfsv4 callback threads. */ +/* + * pg_authenticate method for nfsv4 callback threads. + * + * The authflavor has been negotiated, so an incorrect flavor is a server + * bug. Drop packets with incorrect authflavor. + * + * All other checking done after NFS decoding where the nfs_client can be + * found in nfs4_callback_compound + */ static int nfs_callback_authenticate(struct svc_rqst *rqstp) { - struct nfs_client *clp; - RPC_IFDEBUG(char buf[RPC_MAX_ADDRBUFLEN]); - int ret = SVC_OK; - - /* Don't talk to strangers */ - clp = nfs_cb_find_client(rqstp); - if (clp == NULL) - return SVC_DROP; - - dprintk("%s: %s NFSv4 callback!\n", __func__, - svc_print_addr(rqstp, buf, sizeof(buf))); - switch (rqstp->rq_authop->flavour) { - case RPC_AUTH_NULL: - if (rqstp->rq_proc != CB_NULL) - ret = SVC_DENIED; - break; - case RPC_AUTH_UNIX: - break; - case RPC_AUTH_GSS: - ret = check_gss_callback_principal(clp, rqstp); - break; - default: - ret = SVC_DENIED; + case RPC_AUTH_NULL: + if (rqstp->rq_proc != CB_NULL) + return SVC_DROP; + break; + case RPC_AUTH_GSS: + /* No RPC_AUTH_GSS support yet in NFSv4.1 */ + if (svc_is_backchannel(rqstp)) + return SVC_DROP; } - nfs_put_client(clp); - return ret; + return SVC_OK; } /* diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h index d3b44f9..46d93ce 100644 --- a/fs/nfs/callback.h +++ b/fs/nfs/callback.h @@ -7,6 +7,7 @@ */ #ifndef __LINUX_FS_NFS_CALLBACK_H #define __LINUX_FS_NFS_CALLBACK_H +#include <linux/sunrpc/svc.h> #define NFS4_CALLBACK 0x40000000 #define NFS4_CALLBACK_XDRSIZE 2048 @@ -37,7 +38,6 @@ enum nfs4_callback_opnum { struct cb_process_state { __be32 drc_status; struct nfs_client *clp; - struct nfs4_sessionid *svc_sid; /* v4.1 callback service sessionid */ }; struct cb_compound_hdr_arg { @@ -168,7 +168,7 @@ extern unsigned nfs4_callback_layoutrecall( extern void nfs4_check_drain_bc_complete(struct nfs4_session *ses); extern void nfs4_cb_take_slot(struct nfs_client *clp); #endif /* CONFIG_NFS_V4_1 */ - +extern int check_gss_callback_principal(struct nfs_client *, struct svc_rqst *); extern __be32 nfs4_callback_getattr(struct cb_getattrargs *args, struct cb_getattrres *res, struct cb_process_state *cps); diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c index 4bb91cb..2f41dcce 100644 --- a/fs/nfs/callback_proc.c +++ b/fs/nfs/callback_proc.c @@ -188,10 +188,10 @@ static u32 initiate_bulk_draining(struct nfs_client *clp, rv = NFS4ERR_DELAY; list_del_init(&lo->plh_bulk_recall); spin_unlock(&ino->i_lock); + pnfs_free_lseg_list(&free_me_list); put_layout_hdr(lo); iput(ino); } - pnfs_free_lseg_list(&free_me_list); return rv; } @@ -373,17 +373,11 @@ __be32 nfs4_callback_sequence(struct cb_sequenceargs *args, { struct nfs_client *clp; int i; - __be32 status; + __be32 status = htonl(NFS4ERR_BADSESSION); cps->clp = NULL; - status = htonl(NFS4ERR_BADSESSION); - /* Incoming session must match the callback session */ - if (memcmp(&args->csa_sessionid, cps->svc_sid, NFS4_MAX_SESSIONID_LEN)) - goto out; - - clp = nfs4_find_client_sessionid(args->csa_addr, - &args->csa_sessionid, 1); + clp = nfs4_find_client_sessionid(args->csa_addr, &args->csa_sessionid); if (clp == NULL) goto out; @@ -414,9 +408,9 @@ __be32 nfs4_callback_sequence(struct cb_sequenceargs *args, res->csr_highestslotid = NFS41_BC_MAX_CALLBACKS - 1; res->csr_target_highestslotid = NFS41_BC_MAX_CALLBACKS - 1; nfs4_cb_take_slot(clp); - cps->clp = clp; /* put in nfs4_callback_compound */ out: + cps->clp = clp; /* put in nfs4_callback_compound */ for (i = 0; i < args->csa_nrclists; i++) kfree(args->csa_rclists[i].rcl_refcalls); kfree(args->csa_rclists); diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index 23112c2..14e0f93 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -794,10 +794,9 @@ static __be32 nfs4_callback_compound(struct svc_rqst *rqstp, void *argp, void *r if (hdr_arg.minorversion == 0) { cps.clp = nfs4_find_client_ident(hdr_arg.cb_ident); - if (!cps.clp) + if (!cps.clp || !check_gss_callback_principal(cps.clp, rqstp)) return rpc_drop_reply; - } else - cps.svc_sid = bc_xprt_sid(rqstp); + } hdr_res.taglen = hdr_arg.taglen; hdr_res.tag = hdr_arg.tag; diff --git a/fs/nfs/client.c b/fs/nfs/client.c index 192f2f8..139be96 100644 --- a/fs/nfs/client.c +++ b/fs/nfs/client.c @@ -82,6 +82,11 @@ retry: #endif /* CONFIG_NFS_V4 */ /* + * Turn off NFSv4 uid/gid mapping when using AUTH_SYS + */ +static int nfs4_disable_idmapping = 0; + +/* * RPC cruft for NFS */ static struct rpc_version *nfs_version[5] = { @@ -481,7 +486,12 @@ static struct nfs_client *nfs_match_client(const struct nfs_client_initdata *dat * Look up a client by IP address and protocol version * - creates a new record if one doesn't yet exist */ -static struct nfs_client *nfs_get_client(const struct nfs_client_initdata *cl_init) +static struct nfs_client * +nfs_get_client(const struct nfs_client_initdata *cl_init, + const struct rpc_timeout *timeparms, + const char *ip_addr, + rpc_authflavor_t authflavour, + int noresvport) { struct nfs_client *clp, *new = NULL; int error; @@ -512,6 +522,13 @@ install_client: clp = new; list_add(&clp->cl_share_link, &nfs_client_list); spin_unlock(&nfs_client_lock); + + error = cl_init->rpc_ops->init_client(clp, timeparms, ip_addr, + authflavour, noresvport); + if (error < 0) { + nfs_put_client(clp); + return ERR_PTR(error); + } dprintk("--> nfs_get_client() = %p [new]\n", clp); return clp; @@ -767,9 +784,9 @@ static int nfs_init_server_rpcclient(struct nfs_server *server, /* * Initialise an NFS2 or NFS3 client */ -static int nfs_init_client(struct nfs_client *clp, - const struct rpc_timeout *timeparms, - const struct nfs_parsed_mount_data *data) +int nfs_init_client(struct nfs_client *clp, const struct rpc_timeout *timeparms, + const char *ip_addr, rpc_authflavor_t authflavour, + int noresvport) { int error; @@ -784,7 +801,7 @@ static int nfs_init_client(struct nfs_client *clp, * - RFC 2623, sec 2.3.2 */ error = nfs_create_rpc_client(clp, timeparms, RPC_AUTH_UNIX, - 0, data->flags & NFS_MOUNT_NORESVPORT); + 0, noresvport); if (error < 0) goto error; nfs_mark_client_ready(clp, NFS_CS_READY); @@ -820,19 +837,17 @@ static int nfs_init_server(struct nfs_server *server, cl_init.rpc_ops = &nfs_v3_clientops; #endif + nfs_init_timeout_values(&timeparms, data->nfs_server.protocol, + data->timeo, data->retrans); + /* Allocate or find a client reference we can use */ - clp = nfs_get_client(&cl_init); + clp = nfs_get_client(&cl_init, &timeparms, NULL, RPC_AUTH_UNIX, + data->flags & NFS_MOUNT_NORESVPORT); if (IS_ERR(clp)) { dprintk("<-- nfs_init_server() = error %ld\n", PTR_ERR(clp)); return PTR_ERR(clp); } - nfs_init_timeout_values(&timeparms, data->nfs_server.protocol, - data->timeo, data->retrans); - error = nfs_init_client(clp, &timeparms, data); - if (error < 0) - goto error; - server->nfs_client = clp; /* Initialise the client representation from the mount data */ @@ -1009,14 +1024,19 @@ static void nfs_server_insert_lists(struct nfs_server *server) spin_lock(&nfs_client_lock); list_add_tail_rcu(&server->client_link, &clp->cl_superblocks); list_add_tail(&server->master_link, &nfs_volume_list); + clear_bit(NFS_CS_STOP_RENEW, &clp->cl_res_state); spin_unlock(&nfs_client_lock); } static void nfs_server_remove_lists(struct nfs_server *server) { + struct nfs_client *clp = server->nfs_client; + spin_lock(&nfs_client_lock); list_del_rcu(&server->client_link); + if (clp && list_empty(&clp->cl_superblocks)) + set_bit(NFS_CS_STOP_RENEW, &clp->cl_res_state); list_del(&server->master_link); spin_unlock(&nfs_client_lock); @@ -1206,16 +1226,11 @@ nfs4_find_client_ident(int cb_ident) * For CB_COMPOUND calls, find a client by IP address, protocol version, * minorversion, and sessionID * - * CREATE_SESSION triggers a CB_NULL ping from servers. The callback service - * sessionid can only be set after the CREATE_SESSION return, so a CB_NULL - * can arrive before the callback sessionid is set. For CB_NULL calls, - * find a client by IP address protocol version, and minorversion. - * * Returns NULL if no such client */ struct nfs_client * nfs4_find_client_sessionid(const struct sockaddr *addr, - struct nfs4_sessionid *sid, int is_cb_compound) + struct nfs4_sessionid *sid) { struct nfs_client *clp; @@ -1227,9 +1242,9 @@ nfs4_find_client_sessionid(const struct sockaddr *addr, if (!nfs4_has_session(clp)) continue; - /* Match sessionid unless cb_null call*/ - if (is_cb_compound && (memcmp(clp->cl_session->sess_id.data, - sid->data, NFS4_MAX_SESSIONID_LEN) != 0)) + /* Match sessionid*/ + if (memcmp(clp->cl_session->sess_id.data, + sid->data, NFS4_MAX_SESSIONID_LEN) != 0) continue; atomic_inc(&clp->cl_count); @@ -1244,7 +1259,7 @@ nfs4_find_client_sessionid(const struct sockaddr *addr, struct nfs_client * nfs4_find_client_sessionid(const struct sockaddr *addr, - struct nfs4_sessionid *sid, int is_cb_compound) + struct nfs4_sessionid *sid) { return NULL; } @@ -1312,11 +1327,11 @@ static int nfs4_init_client_minor_version(struct nfs_client *clp) /* * Initialise an NFS4 client record */ -static int nfs4_init_client(struct nfs_client *clp, - const struct rpc_timeout *timeparms, - const char *ip_addr, - rpc_authflavor_t authflavour, - int flags) +int nfs4_init_client(struct nfs_client *clp, + const struct rpc_timeout *timeparms, + const char *ip_addr, + rpc_authflavor_t authflavour, + int noresvport) { int error; @@ -1330,7 +1345,7 @@ static int nfs4_init_client(struct nfs_client *clp, clp->rpc_ops = &nfs_v4_clientops; error = nfs_create_rpc_client(clp, timeparms, authflavour, - 1, flags & NFS_MOUNT_NORESVPORT); + 1, noresvport); if (error < 0) goto error; strlcpy(clp->cl_ipaddr, ip_addr, sizeof(clp->cl_ipaddr)); @@ -1383,27 +1398,71 @@ static int nfs4_set_client(struct nfs_server *server, dprintk("--> nfs4_set_client()\n"); /* Allocate or find a client reference we can use */ - clp = nfs_get_client(&cl_init); + clp = nfs_get_client(&cl_init, timeparms, ip_addr, authflavour, + server->flags & NFS_MOUNT_NORESVPORT); if (IS_ERR(clp)) { error = PTR_ERR(clp); goto error; } - error = nfs4_init_client(clp, timeparms, ip_addr, authflavour, - server->flags); - if (error < 0) - goto error_put; + + /* + * Query for the lease time on clientid setup or renewal + * + * Note that this will be set on nfs_clients that were created + * only for the DS role and did not set this bit, but now will + * serve a dual role. + */ + set_bit(NFS_CS_CHECK_LEASE_TIME, &clp->cl_res_state); server->nfs_client = clp; dprintk("<-- nfs4_set_client() = 0 [new %p]\n", clp); return 0; - -error_put: - nfs_put_client(clp); error: dprintk("<-- nfs4_set_client() = xerror %d\n", error); return error; } +/* + * Set up a pNFS Data Server client. + * + * Return any existing nfs_client that matches server address,port,version + * and minorversion. + * + * For a new nfs_client, use a soft mount (default), a low retrans and a + * low timeout interval so that if a connection is lost, we retry through + * the MDS. + */ +struct nfs_client *nfs4_set_ds_client(struct nfs_client* mds_clp, + const struct sockaddr *ds_addr, + int ds_addrlen, int ds_proto) +{ + struct nfs_client_initdata cl_init = { + .addr = ds_addr, + .addrlen = ds_addrlen, + .rpc_ops = &nfs_v4_clientops, + .proto = ds_proto, + .minorversion = mds_clp->cl_minorversion, + }; + struct rpc_timeout ds_timeout = { + .to_initval = 15 * HZ, + .to_maxval = 15 * HZ, + .to_retries = 1, + .to_exponential = 1, + }; + struct nfs_client *clp; + + /* + * Set an authflavor equual to the MDS value. Use the MDS nfs_client + * cl_ipaddr so as to use the same EXCHANGE_ID co_ownerid as the MDS + * (section 13.1 RFC 5661). + */ + clp = nfs_get_client(&cl_init, &ds_timeout, mds_clp->cl_ipaddr, + mds_clp->cl_rpcclient->cl_auth->au_flavor, 0); + + dprintk("<-- %s %p\n", __func__, clp); + return clp; +} +EXPORT_SYMBOL(nfs4_set_ds_client); /* * Session has been established, and the client marked ready. @@ -1440,6 +1499,10 @@ static int nfs4_server_common_setup(struct nfs_server *server, BUG_ON(!server->nfs_client->rpc_ops); BUG_ON(!server->nfs_client->rpc_ops->file_inode_ops); + /* data servers support only a subset of NFSv4.1 */ + if (is_ds_only_client(server->nfs_client)) + return -EPROTONOSUPPORT; + fattr = nfs_alloc_fattr(); if (fattr == NULL) return -ENOMEM; @@ -1509,6 +1572,13 @@ static int nfs4_init_server(struct nfs_server *server, if (error < 0) goto error; + /* + * Don't use NFS uid/gid mapping if we're using AUTH_SYS or lower + * authentication. + */ + if (nfs4_disable_idmapping && data->auth_flavors[0] == RPC_AUTH_UNIX) + server->caps |= NFS_CAP_UIDGID_NOMAP; + if (data->rsize) server->rsize = nfs_block_size(data->rsize, NULL); if (data->wsize) @@ -1926,3 +1996,7 @@ void nfs_fs_proc_exit(void) } #endif /* CONFIG_PROC_FS */ + +module_param(nfs4_disable_idmapping, bool, 0644); +MODULE_PARM_DESC(nfs4_disable_idmapping, + "Turn off NFSv4 idmapping when using 'sec=sys'"); diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index 364e432..bbbc6bf 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -23,8 +23,6 @@ static void nfs_do_free_delegation(struct nfs_delegation *delegation) { - if (delegation->cred) - put_rpccred(delegation->cred); kfree(delegation); } @@ -37,6 +35,10 @@ static void nfs_free_delegation_callback(struct rcu_head *head) static void nfs_free_delegation(struct nfs_delegation *delegation) { + if (delegation->cred) { + put_rpccred(delegation->cred); + delegation->cred = NULL; + } call_rcu(&delegation->rcu, nfs_free_delegation_callback); } diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 2c3eb33..abdf38d 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -1169,11 +1169,23 @@ static void nfs_dentry_iput(struct dentry *dentry, struct inode *inode) iput(inode); } +static void nfs_d_release(struct dentry *dentry) +{ + /* free cached devname value, if it survived that far */ + if (unlikely(dentry->d_fsdata)) { + if (dentry->d_flags & DCACHE_NFSFS_RENAMED) + WARN_ON(1); + else + kfree(dentry->d_fsdata); + } +} + const struct dentry_operations nfs_dentry_operations = { .d_revalidate = nfs_lookup_revalidate, .d_delete = nfs_dentry_delete, .d_iput = nfs_dentry_iput, .d_automount = nfs_d_automount, + .d_release = nfs_d_release, }; static struct dentry *nfs_lookup(struct inode *dir, struct dentry * dentry, struct nameidata *nd) @@ -1248,6 +1260,7 @@ const struct dentry_operations nfs4_dentry_operations = { .d_delete = nfs_dentry_delete, .d_iput = nfs_dentry_iput, .d_automount = nfs_d_automount, + .d_release = nfs_d_release, }; /* diff --git a/fs/nfs/direct.c b/fs/nfs/direct.c index e6ace0d..8eea253 100644 --- a/fs/nfs/direct.c +++ b/fs/nfs/direct.c @@ -45,6 +45,7 @@ #include <linux/pagemap.h> #include <linux/kref.h> #include <linux/slab.h> +#include <linux/task_io_accounting_ops.h> #include <linux/nfs_fs.h> #include <linux/nfs_page.h> @@ -407,15 +408,18 @@ static ssize_t nfs_direct_read_schedule_iovec(struct nfs_direct_req *dreq, pos += vec->iov_len; } + /* + * If no bytes were started, return the error, and let the + * generic layer handle the completion. + */ + if (requested_bytes == 0) { + nfs_direct_req_release(dreq); + return result < 0 ? result : -EIO; + } + if (put_dreq(dreq)) nfs_direct_complete(dreq); - - if (requested_bytes != 0) - return 0; - - if (result < 0) - return result; - return -EIO; + return 0; } static ssize_t nfs_direct_read(struct kiocb *iocb, const struct iovec *iov, @@ -646,8 +650,7 @@ static void nfs_direct_write_result(struct rpc_task *task, void *calldata) { struct nfs_write_data *data = calldata; - if (nfs_writeback_done(task, data) != 0) - return; + nfs_writeback_done(task, data); } /* @@ -841,15 +844,18 @@ static ssize_t nfs_direct_write_schedule_iovec(struct nfs_direct_req *dreq, pos += vec->iov_len; } + /* + * If no bytes were started, return the error, and let the + * generic layer handle the completion. + */ + if (requested_bytes == 0) { + nfs_direct_req_release(dreq); + return result < 0 ? result : -EIO; + } + if (put_dreq(dreq)) nfs_direct_write_complete(dreq, dreq->inode); - - if (requested_bytes != 0) - return 0; - - if (result < 0) - return result; - return -EIO; + return 0; } static ssize_t nfs_direct_write(struct kiocb *iocb, const struct iovec *iov, @@ -932,6 +938,8 @@ ssize_t nfs_file_direct_read(struct kiocb *iocb, const struct iovec *iov, if (retval) goto out; + task_io_account_read(count); + retval = nfs_direct_read(iocb, iov, nr_segs, pos); if (retval > 0) iocb->ki_pos = pos + retval; @@ -993,6 +1001,8 @@ ssize_t nfs_file_direct_write(struct kiocb *iocb, const struct iovec *iov, if (retval) goto out; + task_io_account_write(count); + retval = nfs_direct_write(iocb, iov, nr_segs, pos, count); if (retval > 0) diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 7bf029e..d85a534 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -387,10 +387,6 @@ static int nfs_write_begin(struct file *file, struct address_space *mapping, file->f_path.dentry->d_name.name, mapping->host->i_ino, len, (long long) pos); - pnfs_update_layout(mapping->host, - nfs_file_open_context(file), - IOMODE_RW); - start: /* * Prevent starvation issues if someone is doing a consistency diff --git a/fs/nfs/getroot.c b/fs/nfs/getroot.c index b5ffe8fa..1084792 100644 --- a/fs/nfs/getroot.c +++ b/fs/nfs/getroot.c @@ -75,18 +75,25 @@ static int nfs_superblock_set_dummy_root(struct super_block *sb, struct inode *i /* * get an NFS2/NFS3 root dentry from the root filehandle */ -struct dentry *nfs_get_root(struct super_block *sb, struct nfs_fh *mntfh) +struct dentry *nfs_get_root(struct super_block *sb, struct nfs_fh *mntfh, + const char *devname) { struct nfs_server *server = NFS_SB(sb); struct nfs_fsinfo fsinfo; struct dentry *ret; struct inode *inode; + void *name = kstrdup(devname, GFP_KERNEL); int error; + if (!name) + return ERR_PTR(-ENOMEM); + /* get the actual root for this mount */ fsinfo.fattr = nfs_alloc_fattr(); - if (fsinfo.fattr == NULL) + if (fsinfo.fattr == NULL) { + kfree(name); return ERR_PTR(-ENOMEM); + } error = server->nfs_client->rpc_ops->getroot(server, mntfh, &fsinfo); if (error < 0) { @@ -119,7 +126,15 @@ struct dentry *nfs_get_root(struct super_block *sb, struct nfs_fh *mntfh) } security_d_instantiate(ret, inode); + spin_lock(&ret->d_lock); + if (IS_ROOT(ret) && !(ret->d_flags & DCACHE_NFSFS_RENAMED)) { + ret->d_fsdata = name; + name = NULL; + } + spin_unlock(&ret->d_lock); out: + if (name) + kfree(name); nfs_free_fattr(fsinfo.fattr); return ret; } @@ -169,27 +184,35 @@ out: /* * get an NFS4 root dentry from the root filehandle */ -struct dentry *nfs4_get_root(struct super_block *sb, struct nfs_fh *mntfh) +struct dentry *nfs4_get_root(struct super_block *sb, struct nfs_fh *mntfh, + const char *devname) { struct nfs_server *server = NFS_SB(sb); struct nfs_fattr *fattr = NULL; struct dentry *ret; struct inode *inode; + void *name = kstrdup(devname, GFP_KERNEL); int error; dprintk("--> nfs4_get_root()\n"); + if (!name) + return ERR_PTR(-ENOMEM); + /* get the info about the server and filesystem */ error = nfs4_server_capabilities(server, mntfh); if (error < 0) { dprintk("nfs_get_root: getcaps error = %d\n", -error); + kfree(name); return ERR_PTR(error); } fattr = nfs_alloc_fattr(); - if (fattr == NULL) - return ERR_PTR(-ENOMEM);; + if (fattr == NULL) { + kfree(name); + return ERR_PTR(-ENOMEM); + } /* get the actual root for this mount */ error = server->nfs_client->rpc_ops->getattr(server, mntfh, fattr); @@ -223,8 +246,15 @@ struct dentry *nfs4_get_root(struct super_block *sb, struct nfs_fh *mntfh) } security_d_instantiate(ret, inode); - + spin_lock(&ret->d_lock); + if (IS_ROOT(ret) && !(ret->d_flags & DCACHE_NFSFS_RENAMED)) { + ret->d_fsdata = name; + name = NULL; + } + spin_unlock(&ret->d_lock); out: + if (name) + kfree(name); nfs_free_fattr(fattr); dprintk("<-- nfs4_get_root()\n"); return ret; diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index 1869688..79664a1 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -33,16 +33,41 @@ * NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. */ +#include <linux/types.h> +#include <linux/string.h> +#include <linux/kernel.h> + +static int nfs_map_string_to_numeric(const char *name, size_t namelen, __u32 *res) +{ + unsigned long val; + char buf[16]; + + if (memchr(name, '@', namelen) != NULL || namelen >= sizeof(buf)) + return 0; + memcpy(buf, name, namelen); + buf[namelen] = '\0'; + if (strict_strtoul(buf, 0, &val) != 0) + return 0; + *res = val; + return 1; +} + +static int nfs_map_numeric_to_string(__u32 id, char *buf, size_t buflen) +{ + return snprintf(buf, buflen, "%u", id); +} #ifdef CONFIG_NFS_USE_NEW_IDMAPPER #include <linux/slab.h> #include <linux/cred.h> +#include <linux/sunrpc/sched.h> +#include <linux/nfs4.h> +#include <linux/nfs_fs_sb.h> #include <linux/nfs_idmap.h> #include <linux/keyctl.h> #include <linux/key-type.h> #include <linux/rcupdate.h> -#include <linux/kernel.h> #include <linux/err.h> #include <keys/user-type.h> @@ -219,23 +244,39 @@ static int nfs_idmap_lookup_id(const char *name, size_t namelen, return ret; } -int nfs_map_name_to_uid(struct nfs_client *clp, const char *name, size_t namelen, __u32 *uid) +int nfs_map_name_to_uid(const struct nfs_server *server, const char *name, size_t namelen, __u32 *uid) { + if (nfs_map_string_to_numeric(name, namelen, uid)) + return 0; return nfs_idmap_lookup_id(name, namelen, "uid", uid); } -int nfs_map_group_to_gid(struct nfs_client *clp, const char *name, size_t namelen, __u32 *gid) +int nfs_map_group_to_gid(const struct nfs_server *server, const char *name, size_t namelen, __u32 *gid) { + if (nfs_map_string_to_numeric(name, namelen, gid)) + return 0; return nfs_idmap_lookup_id(name, namelen, "gid", gid); } -int nfs_map_uid_to_name(struct nfs_client *clp, __u32 uid, char *buf, size_t buflen) +int nfs_map_uid_to_name(const struct nfs_server *server, __u32 uid, char *buf, size_t buflen) { - return nfs_idmap_lookup_name(uid, "user", buf, buflen); + int ret = -EINVAL; + + if (!(server->caps & NFS_CAP_UIDGID_NOMAP)) + ret = nfs_idmap_lookup_name(uid, "user", buf, buflen); + if (ret < 0) + ret = nfs_map_numeric_to_string(uid, buf, buflen); + return ret; } -int nfs_map_gid_to_group(struct nfs_client *clp, __u32 gid, char *buf, size_t buflen) +int nfs_map_gid_to_group(const struct nfs_server *server, __u32 gid, char *buf, size_t buflen) { - return nfs_idmap_lookup_name(gid, "group", buf, buflen); + int ret = -EINVAL; + + if (!(server->caps & NFS_CAP_UIDGID_NOMAP)) + ret = nfs_idmap_lookup_name(gid, "group", buf, buflen); + if (ret < 0) + ret = nfs_map_numeric_to_string(gid, buf, buflen); + return ret; } #else /* CONFIG_NFS_USE_NEW_IDMAPPER not defined */ @@ -243,7 +284,6 @@ int nfs_map_gid_to_group(struct nfs_client *clp, __u32 gid, char *buf, size_t bu #include <linux/module.h> #include <linux/mutex.h> #include <linux/init.h> -#include <linux/types.h> #include <linux/slab.h> #include <linux/socket.h> #include <linux/in.h> @@ -695,31 +735,45 @@ static unsigned int fnvhash32(const void *buf, size_t buflen) return hash; } -int nfs_map_name_to_uid(struct nfs_client *clp, const char *name, size_t namelen, __u32 *uid) +int nfs_map_name_to_uid(const struct nfs_server *server, const char *name, size_t namelen, __u32 *uid) { - struct idmap *idmap = clp->cl_idmap; + struct idmap *idmap = server->nfs_client->cl_idmap; + if (nfs_map_string_to_numeric(name, namelen, uid)) + return 0; return nfs_idmap_id(idmap, &idmap->idmap_user_hash, name, namelen, uid); } -int nfs_map_group_to_gid(struct nfs_client *clp, const char *name, size_t namelen, __u32 *uid) +int nfs_map_group_to_gid(const struct nfs_server *server, const char *name, size_t namelen, __u32 *uid) { - struct idmap *idmap = clp->cl_idmap; + struct idmap *idmap = server->nfs_client->cl_idmap; + if (nfs_map_string_to_numeric(name, namelen, uid)) + return 0; return nfs_idmap_id(idmap, &idmap->idmap_group_hash, name, namelen, uid); } -int nfs_map_uid_to_name(struct nfs_client *clp, __u32 uid, char *buf, size_t buflen) +int nfs_map_uid_to_name(const struct nfs_server *server, __u32 uid, char *buf, size_t buflen) { - struct idmap *idmap = clp->cl_idmap; + struct idmap *idmap = server->nfs_client->cl_idmap; + int ret = -EINVAL; - return nfs_idmap_name(idmap, &idmap->idmap_user_hash, uid, buf); + if (!(server->caps & NFS_CAP_UIDGID_NOMAP)) + ret = nfs_idmap_name(idmap, &idmap->idmap_user_hash, uid, buf); + if (ret < 0) + ret = nfs_map_numeric_to_string(uid, buf, buflen); + return ret; } -int nfs_map_gid_to_group(struct nfs_client *clp, __u32 uid, char *buf, size_t buflen) +int nfs_map_gid_to_group(const struct nfs_server *server, __u32 uid, char *buf, size_t buflen) { - struct idmap *idmap = clp->cl_idmap; + struct idmap *idmap = server->nfs_client->cl_idmap; + int ret = -EINVAL; - return nfs_idmap_name(idmap, &idmap->idmap_group_hash, uid, buf); + if (!(server->caps & NFS_CAP_UIDGID_NOMAP)) + ret = nfs_idmap_name(idmap, &idmap->idmap_group_hash, uid, buf); + if (ret < 0) + ret = nfs_map_numeric_to_string(uid, buf, buflen); + return ret; } #endif /* CONFIG_NFS_USE_NEW_IDMAPPER */ diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index d851242..01768e5 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -37,6 +37,7 @@ #include <linux/inet.h> #include <linux/nfs_xdr.h> #include <linux/slab.h> +#include <linux/compat.h> #include <asm/system.h> #include <asm/uaccess.h> @@ -89,7 +90,11 @@ int nfs_wait_bit_killable(void *word) */ u64 nfs_compat_user_ino64(u64 fileid) { - int ino; +#ifdef CONFIG_COMPAT + compat_ulong_t ino; +#else + unsigned long ino; +#endif if (enable_ino64) return fileid; @@ -881,9 +886,10 @@ out: return ret; } -static void nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr *fattr) +static unsigned long nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr *fattr) { struct nfs_inode *nfsi = NFS_I(inode); + unsigned long ret = 0; if ((fattr->valid & NFS_ATTR_FATTR_PRECHANGE) && (fattr->valid & NFS_ATTR_FATTR_CHANGE) @@ -891,25 +897,32 @@ static void nfs_wcc_update_inode(struct inode *inode, struct nfs_fattr *fattr) nfsi->change_attr = fattr->change_attr; if (S_ISDIR(inode->i_mode)) nfsi->cache_validity |= NFS_INO_INVALID_DATA; + ret |= NFS_INO_INVALID_ATTR; } /* If we have atomic WCC data, we may update some attributes */ if ((fattr->valid & NFS_ATTR_FATTR_PRECTIME) && (fattr->valid & NFS_ATTR_FATTR_CTIME) - && timespec_equal(&inode->i_ctime, &fattr->pre_ctime)) - memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime)); + && timespec_equal(&inode->i_ctime, &fattr->pre_ctime)) { + memcpy(&inode->i_ctime, &fattr->ctime, sizeof(inode->i_ctime)); + ret |= NFS_INO_INVALID_ATTR; + } if ((fattr->valid & NFS_ATTR_FATTR_PREMTIME) && (fattr->valid & NFS_ATTR_FATTR_MTIME) && timespec_equal(&inode->i_mtime, &fattr->pre_mtime)) { - memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime)); - if (S_ISDIR(inode->i_mode)) - nfsi->cache_validity |= NFS_INO_INVALID_DATA; + memcpy(&inode->i_mtime, &fattr->mtime, sizeof(inode->i_mtime)); + if (S_ISDIR(inode->i_mode)) + nfsi->cache_validity |= NFS_INO_INVALID_DATA; + ret |= NFS_INO_INVALID_ATTR; } if ((fattr->valid & NFS_ATTR_FATTR_PRESIZE) && (fattr->valid & NFS_ATTR_FATTR_SIZE) && i_size_read(inode) == nfs_size_to_loff_t(fattr->pre_size) - && nfsi->npages == 0) - i_size_write(inode, nfs_size_to_loff_t(fattr->size)); + && nfsi->npages == 0) { + i_size_write(inode, nfs_size_to_loff_t(fattr->size)); + ret |= NFS_INO_INVALID_ATTR; + } + return ret; } /** @@ -1223,7 +1236,7 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) | NFS_INO_REVAL_PAGECACHE); /* Do atomic weak cache consistency updates */ - nfs_wcc_update_inode(inode, fattr); + invalid |= nfs_wcc_update_inode(inode, fattr); /* More cache consistency checks */ if (fattr->valid & NFS_ATTR_FATTR_CHANGE) { @@ -1505,7 +1518,7 @@ static int nfsiod_start(void) { struct workqueue_struct *wq; dprintk("RPC: creating workqueue nfsiod\n"); - wq = alloc_workqueue("nfsiod", WQ_RESCUER, 0); + wq = alloc_workqueue("nfsiod", WQ_MEM_RECLAIM, 0); if (wq == NULL) return -ENOMEM; nfsiod_workqueue = wq; diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index 4644f04..72e0bdd 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -133,8 +133,7 @@ extern void nfs_put_client(struct nfs_client *); extern struct nfs_client *nfs4_find_client_no_ident(const struct sockaddr *); extern struct nfs_client *nfs4_find_client_ident(int); extern struct nfs_client * -nfs4_find_client_sessionid(const struct sockaddr *, struct nfs4_sessionid *, - int); +nfs4_find_client_sessionid(const struct sockaddr *, struct nfs4_sessionid *); extern struct nfs_server *nfs_create_server( const struct nfs_parsed_mount_data *, struct nfs_fh *); @@ -149,6 +148,9 @@ extern struct nfs_server *nfs_clone_server(struct nfs_server *, struct nfs_fattr *); extern void nfs_mark_client_ready(struct nfs_client *clp, int state); extern int nfs4_check_client_ready(struct nfs_client *clp); +extern struct nfs_client *nfs4_set_ds_client(struct nfs_client* mds_clp, + const struct sockaddr *ds_addr, + int ds_addrlen, int ds_proto); #ifdef CONFIG_PROC_FS extern int __init nfs_fs_proc_init(void); extern void nfs_fs_proc_exit(void); @@ -164,10 +166,10 @@ static inline void nfs_fs_proc_exit(void) /* nfs4namespace.c */ #ifdef CONFIG_NFS_V4 -extern struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry); +extern struct vfsmount *nfs_do_refmount(struct dentry *dentry); #else static inline -struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry) +struct vfsmount *nfs_do_refmount(struct dentry *dentry) { return ERR_PTR(-ENOENT); } @@ -214,8 +216,14 @@ extern const u32 nfs41_maxwrite_overhead; extern struct rpc_procinfo nfs4_procedures[]; #endif +extern int nfs4_init_ds_session(struct nfs_client *clp); + /* proc.c */ void nfs_close_context(struct nfs_open_context *ctx, int is_sync); +extern int nfs_init_client(struct nfs_client *clp, + const struct rpc_timeout *timeparms, + const char *ip_addr, rpc_authflavor_t authflavour, + int noresvport); /* dir.c */ extern int nfs_access_cache_shrinker(struct shrinker *shrink, @@ -248,24 +256,30 @@ extern void nfs_sb_active(struct super_block *sb); extern void nfs_sb_deactive(struct super_block *sb); /* namespace.c */ -extern char *nfs_path(const char *base, - const struct dentry *droot, - const struct dentry *dentry, +extern char *nfs_path(char **p, struct dentry *dentry, char *buffer, ssize_t buflen); extern struct vfsmount *nfs_d_automount(struct path *path); /* getroot.c */ -extern struct dentry *nfs_get_root(struct super_block *, struct nfs_fh *); +extern struct dentry *nfs_get_root(struct super_block *, struct nfs_fh *, + const char *); #ifdef CONFIG_NFS_V4 -extern struct dentry *nfs4_get_root(struct super_block *, struct nfs_fh *); +extern struct dentry *nfs4_get_root(struct super_block *, struct nfs_fh *, + const char *); extern int nfs4_get_rootfh(struct nfs_server *server, struct nfs_fh *mntfh); #endif /* read.c */ +extern int nfs_initiate_read(struct nfs_read_data *data, struct rpc_clnt *clnt, + const struct rpc_call_ops *call_ops); extern void nfs_read_prepare(struct rpc_task *task, void *calldata); /* write.c */ +extern int nfs_initiate_write(struct nfs_write_data *data, + struct rpc_clnt *clnt, + const struct rpc_call_ops *call_ops, + int how); extern void nfs_write_prepare(struct rpc_task *task, void *calldata); #ifdef CONFIG_MIGRATION extern int nfs_migrate_page(struct address_space *, @@ -275,6 +289,13 @@ extern int nfs_migrate_page(struct address_space *, #endif /* nfs4proc.c */ +extern void nfs4_reset_read(struct rpc_task *task, struct nfs_read_data *data); +extern int nfs4_init_client(struct nfs_client *clp, + const struct rpc_timeout *timeparms, + const char *ip_addr, + rpc_authflavor_t authflavour, + int noresvport); +extern void nfs4_reset_write(struct rpc_task *task, struct nfs_write_data *data); extern int _nfs4_call_sync(struct nfs_server *server, struct rpc_message *msg, struct nfs4_sequence_args *args, @@ -289,12 +310,11 @@ extern int _nfs4_call_sync_session(struct nfs_server *server, /* * Determine the device name as a string */ -static inline char *nfs_devname(const struct vfsmount *mnt_parent, - const struct dentry *dentry, +static inline char *nfs_devname(struct dentry *dentry, char *buffer, ssize_t buflen) { - return nfs_path(mnt_parent->mnt_devname, mnt_parent->mnt_root, - dentry, buffer, buflen); + char *dummy; + return nfs_path(&dummy, dentry, buffer, buflen); } /* diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c index f32b860..bf1c680 100644 --- a/fs/nfs/namespace.c +++ b/fs/nfs/namespace.c @@ -25,33 +25,30 @@ static LIST_HEAD(nfs_automount_list); static DECLARE_DELAYED_WORK(nfs_automount_task, nfs_expire_automounts); int nfs_mountpoint_expiry_timeout = 500 * HZ; -static struct vfsmount *nfs_do_submount(const struct vfsmount *mnt_parent, - const struct dentry *dentry, +static struct vfsmount *nfs_do_submount(struct dentry *dentry, struct nfs_fh *fh, struct nfs_fattr *fattr); /* * nfs_path - reconstruct the path given an arbitrary dentry - * @base - arbitrary string to prepend to the path - * @droot - pointer to root dentry for mountpoint + * @base - used to return pointer to the end of devname part of path * @dentry - pointer to dentry * @buffer - result buffer * @buflen - length of buffer * - * Helper function for constructing the path from the - * root dentry to an arbitrary hashed dentry. + * Helper function for constructing the server pathname + * by arbitrary hashed dentry. * * This is mainly for use in figuring out the path on the - * server side when automounting on top of an existing partition. + * server side when automounting on top of an existing partition + * and in generating /proc/mounts and friends. */ -char *nfs_path(const char *base, - const struct dentry *droot, - const struct dentry *dentry, - char *buffer, ssize_t buflen) +char *nfs_path(char **p, struct dentry *dentry, char *buffer, ssize_t buflen) { char *end; int namelen; unsigned seq; + const char *base; rename_retry: end = buffer+buflen; @@ -60,7 +57,10 @@ rename_retry: seq = read_seqbegin(&rename_lock); rcu_read_lock(); - while (!IS_ROOT(dentry) && dentry != droot) { + while (1) { + spin_lock(&dentry->d_lock); + if (IS_ROOT(dentry)) + break; namelen = dentry->d_name.len; buflen -= namelen + 1; if (buflen < 0) @@ -68,27 +68,47 @@ rename_retry: end -= namelen; memcpy(end, dentry->d_name.name, namelen); *--end = '/'; + spin_unlock(&dentry->d_lock); dentry = dentry->d_parent; } - rcu_read_unlock(); - if (read_seqretry(&rename_lock, seq)) + if (read_seqretry(&rename_lock, seq)) { + spin_unlock(&dentry->d_lock); + rcu_read_unlock(); goto rename_retry; + } if (*end != '/') { - if (--buflen < 0) + if (--buflen < 0) { + spin_unlock(&dentry->d_lock); + rcu_read_unlock(); goto Elong; + } *--end = '/'; } + *p = end; + base = dentry->d_fsdata; + if (!base) { + spin_unlock(&dentry->d_lock); + rcu_read_unlock(); + WARN_ON(1); + return end; + } namelen = strlen(base); /* Strip off excess slashes in base string */ while (namelen > 0 && base[namelen - 1] == '/') namelen--; buflen -= namelen; - if (buflen < 0) + if (buflen < 0) { + spin_unlock(&dentry->d_lock); + rcu_read_unlock(); goto Elong; + } end -= namelen; memcpy(end, base, namelen); + spin_unlock(&dentry->d_lock); + rcu_read_unlock(); return end; Elong_unlock: + spin_unlock(&dentry->d_lock); rcu_read_unlock(); if (read_seqretry(&rename_lock, seq)) goto rename_retry; @@ -143,9 +163,9 @@ struct vfsmount *nfs_d_automount(struct path *path) } if (fattr->valid & NFS_ATTR_FATTR_V4_REFERRAL) - mnt = nfs_do_refmount(path->mnt, path->dentry); + mnt = nfs_do_refmount(path->dentry); else - mnt = nfs_do_submount(path->mnt, path->dentry, fh, fattr); + mnt = nfs_do_submount(path->dentry, fh, fattr); if (IS_ERR(mnt)) goto out; @@ -209,19 +229,17 @@ static struct vfsmount *nfs_do_clone_mount(struct nfs_server *server, /** * nfs_do_submount - set up mountpoint when crossing a filesystem boundary - * @mnt_parent - mountpoint of parent directory * @dentry - parent directory * @fh - filehandle for new root dentry * @fattr - attributes for new root inode * */ -static struct vfsmount *nfs_do_submount(const struct vfsmount *mnt_parent, - const struct dentry *dentry, +static struct vfsmount *nfs_do_submount(struct dentry *dentry, struct nfs_fh *fh, struct nfs_fattr *fattr) { struct nfs_clone_mount mountdata = { - .sb = mnt_parent->mnt_sb, + .sb = dentry->d_sb, .dentry = dentry, .fh = fh, .fattr = fattr, @@ -237,11 +255,11 @@ static struct vfsmount *nfs_do_submount(const struct vfsmount *mnt_parent, dentry->d_name.name); if (page == NULL) goto out; - devname = nfs_devname(mnt_parent, dentry, page, PAGE_SIZE); + devname = nfs_devname(dentry, page, PAGE_SIZE); mnt = (struct vfsmount *)devname; if (IS_ERR(devname)) goto free_page; - mnt = nfs_do_clone_mount(NFS_SB(mnt_parent->mnt_sb), devname, &mountdata); + mnt = nfs_do_clone_mount(NFS_SB(dentry->d_sb), devname, &mountdata); free_page: free_page((unsigned long)page); out: diff --git a/fs/nfs/nfs3acl.c b/fs/nfs/nfs3acl.c index 9f88c5f..2743427 100644 --- a/fs/nfs/nfs3acl.c +++ b/fs/nfs/nfs3acl.c @@ -311,8 +311,8 @@ static int nfs3_proc_setacls(struct inode *inode, struct posix_acl *acl, if (!nfs_server_capable(inode, NFS_CAP_ACLS)) goto out; - /* We are doing this here, because XDR marshalling can only - return -ENOMEM. */ + /* We are doing this here because XDR marshalling does not + * return any results, it BUGs. */ status = -ENOSPC; if (acl != NULL && acl->a_count > NFS_ACL_MAX_ENTRIES) goto out; diff --git a/fs/nfs/nfs3proc.c b/fs/nfs/nfs3proc.c index ce939c0..d0c80d8 100644 --- a/fs/nfs/nfs3proc.c +++ b/fs/nfs/nfs3proc.c @@ -885,4 +885,5 @@ const struct nfs_rpc_ops nfs_v3_clientops = { .lock = nfs3_proc_lock, .clear_acl_cache = nfs3_forget_cached_acls, .close_context = nfs_close_context, + .init_client = nfs_init_client, }; diff --git a/fs/nfs/nfs3xdr.c b/fs/nfs/nfs3xdr.c index 01c5e8b..183c6b1 100644 --- a/fs/nfs/nfs3xdr.c +++ b/fs/nfs/nfs3xdr.c @@ -1328,10 +1328,13 @@ static void nfs3_xdr_enc_setacl3args(struct rpc_rqst *req, encode_nfs_fh3(xdr, NFS_FH(args->inode)); encode_uint32(xdr, args->mask); + + base = req->rq_slen; if (args->npages != 0) xdr_write_pages(xdr, args->pages, 0, args->len); + else + xdr_reserve_space(xdr, NFS_ACL_INLINE_BUFSIZE); - base = req->rq_slen; error = nfsacl_encode(xdr->buf, base, args->inode, (args->mask & NFS_ACL) ? args->acl_access : NULL, 1, 0); diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 7a74740..c64be1c 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -252,6 +252,9 @@ static inline struct nfs4_session *nfs4_get_session(const struct nfs_server *ser extern int nfs4_setup_sequence(const struct nfs_server *server, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, int cache_reply, struct rpc_task *task); +extern int nfs41_setup_sequence(struct nfs4_session *session, + struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, + int cache_reply, struct rpc_task *task); extern void nfs4_destroy_session(struct nfs4_session *session); extern struct nfs4_session *nfs4_alloc_session(struct nfs_client *clp); extern int nfs4_proc_create_session(struct nfs_client *); @@ -259,6 +262,19 @@ extern int nfs4_proc_destroy_session(struct nfs4_session *); extern int nfs4_init_session(struct nfs_server *server); extern int nfs4_proc_get_lease_time(struct nfs_client *clp, struct nfs_fsinfo *fsinfo); + +static inline bool +is_ds_only_client(struct nfs_client *clp) +{ + return (clp->cl_exchange_flags & EXCHGID4_FLAG_MASK_PNFS) == + EXCHGID4_FLAG_USE_PNFS_DS; +} + +static inline bool +is_ds_client(struct nfs_client *clp) +{ + return clp->cl_exchange_flags & EXCHGID4_FLAG_USE_PNFS_DS; +} #else /* CONFIG_NFS_v4_1 */ static inline struct nfs4_session *nfs4_get_session(const struct nfs_server *server) { @@ -276,6 +292,18 @@ static inline int nfs4_init_session(struct nfs_server *server) { return 0; } + +static inline bool +is_ds_only_client(struct nfs_client *clp) +{ + return false; +} + +static inline bool +is_ds_client(struct nfs_client *clp) +{ + return false; +} #endif /* CONFIG_NFS_V4_1 */ extern const struct nfs4_minor_version_ops *nfs_v4_minor_ops[]; @@ -298,6 +326,11 @@ struct rpc_cred *nfs4_get_renew_cred_locked(struct nfs_client *clp); #if defined(CONFIG_NFS_V4_1) struct rpc_cred *nfs4_get_machine_cred_locked(struct nfs_client *clp); struct rpc_cred *nfs4_get_exchange_id_cred(struct nfs_client *clp); +extern void nfs4_schedule_session_recovery(struct nfs4_session *); +#else +static inline void nfs4_schedule_session_recovery(struct nfs4_session *session) +{ +} #endif /* CONFIG_NFS_V4_1 */ extern struct nfs4_state_owner * nfs4_get_state_owner(struct nfs_server *, struct rpc_cred *); @@ -307,10 +340,9 @@ extern void nfs4_put_open_state(struct nfs4_state *); extern void nfs4_close_state(struct path *, struct nfs4_state *, fmode_t); extern void nfs4_close_sync(struct path *, struct nfs4_state *, fmode_t); extern void nfs4_state_set_mode_locked(struct nfs4_state *, fmode_t); -extern void nfs4_schedule_state_recovery(struct nfs_client *); +extern void nfs4_schedule_lease_recovery(struct nfs_client *); extern void nfs4_schedule_state_manager(struct nfs_client *); -extern int nfs4_state_mark_reclaim_nograce(struct nfs_client *clp, struct nfs4_state *state); -extern int nfs4_state_mark_reclaim_reboot(struct nfs_client *clp, struct nfs4_state *state); +extern void nfs4_schedule_stateid_recovery(const struct nfs_server *, struct nfs4_state *); extern void nfs41_handle_sequence_flag_errors(struct nfs_client *clp, u32 flags); extern void nfs41_handle_recall_slot(struct nfs_client *clp); extern void nfs4_put_lock_state(struct nfs4_lock_state *lsp); diff --git a/fs/nfs/nfs4filelayout.c b/fs/nfs/nfs4filelayout.c index 23f930c..4285584 100644 --- a/fs/nfs/nfs4filelayout.c +++ b/fs/nfs/nfs4filelayout.c @@ -40,32 +40,309 @@ MODULE_LICENSE("GPL"); MODULE_AUTHOR("Dean Hildebrand <dhildebz@umich.edu>"); MODULE_DESCRIPTION("The NFSv4 file layout driver"); -static int -filelayout_set_layoutdriver(struct nfs_server *nfss) +#define FILELAYOUT_POLL_RETRY_MAX (15*HZ) + +static loff_t +filelayout_get_dense_offset(struct nfs4_filelayout_segment *flseg, + loff_t offset) { - int status = pnfs_alloc_init_deviceid_cache(nfss->nfs_client, - nfs4_fl_free_deviceid_callback); - if (status) { - printk(KERN_WARNING "%s: deviceid cache could not be " - "initialized\n", __func__); - return status; + u32 stripe_width = flseg->stripe_unit * flseg->dsaddr->stripe_count; + u64 tmp; + + offset -= flseg->pattern_offset; + tmp = offset; + do_div(tmp, stripe_width); + + return tmp * flseg->stripe_unit + do_div(offset, flseg->stripe_unit); +} + +/* This function is used by the layout driver to calculate the + * offset of the file on the dserver based on whether the + * layout type is STRIPE_DENSE or STRIPE_SPARSE + */ +static loff_t +filelayout_get_dserver_offset(struct pnfs_layout_segment *lseg, loff_t offset) +{ + struct nfs4_filelayout_segment *flseg = FILELAYOUT_LSEG(lseg); + + switch (flseg->stripe_type) { + case STRIPE_SPARSE: + return offset; + + case STRIPE_DENSE: + return filelayout_get_dense_offset(flseg, offset); } - dprintk("%s: deviceid cache has been initialized successfully\n", - __func__); + + BUG(); +} + +/* For data server errors we don't recover from */ +static void +filelayout_set_lo_fail(struct pnfs_layout_segment *lseg) +{ + if (lseg->pls_range.iomode == IOMODE_RW) { + dprintk("%s Setting layout IOMODE_RW fail bit\n", __func__); + set_bit(lo_fail_bit(IOMODE_RW), &lseg->pls_layout->plh_flags); + } else { + dprintk("%s Setting layout IOMODE_READ fail bit\n", __func__); + set_bit(lo_fail_bit(IOMODE_READ), &lseg->pls_layout->plh_flags); + } +} + +static int filelayout_async_handle_error(struct rpc_task *task, + struct nfs4_state *state, + struct nfs_client *clp, + int *reset) +{ + if (task->tk_status >= 0) + return 0; + + *reset = 0; + + switch (task->tk_status) { + case -NFS4ERR_BADSESSION: + case -NFS4ERR_BADSLOT: + case -NFS4ERR_BAD_HIGH_SLOT: + case -NFS4ERR_DEADSESSION: + case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION: + case -NFS4ERR_SEQ_FALSE_RETRY: + case -NFS4ERR_SEQ_MISORDERED: + dprintk("%s ERROR %d, Reset session. Exchangeid " + "flags 0x%x\n", __func__, task->tk_status, + clp->cl_exchange_flags); + nfs4_schedule_session_recovery(clp->cl_session); + break; + case -NFS4ERR_DELAY: + case -NFS4ERR_GRACE: + case -EKEYEXPIRED: + rpc_delay(task, FILELAYOUT_POLL_RETRY_MAX); + break; + default: + dprintk("%s DS error. Retry through MDS %d\n", __func__, + task->tk_status); + *reset = 1; + break; + } + task->tk_status = 0; + return -EAGAIN; +} + +/* NFS_PROTO call done callback routines */ + +static int filelayout_read_done_cb(struct rpc_task *task, + struct nfs_read_data *data) +{ + struct nfs_client *clp = data->ds_clp; + int reset = 0; + + dprintk("%s DS read\n", __func__); + + if (filelayout_async_handle_error(task, data->args.context->state, + data->ds_clp, &reset) == -EAGAIN) { + dprintk("%s calling restart ds_clp %p ds_clp->cl_session %p\n", + __func__, data->ds_clp, data->ds_clp->cl_session); + if (reset) { + filelayout_set_lo_fail(data->lseg); + nfs4_reset_read(task, data); + clp = NFS_SERVER(data->inode)->nfs_client; + } + nfs_restart_rpc(task, clp); + return -EAGAIN; + } + return 0; } -/* Clear out the layout by destroying its device list */ -static int -filelayout_clear_layoutdriver(struct nfs_server *nfss) +/* + * Call ops for the async read/write cases + * In the case of dense layouts, the offset needs to be reset to its + * original value. + */ +static void filelayout_read_prepare(struct rpc_task *task, void *data) { - dprintk("--> %s\n", __func__); + struct nfs_read_data *rdata = (struct nfs_read_data *)data; + + rdata->read_done_cb = filelayout_read_done_cb; + + if (nfs41_setup_sequence(rdata->ds_clp->cl_session, + &rdata->args.seq_args, &rdata->res.seq_res, + 0, task)) + return; + + rpc_call_start(task); +} + +static void filelayout_read_call_done(struct rpc_task *task, void *data) +{ + struct nfs_read_data *rdata = (struct nfs_read_data *)data; + + dprintk("--> %s task->tk_status %d\n", __func__, task->tk_status); + + /* Note this may cause RPC to be resent */ + rdata->mds_ops->rpc_call_done(task, data); +} + +static void filelayout_read_release(void *data) +{ + struct nfs_read_data *rdata = (struct nfs_read_data *)data; + + rdata->mds_ops->rpc_release(data); +} + +static int filelayout_write_done_cb(struct rpc_task *task, + struct nfs_write_data *data) +{ + int reset = 0; + + if (filelayout_async_handle_error(task, data->args.context->state, + data->ds_clp, &reset) == -EAGAIN) { + struct nfs_client *clp; + + dprintk("%s calling restart ds_clp %p ds_clp->cl_session %p\n", + __func__, data->ds_clp, data->ds_clp->cl_session); + if (reset) { + filelayout_set_lo_fail(data->lseg); + nfs4_reset_write(task, data); + clp = NFS_SERVER(data->inode)->nfs_client; + } else + clp = data->ds_clp; + nfs_restart_rpc(task, clp); + return -EAGAIN; + } - if (nfss->nfs_client->cl_devid_cache) - pnfs_put_deviceid_cache(nfss->nfs_client); return 0; } +static void filelayout_write_prepare(struct rpc_task *task, void *data) +{ + struct nfs_write_data *wdata = (struct nfs_write_data *)data; + + if (nfs41_setup_sequence(wdata->ds_clp->cl_session, + &wdata->args.seq_args, &wdata->res.seq_res, + 0, task)) + return; + + rpc_call_start(task); +} + +static void filelayout_write_call_done(struct rpc_task *task, void *data) +{ + struct nfs_write_data *wdata = (struct nfs_write_data *)data; + + /* Note this may cause RPC to be resent */ + wdata->mds_ops->rpc_call_done(task, data); +} + +static void filelayout_write_release(void *data) +{ + struct nfs_write_data *wdata = (struct nfs_write_data *)data; + + wdata->mds_ops->rpc_release(data); +} + +struct rpc_call_ops filelayout_read_call_ops = { + .rpc_call_prepare = filelayout_read_prepare, + .rpc_call_done = filelayout_read_call_done, + .rpc_release = filelayout_read_release, +}; + +struct rpc_call_ops filelayout_write_call_ops = { + .rpc_call_prepare = filelayout_write_prepare, + .rpc_call_done = filelayout_write_call_done, + .rpc_release = filelayout_write_release, +}; + +static enum pnfs_try_status +filelayout_read_pagelist(struct nfs_read_data *data) +{ + struct pnfs_layout_segment *lseg = data->lseg; + struct nfs4_pnfs_ds *ds; + loff_t offset = data->args.offset; + u32 j, idx; + struct nfs_fh *fh; + int status; + + dprintk("--> %s ino %lu pgbase %u req %Zu@%llu\n", + __func__, data->inode->i_ino, + data->args.pgbase, (size_t)data->args.count, offset); + + /* Retrieve the correct rpc_client for the byte range */ + j = nfs4_fl_calc_j_index(lseg, offset); + idx = nfs4_fl_calc_ds_index(lseg, j); + ds = nfs4_fl_prepare_ds(lseg, idx); + if (!ds) { + /* Either layout fh index faulty, or ds connect failed */ + set_bit(lo_fail_bit(IOMODE_RW), &lseg->pls_layout->plh_flags); + set_bit(lo_fail_bit(IOMODE_READ), &lseg->pls_layout->plh_flags); + return PNFS_NOT_ATTEMPTED; + } + dprintk("%s USE DS:ip %x %hu\n", __func__, + ntohl(ds->ds_ip_addr), ntohs(ds->ds_port)); + + /* No multipath support. Use first DS */ + data->ds_clp = ds->ds_clp; + fh = nfs4_fl_select_ds_fh(lseg, j); + if (fh) + data->args.fh = fh; + + data->args.offset = filelayout_get_dserver_offset(lseg, offset); + data->mds_offset = offset; + + /* Perform an asynchronous read to ds */ + status = nfs_initiate_read(data, ds->ds_clp->cl_rpcclient, + &filelayout_read_call_ops); + BUG_ON(status != 0); + return PNFS_ATTEMPTED; +} + +/* Perform async writes. */ +static enum pnfs_try_status +filelayout_write_pagelist(struct nfs_write_data *data, int sync) +{ + struct pnfs_layout_segment *lseg = data->lseg; + struct nfs4_pnfs_ds *ds; + loff_t offset = data->args.offset; + u32 j, idx; + struct nfs_fh *fh; + int status; + + /* Retrieve the correct rpc_client for the byte range */ + j = nfs4_fl_calc_j_index(lseg, offset); + idx = nfs4_fl_calc_ds_index(lseg, j); + ds = nfs4_fl_prepare_ds(lseg, idx); + if (!ds) { + printk(KERN_ERR "%s: prepare_ds failed, use MDS\n", __func__); + set_bit(lo_fail_bit(IOMODE_RW), &lseg->pls_layout->plh_flags); + set_bit(lo_fail_bit(IOMODE_READ), &lseg->pls_layout->plh_flags); + return PNFS_NOT_ATTEMPTED; + } + dprintk("%s ino %lu sync %d req %Zu@%llu DS:%x:%hu\n", __func__, + data->inode->i_ino, sync, (size_t) data->args.count, offset, + ntohl(ds->ds_ip_addr), ntohs(ds->ds_port)); + + /* We can't handle commit to ds yet */ + if (!FILELAYOUT_LSEG(lseg)->commit_through_mds) + data->args.stable = NFS_FILE_SYNC; + + data->write_done_cb = filelayout_write_done_cb; + data->ds_clp = ds->ds_clp; + fh = nfs4_fl_select_ds_fh(lseg, j); + if (fh) + data->args.fh = fh; + /* + * Get the file offset on the dserver. Set the write offset to + * this offset and save the original offset. + */ + data->args.offset = filelayout_get_dserver_offset(lseg, offset); + data->mds_offset = offset; + + /* Perform an asynchronous write */ + status = nfs_initiate_write(data, ds->ds_clp->cl_rpcclient, + &filelayout_write_call_ops, sync); + BUG_ON(status != 0); + return PNFS_ATTEMPTED; +} + /* * filelayout_check_layout() * @@ -92,14 +369,14 @@ filelayout_check_layout(struct pnfs_layout_hdr *lo, goto out; } - if (fl->stripe_unit % PAGE_SIZE) { - dprintk("%s Stripe unit (%u) not page aligned\n", + if (!fl->stripe_unit || fl->stripe_unit % PAGE_SIZE) { + dprintk("%s Invalid stripe unit (%u)\n", __func__, fl->stripe_unit); goto out; } /* find and reference the deviceid */ - dsaddr = nfs4_fl_find_get_deviceid(nfss->nfs_client, id); + dsaddr = nfs4_fl_find_get_deviceid(id); if (dsaddr == NULL) { dsaddr = get_device_info(lo->plh_inode, id); if (dsaddr == NULL) @@ -134,7 +411,7 @@ out: dprintk("--> %s returns %d\n", __func__, status); return status; out_put: - pnfs_put_deviceid(nfss->nfs_client->cl_devid_cache, &dsaddr->deviceid); + nfs4_fl_put_deviceid(dsaddr); goto out; } @@ -243,23 +520,47 @@ filelayout_alloc_lseg(struct pnfs_layout_hdr *layoutid, static void filelayout_free_lseg(struct pnfs_layout_segment *lseg) { - struct nfs_server *nfss = NFS_SERVER(lseg->pls_layout->plh_inode); struct nfs4_filelayout_segment *fl = FILELAYOUT_LSEG(lseg); dprintk("--> %s\n", __func__); - pnfs_put_deviceid(nfss->nfs_client->cl_devid_cache, - &fl->dsaddr->deviceid); + nfs4_fl_put_deviceid(fl->dsaddr); _filelayout_free_lseg(fl); } +/* + * filelayout_pg_test(). Called by nfs_can_coalesce_requests() + * + * return 1 : coalesce page + * return 0 : don't coalesce page + */ +int +filelayout_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev, + struct nfs_page *req) +{ + u64 p_stripe, r_stripe; + u32 stripe_unit; + + if (!pgio->pg_lseg) + return 1; + p_stripe = (u64)prev->wb_index << PAGE_CACHE_SHIFT; + r_stripe = (u64)req->wb_index << PAGE_CACHE_SHIFT; + stripe_unit = FILELAYOUT_LSEG(pgio->pg_lseg)->stripe_unit; + + do_div(p_stripe, stripe_unit); + do_div(r_stripe, stripe_unit); + + return (p_stripe == r_stripe); +} + static struct pnfs_layoutdriver_type filelayout_type = { - .id = LAYOUT_NFSV4_1_FILES, - .name = "LAYOUT_NFSV4_1_FILES", - .owner = THIS_MODULE, - .set_layoutdriver = filelayout_set_layoutdriver, - .clear_layoutdriver = filelayout_clear_layoutdriver, - .alloc_lseg = filelayout_alloc_lseg, - .free_lseg = filelayout_free_lseg, + .id = LAYOUT_NFSV4_1_FILES, + .name = "LAYOUT_NFSV4_1_FILES", + .owner = THIS_MODULE, + .alloc_lseg = filelayout_alloc_lseg, + .free_lseg = filelayout_free_lseg, + .pg_test = filelayout_pg_test, + .read_pagelist = filelayout_read_pagelist, + .write_pagelist = filelayout_write_pagelist, }; static int __init nfs4filelayout_init(void) diff --git a/fs/nfs/nfs4filelayout.h b/fs/nfs/nfs4filelayout.h index bbf60dd..ee0c907 100644 --- a/fs/nfs/nfs4filelayout.h +++ b/fs/nfs/nfs4filelayout.h @@ -55,8 +55,14 @@ struct nfs4_pnfs_ds { atomic_t ds_count; }; +/* nfs4_file_layout_dsaddr flags */ +#define NFS4_DEVICE_ID_NEG_ENTRY 0x00000001 + struct nfs4_file_layout_dsaddr { - struct pnfs_deviceid_node deviceid; + struct hlist_node node; + struct nfs4_deviceid deviceid; + atomic_t ref; + unsigned long flags; u32 stripe_count; u8 *stripe_indices; u32 ds_num; @@ -83,11 +89,18 @@ FILELAYOUT_LSEG(struct pnfs_layout_segment *lseg) generic_hdr); } -extern void nfs4_fl_free_deviceid_callback(struct pnfs_deviceid_node *); +extern struct nfs_fh * +nfs4_fl_select_ds_fh(struct pnfs_layout_segment *lseg, u32 j); + extern void print_ds(struct nfs4_pnfs_ds *ds); extern void print_deviceid(struct nfs4_deviceid *dev_id); +u32 nfs4_fl_calc_j_index(struct pnfs_layout_segment *lseg, loff_t offset); +u32 nfs4_fl_calc_ds_index(struct pnfs_layout_segment *lseg, u32 j); +struct nfs4_pnfs_ds *nfs4_fl_prepare_ds(struct pnfs_layout_segment *lseg, + u32 ds_idx); extern struct nfs4_file_layout_dsaddr * -nfs4_fl_find_get_deviceid(struct nfs_client *, struct nfs4_deviceid *dev_id); +nfs4_fl_find_get_deviceid(struct nfs4_deviceid *dev_id); +extern void nfs4_fl_put_deviceid(struct nfs4_file_layout_dsaddr *dsaddr); struct nfs4_file_layout_dsaddr * get_device_info(struct inode *inode, struct nfs4_deviceid *dev_id); diff --git a/fs/nfs/nfs4filelayoutdev.c b/fs/nfs/nfs4filelayoutdev.c index 51fe64ac..68143c1 100644 --- a/fs/nfs/nfs4filelayoutdev.c +++ b/fs/nfs/nfs4filelayoutdev.c @@ -37,6 +37,30 @@ #define NFSDBG_FACILITY NFSDBG_PNFS_LD /* + * Device ID RCU cache. A device ID is unique per client ID and layout type. + */ +#define NFS4_FL_DEVICE_ID_HASH_BITS 5 +#define NFS4_FL_DEVICE_ID_HASH_SIZE (1 << NFS4_FL_DEVICE_ID_HASH_BITS) +#define NFS4_FL_DEVICE_ID_HASH_MASK (NFS4_FL_DEVICE_ID_HASH_SIZE - 1) + +static inline u32 +nfs4_fl_deviceid_hash(struct nfs4_deviceid *id) +{ + unsigned char *cptr = (unsigned char *)id->data; + unsigned int nbytes = NFS4_DEVICEID4_SIZE; + u32 x = 0; + + while (nbytes--) { + x *= 37; + x += *cptr++; + } + return x & NFS4_FL_DEVICE_ID_HASH_MASK; +} + +static struct hlist_head filelayout_deviceid_cache[NFS4_FL_DEVICE_ID_HASH_SIZE]; +static DEFINE_SPINLOCK(filelayout_deviceid_lock); + +/* * Data server cache * * Data servers can be mapped to different device ids. @@ -104,6 +128,67 @@ _data_server_lookup_locked(u32 ip_addr, u32 port) return NULL; } +/* + * Create an rpc connection to the nfs4_pnfs_ds data server + * Currently only support IPv4 + */ +static int +nfs4_ds_connect(struct nfs_server *mds_srv, struct nfs4_pnfs_ds *ds) +{ + struct nfs_client *clp; + struct sockaddr_in sin; + int status = 0; + + dprintk("--> %s ip:port %x:%hu au_flavor %d\n", __func__, + ntohl(ds->ds_ip_addr), ntohs(ds->ds_port), + mds_srv->nfs_client->cl_rpcclient->cl_auth->au_flavor); + + sin.sin_family = AF_INET; + sin.sin_addr.s_addr = ds->ds_ip_addr; + sin.sin_port = ds->ds_port; + + clp = nfs4_set_ds_client(mds_srv->nfs_client, (struct sockaddr *)&sin, + sizeof(sin), IPPROTO_TCP); + if (IS_ERR(clp)) { + status = PTR_ERR(clp); + goto out; + } + + if ((clp->cl_exchange_flags & EXCHGID4_FLAG_MASK_PNFS) != 0) { + if (!is_ds_client(clp)) { + status = -ENODEV; + goto out_put; + } + ds->ds_clp = clp; + dprintk("%s [existing] ip=%x, port=%hu\n", __func__, + ntohl(ds->ds_ip_addr), ntohs(ds->ds_port)); + goto out; + } + + /* + * Do not set NFS_CS_CHECK_LEASE_TIME instead set the DS lease to + * be equal to the MDS lease. Renewal is scheduled in create_session. + */ + spin_lock(&mds_srv->nfs_client->cl_lock); + clp->cl_lease_time = mds_srv->nfs_client->cl_lease_time; + spin_unlock(&mds_srv->nfs_client->cl_lock); + clp->cl_last_renewal = jiffies; + + /* New nfs_client */ + status = nfs4_init_ds_session(clp); + if (status) + goto out_put; + + ds->ds_clp = clp; + dprintk("%s [new] ip=%x, port=%hu\n", __func__, ntohl(ds->ds_ip_addr), + ntohs(ds->ds_port)); +out: + return status; +out_put: + nfs_put_client(clp); + goto out; +} + static void destroy_ds(struct nfs4_pnfs_ds *ds) { @@ -122,7 +207,7 @@ nfs4_fl_free_deviceid(struct nfs4_file_layout_dsaddr *dsaddr) struct nfs4_pnfs_ds *ds; int i; - print_deviceid(&dsaddr->deviceid.de_id); + print_deviceid(&dsaddr->deviceid); for (i = 0; i < dsaddr->ds_num; i++) { ds = dsaddr->ds_list[i]; @@ -139,15 +224,6 @@ nfs4_fl_free_deviceid(struct nfs4_file_layout_dsaddr *dsaddr) kfree(dsaddr); } -void -nfs4_fl_free_deviceid_callback(struct pnfs_deviceid_node *device) -{ - struct nfs4_file_layout_dsaddr *dsaddr = - container_of(device, struct nfs4_file_layout_dsaddr, deviceid); - - nfs4_fl_free_deviceid(dsaddr); -} - static struct nfs4_pnfs_ds * nfs4_pnfs_ds_add(struct inode *inode, u32 ip_addr, u32 port) { @@ -214,17 +290,26 @@ decode_and_add_ds(__be32 **pp, struct inode *inode) /* ipv6 length plus port is legal */ if (rlen > INET6_ADDRSTRLEN + 8) { - dprintk("%s Invalid address, length %d\n", __func__, + dprintk("%s: Invalid address, length %d\n", __func__, rlen); goto out_err; } buf = kmalloc(rlen + 1, GFP_KERNEL); + if (!buf) { + dprintk("%s: Not enough memory\n", __func__); + goto out_err; + } buf[rlen] = '\0'; memcpy(buf, r_addr, rlen); /* replace the port dots with dashes for the in4_pton() delimiter*/ for (i = 0; i < 2; i++) { char *res = strrchr(buf, '.'); + if (!res) { + dprintk("%s: Failed finding expected dots in port\n", + __func__); + goto out_free; + } *res = '-'; } @@ -240,7 +325,7 @@ decode_and_add_ds(__be32 **pp, struct inode *inode) port = htons((tmp[0] << 8) | (tmp[1])); ds = nfs4_pnfs_ds_add(inode, ip_addr, port); - dprintk("%s Decoded address and port %s\n", __func__, buf); + dprintk("%s: Decoded address and port %s\n", __func__, buf); out_free: kfree(buf); out_err: @@ -291,7 +376,7 @@ decode_device(struct inode *ino, struct pnfs_device *pdev) dsaddr->stripe_count = cnt; dsaddr->ds_num = num; - memcpy(&dsaddr->deviceid.de_id, &pdev->dev_id, sizeof(pdev->dev_id)); + memcpy(&dsaddr->deviceid, &pdev->dev_id, sizeof(pdev->dev_id)); /* Go back an read stripe indices */ p = indicesp; @@ -341,28 +426,37 @@ out_err: } /* - * Decode the opaque device specified in 'dev' - * and add it to the list of available devices. - * If the deviceid is already cached, nfs4_add_deviceid will return - * a pointer to the cached struct and throw away the new. + * Decode the opaque device specified in 'dev' and add it to the cache of + * available devices. */ -static struct nfs4_file_layout_dsaddr* +static struct nfs4_file_layout_dsaddr * decode_and_add_device(struct inode *inode, struct pnfs_device *dev) { - struct nfs4_file_layout_dsaddr *dsaddr; - struct pnfs_deviceid_node *d; + struct nfs4_file_layout_dsaddr *d, *new; + long hash; - dsaddr = decode_device(inode, dev); - if (!dsaddr) { + new = decode_device(inode, dev); + if (!new) { printk(KERN_WARNING "%s: Could not decode or add device\n", __func__); return NULL; } - d = pnfs_add_deviceid(NFS_SERVER(inode)->nfs_client->cl_devid_cache, - &dsaddr->deviceid); + spin_lock(&filelayout_deviceid_lock); + d = nfs4_fl_find_get_deviceid(&new->deviceid); + if (d) { + spin_unlock(&filelayout_deviceid_lock); + nfs4_fl_free_deviceid(new); + return d; + } - return container_of(d, struct nfs4_file_layout_dsaddr, deviceid); + INIT_HLIST_NODE(&new->node); + atomic_set(&new->ref, 1); + hash = nfs4_fl_deviceid_hash(&new->deviceid); + hlist_add_head_rcu(&new->node, &filelayout_deviceid_cache[hash]); + spin_unlock(&filelayout_deviceid_lock); + + return new; } /* @@ -437,12 +531,123 @@ out_free: return dsaddr; } +void +nfs4_fl_put_deviceid(struct nfs4_file_layout_dsaddr *dsaddr) +{ + if (atomic_dec_and_lock(&dsaddr->ref, &filelayout_deviceid_lock)) { + hlist_del_rcu(&dsaddr->node); + spin_unlock(&filelayout_deviceid_lock); + + synchronize_rcu(); + nfs4_fl_free_deviceid(dsaddr); + } +} + struct nfs4_file_layout_dsaddr * -nfs4_fl_find_get_deviceid(struct nfs_client *clp, struct nfs4_deviceid *id) +nfs4_fl_find_get_deviceid(struct nfs4_deviceid *id) +{ + struct nfs4_file_layout_dsaddr *d; + struct hlist_node *n; + long hash = nfs4_fl_deviceid_hash(id); + + + rcu_read_lock(); + hlist_for_each_entry_rcu(d, n, &filelayout_deviceid_cache[hash], node) { + if (!memcmp(&d->deviceid, id, sizeof(*id))) { + if (!atomic_inc_not_zero(&d->ref)) + goto fail; + rcu_read_unlock(); + return d; + } + } +fail: + rcu_read_unlock(); + return NULL; +} + +/* + * Want res = (offset - layout->pattern_offset)/ layout->stripe_unit + * Then: ((res + fsi) % dsaddr->stripe_count) + */ +u32 +nfs4_fl_calc_j_index(struct pnfs_layout_segment *lseg, loff_t offset) +{ + struct nfs4_filelayout_segment *flseg = FILELAYOUT_LSEG(lseg); + u64 tmp; + + tmp = offset - flseg->pattern_offset; + do_div(tmp, flseg->stripe_unit); + tmp += flseg->first_stripe_index; + return do_div(tmp, flseg->dsaddr->stripe_count); +} + +u32 +nfs4_fl_calc_ds_index(struct pnfs_layout_segment *lseg, u32 j) +{ + return FILELAYOUT_LSEG(lseg)->dsaddr->stripe_indices[j]; +} + +struct nfs_fh * +nfs4_fl_select_ds_fh(struct pnfs_layout_segment *lseg, u32 j) +{ + struct nfs4_filelayout_segment *flseg = FILELAYOUT_LSEG(lseg); + u32 i; + + if (flseg->stripe_type == STRIPE_SPARSE) { + if (flseg->num_fh == 1) + i = 0; + else if (flseg->num_fh == 0) + /* Use the MDS OPEN fh set in nfs_read_rpcsetup */ + return NULL; + else + i = nfs4_fl_calc_ds_index(lseg, j); + } else + i = j; + return flseg->fh_array[i]; +} + +static void +filelayout_mark_devid_negative(struct nfs4_file_layout_dsaddr *dsaddr, + int err, u32 ds_addr) +{ + u32 *p = (u32 *)&dsaddr->deviceid; + + printk(KERN_ERR "NFS: data server %x connection error %d." + " Deviceid [%x%x%x%x] marked out of use.\n", + ds_addr, err, p[0], p[1], p[2], p[3]); + + spin_lock(&filelayout_deviceid_lock); + dsaddr->flags |= NFS4_DEVICE_ID_NEG_ENTRY; + spin_unlock(&filelayout_deviceid_lock); +} + +struct nfs4_pnfs_ds * +nfs4_fl_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx) { - struct pnfs_deviceid_node *d; + struct nfs4_file_layout_dsaddr *dsaddr = FILELAYOUT_LSEG(lseg)->dsaddr; + struct nfs4_pnfs_ds *ds = dsaddr->ds_list[ds_idx]; + + if (ds == NULL) { + printk(KERN_ERR "%s: No data server for offset index %d\n", + __func__, ds_idx); + return NULL; + } - d = pnfs_find_get_deviceid(clp->cl_devid_cache, id); - return (d == NULL) ? NULL : - container_of(d, struct nfs4_file_layout_dsaddr, deviceid); + if (!ds->ds_clp) { + struct nfs_server *s = NFS_SERVER(lseg->pls_layout->plh_inode); + int err; + + if (dsaddr->flags & NFS4_DEVICE_ID_NEG_ENTRY) { + /* Already tried to connect, don't try again */ + dprintk("%s Deviceid marked out of use\n", __func__); + return NULL; + } + err = nfs4_ds_connect(s, ds); + if (err) { + filelayout_mark_devid_negative(dsaddr, err, + ntohl(ds->ds_ip_addr)); + return NULL; + } + } + return ds; } diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c index 3c2a172..bb80c49 100644 --- a/fs/nfs/nfs4namespace.c +++ b/fs/nfs/nfs4namespace.c @@ -54,33 +54,29 @@ Elong: /* * Determine the mount path as a string */ -static char *nfs4_path(const struct vfsmount *mnt_parent, - const struct dentry *dentry, - char *buffer, ssize_t buflen) +static char *nfs4_path(struct dentry *dentry, char *buffer, ssize_t buflen) { - const char *srvpath; - - srvpath = strchr(mnt_parent->mnt_devname, ':'); - if (srvpath) - srvpath++; - else - srvpath = mnt_parent->mnt_devname; - - return nfs_path(srvpath, mnt_parent->mnt_root, dentry, buffer, buflen); + char *limit; + char *path = nfs_path(&limit, dentry, buffer, buflen); + if (!IS_ERR(path)) { + char *colon = strchr(path, ':'); + if (colon && colon < limit) + path = colon + 1; + } + return path; } /* * Check that fs_locations::fs_root [RFC3530 6.3] is a prefix for what we * believe to be the server path to this dentry */ -static int nfs4_validate_fspath(const struct vfsmount *mnt_parent, - const struct dentry *dentry, +static int nfs4_validate_fspath(struct dentry *dentry, const struct nfs4_fs_locations *locations, char *page, char *page2) { const char *path, *fs_path; - path = nfs4_path(mnt_parent, dentry, page, PAGE_SIZE); + path = nfs4_path(dentry, page, PAGE_SIZE); if (IS_ERR(path)) return PTR_ERR(path); @@ -165,20 +161,18 @@ static struct vfsmount *try_location(struct nfs_clone_mount *mountdata, /** * nfs_follow_referral - set up mountpoint when hitting a referral on moved error - * @mnt_parent - mountpoint of parent directory * @dentry - parent directory * @locations - array of NFSv4 server location information * */ -static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent, - const struct dentry *dentry, +static struct vfsmount *nfs_follow_referral(struct dentry *dentry, const struct nfs4_fs_locations *locations) { struct vfsmount *mnt = ERR_PTR(-ENOENT); struct nfs_clone_mount mountdata = { - .sb = mnt_parent->mnt_sb, + .sb = dentry->d_sb, .dentry = dentry, - .authflavor = NFS_SB(mnt_parent->mnt_sb)->client->cl_auth->au_flavor, + .authflavor = NFS_SB(dentry->d_sb)->client->cl_auth->au_flavor, }; char *page = NULL, *page2 = NULL; int loc, error; @@ -198,7 +192,7 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent, goto out; /* Ensure fs path is a prefix of current dentry path */ - error = nfs4_validate_fspath(mnt_parent, dentry, locations, page, page2); + error = nfs4_validate_fspath(dentry, locations, page, page2); if (error < 0) { mnt = ERR_PTR(error); goto out; @@ -225,11 +219,10 @@ out: /* * nfs_do_refmount - handle crossing a referral on server - * @mnt_parent - mountpoint of referral * @dentry - dentry of referral * */ -struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry) +struct vfsmount *nfs_do_refmount(struct dentry *dentry) { struct vfsmount *mnt = ERR_PTR(-ENOMEM); struct dentry *parent; @@ -262,7 +255,7 @@ struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentr fs_locations->fs_path.ncomponents <= 0) goto out_free; - mnt = nfs_follow_referral(mnt_parent, dentry, fs_locations); + mnt = nfs_follow_referral(dentry, fs_locations); out_free: __free_page(page); kfree(fs_locations); diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index 9d992b0..1d84e70 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -50,6 +50,7 @@ #include <linux/module.h> #include <linux/sunrpc/bc_xprt.h> #include <linux/xattr.h> +#include <linux/utsname.h> #include "nfs4_fs.h" #include "delegation.h" @@ -84,6 +85,9 @@ static int nfs4_map_errors(int err) switch (err) { case -NFS4ERR_RESOURCE: return -EREMOTEIO; + case -NFS4ERR_BADOWNER: + case -NFS4ERR_BADNAME: + return -EINVAL; default: dprintk("%s could not handle NFSv4 error %d\n", __func__, -err); @@ -240,7 +244,7 @@ static int nfs4_delay(struct rpc_clnt *clnt, long *timeout) /* This is the error handling routine for processes that are allowed * to sleep. */ -static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, struct nfs4_exception *exception) +static int nfs4_handle_exception(struct nfs_server *server, int errorcode, struct nfs4_exception *exception) { struct nfs_client *clp = server->nfs_client; struct nfs4_state *state = exception->state; @@ -255,12 +259,13 @@ static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, case -NFS4ERR_OPENMODE: if (state == NULL) break; - nfs4_state_mark_reclaim_nograce(clp, state); - goto do_state_recovery; + nfs4_schedule_stateid_recovery(server, state); + goto wait_on_recovery; case -NFS4ERR_STALE_STATEID: case -NFS4ERR_STALE_CLIENTID: case -NFS4ERR_EXPIRED: - goto do_state_recovery; + nfs4_schedule_lease_recovery(clp); + goto wait_on_recovery; #if defined(CONFIG_NFS_V4_1) case -NFS4ERR_BADSESSION: case -NFS4ERR_BADSLOT: @@ -271,7 +276,7 @@ static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, case -NFS4ERR_SEQ_MISORDERED: dprintk("%s ERROR: %d Reset session\n", __func__, errorcode); - nfs4_schedule_state_recovery(clp); + nfs4_schedule_session_recovery(clp->cl_session); exception->retry = 1; break; #endif /* defined(CONFIG_NFS_V4_1) */ @@ -291,11 +296,23 @@ static int nfs4_handle_exception(const struct nfs_server *server, int errorcode, break; case -NFS4ERR_OLD_STATEID: exception->retry = 1; + break; + case -NFS4ERR_BADOWNER: + /* The following works around a Linux server bug! */ + case -NFS4ERR_BADNAME: + if (server->caps & NFS_CAP_UIDGID_NOMAP) { + server->caps &= ~NFS_CAP_UIDGID_NOMAP; + exception->retry = 1; + printk(KERN_WARNING "NFS: v4 server %s " + "does not accept raw " + "uid/gids. " + "Reenabling the idmapper.\n", + server->nfs_client->cl_hostname); + } } /* We failed to handle the error */ return nfs4_map_errors(ret); -do_state_recovery: - nfs4_schedule_state_recovery(clp); +wait_on_recovery: ret = nfs4_wait_clnt_recover(clp); if (ret == 0) exception->retry = 1; @@ -434,8 +451,8 @@ static int nfs41_sequence_done(struct rpc_task *task, struct nfs4_sequence_res * clp = res->sr_session->clp; do_renew_lease(clp, timestamp); /* Check sequence flags */ - if (atomic_read(&clp->cl_count) > 1) - nfs41_handle_sequence_flag_errors(clp, res->sr_status_flags); + if (res->sr_status_flags != 0) + nfs4_schedule_lease_recovery(clp); break; case -NFS4ERR_DELAY: /* The server detected a resend of the RPC call and @@ -504,7 +521,7 @@ out: return ret_id; } -static int nfs41_setup_sequence(struct nfs4_session *session, +int nfs41_setup_sequence(struct nfs4_session *session, struct nfs4_sequence_args *args, struct nfs4_sequence_res *res, int cache_reply, @@ -570,6 +587,7 @@ static int nfs41_setup_sequence(struct nfs4_session *session, res->sr_status = 1; return 0; } +EXPORT_SYMBOL_GPL(nfs41_setup_sequence); int nfs4_setup_sequence(const struct nfs_server *server, struct nfs4_sequence_args *args, @@ -1254,14 +1272,13 @@ int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state case -NFS4ERR_BAD_HIGH_SLOT: case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION: case -NFS4ERR_DEADSESSION: - nfs4_schedule_state_recovery( - server->nfs_client); + nfs4_schedule_session_recovery(server->nfs_client->cl_session); goto out; case -NFS4ERR_STALE_CLIENTID: case -NFS4ERR_STALE_STATEID: case -NFS4ERR_EXPIRED: /* Don't recall a delegation if it was lost */ - nfs4_schedule_state_recovery(server->nfs_client); + nfs4_schedule_lease_recovery(server->nfs_client); goto out; case -ERESTARTSYS: /* @@ -1270,7 +1287,7 @@ int nfs4_open_delegation_recall(struct nfs_open_context *ctx, struct nfs4_state */ case -NFS4ERR_ADMIN_REVOKED: case -NFS4ERR_BAD_STATEID: - nfs4_state_mark_reclaim_nograce(server->nfs_client, state); + nfs4_schedule_stateid_recovery(server, state); case -EKEYEXPIRED: /* * User RPCSEC_GSS context has expired. @@ -1573,9 +1590,8 @@ static int _nfs4_proc_open(struct nfs4_opendata *data) return 0; } -static int nfs4_recover_expired_lease(struct nfs_server *server) +static int nfs4_client_recover_expired_lease(struct nfs_client *clp) { - struct nfs_client *clp = server->nfs_client; unsigned int loop; int ret; @@ -1586,12 +1602,17 @@ static int nfs4_recover_expired_lease(struct nfs_server *server) if (!test_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) && !test_bit(NFS4CLNT_CHECK_LEASE,&clp->cl_state)) break; - nfs4_schedule_state_recovery(clp); + nfs4_schedule_state_manager(clp); ret = -EIO; } return ret; } +static int nfs4_recover_expired_lease(struct nfs_server *server) +{ + return nfs4_client_recover_expired_lease(server->nfs_client); +} + /* * OPEN_EXPIRED: * reclaim state on the server after a network partition. @@ -3069,15 +3090,10 @@ static int nfs4_proc_pathconf(struct nfs_server *server, struct nfs_fh *fhandle, return err; } -static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data) +static int nfs4_read_done_cb(struct rpc_task *task, struct nfs_read_data *data) { struct nfs_server *server = NFS_SERVER(data->inode); - dprintk("--> %s\n", __func__); - - if (!nfs4_sequence_done(task, &data->res.seq_res)) - return -EAGAIN; - if (nfs4_async_handle_error(task, server, data->args.context->state) == -EAGAIN) { nfs_restart_rpc(task, server->nfs_client); return -EAGAIN; @@ -3089,19 +3105,44 @@ static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data) return 0; } +static int nfs4_read_done(struct rpc_task *task, struct nfs_read_data *data) +{ + + dprintk("--> %s\n", __func__); + + if (!nfs4_sequence_done(task, &data->res.seq_res)) + return -EAGAIN; + + return data->read_done_cb(task, data); +} + static void nfs4_proc_read_setup(struct nfs_read_data *data, struct rpc_message *msg) { data->timestamp = jiffies; + data->read_done_cb = nfs4_read_done_cb; msg->rpc_proc = &nfs4_procedures[NFSPROC4_CLNT_READ]; } -static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data) +/* Reset the the nfs_read_data to send the read to the MDS. */ +void nfs4_reset_read(struct rpc_task *task, struct nfs_read_data *data) +{ + dprintk("%s Reset task for i/o through\n", __func__); + put_lseg(data->lseg); + data->lseg = NULL; + /* offsets will differ in the dense stripe case */ + data->args.offset = data->mds_offset; + data->ds_clp = NULL; + data->args.fh = NFS_FH(data->inode); + data->read_done_cb = nfs4_read_done_cb; + task->tk_ops = data->mds_ops; + rpc_task_reset_client(task, NFS_CLIENT(data->inode)); +} +EXPORT_SYMBOL_GPL(nfs4_reset_read); + +static int nfs4_write_done_cb(struct rpc_task *task, struct nfs_write_data *data) { struct inode *inode = data->inode; - if (!nfs4_sequence_done(task, &data->res.seq_res)) - return -EAGAIN; - if (nfs4_async_handle_error(task, NFS_SERVER(inode), data->args.context->state) == -EAGAIN) { nfs_restart_rpc(task, NFS_SERVER(inode)->nfs_client); return -EAGAIN; @@ -3113,11 +3154,41 @@ static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data) return 0; } +static int nfs4_write_done(struct rpc_task *task, struct nfs_write_data *data) +{ + if (!nfs4_sequence_done(task, &data->res.seq_res)) + return -EAGAIN; + return data->write_done_cb(task, data); +} + +/* Reset the the nfs_write_data to send the write to the MDS. */ +void nfs4_reset_write(struct rpc_task *task, struct nfs_write_data *data) +{ + dprintk("%s Reset task for i/o through\n", __func__); + put_lseg(data->lseg); + data->lseg = NULL; + data->ds_clp = NULL; + data->write_done_cb = nfs4_write_done_cb; + data->args.fh = NFS_FH(data->inode); + data->args.bitmask = data->res.server->cache_consistency_bitmask; + data->args.offset = data->mds_offset; + data->res.fattr = &data->fattr; + task->tk_ops = data->mds_ops; + rpc_task_reset_client(task, NFS_CLIENT(data->inode)); +} +EXPORT_SYMBOL_GPL(nfs4_reset_write); + static void nfs4_proc_write_setup(struct nfs_write_data *data, struct rpc_message *msg) { struct nfs_server *server = NFS_SERVER(data->inode); - data->args.bitmask = server->cache_consistency_bitmask; + if (data->lseg) { + data->args.bitmask = NULL; + data->res.fattr = NULL; + } else + data->args.bitmask = server->cache_consistency_bitmask; + if (!data->write_done_cb) + data->write_done_cb = nfs4_write_done_cb; data->res.server = server; data->timestamp = jiffies; @@ -3177,7 +3248,7 @@ static void nfs4_renew_done(struct rpc_task *task, void *calldata) if (task->tk_status < 0) { /* Unless we're shutting down, schedule state recovery! */ if (test_bit(NFS_CS_RENEWD, &clp->cl_res_state) != 0) - nfs4_schedule_state_recovery(clp); + nfs4_schedule_lease_recovery(clp); return; } do_renew_lease(clp, timestamp); @@ -3251,6 +3322,35 @@ static void buf_to_pages(const void *buf, size_t buflen, } } +static int buf_to_pages_noslab(const void *buf, size_t buflen, + struct page **pages, unsigned int *pgbase) +{ + struct page *newpage, **spages; + int rc = 0; + size_t len; + spages = pages; + + do { + len = min_t(size_t, PAGE_CACHE_SIZE, buflen); + newpage = alloc_page(GFP_KERNEL); + + if (newpage == NULL) + goto unwind; + memcpy(page_address(newpage), buf, len); + buf += len; + buflen -= len; + *pages++ = newpage; + rc++; + } while (buflen != 0); + + return rc; + +unwind: + for(; rc > 0; rc--) + __free_page(spages[rc-1]); + return -ENOMEM; +} + struct nfs4_cached_acl { int cached; size_t len; @@ -3419,13 +3519,23 @@ static int __nfs4_proc_set_acl(struct inode *inode, const void *buf, size_t bufl .rpc_argp = &arg, .rpc_resp = &res, }; - int ret; + int ret, i; if (!nfs4_server_supports_acls(server)) return -EOPNOTSUPP; + i = buf_to_pages_noslab(buf, buflen, arg.acl_pages, &arg.acl_pgbase); + if (i < 0) + return i; nfs_inode_return_delegation(inode); - buf_to_pages(buf, buflen, arg.acl_pages, &arg.acl_pgbase); ret = nfs4_call_sync(server, &msg, &arg, &res, 1); + + /* + * Free each page after tx, so the only ref left is + * held by the network stack + */ + for (; i > 0; i--) + put_page(pages[i-1]); + /* * Acl update can result in inode attribute update. * so mark the attribute cache invalid. @@ -3463,12 +3573,13 @@ nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, case -NFS4ERR_OPENMODE: if (state == NULL) break; - nfs4_state_mark_reclaim_nograce(clp, state); - goto do_state_recovery; + nfs4_schedule_stateid_recovery(server, state); + goto wait_on_recovery; case -NFS4ERR_STALE_STATEID: case -NFS4ERR_STALE_CLIENTID: case -NFS4ERR_EXPIRED: - goto do_state_recovery; + nfs4_schedule_lease_recovery(clp); + goto wait_on_recovery; #if defined(CONFIG_NFS_V4_1) case -NFS4ERR_BADSESSION: case -NFS4ERR_BADSLOT: @@ -3479,7 +3590,7 @@ nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, case -NFS4ERR_SEQ_MISORDERED: dprintk("%s ERROR %d, Reset session\n", __func__, task->tk_status); - nfs4_schedule_state_recovery(clp); + nfs4_schedule_session_recovery(clp->cl_session); task->tk_status = 0; return -EAGAIN; #endif /* CONFIG_NFS_V4_1 */ @@ -3496,9 +3607,8 @@ nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, } task->tk_status = nfs4_map_errors(task->tk_status); return 0; -do_state_recovery: +wait_on_recovery: rpc_sleep_on(&clp->cl_rpcwaitq, task, NULL); - nfs4_schedule_state_recovery(clp); if (test_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state) == 0) rpc_wake_up_queued_task(&clp->cl_rpcwaitq, task); task->tk_status = 0; @@ -4109,7 +4219,7 @@ static void nfs4_lock_release(void *calldata) task = nfs4_do_unlck(&data->fl, data->ctx, data->lsp, data->arg.lock_seqid); if (!IS_ERR(task)) - rpc_put_task(task); + rpc_put_task_async(task); dprintk("%s: cancelling lock!\n", __func__); } else nfs_free_seqid(data->arg.lock_seqid); @@ -4133,23 +4243,18 @@ static const struct rpc_call_ops nfs4_recover_lock_ops = { static void nfs4_handle_setlk_error(struct nfs_server *server, struct nfs4_lock_state *lsp, int new_lock_owner, int error) { - struct nfs_client *clp = server->nfs_client; - struct nfs4_state *state = lsp->ls_state; - switch (error) { case -NFS4ERR_ADMIN_REVOKED: case -NFS4ERR_BAD_STATEID: - case -NFS4ERR_EXPIRED: + lsp->ls_seqid.flags &= ~NFS_SEQID_CONFIRMED; if (new_lock_owner != 0 || (lsp->ls_flags & NFS_LOCK_INITIALIZED) != 0) - nfs4_state_mark_reclaim_nograce(clp, state); - lsp->ls_seqid.flags &= ~NFS_SEQID_CONFIRMED; + nfs4_schedule_stateid_recovery(server, lsp->ls_state); break; case -NFS4ERR_STALE_STATEID: - if (new_lock_owner != 0 || - (lsp->ls_flags & NFS_LOCK_INITIALIZED) != 0) - nfs4_state_mark_reclaim_reboot(clp, state); lsp->ls_seqid.flags &= ~NFS_SEQID_CONFIRMED; + case -NFS4ERR_EXPIRED: + nfs4_schedule_lease_recovery(server->nfs_client); }; } @@ -4365,12 +4470,14 @@ int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl) case -NFS4ERR_EXPIRED: case -NFS4ERR_STALE_CLIENTID: case -NFS4ERR_STALE_STATEID: + nfs4_schedule_lease_recovery(server->nfs_client); + goto out; case -NFS4ERR_BADSESSION: case -NFS4ERR_BADSLOT: case -NFS4ERR_BAD_HIGH_SLOT: case -NFS4ERR_CONN_NOT_BOUND_TO_SESSION: case -NFS4ERR_DEADSESSION: - nfs4_schedule_state_recovery(server->nfs_client); + nfs4_schedule_session_recovery(server->nfs_client->cl_session); goto out; case -ERESTARTSYS: /* @@ -4380,7 +4487,7 @@ int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl) case -NFS4ERR_ADMIN_REVOKED: case -NFS4ERR_BAD_STATEID: case -NFS4ERR_OPENMODE: - nfs4_state_mark_reclaim_nograce(server->nfs_client, state); + nfs4_schedule_stateid_recovery(server, state); err = 0; goto out; case -EKEYEXPIRED: @@ -4572,27 +4679,16 @@ int nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred) *p = htonl((u32)clp->cl_boot_time.tv_nsec); args.verifier = &verifier; - while (1) { - args.id_len = scnprintf(args.id, sizeof(args.id), - "%s/%s %u", - clp->cl_ipaddr, - rpc_peeraddr2str(clp->cl_rpcclient, - RPC_DISPLAY_ADDR), - clp->cl_id_uniquifier); - - status = rpc_call_sync(clp->cl_rpcclient, &msg, 0); - - if (status != -NFS4ERR_CLID_INUSE) - break; - - if (signalled()) - break; - - if (++clp->cl_id_uniquifier == 0) - break; - } + args.id_len = scnprintf(args.id, sizeof(args.id), + "%s/%s.%s/%u", + clp->cl_ipaddr, + init_utsname()->nodename, + init_utsname()->domainname, + clp->cl_rpcclient->cl_auth->au_flavor); - status = nfs4_check_cl_exchange_flags(clp->cl_exchange_flags); + status = rpc_call_sync(clp->cl_rpcclient, &msg, 0); + if (!status) + status = nfs4_check_cl_exchange_flags(clp->cl_exchange_flags); dprintk("<-- %s status= %d\n", __func__, status); return status; } @@ -4998,10 +5094,20 @@ int nfs4_proc_create_session(struct nfs_client *clp) int status; unsigned *ptr; struct nfs4_session *session = clp->cl_session; + long timeout = 0; + int err; dprintk("--> %s clp=%p session=%p\n", __func__, clp, session); - status = _nfs4_proc_create_session(clp); + do { + status = _nfs4_proc_create_session(clp); + if (status == -NFS4ERR_DELAY) { + err = nfs4_delay(clp->cl_rpcclient, &timeout); + if (err) + status = err; + } + } while (status == -NFS4ERR_DELAY); + if (status) goto out; @@ -5083,6 +5189,27 @@ int nfs4_init_session(struct nfs_server *server) return ret; } +int nfs4_init_ds_session(struct nfs_client *clp) +{ + struct nfs4_session *session = clp->cl_session; + int ret; + + if (!test_and_clear_bit(NFS4_SESSION_INITING, &session->session_state)) + return 0; + + ret = nfs4_client_recover_expired_lease(clp); + if (!ret) + /* Test for the DS role */ + if (!is_ds_client(clp)) + ret = -ENODEV; + if (!ret) + ret = nfs4_check_client_ready(clp); + return ret; + +} +EXPORT_SYMBOL_GPL(nfs4_init_ds_session); + + /* * Renew the cl_session lease. */ @@ -5110,7 +5237,7 @@ static int nfs41_sequence_handle_errors(struct rpc_task *task, struct nfs_client rpc_delay(task, NFS4_POLL_RETRY_MAX); return -EAGAIN; default: - nfs4_schedule_state_recovery(clp); + nfs4_schedule_lease_recovery(clp); } return 0; } @@ -5197,7 +5324,7 @@ static int nfs41_proc_async_sequence(struct nfs_client *clp, struct rpc_cred *cr if (IS_ERR(task)) ret = PTR_ERR(task); else - rpc_put_task(task); + rpc_put_task_async(task); dprintk("<-- %s status=%d\n", __func__, ret); return ret; } @@ -5213,8 +5340,13 @@ static int nfs4_proc_sequence(struct nfs_client *clp, struct rpc_cred *cred) goto out; } ret = rpc_wait_for_completion_task(task); - if (!ret) + if (!ret) { + struct nfs4_sequence_res *res = task->tk_msg.rpc_resp; + + if (task->tk_status == 0) + nfs41_handle_sequence_flag_errors(clp, res->sr_status_flags); ret = task->tk_status; + } rpc_put_task(task); out: dprintk("<-- %s status=%d\n", __func__, ret); @@ -5251,7 +5383,7 @@ static int nfs41_reclaim_complete_handle_errors(struct rpc_task *task, struct nf rpc_delay(task, NFS4_POLL_RETRY_MAX); return -EAGAIN; default: - nfs4_schedule_state_recovery(clp); + nfs4_schedule_lease_recovery(clp); } return 0; } @@ -5319,6 +5451,9 @@ static int nfs41_proc_reclaim_complete(struct nfs_client *clp) status = PTR_ERR(task); goto out; } + status = nfs4_wait_for_completion_rpc_task(task); + if (status == 0) + status = task->tk_status; rpc_put_task(task); return 0; out: @@ -5605,6 +5740,7 @@ const struct nfs_rpc_ops nfs_v4_clientops = { .clear_acl_cache = nfs4_zap_acl_attr, .close_context = nfs4_close_context, .open_context = nfs4_atomic_open, + .init_client = nfs4_init_client, }; static const struct xattr_handler nfs4_xattr_nfs4_acl_handler = { diff --git a/fs/nfs/nfs4renewd.c b/fs/nfs/nfs4renewd.c index 402143d..df8e7f3 100644 --- a/fs/nfs/nfs4renewd.c +++ b/fs/nfs/nfs4renewd.c @@ -64,12 +64,8 @@ nfs4_renew_state(struct work_struct *work) ops = clp->cl_mvops->state_renewal_ops; dprintk("%s: start\n", __func__); - rcu_read_lock(); - if (list_empty(&clp->cl_superblocks)) { - rcu_read_unlock(); + if (test_bit(NFS_CS_STOP_RENEW, &clp->cl_res_state)) goto out; - } - rcu_read_unlock(); spin_lock(&clp->cl_lock); lease = clp->cl_lease_time; diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index 2336d53..ab1bf5b 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -153,6 +153,11 @@ static int nfs41_setup_state_renewal(struct nfs_client *clp) int status; struct nfs_fsinfo fsinfo; + if (!test_bit(NFS_CS_CHECK_LEASE_TIME, &clp->cl_res_state)) { + nfs4_schedule_state_renewal(clp); + return 0; + } + status = nfs4_proc_get_lease_time(clp, &fsinfo); if (status == 0) { /* Update lease time and schedule renewal */ @@ -232,12 +237,6 @@ int nfs41_init_clientid(struct nfs_client *clp, struct rpc_cred *cred) status = nfs4_proc_create_session(clp); if (status != 0) goto out; - status = nfs4_set_callback_sessionid(clp); - if (status != 0) { - printk(KERN_WARNING "Sessionid not set. No callback service\n"); - nfs_callback_down(1); - status = 0; - } nfs41_setup_state_renewal(clp); nfs_mark_client_ready(clp, NFS_CS_READY); out: @@ -1013,9 +1012,9 @@ void nfs4_schedule_state_manager(struct nfs_client *clp) } /* - * Schedule a state recovery attempt + * Schedule a lease recovery attempt */ -void nfs4_schedule_state_recovery(struct nfs_client *clp) +void nfs4_schedule_lease_recovery(struct nfs_client *clp) { if (!clp) return; @@ -1024,7 +1023,7 @@ void nfs4_schedule_state_recovery(struct nfs_client *clp) nfs4_schedule_state_manager(clp); } -int nfs4_state_mark_reclaim_reboot(struct nfs_client *clp, struct nfs4_state *state) +static int nfs4_state_mark_reclaim_reboot(struct nfs_client *clp, struct nfs4_state *state) { set_bit(NFS_STATE_RECLAIM_REBOOT, &state->flags); @@ -1038,7 +1037,7 @@ int nfs4_state_mark_reclaim_reboot(struct nfs_client *clp, struct nfs4_state *st return 1; } -int nfs4_state_mark_reclaim_nograce(struct nfs_client *clp, struct nfs4_state *state) +static int nfs4_state_mark_reclaim_nograce(struct nfs_client *clp, struct nfs4_state *state) { set_bit(NFS_STATE_RECLAIM_NOGRACE, &state->flags); clear_bit(NFS_STATE_RECLAIM_REBOOT, &state->flags); @@ -1047,6 +1046,14 @@ int nfs4_state_mark_reclaim_nograce(struct nfs_client *clp, struct nfs4_state *s return 1; } +void nfs4_schedule_stateid_recovery(const struct nfs_server *server, struct nfs4_state *state) +{ + struct nfs_client *clp = server->nfs_client; + + nfs4_state_mark_reclaim_nograce(clp, state); + nfs4_schedule_state_manager(clp); +} + static int nfs4_reclaim_locks(struct nfs4_state *state, const struct nfs4_state_recovery_ops *ops) { struct inode *inode = state->inode; @@ -1442,10 +1449,16 @@ static int nfs4_reclaim_lease(struct nfs_client *clp) } #ifdef CONFIG_NFS_V4_1 +void nfs4_schedule_session_recovery(struct nfs4_session *session) +{ + nfs4_schedule_lease_recovery(session->clp); +} +EXPORT_SYMBOL_GPL(nfs4_schedule_session_recovery); + void nfs41_handle_recall_slot(struct nfs_client *clp) { set_bit(NFS4CLNT_RECALL_SLOT, &clp->cl_state); - nfs4_schedule_state_recovery(clp); + nfs4_schedule_state_manager(clp); } static void nfs4_reset_all_state(struct nfs_client *clp) @@ -1453,7 +1466,7 @@ static void nfs4_reset_all_state(struct nfs_client *clp) if (test_and_set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) == 0) { clp->cl_boot_time = CURRENT_TIME; nfs4_state_start_reclaim_nograce(clp); - nfs4_schedule_state_recovery(clp); + nfs4_schedule_state_manager(clp); } } @@ -1461,7 +1474,7 @@ static void nfs41_handle_server_reboot(struct nfs_client *clp) { if (test_and_set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state) == 0) { nfs4_state_start_reclaim_reboot(clp); - nfs4_schedule_state_recovery(clp); + nfs4_schedule_state_manager(clp); } } @@ -1481,7 +1494,7 @@ static void nfs41_handle_cb_path_down(struct nfs_client *clp) { nfs_expire_all_delegations(clp); if (test_and_set_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state) == 0) - nfs4_schedule_state_recovery(clp); + nfs4_schedule_state_manager(clp); } void nfs41_handle_sequence_flag_errors(struct nfs_client *clp, u32 flags) diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index 2ab8e5c..0cf560f 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -844,7 +844,7 @@ static void encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const if (iap->ia_valid & ATTR_MODE) len += 4; if (iap->ia_valid & ATTR_UID) { - owner_namelen = nfs_map_uid_to_name(server->nfs_client, iap->ia_uid, owner_name, IDMAP_NAMESZ); + owner_namelen = nfs_map_uid_to_name(server, iap->ia_uid, owner_name, IDMAP_NAMESZ); if (owner_namelen < 0) { dprintk("nfs: couldn't resolve uid %d to string\n", iap->ia_uid); @@ -856,7 +856,7 @@ static void encode_attrs(struct xdr_stream *xdr, const struct iattr *iap, const len += 4 + (XDR_QUADLEN(owner_namelen) << 2); } if (iap->ia_valid & ATTR_GID) { - owner_grouplen = nfs_map_gid_to_group(server->nfs_client, iap->ia_gid, owner_group, IDMAP_NAMESZ); + owner_grouplen = nfs_map_gid_to_group(server, iap->ia_gid, owner_group, IDMAP_NAMESZ); if (owner_grouplen < 0) { dprintk("nfs: couldn't resolve gid %d to string\n", iap->ia_gid); @@ -1384,7 +1384,7 @@ static void encode_putrootfh(struct xdr_stream *xdr, struct compound_hdr *hdr) hdr->replen += decode_putrootfh_maxsz; } -static void encode_stateid(struct xdr_stream *xdr, const struct nfs_open_context *ctx, const struct nfs_lock_context *l_ctx) +static void encode_stateid(struct xdr_stream *xdr, const struct nfs_open_context *ctx, const struct nfs_lock_context *l_ctx, int zero_seqid) { nfs4_stateid stateid; __be32 *p; @@ -1392,6 +1392,8 @@ static void encode_stateid(struct xdr_stream *xdr, const struct nfs_open_context p = reserve_space(xdr, NFS4_STATEID_SIZE); if (ctx->state != NULL) { nfs4_copy_stateid(&stateid, ctx->state, l_ctx->lockowner, l_ctx->pid); + if (zero_seqid) + stateid.stateid.seqid = 0; xdr_encode_opaque_fixed(p, stateid.data, NFS4_STATEID_SIZE); } else xdr_encode_opaque_fixed(p, zero_stateid.data, NFS4_STATEID_SIZE); @@ -1404,7 +1406,8 @@ static void encode_read(struct xdr_stream *xdr, const struct nfs_readargs *args, p = reserve_space(xdr, 4); *p = cpu_to_be32(OP_READ); - encode_stateid(xdr, args->context, args->lock_context); + encode_stateid(xdr, args->context, args->lock_context, + hdr->minorversion); p = reserve_space(xdr, 12); p = xdr_encode_hyper(p, args->offset); @@ -1592,7 +1595,8 @@ static void encode_write(struct xdr_stream *xdr, const struct nfs_writeargs *arg p = reserve_space(xdr, 4); *p = cpu_to_be32(OP_WRITE); - encode_stateid(xdr, args->context, args->lock_context); + encode_stateid(xdr, args->context, args->lock_context, + hdr->minorversion); p = reserve_space(xdr, 16); p = xdr_encode_hyper(p, args->offset); @@ -1660,7 +1664,7 @@ static void encode_create_session(struct xdr_stream *xdr, p = reserve_space(xdr, 20 + 2*28 + 20 + len + 12); *p++ = cpu_to_be32(OP_CREATE_SESSION); - p = xdr_encode_hyper(p, clp->cl_ex_clid); + p = xdr_encode_hyper(p, clp->cl_clientid); *p++ = cpu_to_be32(clp->cl_seqid); /*Sequence id */ *p++ = cpu_to_be32(args->flags); /*flags */ @@ -2271,7 +2275,8 @@ static void nfs4_xdr_enc_write(struct rpc_rqst *req, struct xdr_stream *xdr, encode_putfh(xdr, args->fh, &hdr); encode_write(xdr, args, &hdr); req->rq_snd_buf.flags |= XDRBUF_WRITE; - encode_getfattr(xdr, args->bitmask, &hdr); + if (args->bitmask) + encode_getfattr(xdr, args->bitmask, &hdr); encode_nops(&hdr); } @@ -3382,7 +3387,7 @@ out_overflow: } static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, - struct nfs_client *clp, uint32_t *uid, int may_sleep) + const struct nfs_server *server, uint32_t *uid, int may_sleep) { uint32_t len; __be32 *p; @@ -3402,7 +3407,7 @@ static int decode_attr_owner(struct xdr_stream *xdr, uint32_t *bitmap, if (!may_sleep) { /* do nothing */ } else if (len < XDR_MAX_NETOBJ) { - if (nfs_map_name_to_uid(clp, (char *)p, len, uid) == 0) + if (nfs_map_name_to_uid(server, (char *)p, len, uid) == 0) ret = NFS_ATTR_FATTR_OWNER; else dprintk("%s: nfs_map_name_to_uid failed!\n", @@ -3420,7 +3425,7 @@ out_overflow: } static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, - struct nfs_client *clp, uint32_t *gid, int may_sleep) + const struct nfs_server *server, uint32_t *gid, int may_sleep) { uint32_t len; __be32 *p; @@ -3440,7 +3445,7 @@ static int decode_attr_group(struct xdr_stream *xdr, uint32_t *bitmap, if (!may_sleep) { /* do nothing */ } else if (len < XDR_MAX_NETOBJ) { - if (nfs_map_group_to_gid(clp, (char *)p, len, gid) == 0) + if (nfs_map_group_to_gid(server, (char *)p, len, gid) == 0) ret = NFS_ATTR_FATTR_GROUP; else dprintk("%s: nfs_map_group_to_gid failed!\n", @@ -3939,14 +3944,12 @@ static int decode_getfattr_attrs(struct xdr_stream *xdr, uint32_t *bitmap, goto xdr_error; fattr->valid |= status; - status = decode_attr_owner(xdr, bitmap, server->nfs_client, - &fattr->uid, may_sleep); + status = decode_attr_owner(xdr, bitmap, server, &fattr->uid, may_sleep); if (status < 0) goto xdr_error; fattr->valid |= status; - status = decode_attr_group(xdr, bitmap, server->nfs_client, - &fattr->gid, may_sleep); + status = decode_attr_group(xdr, bitmap, server, &fattr->gid, may_sleep); if (status < 0) goto xdr_error; fattr->valid |= status; @@ -4694,7 +4697,7 @@ static int decode_exchange_id(struct xdr_stream *xdr, p = xdr_inline_decode(xdr, 8); if (unlikely(!p)) goto out_overflow; - xdr_decode_hyper(p, &clp->cl_ex_clid); + xdr_decode_hyper(p, &clp->cl_clientid); p = xdr_inline_decode(xdr, 12); if (unlikely(!p)) goto out_overflow; @@ -5690,8 +5693,9 @@ static int nfs4_xdr_dec_write(struct rpc_rqst *rqstp, struct xdr_stream *xdr, status = decode_write(xdr, res); if (status) goto out; - decode_getfattr(xdr, res->fattr, res->server, - !RPC_IS_ASYNC(rqstp->rq_task)); + if (res->fattr) + decode_getfattr(xdr, res->fattr, res->server, + !RPC_IS_ASYNC(rqstp->rq_task)); if (!status) status = res->count; out: @@ -6086,11 +6090,11 @@ int nfs4_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, __be32 *p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) goto out_overflow; - if (!ntohl(*p++)) { + if (*p == xdr_zero) { p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) goto out_overflow; - if (!ntohl(*p++)) + if (*p == xdr_zero) return -EAGAIN; entry->eof = 1; return -EBADCOOKIE; @@ -6101,7 +6105,7 @@ int nfs4_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, goto out_overflow; entry->prev_cookie = entry->cookie; p = xdr_decode_hyper(p, &entry->cookie); - entry->len = ntohl(*p++); + entry->len = be32_to_cpup(p); p = xdr_inline_decode(xdr, entry->len); if (unlikely(!p)) @@ -6132,9 +6136,6 @@ int nfs4_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, if (entry->fattr->valid & NFS_ATTR_FATTR_TYPE) entry->d_type = nfs_umode_to_dtype(entry->fattr->mode); - if (verify_attr_len(xdr, p, len) < 0) - goto out_overflow; - return 0; out_overflow: @@ -6170,8 +6171,6 @@ static struct { { NFS4ERR_DQUOT, -EDQUOT }, { NFS4ERR_STALE, -ESTALE }, { NFS4ERR_BADHANDLE, -EBADHANDLE }, - { NFS4ERR_BADOWNER, -EINVAL }, - { NFS4ERR_BADNAME, -EINVAL }, { NFS4ERR_BAD_COOKIE, -EBADCOOKIE }, { NFS4ERR_NOTSUPP, -ENOTSUPP }, { NFS4ERR_TOOSMALL, -ETOOSMALL }, diff --git a/fs/nfs/nfsroot.c b/fs/nfs/nfsroot.c index 903908a..c541093 100644 --- a/fs/nfs/nfsroot.c +++ b/fs/nfs/nfsroot.c @@ -86,11 +86,14 @@ /* Default path we try to mount. "%s" gets replaced by our IP address */ #define NFS_ROOT "/tftpboot/%s" +/* Default NFSROOT mount options. */ +#define NFS_DEF_OPTIONS "udp" + /* Parameters passed from the kernel command line */ static char nfs_root_parms[256] __initdata = ""; /* Text-based mount options passed to super.c */ -static char nfs_root_options[256] __initdata = ""; +static char nfs_root_options[256] __initdata = NFS_DEF_OPTIONS; /* Address of NFS server */ static __be32 servaddr __initdata = htonl(INADDR_NONE); @@ -160,8 +163,14 @@ static int __init root_nfs_copy(char *dest, const char *src, } static int __init root_nfs_cat(char *dest, const char *src, - const size_t destlen) + const size_t destlen) { + size_t len = strlen(dest); + + if (len && dest[len - 1] != ',') + if (strlcat(dest, ",", destlen) > destlen) + return -1; + if (strlcat(dest, src, destlen) > destlen) return -1; return 0; @@ -194,16 +203,6 @@ static int __init root_nfs_parse_options(char *incoming, char *exppath, if (root_nfs_cat(nfs_root_options, incoming, sizeof(nfs_root_options))) return -1; - - /* - * Possibly prepare for more options to be appended - */ - if (nfs_root_options[0] != '\0' && - nfs_root_options[strlen(nfs_root_options)] != ',') - if (root_nfs_cat(nfs_root_options, ",", - sizeof(nfs_root_options))) - return -1; - return 0; } @@ -217,7 +216,7 @@ static int __init root_nfs_parse_options(char *incoming, char *exppath, */ static int __init root_nfs_data(char *cmdline) { - char addr_option[sizeof("nolock,addr=") + INET_ADDRSTRLEN + 1]; + char mand_options[sizeof("nolock,addr=") + INET_ADDRSTRLEN + 1]; int len, retval = -1; char *tmp = NULL; const size_t tmplen = sizeof(nfs_export_path); @@ -244,9 +243,9 @@ static int __init root_nfs_data(char *cmdline) * Append mandatory options for nfsroot so they override * what has come before */ - snprintf(addr_option, sizeof(addr_option), "nolock,addr=%pI4", + snprintf(mand_options, sizeof(mand_options), "nolock,addr=%pI4", &servaddr); - if (root_nfs_cat(nfs_root_options, addr_option, + if (root_nfs_cat(nfs_root_options, mand_options, sizeof(nfs_root_options))) goto out_optionstoolong; diff --git a/fs/nfs/pagelist.c b/fs/nfs/pagelist.c index e1164e3..23e7944 100644 --- a/fs/nfs/pagelist.c +++ b/fs/nfs/pagelist.c @@ -20,6 +20,7 @@ #include <linux/nfs_mount.h> #include "internal.h" +#include "pnfs.h" static struct kmem_cache *nfs_page_cachep; @@ -213,7 +214,7 @@ nfs_wait_on_request(struct nfs_page *req) */ void nfs_pageio_init(struct nfs_pageio_descriptor *desc, struct inode *inode, - int (*doio)(struct inode *, struct list_head *, unsigned int, size_t, int), + int (*doio)(struct nfs_pageio_descriptor *), size_t bsize, int io_flags) { @@ -226,6 +227,7 @@ void nfs_pageio_init(struct nfs_pageio_descriptor *desc, desc->pg_doio = doio; desc->pg_ioflags = io_flags; desc->pg_error = 0; + desc->pg_lseg = NULL; } /** @@ -240,7 +242,8 @@ void nfs_pageio_init(struct nfs_pageio_descriptor *desc, * Return 'true' if this is the case, else return 'false'. */ static int nfs_can_coalesce_requests(struct nfs_page *prev, - struct nfs_page *req) + struct nfs_page *req, + struct nfs_pageio_descriptor *pgio) { if (req->wb_context->cred != prev->wb_context->cred) return 0; @@ -254,6 +257,12 @@ static int nfs_can_coalesce_requests(struct nfs_page *prev, return 0; if (prev->wb_pgbase + prev->wb_bytes != PAGE_CACHE_SIZE) return 0; + /* + * Non-whole file layouts need to check that req is inside of + * pgio->pg_lseg. + */ + if (pgio->pg_test && !pgio->pg_test(pgio, prev, req)) + return 0; return 1; } @@ -286,7 +295,7 @@ static int nfs_pageio_do_add_request(struct nfs_pageio_descriptor *desc, if (newlen > desc->pg_bsize) return 0; prev = nfs_list_entry(desc->pg_list.prev); - if (!nfs_can_coalesce_requests(prev, req)) + if (!nfs_can_coalesce_requests(prev, req, desc)) return 0; } else desc->pg_base = req->wb_pgbase; @@ -302,12 +311,7 @@ static int nfs_pageio_do_add_request(struct nfs_pageio_descriptor *desc, static void nfs_pageio_doio(struct nfs_pageio_descriptor *desc) { if (!list_empty(&desc->pg_list)) { - int error = desc->pg_doio(desc->pg_inode, - &desc->pg_list, - nfs_page_array_len(desc->pg_base, - desc->pg_count), - desc->pg_count, - desc->pg_ioflags); + int error = desc->pg_doio(desc); if (error < 0) desc->pg_error = error; else diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index bc40897..f38813a 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -30,6 +30,7 @@ #include <linux/nfs_fs.h> #include "internal.h" #include "pnfs.h" +#include "iostat.h" #define NFSDBG_FACILITY NFSDBG_PNFS @@ -74,10 +75,8 @@ find_pnfs_driver(u32 id) void unset_pnfs_layoutdriver(struct nfs_server *nfss) { - if (nfss->pnfs_curr_ld) { - nfss->pnfs_curr_ld->clear_layoutdriver(nfss); + if (nfss->pnfs_curr_ld) module_put(nfss->pnfs_curr_ld->owner); - } nfss->pnfs_curr_ld = NULL; } @@ -115,13 +114,7 @@ set_pnfs_layoutdriver(struct nfs_server *server, u32 id) goto out_no_driver; } server->pnfs_curr_ld = ld_type; - if (ld_type->set_layoutdriver(server)) { - printk(KERN_ERR - "%s: Error initializing mount point for layout driver %u.\n", - __func__, id); - module_put(ld_type->owner); - goto out_no_driver; - } + dprintk("%s: pNFS module for %u set\n", __func__, id); return; @@ -230,37 +223,41 @@ static void free_lseg(struct pnfs_layout_segment *lseg) put_layout_hdr(NFS_I(ino)->layout); } -/* The use of tmp_list is necessary because pnfs_curr_ld->free_lseg - * could sleep, so must be called outside of the lock. - * Returns 1 if object was removed, otherwise return 0. - */ -static int -put_lseg_locked(struct pnfs_layout_segment *lseg, - struct list_head *tmp_list) +static void +put_lseg_common(struct pnfs_layout_segment *lseg) +{ + struct inode *inode = lseg->pls_layout->plh_inode; + + BUG_ON(test_bit(NFS_LSEG_VALID, &lseg->pls_flags)); + list_del_init(&lseg->pls_list); + if (list_empty(&lseg->pls_layout->plh_segs)) { + set_bit(NFS_LAYOUT_DESTROYED, &lseg->pls_layout->plh_flags); + /* Matched by initial refcount set in alloc_init_layout_hdr */ + put_layout_hdr_locked(lseg->pls_layout); + } + rpc_wake_up(&NFS_SERVER(inode)->roc_rpcwaitq); +} + +void +put_lseg(struct pnfs_layout_segment *lseg) { + struct inode *inode; + + if (!lseg) + return; + dprintk("%s: lseg %p ref %d valid %d\n", __func__, lseg, atomic_read(&lseg->pls_refcount), test_bit(NFS_LSEG_VALID, &lseg->pls_flags)); - if (atomic_dec_and_test(&lseg->pls_refcount)) { - struct inode *ino = lseg->pls_layout->plh_inode; + inode = lseg->pls_layout->plh_inode; + if (atomic_dec_and_lock(&lseg->pls_refcount, &inode->i_lock)) { + LIST_HEAD(free_me); - BUG_ON(test_bit(NFS_LSEG_VALID, &lseg->pls_flags)); - list_del(&lseg->pls_list); - if (list_empty(&lseg->pls_layout->plh_segs)) { - struct nfs_client *clp; - - clp = NFS_SERVER(ino)->nfs_client; - spin_lock(&clp->cl_lock); - /* List does not take a reference, so no need for put here */ - list_del_init(&lseg->pls_layout->plh_layouts); - spin_unlock(&clp->cl_lock); - clear_bit(NFS_LAYOUT_BULK_RECALL, &lseg->pls_layout->plh_flags); - } - rpc_wake_up(&NFS_SERVER(ino)->roc_rpcwaitq); - list_add(&lseg->pls_list, tmp_list); - return 1; + put_lseg_common(lseg); + list_add(&lseg->pls_list, &free_me); + spin_unlock(&inode->i_lock); + pnfs_free_lseg_list(&free_me); } - return 0; } static bool @@ -281,7 +278,13 @@ static int mark_lseg_invalid(struct pnfs_layout_segment *lseg, * list. It will now be removed when all * outstanding io is finished. */ - rv = put_lseg_locked(lseg, tmp_list); + dprintk("%s: lseg %p ref %d\n", __func__, lseg, + atomic_read(&lseg->pls_refcount)); + if (atomic_dec_and_test(&lseg->pls_refcount)) { + put_lseg_common(lseg); + list_add(&lseg->pls_list, tmp_list); + rv = 1; + } } return rv; } @@ -299,6 +302,11 @@ mark_matching_lsegs_invalid(struct pnfs_layout_hdr *lo, dprintk("%s:Begin lo %p\n", __func__, lo); + if (list_empty(&lo->plh_segs)) { + if (!test_and_set_bit(NFS_LAYOUT_DESTROYED, &lo->plh_flags)) + put_layout_hdr_locked(lo); + return 0; + } list_for_each_entry_safe(lseg, next, &lo->plh_segs, pls_list) if (should_free_lseg(lseg->pls_range.iomode, iomode)) { dprintk("%s: freeing lseg %p iomode %d " @@ -312,11 +320,27 @@ mark_matching_lsegs_invalid(struct pnfs_layout_hdr *lo, return invalid - removed; } +/* note free_me must contain lsegs from a single layout_hdr */ void pnfs_free_lseg_list(struct list_head *free_me) { struct pnfs_layout_segment *lseg, *tmp; + struct pnfs_layout_hdr *lo; + + if (list_empty(free_me)) + return; + lo = list_first_entry(free_me, struct pnfs_layout_segment, + pls_list)->pls_layout; + + if (test_bit(NFS_LAYOUT_DESTROYED, &lo->plh_flags)) { + struct nfs_client *clp; + + clp = NFS_SERVER(lo->plh_inode)->nfs_client; + spin_lock(&clp->cl_lock); + list_del_init(&lo->plh_layouts); + spin_unlock(&clp->cl_lock); + } list_for_each_entry_safe(lseg, tmp, free_me, pls_list) { list_del(&lseg->pls_list); free_lseg(lseg); @@ -332,10 +356,8 @@ pnfs_destroy_layout(struct nfs_inode *nfsi) spin_lock(&nfsi->vfs_inode.i_lock); lo = nfsi->layout; if (lo) { - set_bit(NFS_LAYOUT_DESTROYED, &nfsi->layout->plh_flags); + lo->plh_block_lgets++; /* permanently block new LAYOUTGETs */ mark_matching_lsegs_invalid(lo, &tmp_list, IOMODE_ANY); - /* Matched by refcount set to 1 in alloc_init_layout_hdr */ - put_layout_hdr_locked(lo); } spin_unlock(&nfsi->vfs_inode.i_lock); pnfs_free_lseg_list(&tmp_list); @@ -403,6 +425,7 @@ pnfs_layoutgets_blocked(struct pnfs_layout_hdr *lo, nfs4_stateid *stateid, (int)(lo->plh_barrier - be32_to_cpu(stateid->stateid.seqid)) >= 0) return true; return lo->plh_block_lgets || + test_bit(NFS_LAYOUT_DESTROYED, &lo->plh_flags) || test_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags) || (list_empty(&lo->plh_segs) && (atomic_read(&lo->plh_outstanding) > lget)); @@ -674,7 +697,7 @@ pnfs_find_lseg(struct pnfs_layout_hdr *lo, u32 iomode) list_for_each_entry(lseg, &lo->plh_segs, pls_list) { if (test_bit(NFS_LSEG_VALID, &lseg->pls_flags) && is_matching_lseg(lseg, iomode)) { - ret = lseg; + ret = get_lseg(lseg); break; } if (cmp_layout(iomode, lseg->pls_range.iomode) > 0) @@ -699,6 +722,7 @@ pnfs_update_layout(struct inode *ino, struct nfs_client *clp = NFS_SERVER(ino)->nfs_client; struct pnfs_layout_hdr *lo; struct pnfs_layout_segment *lseg = NULL; + bool first = false; if (!pnfs_enabled_sb(NFS_SERVER(ino))) return NULL; @@ -715,21 +739,25 @@ pnfs_update_layout(struct inode *ino, dprintk("%s matches recall, use MDS\n", __func__); goto out_unlock; } - /* Check to see if the layout for the given range already exists */ - lseg = pnfs_find_lseg(lo, iomode); - if (lseg) - goto out_unlock; /* if LAYOUTGET already failed once we don't try again */ if (test_bit(lo_fail_bit(iomode), &nfsi->layout->plh_flags)) goto out_unlock; + /* Check to see if the layout for the given range already exists */ + lseg = pnfs_find_lseg(lo, iomode); + if (lseg) + goto out_unlock; + if (pnfs_layoutgets_blocked(lo, NULL, 0)) goto out_unlock; atomic_inc(&lo->plh_outstanding); get_layout_hdr(lo); - if (list_empty(&lo->plh_segs)) { + if (list_empty(&lo->plh_segs)) + first = true; + spin_unlock(&ino->i_lock); + if (first) { /* The lo must be on the clp list if there is any * chance of a CB_LAYOUTRECALL(FILE) coming in. */ @@ -738,24 +766,18 @@ pnfs_update_layout(struct inode *ino, list_add_tail(&lo->plh_layouts, &clp->cl_layouts); spin_unlock(&clp->cl_lock); } - spin_unlock(&ino->i_lock); lseg = send_layoutget(lo, ctx, iomode); - if (!lseg) { - spin_lock(&ino->i_lock); - if (list_empty(&lo->plh_segs)) { - spin_lock(&clp->cl_lock); - list_del_init(&lo->plh_layouts); - spin_unlock(&clp->cl_lock); - clear_bit(NFS_LAYOUT_BULK_RECALL, &lo->plh_flags); - } - spin_unlock(&ino->i_lock); + if (!lseg && first) { + spin_lock(&clp->cl_lock); + list_del_init(&lo->plh_layouts); + spin_unlock(&clp->cl_lock); } atomic_dec(&lo->plh_outstanding); put_layout_hdr(lo); out: dprintk("%s end, state 0x%lx lseg %p\n", __func__, - nfsi->layout->plh_flags, lseg); + nfsi->layout ? nfsi->layout->plh_flags : -1, lseg); return lseg; out_unlock: spin_unlock(&ino->i_lock); @@ -808,7 +830,7 @@ pnfs_layout_process(struct nfs4_layoutget *lgp) } init_lseg(lo, lseg); lseg->pls_range = res->range; - *lgp->lsegpp = lseg; + *lgp->lsegpp = get_lseg(lseg); pnfs_insert_layout(lo, lseg); if (res->return_on_close) { @@ -829,137 +851,97 @@ out_forget_reply: goto out; } -/* - * Device ID cache. Currently supports one layout type per struct nfs_client. - * Add layout type to the lookup key to expand to support multiple types. - */ -int -pnfs_alloc_init_deviceid_cache(struct nfs_client *clp, - void (*free_callback)(struct pnfs_deviceid_node *)) +static int pnfs_read_pg_test(struct nfs_pageio_descriptor *pgio, + struct nfs_page *prev, + struct nfs_page *req) { - struct pnfs_deviceid_cache *c; - - c = kzalloc(sizeof(struct pnfs_deviceid_cache), GFP_KERNEL); - if (!c) - return -ENOMEM; - spin_lock(&clp->cl_lock); - if (clp->cl_devid_cache != NULL) { - atomic_inc(&clp->cl_devid_cache->dc_ref); - dprintk("%s [kref [%d]]\n", __func__, - atomic_read(&clp->cl_devid_cache->dc_ref)); - kfree(c); - } else { - /* kzalloc initializes hlists */ - spin_lock_init(&c->dc_lock); - atomic_set(&c->dc_ref, 1); - c->dc_free_callback = free_callback; - clp->cl_devid_cache = c; - dprintk("%s [new]\n", __func__); + if (pgio->pg_count == prev->wb_bytes) { + /* This is first coelesce call for a series of nfs_pages */ + pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode, + prev->wb_context, + IOMODE_READ); } - spin_unlock(&clp->cl_lock); - return 0; + return NFS_SERVER(pgio->pg_inode)->pnfs_curr_ld->pg_test(pgio, prev, req); } -EXPORT_SYMBOL_GPL(pnfs_alloc_init_deviceid_cache); -/* - * Called from pnfs_layoutdriver_type->free_lseg - * last layout segment reference frees deviceid - */ void -pnfs_put_deviceid(struct pnfs_deviceid_cache *c, - struct pnfs_deviceid_node *devid) +pnfs_pageio_init_read(struct nfs_pageio_descriptor *pgio, struct inode *inode) { - struct nfs4_deviceid *id = &devid->de_id; - struct pnfs_deviceid_node *d; - struct hlist_node *n; - long h = nfs4_deviceid_hash(id); + struct pnfs_layoutdriver_type *ld; - dprintk("%s [%d]\n", __func__, atomic_read(&devid->de_ref)); - if (!atomic_dec_and_lock(&devid->de_ref, &c->dc_lock)) - return; + ld = NFS_SERVER(inode)->pnfs_curr_ld; + pgio->pg_test = (ld && ld->pg_test) ? pnfs_read_pg_test : NULL; +} - hlist_for_each_entry_rcu(d, n, &c->dc_deviceids[h], de_node) - if (!memcmp(&d->de_id, id, sizeof(*id))) { - hlist_del_rcu(&d->de_node); - spin_unlock(&c->dc_lock); - synchronize_rcu(); - c->dc_free_callback(devid); - return; - } - spin_unlock(&c->dc_lock); - /* Why wasn't it found in the list? */ - BUG(); -} -EXPORT_SYMBOL_GPL(pnfs_put_deviceid); - -/* Find and reference a deviceid */ -struct pnfs_deviceid_node * -pnfs_find_get_deviceid(struct pnfs_deviceid_cache *c, struct nfs4_deviceid *id) -{ - struct pnfs_deviceid_node *d; - struct hlist_node *n; - long hash = nfs4_deviceid_hash(id); - - dprintk("--> %s hash %ld\n", __func__, hash); - rcu_read_lock(); - hlist_for_each_entry_rcu(d, n, &c->dc_deviceids[hash], de_node) { - if (!memcmp(&d->de_id, id, sizeof(*id))) { - if (!atomic_inc_not_zero(&d->de_ref)) { - goto fail; - } else { - rcu_read_unlock(); - return d; - } - } +static int pnfs_write_pg_test(struct nfs_pageio_descriptor *pgio, + struct nfs_page *prev, + struct nfs_page *req) +{ + if (pgio->pg_count == prev->wb_bytes) { + /* This is first coelesce call for a series of nfs_pages */ + pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode, + prev->wb_context, + IOMODE_RW); } -fail: - rcu_read_unlock(); - return NULL; + return NFS_SERVER(pgio->pg_inode)->pnfs_curr_ld->pg_test(pgio, prev, req); +} + +void +pnfs_pageio_init_write(struct nfs_pageio_descriptor *pgio, struct inode *inode) +{ + struct pnfs_layoutdriver_type *ld; + + ld = NFS_SERVER(inode)->pnfs_curr_ld; + pgio->pg_test = (ld && ld->pg_test) ? pnfs_write_pg_test : NULL; +} + +enum pnfs_try_status +pnfs_try_to_write_data(struct nfs_write_data *wdata, + const struct rpc_call_ops *call_ops, int how) +{ + struct inode *inode = wdata->inode; + enum pnfs_try_status trypnfs; + struct nfs_server *nfss = NFS_SERVER(inode); + + wdata->mds_ops = call_ops; + + dprintk("%s: Writing ino:%lu %u@%llu (how %d)\n", __func__, + inode->i_ino, wdata->args.count, wdata->args.offset, how); + + trypnfs = nfss->pnfs_curr_ld->write_pagelist(wdata, how); + if (trypnfs == PNFS_NOT_ATTEMPTED) { + put_lseg(wdata->lseg); + wdata->lseg = NULL; + } else + nfs_inc_stats(inode, NFSIOS_PNFS_WRITE); + + dprintk("%s End (trypnfs:%d)\n", __func__, trypnfs); + return trypnfs; } -EXPORT_SYMBOL_GPL(pnfs_find_get_deviceid); /* - * Add a deviceid to the cache. - * GETDEVICEINFOs for same deviceid can race. If deviceid is found, discard new + * Call the appropriate parallel I/O subsystem read function. */ -struct pnfs_deviceid_node * -pnfs_add_deviceid(struct pnfs_deviceid_cache *c, struct pnfs_deviceid_node *new) -{ - struct pnfs_deviceid_node *d; - long hash = nfs4_deviceid_hash(&new->de_id); - - dprintk("--> %s hash %ld\n", __func__, hash); - spin_lock(&c->dc_lock); - d = pnfs_find_get_deviceid(c, &new->de_id); - if (d) { - spin_unlock(&c->dc_lock); - dprintk("%s [discard]\n", __func__); - c->dc_free_callback(new); - return d; - } - INIT_HLIST_NODE(&new->de_node); - atomic_set(&new->de_ref, 1); - hlist_add_head_rcu(&new->de_node, &c->dc_deviceids[hash]); - spin_unlock(&c->dc_lock); - dprintk("%s [new]\n", __func__); - return new; -} -EXPORT_SYMBOL_GPL(pnfs_add_deviceid); - -void -pnfs_put_deviceid_cache(struct nfs_client *clp) +enum pnfs_try_status +pnfs_try_to_read_data(struct nfs_read_data *rdata, + const struct rpc_call_ops *call_ops) { - struct pnfs_deviceid_cache *local = clp->cl_devid_cache; + struct inode *inode = rdata->inode; + struct nfs_server *nfss = NFS_SERVER(inode); + enum pnfs_try_status trypnfs; - dprintk("--> %s cl_devid_cache %p\n", __func__, clp->cl_devid_cache); - if (atomic_dec_and_lock(&local->dc_ref, &clp->cl_lock)) { - int i; - /* Verify cache is empty */ - for (i = 0; i < NFS4_DEVICE_ID_HASH_SIZE; i++) - BUG_ON(!hlist_empty(&local->dc_deviceids[i])); - clp->cl_devid_cache = NULL; - spin_unlock(&clp->cl_lock); - kfree(local); + rdata->mds_ops = call_ops; + + dprintk("%s: Reading ino:%lu %u@%llu\n", + __func__, inode->i_ino, rdata->args.count, rdata->args.offset); + + trypnfs = nfss->pnfs_curr_ld->read_pagelist(rdata); + if (trypnfs == PNFS_NOT_ATTEMPTED) { + put_lseg(rdata->lseg); + rdata->lseg = NULL; + } else { + nfs_inc_stats(inode, NFSIOS_PNFS_READ); } + dprintk("%s End (trypnfs:%d)\n", __func__, trypnfs); + return trypnfs; } -EXPORT_SYMBOL_GPL(pnfs_put_deviceid_cache); diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h index e2612ea..6380b94 100644 --- a/fs/nfs/pnfs.h +++ b/fs/nfs/pnfs.h @@ -30,6 +30,8 @@ #ifndef FS_NFS_PNFS_H #define FS_NFS_PNFS_H +#include <linux/nfs_page.h> + enum { NFS_LSEG_VALID = 0, /* cleared when lseg is recalled/returned */ NFS_LSEG_ROC, /* roc bit received from server */ @@ -43,6 +45,11 @@ struct pnfs_layout_segment { struct pnfs_layout_hdr *pls_layout; }; +enum pnfs_try_status { + PNFS_ATTEMPTED = 0, + PNFS_NOT_ATTEMPTED = 1, +}; + #ifdef CONFIG_NFS_V4_1 #define LAYOUT_NFSV4_1_MODULE_PREFIX "nfs-layouttype4" @@ -61,10 +68,18 @@ struct pnfs_layoutdriver_type { const u32 id; const char *name; struct module *owner; - int (*set_layoutdriver) (struct nfs_server *); - int (*clear_layoutdriver) (struct nfs_server *); struct pnfs_layout_segment * (*alloc_lseg) (struct pnfs_layout_hdr *layoutid, struct nfs4_layoutget_res *lgr); void (*free_lseg) (struct pnfs_layout_segment *lseg); + + /* test for nfs page cache coalescing */ + int (*pg_test)(struct nfs_pageio_descriptor *, struct nfs_page *, struct nfs_page *); + + /* + * Return PNFS_ATTEMPTED to indicate the layout code has attempted + * I/O, else return PNFS_NOT_ATTEMPTED to fall back to normal NFS + */ + enum pnfs_try_status (*read_pagelist) (struct nfs_read_data *nfs_data); + enum pnfs_try_status (*write_pagelist) (struct nfs_write_data *nfs_data, int how); }; struct pnfs_layout_hdr { @@ -90,52 +105,6 @@ struct pnfs_device { unsigned int pglen; }; -/* - * Device ID RCU cache. A device ID is unique per client ID and layout type. - */ -#define NFS4_DEVICE_ID_HASH_BITS 5 -#define NFS4_DEVICE_ID_HASH_SIZE (1 << NFS4_DEVICE_ID_HASH_BITS) -#define NFS4_DEVICE_ID_HASH_MASK (NFS4_DEVICE_ID_HASH_SIZE - 1) - -static inline u32 -nfs4_deviceid_hash(struct nfs4_deviceid *id) -{ - unsigned char *cptr = (unsigned char *)id->data; - unsigned int nbytes = NFS4_DEVICEID4_SIZE; - u32 x = 0; - - while (nbytes--) { - x *= 37; - x += *cptr++; - } - return x & NFS4_DEVICE_ID_HASH_MASK; -} - -struct pnfs_deviceid_node { - struct hlist_node de_node; - struct nfs4_deviceid de_id; - atomic_t de_ref; -}; - -struct pnfs_deviceid_cache { - spinlock_t dc_lock; - atomic_t dc_ref; - void (*dc_free_callback)(struct pnfs_deviceid_node *); - struct hlist_head dc_deviceids[NFS4_DEVICE_ID_HASH_SIZE]; -}; - -extern int pnfs_alloc_init_deviceid_cache(struct nfs_client *, - void (*free_callback)(struct pnfs_deviceid_node *)); -extern void pnfs_put_deviceid_cache(struct nfs_client *); -extern struct pnfs_deviceid_node *pnfs_find_get_deviceid( - struct pnfs_deviceid_cache *, - struct nfs4_deviceid *); -extern struct pnfs_deviceid_node *pnfs_add_deviceid( - struct pnfs_deviceid_cache *, - struct pnfs_deviceid_node *); -extern void pnfs_put_deviceid(struct pnfs_deviceid_cache *c, - struct pnfs_deviceid_node *devid); - extern int pnfs_register_layoutdriver(struct pnfs_layoutdriver_type *); extern void pnfs_unregister_layoutdriver(struct pnfs_layoutdriver_type *); @@ -146,11 +115,18 @@ extern int nfs4_proc_layoutget(struct nfs4_layoutget *lgp); /* pnfs.c */ void get_layout_hdr(struct pnfs_layout_hdr *lo); +void put_lseg(struct pnfs_layout_segment *lseg); struct pnfs_layout_segment * pnfs_update_layout(struct inode *ino, struct nfs_open_context *ctx, enum pnfs_iomode access_type); void set_pnfs_layoutdriver(struct nfs_server *, u32 id); void unset_pnfs_layoutdriver(struct nfs_server *); +enum pnfs_try_status pnfs_try_to_write_data(struct nfs_write_data *, + const struct rpc_call_ops *, int); +enum pnfs_try_status pnfs_try_to_read_data(struct nfs_read_data *, + const struct rpc_call_ops *); +void pnfs_pageio_init_read(struct nfs_pageio_descriptor *, struct inode *); +void pnfs_pageio_init_write(struct nfs_pageio_descriptor *, struct inode *); int pnfs_layout_process(struct nfs4_layoutget *lgp); void pnfs_free_lseg_list(struct list_head *tmp_list); void pnfs_destroy_layout(struct nfs_inode *); @@ -177,6 +153,16 @@ static inline int lo_fail_bit(u32 iomode) NFS_LAYOUT_RW_FAILED : NFS_LAYOUT_RO_FAILED; } +static inline struct pnfs_layout_segment * +get_lseg(struct pnfs_layout_segment *lseg) +{ + if (lseg) { + atomic_inc(&lseg->pls_refcount); + smp_mb__after_atomic_inc(); + } + return lseg; +} + /* Return true if a layout driver is being used for this mountpoint */ static inline int pnfs_enabled_sb(struct nfs_server *nfss) { @@ -194,12 +180,36 @@ static inline void pnfs_destroy_layout(struct nfs_inode *nfsi) } static inline struct pnfs_layout_segment * +get_lseg(struct pnfs_layout_segment *lseg) +{ + return NULL; +} + +static inline void put_lseg(struct pnfs_layout_segment *lseg) +{ +} + +static inline struct pnfs_layout_segment * pnfs_update_layout(struct inode *ino, struct nfs_open_context *ctx, enum pnfs_iomode access_type) { return NULL; } +static inline enum pnfs_try_status +pnfs_try_to_read_data(struct nfs_read_data *data, + const struct rpc_call_ops *call_ops) +{ + return PNFS_NOT_ATTEMPTED; +} + +static inline enum pnfs_try_status +pnfs_try_to_write_data(struct nfs_write_data *data, + const struct rpc_call_ops *call_ops, int how) +{ + return PNFS_NOT_ATTEMPTED; +} + static inline bool pnfs_roc(struct inode *ino) { @@ -230,6 +240,18 @@ static inline void unset_pnfs_layoutdriver(struct nfs_server *s) { } +static inline void +pnfs_pageio_init_read(struct nfs_pageio_descriptor *pgio, struct inode *ino) +{ + pgio->pg_test = NULL; +} + +static inline void +pnfs_pageio_init_write(struct nfs_pageio_descriptor *pgio, struct inode *ino) +{ + pgio->pg_test = NULL; +} + #endif /* CONFIG_NFS_V4_1 */ #endif /* FS_NFS_PNFS_H */ diff --git a/fs/nfs/proc.c b/fs/nfs/proc.c index 77d5e21..b8ec170 100644 --- a/fs/nfs/proc.c +++ b/fs/nfs/proc.c @@ -741,4 +741,5 @@ const struct nfs_rpc_ops nfs_v2_clientops = { .lock = nfs_proc_lock, .lock_check_bounds = nfs_lock_check_bounds, .close_context = nfs_close_context, + .init_client = nfs_init_client, }; diff --git a/fs/nfs/read.c b/fs/nfs/read.c index aedcaa7..7cded2b 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -18,19 +18,20 @@ #include <linux/sunrpc/clnt.h> #include <linux/nfs_fs.h> #include <linux/nfs_page.h> +#include <linux/module.h> #include <asm/system.h> +#include "pnfs.h" #include "nfs4_fs.h" #include "internal.h" #include "iostat.h" #include "fscache.h" -#include "pnfs.h" #define NFSDBG_FACILITY NFSDBG_PAGECACHE -static int nfs_pagein_multi(struct inode *, struct list_head *, unsigned int, size_t, int); -static int nfs_pagein_one(struct inode *, struct list_head *, unsigned int, size_t, int); +static int nfs_pagein_multi(struct nfs_pageio_descriptor *desc); +static int nfs_pagein_one(struct nfs_pageio_descriptor *desc); static const struct rpc_call_ops nfs_read_partial_ops; static const struct rpc_call_ops nfs_read_full_ops; @@ -69,6 +70,7 @@ void nfs_readdata_free(struct nfs_read_data *p) static void nfs_readdata_release(struct nfs_read_data *rdata) { + put_lseg(rdata->lseg); put_nfs_open_context(rdata->args.context); nfs_readdata_free(rdata); } @@ -114,14 +116,13 @@ static void nfs_readpage_truncate_uninitialised_page(struct nfs_read_data *data) int nfs_readpage_async(struct nfs_open_context *ctx, struct inode *inode, struct page *page) { - LIST_HEAD(one_request); struct nfs_page *new; unsigned int len; + struct nfs_pageio_descriptor pgio; len = nfs_page_length(page); if (len == 0) return nfs_return_empty_page(page); - pnfs_update_layout(inode, ctx, IOMODE_READ); new = nfs_create_request(ctx, inode, page, 0, len); if (IS_ERR(new)) { unlock_page(page); @@ -130,11 +131,14 @@ int nfs_readpage_async(struct nfs_open_context *ctx, struct inode *inode, if (len < PAGE_CACHE_SIZE) zero_user_segment(page, len, PAGE_CACHE_SIZE); - nfs_list_add_request(new, &one_request); + nfs_pageio_init(&pgio, inode, NULL, 0, 0); + nfs_list_add_request(new, &pgio.pg_list); + pgio.pg_count = len; + if (NFS_SERVER(inode)->rsize < PAGE_CACHE_SIZE) - nfs_pagein_multi(inode, &one_request, 1, len, 0); + nfs_pagein_multi(&pgio); else - nfs_pagein_one(inode, &one_request, 1, len, 0); + nfs_pagein_one(&pgio); return 0; } @@ -155,24 +159,20 @@ static void nfs_readpage_release(struct nfs_page *req) nfs_release_request(req); } -/* - * Set up the NFS read request struct - */ -static int nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data, - const struct rpc_call_ops *call_ops, - unsigned int count, unsigned int offset) +int nfs_initiate_read(struct nfs_read_data *data, struct rpc_clnt *clnt, + const struct rpc_call_ops *call_ops) { - struct inode *inode = req->wb_context->path.dentry->d_inode; + struct inode *inode = data->inode; int swap_flags = IS_SWAPFILE(inode) ? NFS_RPC_SWAPFLAGS : 0; struct rpc_task *task; struct rpc_message msg = { .rpc_argp = &data->args, .rpc_resp = &data->res, - .rpc_cred = req->wb_context->cred, + .rpc_cred = data->cred, }; struct rpc_task_setup task_setup_data = { .task = &data->task, - .rpc_client = NFS_CLIENT(inode), + .rpc_client = clnt, .rpc_message = &msg, .callback_ops = call_ops, .callback_data = data, @@ -180,9 +180,39 @@ static int nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data, .flags = RPC_TASK_ASYNC | swap_flags, }; + /* Set up the initial task struct. */ + NFS_PROTO(inode)->read_setup(data, &msg); + + dprintk("NFS: %5u initiated read call (req %s/%lld, %u bytes @ " + "offset %llu)\n", + data->task.tk_pid, + inode->i_sb->s_id, + (long long)NFS_FILEID(inode), + data->args.count, + (unsigned long long)data->args.offset); + + task = rpc_run_task(&task_setup_data); + if (IS_ERR(task)) + return PTR_ERR(task); + rpc_put_task(task); + return 0; +} +EXPORT_SYMBOL_GPL(nfs_initiate_read); + +/* + * Set up the NFS read request struct + */ +static int nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data, + const struct rpc_call_ops *call_ops, + unsigned int count, unsigned int offset, + struct pnfs_layout_segment *lseg) +{ + struct inode *inode = req->wb_context->path.dentry->d_inode; + data->req = req; data->inode = inode; - data->cred = msg.rpc_cred; + data->cred = req->wb_context->cred; + data->lseg = get_lseg(lseg); data->args.fh = NFS_FH(inode); data->args.offset = req_offset(req) + offset; @@ -197,21 +227,11 @@ static int nfs_read_rpcsetup(struct nfs_page *req, struct nfs_read_data *data, data->res.eof = 0; nfs_fattr_init(&data->fattr); - /* Set up the initial task struct. */ - NFS_PROTO(inode)->read_setup(data, &msg); - - dprintk("NFS: %5u initiated read call (req %s/%Ld, %u bytes @ offset %Lu)\n", - data->task.tk_pid, - inode->i_sb->s_id, - (long long)NFS_FILEID(inode), - count, - (unsigned long long)data->args.offset); + if (data->lseg && + (pnfs_try_to_read_data(data, call_ops) == PNFS_ATTEMPTED)) + return 0; - task = rpc_run_task(&task_setup_data); - if (IS_ERR(task)) - return PTR_ERR(task); - rpc_put_task(task); - return 0; + return nfs_initiate_read(data, NFS_CLIENT(inode), call_ops); } static void @@ -240,20 +260,21 @@ nfs_async_read_error(struct list_head *head) * won't see the new data until our attribute cache is updated. This is more * or less conventional NFS client behavior. */ -static int nfs_pagein_multi(struct inode *inode, struct list_head *head, unsigned int npages, size_t count, int flags) +static int nfs_pagein_multi(struct nfs_pageio_descriptor *desc) { - struct nfs_page *req = nfs_list_entry(head->next); + struct nfs_page *req = nfs_list_entry(desc->pg_list.next); struct page *page = req->wb_page; struct nfs_read_data *data; - size_t rsize = NFS_SERVER(inode)->rsize, nbytes; + size_t rsize = NFS_SERVER(desc->pg_inode)->rsize, nbytes; unsigned int offset; int requests = 0; int ret = 0; + struct pnfs_layout_segment *lseg; LIST_HEAD(list); nfs_list_remove_request(req); - nbytes = count; + nbytes = desc->pg_count; do { size_t len = min(nbytes,rsize); @@ -266,9 +287,11 @@ static int nfs_pagein_multi(struct inode *inode, struct list_head *head, unsigne } while(nbytes != 0); atomic_set(&req->wb_complete, requests); + BUG_ON(desc->pg_lseg != NULL); + lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_READ); ClearPageError(page); offset = 0; - nbytes = count; + nbytes = desc->pg_count; do { int ret2; @@ -280,12 +303,14 @@ static int nfs_pagein_multi(struct inode *inode, struct list_head *head, unsigne if (nbytes < rsize) rsize = nbytes; ret2 = nfs_read_rpcsetup(req, data, &nfs_read_partial_ops, - rsize, offset); + rsize, offset, lseg); if (ret == 0) ret = ret2; offset += rsize; nbytes -= rsize; } while (nbytes != 0); + put_lseg(lseg); + desc->pg_lseg = NULL; return ret; @@ -300,16 +325,21 @@ out_bad: return -ENOMEM; } -static int nfs_pagein_one(struct inode *inode, struct list_head *head, unsigned int npages, size_t count, int flags) +static int nfs_pagein_one(struct nfs_pageio_descriptor *desc) { struct nfs_page *req; struct page **pages; struct nfs_read_data *data; + struct list_head *head = &desc->pg_list; + struct pnfs_layout_segment *lseg = desc->pg_lseg; int ret = -ENOMEM; - data = nfs_readdata_alloc(npages); - if (!data) - goto out_bad; + data = nfs_readdata_alloc(nfs_page_array_len(desc->pg_base, + desc->pg_count)); + if (!data) { + nfs_async_read_error(head); + goto out; + } pages = data->pagevec; while (!list_empty(head)) { @@ -320,10 +350,14 @@ static int nfs_pagein_one(struct inode *inode, struct list_head *head, unsigned *pages++ = req->wb_page; } req = nfs_list_entry(data->pages.next); + if ((!lseg) && list_is_singular(&data->pages)) + lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_READ); - return nfs_read_rpcsetup(req, data, &nfs_read_full_ops, count, 0); -out_bad: - nfs_async_read_error(head); + ret = nfs_read_rpcsetup(req, data, &nfs_read_full_ops, desc->pg_count, + 0, lseg); +out: + put_lseg(lseg); + desc->pg_lseg = NULL; return ret; } @@ -366,6 +400,7 @@ static void nfs_readpage_retry(struct rpc_task *task, struct nfs_read_data *data return; /* Yes, so retry the read at the end of the data */ + data->mds_offset += resp->count; argp->offset += resp->count; argp->pgbase += resp->count; argp->count -= resp->count; @@ -625,7 +660,7 @@ int nfs_readpages(struct file *filp, struct address_space *mapping, if (ret == 0) goto read_complete; /* all pages were read */ - pnfs_update_layout(inode, desc.ctx, IOMODE_READ); + pnfs_pageio_init_read(&pgio, inode); if (rsize < PAGE_CACHE_SIZE) nfs_pageio_init(&pgio, inode, nfs_pagein_multi, rsize, 0); else diff --git a/fs/nfs/super.c b/fs/nfs/super.c index b68c860..2b8e9a5 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -263,8 +263,11 @@ static match_table_t nfs_local_lock_tokens = { static void nfs_umount_begin(struct super_block *); static int nfs_statfs(struct dentry *, struct kstatfs *); static int nfs_show_options(struct seq_file *, struct vfsmount *); +static int nfs_show_devname(struct seq_file *, struct vfsmount *); +static int nfs_show_path(struct seq_file *, struct vfsmount *); static int nfs_show_stats(struct seq_file *, struct vfsmount *); -static int nfs_get_sb(struct file_system_type *, int, const char *, void *, struct vfsmount *); +static struct dentry *nfs_fs_mount(struct file_system_type *, + int, const char *, void *); static struct dentry *nfs_xdev_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *raw_data); static void nfs_put_super(struct super_block *); @@ -274,7 +277,7 @@ static int nfs_remount(struct super_block *sb, int *flags, char *raw_data); static struct file_system_type nfs_fs_type = { .owner = THIS_MODULE, .name = "nfs", - .get_sb = nfs_get_sb, + .mount = nfs_fs_mount, .kill_sb = nfs_kill_super, .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, }; @@ -296,6 +299,8 @@ static const struct super_operations nfs_sops = { .evict_inode = nfs_evict_inode, .umount_begin = nfs_umount_begin, .show_options = nfs_show_options, + .show_devname = nfs_show_devname, + .show_path = nfs_show_path, .show_stats = nfs_show_stats, .remount_fs = nfs_remount, }; @@ -303,16 +308,16 @@ static const struct super_operations nfs_sops = { #ifdef CONFIG_NFS_V4 static int nfs4_validate_text_mount_data(void *options, struct nfs_parsed_mount_data *args, const char *dev_name); -static int nfs4_try_mount(int flags, const char *dev_name, - struct nfs_parsed_mount_data *data, struct vfsmount *mnt); -static int nfs4_get_sb(struct file_system_type *fs_type, - int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); +static struct dentry *nfs4_try_mount(int flags, const char *dev_name, + struct nfs_parsed_mount_data *data); +static struct dentry *nfs4_mount(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data); static struct dentry *nfs4_remote_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *raw_data); static struct dentry *nfs4_xdev_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *raw_data); -static int nfs4_referral_get_sb(struct file_system_type *fs_type, - int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt); +static struct dentry *nfs4_referral_mount(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data); static struct dentry *nfs4_remote_referral_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *raw_data); static void nfs4_kill_super(struct super_block *sb); @@ -320,7 +325,7 @@ static void nfs4_kill_super(struct super_block *sb); static struct file_system_type nfs4_fs_type = { .owner = THIS_MODULE, .name = "nfs4", - .get_sb = nfs4_get_sb, + .mount = nfs4_mount, .kill_sb = nfs4_kill_super, .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, }; @@ -352,7 +357,7 @@ static struct file_system_type nfs4_remote_referral_fs_type = { struct file_system_type nfs4_referral_fs_type = { .owner = THIS_MODULE, .name = "nfs4", - .get_sb = nfs4_referral_get_sb, + .mount = nfs4_referral_mount, .kill_sb = nfs4_kill_super, .fs_flags = FS_RENAME_DOES_D_MOVE|FS_REVAL_DOT|FS_BINARY_MOUNTDATA, }; @@ -366,6 +371,8 @@ static const struct super_operations nfs4_sops = { .evict_inode = nfs4_evict_inode, .umount_begin = nfs_umount_begin, .show_options = nfs_show_options, + .show_devname = nfs_show_devname, + .show_path = nfs_show_path, .show_stats = nfs_show_stats, .remount_fs = nfs_remount, }; @@ -726,6 +733,28 @@ static int nfs_show_options(struct seq_file *m, struct vfsmount *mnt) return 0; } +static int nfs_show_devname(struct seq_file *m, struct vfsmount *mnt) +{ + char *page = (char *) __get_free_page(GFP_KERNEL); + char *devname, *dummy; + int err = 0; + if (!page) + return -ENOMEM; + devname = nfs_path(&dummy, mnt->mnt_root, page, PAGE_SIZE); + if (IS_ERR(devname)) + err = PTR_ERR(devname); + else + seq_escape(m, devname, " \t\n\\"); + free_page((unsigned long)page); + return err; +} + +static int nfs_show_path(struct seq_file *m, struct vfsmount *mnt) +{ + seq_puts(m, "/"); + return 0; +} + /* * Present statistical information for this VFS mountpoint */ @@ -979,6 +1008,27 @@ static int nfs_parse_security_flavors(char *value, return 1; } +static int nfs_get_option_str(substring_t args[], char **option) +{ + kfree(*option); + *option = match_strdup(args); + return !option; +} + +static int nfs_get_option_ul(substring_t args[], unsigned long *option) +{ + int rc; + char *string; + + string = match_strdup(args); + if (string == NULL) + return -ENOMEM; + rc = strict_strtoul(string, 10, option); + kfree(string); + + return rc; +} + /* * Error-check and convert a string of mount options from user space into * a data structure. The whole mount string is processed; bad options are @@ -1127,155 +1177,82 @@ static int nfs_parse_mount_options(char *raw, * options that take numeric values */ case Opt_port: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0 || option > USHRT_MAX) + if (nfs_get_option_ul(args, &option) || + option > USHRT_MAX) goto out_invalid_value; mnt->nfs_server.port = option; break; case Opt_rsize: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0) + if (nfs_get_option_ul(args, &option)) goto out_invalid_value; mnt->rsize = option; break; case Opt_wsize: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0) + if (nfs_get_option_ul(args, &option)) goto out_invalid_value; mnt->wsize = option; break; case Opt_bsize: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0) + if (nfs_get_option_ul(args, &option)) goto out_invalid_value; mnt->bsize = option; break; case Opt_timeo: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0 || option == 0) + if (nfs_get_option_ul(args, &option) || option == 0) goto out_invalid_value; mnt->timeo = option; break; case Opt_retrans: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0 || option == 0) + if (nfs_get_option_ul(args, &option) || option == 0) goto out_invalid_value; mnt->retrans = option; break; case Opt_acregmin: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0) + if (nfs_get_option_ul(args, &option)) goto out_invalid_value; mnt->acregmin = option; break; case Opt_acregmax: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0) + if (nfs_get_option_ul(args, &option)) goto out_invalid_value; mnt->acregmax = option; break; case Opt_acdirmin: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0) + if (nfs_get_option_ul(args, &option)) goto out_invalid_value; mnt->acdirmin = option; break; case Opt_acdirmax: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0) + if (nfs_get_option_ul(args, &option)) goto out_invalid_value; mnt->acdirmax = option; break; case Opt_actimeo: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0) + if (nfs_get_option_ul(args, &option)) goto out_invalid_value; mnt->acregmin = mnt->acregmax = mnt->acdirmin = mnt->acdirmax = option; break; case Opt_namelen: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0) + if (nfs_get_option_ul(args, &option)) goto out_invalid_value; mnt->namlen = option; break; case Opt_mountport: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0 || option > USHRT_MAX) + if (nfs_get_option_ul(args, &option) || + option > USHRT_MAX) goto out_invalid_value; mnt->mount_server.port = option; break; case Opt_mountvers: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0 || + if (nfs_get_option_ul(args, &option) || option < NFS_MNT_VERSION || option > NFS_MNT3_VERSION) goto out_invalid_value; mnt->mount_server.version = option; break; case Opt_nfsvers: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0) + if (nfs_get_option_ul(args, &option)) goto out_invalid_value; switch (option) { case NFS2_VERSION: @@ -1295,12 +1272,7 @@ static int nfs_parse_mount_options(char *raw, } break; case Opt_minorversion: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - rc = strict_strtoul(string, 10, &option); - kfree(string); - if (rc != 0) + if (nfs_get_option_ul(args, &option)) goto out_invalid_value; if (option > NFS4_MAX_MINOR_VERSION) goto out_invalid_value; @@ -1336,21 +1308,18 @@ static int nfs_parse_mount_options(char *raw, case Opt_xprt_udp: mnt->flags &= ~NFS_MOUNT_TCP; mnt->nfs_server.protocol = XPRT_TRANSPORT_UDP; - kfree(string); break; case Opt_xprt_tcp6: protofamily = AF_INET6; case Opt_xprt_tcp: mnt->flags |= NFS_MOUNT_TCP; mnt->nfs_server.protocol = XPRT_TRANSPORT_TCP; - kfree(string); break; case Opt_xprt_rdma: /* vector side protocols to TCP */ mnt->flags |= NFS_MOUNT_TCP; mnt->nfs_server.protocol = XPRT_TRANSPORT_RDMA; xprt_load_transport(string); - kfree(string); break; default: dfprintk(MOUNT, "NFS: unrecognized " @@ -1358,6 +1327,7 @@ static int nfs_parse_mount_options(char *raw, kfree(string); return 0; } + kfree(string); break; case Opt_mountproto: string = match_strdup(args); @@ -1400,18 +1370,13 @@ static int nfs_parse_mount_options(char *raw, goto out_invalid_address; break; case Opt_clientaddr: - string = match_strdup(args); - if (string == NULL) + if (nfs_get_option_str(args, &mnt->client_address)) goto out_nomem; - kfree(mnt->client_address); - mnt->client_address = string; break; case Opt_mounthost: - string = match_strdup(args); - if (string == NULL) + if (nfs_get_option_str(args, + &mnt->mount_server.hostname)) goto out_nomem; - kfree(mnt->mount_server.hostname); - mnt->mount_server.hostname = string; break; case Opt_mountaddr: string = match_strdup(args); @@ -1451,11 +1416,8 @@ static int nfs_parse_mount_options(char *raw, }; break; case Opt_fscache_uniq: - string = match_strdup(args); - if (string == NULL) + if (nfs_get_option_str(args, &mnt->fscache_uniq)) goto out_nomem; - kfree(mnt->fscache_uniq); - mnt->fscache_uniq = string; mnt->options |= NFS_OPTION_FSCACHE; break; case Opt_local_lock: @@ -1665,99 +1627,59 @@ static int nfs_try_mount(struct nfs_parsed_mount_data *args, return nfs_walk_authlist(args, &request); } -static int nfs_parse_simple_hostname(const char *dev_name, - char **hostname, size_t maxnamlen, - char **export_path, size_t maxpathlen) +/* + * Split "dev_name" into "hostname:export_path". + * + * The leftmost colon demarks the split between the server's hostname + * and the export path. If the hostname starts with a left square + * bracket, then it may contain colons. + * + * Note: caller frees hostname and export path, even on error. + */ +static int nfs_parse_devname(const char *dev_name, + char **hostname, size_t maxnamlen, + char **export_path, size_t maxpathlen) { size_t len; - char *colon, *comma; - - colon = strchr(dev_name, ':'); - if (colon == NULL) - goto out_bad_devname; - - len = colon - dev_name; - if (len > maxnamlen) - goto out_hostname; + char *end; - /* N.B. caller will free nfs_server.hostname in all cases */ - *hostname = kstrndup(dev_name, len, GFP_KERNEL); - if (!*hostname) - goto out_nomem; - - /* kill possible hostname list: not supported */ - comma = strchr(*hostname, ','); - if (comma != NULL) { - if (comma == *hostname) + /* Is the host name protected with square brakcets? */ + if (*dev_name == '[') { + end = strchr(++dev_name, ']'); + if (end == NULL || end[1] != ':') goto out_bad_devname; - *comma = '\0'; - } - colon++; - len = strlen(colon); - if (len > maxpathlen) - goto out_path; - *export_path = kstrndup(colon, len, GFP_KERNEL); - if (!*export_path) - goto out_nomem; - - dfprintk(MOUNT, "NFS: MNTPATH: '%s'\n", *export_path); - return 0; - -out_bad_devname: - dfprintk(MOUNT, "NFS: device name not in host:path format\n"); - return -EINVAL; - -out_nomem: - dfprintk(MOUNT, "NFS: not enough memory to parse device name\n"); - return -ENOMEM; - -out_hostname: - dfprintk(MOUNT, "NFS: server hostname too long\n"); - return -ENAMETOOLONG; - -out_path: - dfprintk(MOUNT, "NFS: export pathname too long\n"); - return -ENAMETOOLONG; -} - -/* - * Hostname has square brackets around it because it contains one or - * more colons. We look for the first closing square bracket, and a - * colon must follow it. - */ -static int nfs_parse_protected_hostname(const char *dev_name, - char **hostname, size_t maxnamlen, - char **export_path, size_t maxpathlen) -{ - size_t len; - char *start, *end; + len = end - dev_name; + end++; + } else { + char *comma; - start = (char *)(dev_name + 1); + end = strchr(dev_name, ':'); + if (end == NULL) + goto out_bad_devname; + len = end - dev_name; - end = strchr(start, ']'); - if (end == NULL) - goto out_bad_devname; - if (*(end + 1) != ':') - goto out_bad_devname; + /* kill possible hostname list: not supported */ + comma = strchr(dev_name, ','); + if (comma != NULL && comma < end) + *comma = 0; + } - len = end - start; if (len > maxnamlen) goto out_hostname; /* N.B. caller will free nfs_server.hostname in all cases */ - *hostname = kstrndup(start, len, GFP_KERNEL); + *hostname = kstrndup(dev_name, len, GFP_KERNEL); if (*hostname == NULL) goto out_nomem; - - end += 2; - len = strlen(end); + len = strlen(++end); if (len > maxpathlen) goto out_path; *export_path = kstrndup(end, len, GFP_KERNEL); if (!*export_path) goto out_nomem; + dfprintk(MOUNT, "NFS: MNTPATH: '%s'\n", *export_path); return 0; out_bad_devname: @@ -1778,29 +1700,6 @@ out_path: } /* - * Split "dev_name" into "hostname:export_path". - * - * The leftmost colon demarks the split between the server's hostname - * and the export path. If the hostname starts with a left square - * bracket, then it may contain colons. - * - * Note: caller frees hostname and export path, even on error. - */ -static int nfs_parse_devname(const char *dev_name, - char **hostname, size_t maxnamlen, - char **export_path, size_t maxpathlen) -{ - if (*dev_name == '[') - return nfs_parse_protected_hostname(dev_name, - hostname, maxnamlen, - export_path, maxpathlen); - - return nfs_parse_simple_hostname(dev_name, - hostname, maxnamlen, - export_path, maxpathlen); -} - -/* * Validate the NFS2/NFS3 mount data * - fills in the mount root filehandle * @@ -2267,19 +2166,19 @@ static int nfs_bdi_register(struct nfs_server *server) return bdi_register_dev(&server->backing_dev_info, server->s_dev); } -static int nfs_get_sb(struct file_system_type *fs_type, - int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt) +static struct dentry *nfs_fs_mount(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data) { struct nfs_server *server = NULL; struct super_block *s; struct nfs_parsed_mount_data *data; struct nfs_fh *mntfh; - struct dentry *mntroot; + struct dentry *mntroot = ERR_PTR(-ENOMEM); int (*compare_super)(struct super_block *, void *) = nfs_compare_super; struct nfs_sb_mountdata sb_mntdata = { .mntflags = flags, }; - int error = -ENOMEM; + int error; data = nfs_alloc_parsed_mount_data(NFS_DEFAULT_VERSION); mntfh = nfs_alloc_fhandle(); @@ -2290,12 +2189,14 @@ static int nfs_get_sb(struct file_system_type *fs_type, /* Validate the mount data */ error = nfs_validate_mount_data(raw_data, data, mntfh, dev_name); - if (error < 0) + if (error < 0) { + mntroot = ERR_PTR(error); goto out; + } #ifdef CONFIG_NFS_V4 if (data->version == 4) { - error = nfs4_try_mount(flags, dev_name, data, mnt); + mntroot = nfs4_try_mount(flags, dev_name, data); kfree(data->client_address); kfree(data->nfs_server.export_path); goto out; @@ -2305,7 +2206,7 @@ static int nfs_get_sb(struct file_system_type *fs_type, /* Get a volume representation */ server = nfs_create_server(data, mntfh); if (IS_ERR(server)) { - error = PTR_ERR(server); + mntroot = ERR_CAST(server); goto out; } sb_mntdata.server = server; @@ -2316,7 +2217,7 @@ static int nfs_get_sb(struct file_system_type *fs_type, /* Get a superblock - note that we may end up sharing one that already exists */ s = sget(fs_type, compare_super, nfs_set_super, &sb_mntdata); if (IS_ERR(s)) { - error = PTR_ERR(s); + mntroot = ERR_CAST(s); goto out_err_nosb; } @@ -2325,8 +2226,10 @@ static int nfs_get_sb(struct file_system_type *fs_type, server = NULL; } else { error = nfs_bdi_register(server); - if (error) + if (error) { + mntroot = ERR_PTR(error); goto error_splat_bdi; + } } if (!s->s_root) { @@ -2336,20 +2239,15 @@ static int nfs_get_sb(struct file_system_type *fs_type, s, data ? data->fscache_uniq : NULL, NULL); } - mntroot = nfs_get_root(s, mntfh); - if (IS_ERR(mntroot)) { - error = PTR_ERR(mntroot); + mntroot = nfs_get_root(s, mntfh, dev_name); + if (IS_ERR(mntroot)) goto error_splat_super; - } error = security_sb_set_mnt_opts(s, &data->lsm_opts); if (error) goto error_splat_root; s->s_flags |= MS_ACTIVE; - mnt->mnt_sb = s; - mnt->mnt_root = mntroot; - error = 0; out: kfree(data->nfs_server.hostname); @@ -2359,7 +2257,7 @@ out: out_free_fh: nfs_free_fhandle(mntfh); kfree(data); - return error; + return mntroot; out_err_nosb: nfs_free_server(server); @@ -2367,6 +2265,7 @@ out_err_nosb: error_splat_root: dput(mntroot); + mntroot = ERR_PTR(error); error_splat_super: if (server && !s->s_root) bdi_unregister(&server->backing_dev_info); @@ -2450,7 +2349,7 @@ nfs_xdev_mount(struct file_system_type *fs_type, int flags, nfs_fscache_get_super_cookie(s, NULL, data); } - mntroot = nfs_get_root(s, data->fh); + mntroot = nfs_get_root(s, data->fh, dev_name); if (IS_ERR(mntroot)) { error = PTR_ERR(mntroot); goto error_splat_super; @@ -2718,7 +2617,7 @@ nfs4_remote_mount(struct file_system_type *fs_type, int flags, s, data ? data->fscache_uniq : NULL, NULL); } - mntroot = nfs4_get_root(s, mntfh); + mntroot = nfs4_get_root(s, mntfh, dev_name); if (IS_ERR(mntroot)) { error = PTR_ERR(mntroot); goto error_splat_super; @@ -2771,27 +2670,6 @@ static struct vfsmount *nfs_do_root_mount(struct file_system_type *fs_type, return root_mnt; } -static void nfs_fix_devname(const struct path *path, struct vfsmount *mnt) -{ - char *page = (char *) __get_free_page(GFP_KERNEL); - char *devname, *tmp; - - if (page == NULL) - return; - devname = nfs_path(path->mnt->mnt_devname, - path->mnt->mnt_root, path->dentry, - page, PAGE_SIZE); - if (IS_ERR(devname)) - goto out_freepage; - tmp = kstrdup(devname, GFP_KERNEL); - if (tmp == NULL) - goto out_freepage; - kfree(mnt->mnt_devname); - mnt->mnt_devname = tmp; -out_freepage: - free_page((unsigned long)page); -} - struct nfs_referral_count { struct list_head list; const struct task_struct *task; @@ -2858,17 +2736,18 @@ static void nfs_referral_loop_unprotect(void) kfree(p); } -static int nfs_follow_remote_path(struct vfsmount *root_mnt, - const char *export_path, struct vfsmount *mnt_target) +static struct dentry *nfs_follow_remote_path(struct vfsmount *root_mnt, + const char *export_path) { struct nameidata *nd = NULL; struct mnt_namespace *ns_private; struct super_block *s; + struct dentry *dentry; int ret; nd = kmalloc(sizeof(*nd), GFP_KERNEL); if (nd == NULL) - return -ENOMEM; + return ERR_PTR(-ENOMEM); ns_private = create_mnt_ns(root_mnt); ret = PTR_ERR(ns_private); @@ -2890,32 +2769,27 @@ static int nfs_follow_remote_path(struct vfsmount *root_mnt, s = nd->path.mnt->mnt_sb; atomic_inc(&s->s_active); - mnt_target->mnt_sb = s; - mnt_target->mnt_root = dget(nd->path.dentry); - - /* Correct the device pathname */ - nfs_fix_devname(&nd->path, mnt_target); + dentry = dget(nd->path.dentry); path_put(&nd->path); kfree(nd); down_write(&s->s_umount); - return 0; + return dentry; out_put_mnt_ns: put_mnt_ns(ns_private); out_mntput: mntput(root_mnt); out_err: kfree(nd); - return ret; + return ERR_PTR(ret); } -static int nfs4_try_mount(int flags, const char *dev_name, - struct nfs_parsed_mount_data *data, - struct vfsmount *mnt) +static struct dentry *nfs4_try_mount(int flags, const char *dev_name, + struct nfs_parsed_mount_data *data) { char *export_path; struct vfsmount *root_mnt; - int error; + struct dentry *res; dfprintk(MOUNT, "--> nfs4_try_mount()\n"); @@ -2925,26 +2799,25 @@ static int nfs4_try_mount(int flags, const char *dev_name, data->nfs_server.hostname); data->nfs_server.export_path = export_path; - error = PTR_ERR(root_mnt); - if (IS_ERR(root_mnt)) - goto out; - - error = nfs_follow_remote_path(root_mnt, export_path, mnt); + res = ERR_CAST(root_mnt); + if (!IS_ERR(root_mnt)) + res = nfs_follow_remote_path(root_mnt, export_path); -out: - dfprintk(MOUNT, "<-- nfs4_try_mount() = %d%s\n", error, - error != 0 ? " [error]" : ""); - return error; + dfprintk(MOUNT, "<-- nfs4_try_mount() = %ld%s\n", + IS_ERR(res) ? PTR_ERR(res) : 0, + IS_ERR(res) ? " [error]" : ""); + return res; } /* * Get the superblock for an NFS4 mountpoint */ -static int nfs4_get_sb(struct file_system_type *fs_type, - int flags, const char *dev_name, void *raw_data, struct vfsmount *mnt) +static struct dentry *nfs4_mount(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data) { struct nfs_parsed_mount_data *data; int error = -ENOMEM; + struct dentry *res = ERR_PTR(-ENOMEM); data = nfs_alloc_parsed_mount_data(4); if (data == NULL) @@ -2952,10 +2825,14 @@ static int nfs4_get_sb(struct file_system_type *fs_type, /* Validate the mount data */ error = nfs4_validate_mount_data(raw_data, data, dev_name); - if (error < 0) + if (error < 0) { + res = ERR_PTR(error); goto out; + } - error = nfs4_try_mount(flags, dev_name, data, mnt); + res = nfs4_try_mount(flags, dev_name, data); + if (IS_ERR(res)) + error = PTR_ERR(res); out: kfree(data->client_address); @@ -2964,9 +2841,9 @@ out: kfree(data->fscache_uniq); out_free_data: kfree(data); - dprintk("<-- nfs4_get_sb() = %d%s\n", error, + dprintk("<-- nfs4_mount() = %d%s\n", error, error != 0 ? " [error]" : ""); - return error; + return res; } static void nfs4_kill_super(struct super_block *sb) @@ -3033,7 +2910,7 @@ nfs4_xdev_mount(struct file_system_type *fs_type, int flags, nfs_fscache_get_super_cookie(s, NULL, data); } - mntroot = nfs4_get_root(s, data->fh); + mntroot = nfs4_get_root(s, data->fh, dev_name); if (IS_ERR(mntroot)) { error = PTR_ERR(mntroot); goto error_splat_super; @@ -3120,7 +2997,7 @@ nfs4_remote_referral_mount(struct file_system_type *fs_type, int flags, nfs_fscache_get_super_cookie(s, NULL, data); } - mntroot = nfs4_get_root(s, mntfh); + mntroot = nfs4_get_root(s, mntfh, dev_name); if (IS_ERR(mntroot)) { error = PTR_ERR(mntroot); goto error_splat_super; @@ -3160,16 +3037,15 @@ error_splat_bdi: /* * Create an NFS4 server record on referral traversal */ -static int nfs4_referral_get_sb(struct file_system_type *fs_type, - int flags, const char *dev_name, void *raw_data, - struct vfsmount *mnt) +static struct dentry *nfs4_referral_mount(struct file_system_type *fs_type, + int flags, const char *dev_name, void *raw_data) { struct nfs_clone_mount *data = raw_data; char *export_path; struct vfsmount *root_mnt; - int error; + struct dentry *res; - dprintk("--> nfs4_referral_get_sb()\n"); + dprintk("--> nfs4_referral_mount()\n"); export_path = data->mnt_path; data->mnt_path = "/"; @@ -3178,15 +3054,13 @@ static int nfs4_referral_get_sb(struct file_system_type *fs_type, flags, data, data->hostname); data->mnt_path = export_path; - error = PTR_ERR(root_mnt); - if (IS_ERR(root_mnt)) - goto out; - - error = nfs_follow_remote_path(root_mnt, export_path, mnt); -out: - dprintk("<-- nfs4_referral_get_sb() = %d%s\n", error, - error != 0 ? " [error]" : ""); - return error; + res = ERR_CAST(root_mnt); + if (!IS_ERR(root_mnt)) + res = nfs_follow_remote_path(root_mnt, export_path); + dprintk("<-- nfs4_referral_mount() = %ld%s\n", + IS_ERR(res) ? PTR_ERR(res) : 0, + IS_ERR(res) ? " [error]" : ""); + return res; } #endif /* CONFIG_NFS_V4 */ diff --git a/fs/nfs/unlink.c b/fs/nfs/unlink.c index e313a51..8d6864c 100644 --- a/fs/nfs/unlink.c +++ b/fs/nfs/unlink.c @@ -148,6 +148,7 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n alias = d_lookup(parent, &data->args.name); if (alias != NULL) { int ret = 0; + void *devname_garbage = NULL; /* * Hey, we raced with lookup... See if we need to transfer @@ -157,6 +158,7 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n spin_lock(&alias->d_lock); if (alias->d_inode != NULL && !(alias->d_flags & DCACHE_NFSFS_RENAMED)) { + devname_garbage = alias->d_fsdata; alias->d_fsdata = data; alias->d_flags |= DCACHE_NFSFS_RENAMED; ret = 1; @@ -164,6 +166,13 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n spin_unlock(&alias->d_lock); nfs_dec_sillycount(dir); dput(alias); + /* + * If we'd displaced old cached devname, free it. At that + * point dentry is definitely not a root, so we won't need + * that anymore. + */ + if (devname_garbage) + kfree(devname_garbage); return ret; } data->dir = igrab(dir); @@ -180,7 +189,7 @@ static int nfs_do_call_unlink(struct dentry *parent, struct inode *dir, struct n task_setup_data.rpc_client = NFS_CLIENT(dir); task = rpc_run_task(&task_setup_data); if (!IS_ERR(task)) - rpc_put_task(task); + rpc_put_task_async(task); return 1; } @@ -252,6 +261,7 @@ nfs_async_unlink(struct inode *dir, struct dentry *dentry) { struct nfs_unlinkdata *data; int status = -ENOMEM; + void *devname_garbage = NULL; data = kzalloc(sizeof(*data), GFP_KERNEL); if (data == NULL) @@ -269,8 +279,16 @@ nfs_async_unlink(struct inode *dir, struct dentry *dentry) if (dentry->d_flags & DCACHE_NFSFS_RENAMED) goto out_unlock; dentry->d_flags |= DCACHE_NFSFS_RENAMED; + devname_garbage = dentry->d_fsdata; dentry->d_fsdata = data; spin_unlock(&dentry->d_lock); + /* + * If we'd displaced old cached devname, free it. At that + * point dentry is definitely not a root, so we won't need + * that anymore. + */ + if (devname_garbage) + kfree(devname_garbage); return 0; out_unlock: spin_unlock(&dentry->d_lock); @@ -299,6 +317,7 @@ nfs_complete_unlink(struct dentry *dentry, struct inode *inode) if (dentry->d_flags & DCACHE_NFSFS_RENAMED) { dentry->d_flags &= ~DCACHE_NFSFS_RENAMED; data = dentry->d_fsdata; + dentry->d_fsdata = NULL; } spin_unlock(&dentry->d_lock); @@ -315,6 +334,7 @@ nfs_cancel_async_unlink(struct dentry *dentry) struct nfs_unlinkdata *data = dentry->d_fsdata; dentry->d_flags &= ~DCACHE_NFSFS_RENAMED; + dentry->d_fsdata = NULL; spin_unlock(&dentry->d_lock); nfs_free_unlinkdata(data); return; diff --git a/fs/nfs/write.c b/fs/nfs/write.c index 10d648ea..47a3ad6 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -28,6 +28,7 @@ #include "iostat.h" #include "nfs4_fs.h" #include "fscache.h" +#include "pnfs.h" #define NFSDBG_FACILITY NFSDBG_PAGECACHE @@ -96,6 +97,7 @@ void nfs_writedata_free(struct nfs_write_data *p) static void nfs_writedata_release(struct nfs_write_data *wdata) { + put_lseg(wdata->lseg); put_nfs_open_context(wdata->args.context); nfs_writedata_free(wdata); } @@ -781,25 +783,21 @@ static int flush_task_priority(int how) return RPC_PRIORITY_NORMAL; } -/* - * Set up the argument/result storage required for the RPC call. - */ -static int nfs_write_rpcsetup(struct nfs_page *req, - struct nfs_write_data *data, - const struct rpc_call_ops *call_ops, - unsigned int count, unsigned int offset, - int how) +int nfs_initiate_write(struct nfs_write_data *data, + struct rpc_clnt *clnt, + const struct rpc_call_ops *call_ops, + int how) { - struct inode *inode = req->wb_context->path.dentry->d_inode; + struct inode *inode = data->inode; int priority = flush_task_priority(how); struct rpc_task *task; struct rpc_message msg = { .rpc_argp = &data->args, .rpc_resp = &data->res, - .rpc_cred = req->wb_context->cred, + .rpc_cred = data->cred, }; struct rpc_task_setup task_setup_data = { - .rpc_client = NFS_CLIENT(inode), + .rpc_client = clnt, .task = &data->task, .rpc_message = &msg, .callback_ops = call_ops, @@ -810,12 +808,52 @@ static int nfs_write_rpcsetup(struct nfs_page *req, }; int ret = 0; + /* Set up the initial task struct. */ + NFS_PROTO(inode)->write_setup(data, &msg); + + dprintk("NFS: %5u initiated write call " + "(req %s/%lld, %u bytes @ offset %llu)\n", + data->task.tk_pid, + inode->i_sb->s_id, + (long long)NFS_FILEID(inode), + data->args.count, + (unsigned long long)data->args.offset); + + task = rpc_run_task(&task_setup_data); + if (IS_ERR(task)) { + ret = PTR_ERR(task); + goto out; + } + if (how & FLUSH_SYNC) { + ret = rpc_wait_for_completion_task(task); + if (ret == 0) + ret = task->tk_status; + } + rpc_put_task(task); +out: + return ret; +} +EXPORT_SYMBOL_GPL(nfs_initiate_write); + +/* + * Set up the argument/result storage required for the RPC call. + */ +static int nfs_write_rpcsetup(struct nfs_page *req, + struct nfs_write_data *data, + const struct rpc_call_ops *call_ops, + unsigned int count, unsigned int offset, + struct pnfs_layout_segment *lseg, + int how) +{ + struct inode *inode = req->wb_context->path.dentry->d_inode; + /* Set up the RPC argument and reply structs * NB: take care not to mess about with data->commit et al. */ data->req = req; data->inode = inode = req->wb_context->path.dentry->d_inode; - data->cred = msg.rpc_cred; + data->cred = req->wb_context->cred; + data->lseg = get_lseg(lseg); data->args.fh = NFS_FH(inode); data->args.offset = req_offset(req) + offset; @@ -836,30 +874,11 @@ static int nfs_write_rpcsetup(struct nfs_page *req, data->res.verf = &data->verf; nfs_fattr_init(&data->fattr); - /* Set up the initial task struct. */ - NFS_PROTO(inode)->write_setup(data, &msg); - - dprintk("NFS: %5u initiated write call " - "(req %s/%lld, %u bytes @ offset %llu)\n", - data->task.tk_pid, - inode->i_sb->s_id, - (long long)NFS_FILEID(inode), - count, - (unsigned long long)data->args.offset); + if (data->lseg && + (pnfs_try_to_write_data(data, call_ops, how) == PNFS_ATTEMPTED)) + return 0; - task = rpc_run_task(&task_setup_data); - if (IS_ERR(task)) { - ret = PTR_ERR(task); - goto out; - } - if (how & FLUSH_SYNC) { - ret = rpc_wait_for_completion_task(task); - if (ret == 0) - ret = task->tk_status; - } - rpc_put_task(task); -out: - return ret; + return nfs_initiate_write(data, NFS_CLIENT(inode), call_ops, how); } /* If a nfs_flush_* function fails, it should remove reqs from @head and @@ -879,20 +898,21 @@ static void nfs_redirty_request(struct nfs_page *req) * Generate multiple small requests to write out a single * contiguous dirty area on one page. */ -static int nfs_flush_multi(struct inode *inode, struct list_head *head, unsigned int npages, size_t count, int how) +static int nfs_flush_multi(struct nfs_pageio_descriptor *desc) { - struct nfs_page *req = nfs_list_entry(head->next); + struct nfs_page *req = nfs_list_entry(desc->pg_list.next); struct page *page = req->wb_page; struct nfs_write_data *data; - size_t wsize = NFS_SERVER(inode)->wsize, nbytes; + size_t wsize = NFS_SERVER(desc->pg_inode)->wsize, nbytes; unsigned int offset; int requests = 0; int ret = 0; + struct pnfs_layout_segment *lseg; LIST_HEAD(list); nfs_list_remove_request(req); - nbytes = count; + nbytes = desc->pg_count; do { size_t len = min(nbytes, wsize); @@ -905,9 +925,11 @@ static int nfs_flush_multi(struct inode *inode, struct list_head *head, unsigned } while (nbytes != 0); atomic_set(&req->wb_complete, requests); + BUG_ON(desc->pg_lseg); + lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_RW); ClearPageError(page); offset = 0; - nbytes = count; + nbytes = desc->pg_count; do { int ret2; @@ -919,20 +941,22 @@ static int nfs_flush_multi(struct inode *inode, struct list_head *head, unsigned if (nbytes < wsize) wsize = nbytes; ret2 = nfs_write_rpcsetup(req, data, &nfs_write_partial_ops, - wsize, offset, how); + wsize, offset, lseg, desc->pg_ioflags); if (ret == 0) ret = ret2; offset += wsize; nbytes -= wsize; } while (nbytes != 0); + put_lseg(lseg); + desc->pg_lseg = NULL; return ret; out_bad: while (!list_empty(&list)) { data = list_entry(list.next, struct nfs_write_data, pages); list_del(&data->pages); - nfs_writedata_release(data); + nfs_writedata_free(data); } nfs_redirty_request(req); return -ENOMEM; @@ -946,16 +970,26 @@ out_bad: * This is the case if nfs_updatepage detects a conflicting request * that has been written but not committed. */ -static int nfs_flush_one(struct inode *inode, struct list_head *head, unsigned int npages, size_t count, int how) +static int nfs_flush_one(struct nfs_pageio_descriptor *desc) { struct nfs_page *req; struct page **pages; struct nfs_write_data *data; + struct list_head *head = &desc->pg_list; + struct pnfs_layout_segment *lseg = desc->pg_lseg; + int ret; - data = nfs_writedata_alloc(npages); - if (!data) - goto out_bad; - + data = nfs_writedata_alloc(nfs_page_array_len(desc->pg_base, + desc->pg_count)); + if (!data) { + while (!list_empty(head)) { + req = nfs_list_entry(head->next); + nfs_list_remove_request(req); + nfs_redirty_request(req); + } + ret = -ENOMEM; + goto out; + } pages = data->pagevec; while (!list_empty(head)) { req = nfs_list_entry(head->next); @@ -965,16 +999,15 @@ static int nfs_flush_one(struct inode *inode, struct list_head *head, unsigned i *pages++ = req->wb_page; } req = nfs_list_entry(data->pages.next); + if ((!lseg) && list_is_singular(&data->pages)) + lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_RW); /* Set up the argument struct */ - return nfs_write_rpcsetup(req, data, &nfs_write_full_ops, count, 0, how); - out_bad: - while (!list_empty(head)) { - req = nfs_list_entry(head->next); - nfs_list_remove_request(req); - nfs_redirty_request(req); - } - return -ENOMEM; + ret = nfs_write_rpcsetup(req, data, &nfs_write_full_ops, desc->pg_count, 0, lseg, desc->pg_ioflags); +out: + put_lseg(lseg); /* Cleans any gotten in ->pg_test */ + desc->pg_lseg = NULL; + return ret; } static void nfs_pageio_init_write(struct nfs_pageio_descriptor *pgio, @@ -982,6 +1015,8 @@ static void nfs_pageio_init_write(struct nfs_pageio_descriptor *pgio, { size_t wsize = NFS_SERVER(inode)->wsize; + pnfs_pageio_init_write(pgio, inode); + if (wsize < PAGE_CACHE_SIZE) nfs_pageio_init(pgio, inode, nfs_flush_multi, wsize, ioflags); else @@ -1132,7 +1167,7 @@ static const struct rpc_call_ops nfs_write_full_ops = { /* * This function is called when the WRITE call is complete. */ -int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) +void nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) { struct nfs_writeargs *argp = &data->args; struct nfs_writeres *resp = &data->res; @@ -1151,7 +1186,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) */ status = NFS_PROTO(data->inode)->write_done(task, data); if (status != 0) - return status; + return; nfs_add_stats(data->inode, NFSIOS_SERVERWRITTENBYTES, resp->count); #if defined(CONFIG_NFS_V3) || defined(CONFIG_NFS_V4) @@ -1166,6 +1201,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) */ static unsigned long complain; + /* Note this will print the MDS for a DS write */ if (time_before(complain, jiffies)) { dprintk("NFS: faulty NFS server %s:" " (committed = %d) != (stable = %d)\n", @@ -1186,6 +1222,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) /* Was this an NFSv2 write or an NFSv3 stable write? */ if (resp->verf->committed != NFS_UNSTABLE) { /* Resend from where the server left off */ + data->mds_offset += resp->count; argp->offset += resp->count; argp->pgbase += resp->count; argp->count -= resp->count; @@ -1196,7 +1233,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) argp->stable = NFS_FILE_SYNC; } nfs_restart_rpc(task, server->nfs_client); - return -EAGAIN; + return; } if (time_before(complain, jiffies)) { printk(KERN_WARNING @@ -1207,7 +1244,7 @@ int nfs_writeback_done(struct rpc_task *task, struct nfs_write_data *data) /* Can't do anything about it except throw an error. */ task->tk_status = -EIO; } - return 0; + return; } @@ -1292,6 +1329,8 @@ static int nfs_commit_rpcsetup(struct list_head *head, task = rpc_run_task(&task_setup_data); if (IS_ERR(task)) return PTR_ERR(task); + if (how & FLUSH_SYNC) + rpc_wait_for_completion_task(task); rpc_put_task(task); return 0; } diff --git a/fs/nfs_common/nfsacl.c b/fs/nfs_common/nfsacl.c index fc1c525..84c27d6 100644 --- a/fs/nfs_common/nfsacl.c +++ b/fs/nfs_common/nfsacl.c @@ -42,6 +42,11 @@ struct nfsacl_encode_desc { gid_t gid; }; +struct nfsacl_simple_acl { + struct posix_acl acl; + struct posix_acl_entry ace[4]; +}; + static int xdr_nfsace_encode(struct xdr_array2_desc *desc, void *elem) { @@ -72,9 +77,20 @@ xdr_nfsace_encode(struct xdr_array2_desc *desc, void *elem) return 0; } -unsigned int -nfsacl_encode(struct xdr_buf *buf, unsigned int base, struct inode *inode, - struct posix_acl *acl, int encode_entries, int typeflag) +/** + * nfsacl_encode - Encode an NFSv3 ACL + * + * @buf: destination xdr_buf to contain XDR encoded ACL + * @base: byte offset in xdr_buf where XDR'd ACL begins + * @inode: inode of file whose ACL this is + * @acl: posix_acl to encode + * @encode_entries: whether to encode ACEs as well + * @typeflag: ACL type: NFS_ACL_DEFAULT or zero + * + * Returns size of encoded ACL in bytes or a negative errno value. + */ +int nfsacl_encode(struct xdr_buf *buf, unsigned int base, struct inode *inode, + struct posix_acl *acl, int encode_entries, int typeflag) { int entries = (acl && acl->a_count) ? max_t(int, acl->a_count, 4) : 0; struct nfsacl_encode_desc nfsacl_desc = { @@ -88,17 +104,22 @@ nfsacl_encode(struct xdr_buf *buf, unsigned int base, struct inode *inode, .uid = inode->i_uid, .gid = inode->i_gid, }; + struct nfsacl_simple_acl aclbuf; int err; - struct posix_acl *acl2 = NULL; if (entries > NFS_ACL_MAX_ENTRIES || xdr_encode_word(buf, base, entries)) return -EINVAL; if (encode_entries && acl && acl->a_count == 3) { - /* Fake up an ACL_MASK entry. */ - acl2 = posix_acl_alloc(4, GFP_KERNEL); - if (!acl2) - return -ENOMEM; + struct posix_acl *acl2 = &aclbuf.acl; + + /* Avoid the use of posix_acl_alloc(). nfsacl_encode() is + * invoked in contexts where a memory allocation failure is + * fatal. Fortunately this fake ACL is small enough to + * construct on the stack. */ + memset(acl2, 0, sizeof(acl2)); + posix_acl_init(acl2, 4); + /* Insert entries in canonical order: other orders seem to confuse Solaris VxFS. */ acl2->a_entries[0] = acl->a_entries[0]; /* ACL_USER_OBJ */ @@ -109,8 +130,6 @@ nfsacl_encode(struct xdr_buf *buf, unsigned int base, struct inode *inode, nfsacl_desc.acl = acl2; } err = xdr_encode_array2(buf, base + 4, &nfsacl_desc.desc); - if (acl2) - posix_acl_release(acl2); if (!err) err = 8 + nfsacl_desc.desc.elem_size * nfsacl_desc.desc.array_len; @@ -224,9 +243,18 @@ posix_acl_from_nfsacl(struct posix_acl *acl) return 0; } -unsigned int -nfsacl_decode(struct xdr_buf *buf, unsigned int base, unsigned int *aclcnt, - struct posix_acl **pacl) +/** + * nfsacl_decode - Decode an NFSv3 ACL + * + * @buf: xdr_buf containing XDR'd ACL data to decode + * @base: byte offset in xdr_buf where XDR'd ACL begins + * @aclcnt: count of ACEs in decoded posix_acl + * @pacl: buffer in which to place decoded posix_acl + * + * Returns the length of the decoded ACL in bytes, or a negative errno value. + */ +int nfsacl_decode(struct xdr_buf *buf, unsigned int base, unsigned int *aclcnt, + struct posix_acl **pacl) { struct nfsacl_decode_desc nfsacl_desc = { .desc = { diff --git a/fs/nfsctl.c b/fs/nfsctl.c index bf9cbd2..124e8fc 100644 --- a/fs/nfsctl.c +++ b/fs/nfsctl.c @@ -22,30 +22,17 @@ static struct file *do_open(char *name, int flags) { - struct nameidata nd; struct vfsmount *mnt; - int error; + struct file *file; mnt = do_kern_mount("nfsd", 0, "nfsd", NULL); if (IS_ERR(mnt)) return (struct file *)mnt; - error = vfs_path_lookup(mnt->mnt_root, mnt, name, 0, &nd); - mntput(mnt); /* drop do_kern_mount reference */ - if (error) - return ERR_PTR(error); - - if (flags == O_RDWR) - error = may_open(&nd.path, MAY_READ|MAY_WRITE, flags); - else - error = may_open(&nd.path, MAY_WRITE, flags); + file = file_open_root(mnt->mnt_root, mnt, name, flags); - if (!error) - return dentry_open(nd.path.dentry, nd.path.mnt, flags, - current_cred()); - - path_put(&nd.path); - return ERR_PTR(error); + mntput(mnt); /* drop do_kern_mount reference */ + return file; } static struct { diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c index 3be975e..02eb4ed 100644 --- a/fs/nfsd/nfs4callback.c +++ b/fs/nfsd/nfs4callback.c @@ -432,7 +432,7 @@ static int decode_cb_sequence4resok(struct xdr_stream *xdr, * If the server returns different values for sessionID, slotID or * sequence number, the server is looney tunes. */ - p = xdr_inline_decode(xdr, NFS4_MAX_SESSIONID_LEN + 4 + 4); + p = xdr_inline_decode(xdr, NFS4_MAX_SESSIONID_LEN + 4 + 4 + 4 + 4); if (unlikely(p == NULL)) goto out_overflow; memcpy(id.data, p, NFS4_MAX_SESSIONID_LEN); @@ -484,7 +484,7 @@ static int decode_cb_sequence4res(struct xdr_stream *xdr, out: return status; out_default: - return nfs_cb_stat_to_errno(status); + return nfs_cb_stat_to_errno(nfserr); } /* @@ -564,11 +564,9 @@ static int nfs4_xdr_dec_cb_recall(struct rpc_rqst *rqstp, if (unlikely(status)) goto out; if (unlikely(nfserr != NFS4_OK)) - goto out_default; + status = nfs_cb_stat_to_errno(nfserr); out: return status; -out_default: - return nfs_cb_stat_to_errno(status); } /* diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index d98d021..7b566ec 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -230,9 +230,6 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f dp->dl_client = clp; get_nfs4_file(fp); dp->dl_file = fp; - dp->dl_vfs_file = find_readable_file(fp); - get_file(dp->dl_vfs_file); - dp->dl_flock = NULL; dp->dl_type = type; dp->dl_stateid.si_boot = boot_time; dp->dl_stateid.si_stateownerid = current_delegid++; @@ -241,8 +238,6 @@ alloc_init_deleg(struct nfs4_client *clp, struct nfs4_stateid *stp, struct svc_f fh_copy_shallow(&dp->dl_fh, ¤t_fh->fh_handle); dp->dl_time = 0; atomic_set(&dp->dl_count, 1); - list_add(&dp->dl_perfile, &fp->fi_delegations); - list_add(&dp->dl_perclnt, &clp->cl_delegations); INIT_WORK(&dp->dl_recall.cb_work, nfsd4_do_callback_rpc); return dp; } @@ -253,36 +248,30 @@ nfs4_put_delegation(struct nfs4_delegation *dp) if (atomic_dec_and_test(&dp->dl_count)) { dprintk("NFSD: freeing dp %p\n",dp); put_nfs4_file(dp->dl_file); - fput(dp->dl_vfs_file); kmem_cache_free(deleg_slab, dp); num_delegations--; } } -/* Remove the associated file_lock first, then remove the delegation. - * lease_modify() is called to remove the FS_LEASE file_lock from - * the i_flock list, eventually calling nfsd's lock_manager - * fl_release_callback. - */ -static void -nfs4_close_delegation(struct nfs4_delegation *dp) +static void nfs4_put_deleg_lease(struct nfs4_file *fp) { - dprintk("NFSD: close_delegation dp %p\n",dp); - /* XXX: do we even need this check?: */ - if (dp->dl_flock) - vfs_setlease(dp->dl_vfs_file, F_UNLCK, &dp->dl_flock); + if (atomic_dec_and_test(&fp->fi_delegees)) { + vfs_setlease(fp->fi_deleg_file, F_UNLCK, &fp->fi_lease); + fp->fi_lease = NULL; + fp->fi_deleg_file = NULL; + } } /* Called under the state lock. */ static void unhash_delegation(struct nfs4_delegation *dp) { - list_del_init(&dp->dl_perfile); list_del_init(&dp->dl_perclnt); spin_lock(&recall_lock); + list_del_init(&dp->dl_perfile); list_del_init(&dp->dl_recall_lru); spin_unlock(&recall_lock); - nfs4_close_delegation(dp); + nfs4_put_deleg_lease(dp->dl_file); nfs4_put_delegation(dp); } @@ -958,8 +947,6 @@ expire_client(struct nfs4_client *clp) spin_lock(&recall_lock); while (!list_empty(&clp->cl_delegations)) { dp = list_entry(clp->cl_delegations.next, struct nfs4_delegation, dl_perclnt); - dprintk("NFSD: expire client. dp %p, fp %p\n", dp, - dp->dl_flock); list_del_init(&dp->dl_perclnt); list_move(&dp->dl_recall_lru, &reaplist); } @@ -2078,6 +2065,7 @@ alloc_init_file(struct inode *ino) fp->fi_inode = igrab(ino); fp->fi_id = current_fileid++; fp->fi_had_conflict = false; + fp->fi_lease = NULL; memset(fp->fi_fds, 0, sizeof(fp->fi_fds)); memset(fp->fi_access, 0, sizeof(fp->fi_access)); spin_lock(&recall_lock); @@ -2329,23 +2317,8 @@ nfs4_file_downgrade(struct nfs4_file *fp, unsigned int share_access) nfs4_file_put_access(fp, O_RDONLY); } -/* - * Spawn a thread to perform a recall on the delegation represented - * by the lease (file_lock) - * - * Called from break_lease() with lock_flocks() held. - * Note: we assume break_lease will only call this *once* for any given - * lease. - */ -static -void nfsd_break_deleg_cb(struct file_lock *fl) +static void nfsd_break_one_deleg(struct nfs4_delegation *dp) { - struct nfs4_delegation *dp = (struct nfs4_delegation *)fl->fl_owner; - - dprintk("NFSD nfsd_break_deleg_cb: dp %p fl %p\n",dp,fl); - if (!dp) - return; - /* We're assuming the state code never drops its reference * without first removing the lease. Since we're in this lease * callback (and since the lease code is serialized by the kernel @@ -2353,22 +2326,35 @@ void nfsd_break_deleg_cb(struct file_lock *fl) * it's safe to take a reference: */ atomic_inc(&dp->dl_count); - spin_lock(&recall_lock); list_add_tail(&dp->dl_recall_lru, &del_recall_lru); - spin_unlock(&recall_lock); /* only place dl_time is set. protected by lock_flocks*/ dp->dl_time = get_seconds(); + nfsd4_cb_recall(dp); +} + +/* Called from break_lease() with lock_flocks() held. */ +static void nfsd_break_deleg_cb(struct file_lock *fl) +{ + struct nfs4_file *fp = (struct nfs4_file *)fl->fl_owner; + struct nfs4_delegation *dp; + + BUG_ON(!fp); + /* We assume break_lease is only called once per lease: */ + BUG_ON(fp->fi_had_conflict); /* * We don't want the locks code to timeout the lease for us; - * we'll remove it ourself if the delegation isn't returned - * in time. + * we'll remove it ourself if a delegation isn't returned + * in time: */ fl->fl_break_time = 0; - dp->dl_file->fi_had_conflict = true; - nfsd4_cb_recall(dp); + spin_lock(&recall_lock); + fp->fi_had_conflict = true; + list_for_each_entry(dp, &fp->fi_delegations, dl_perfile) + nfsd_break_one_deleg(dp); + spin_unlock(&recall_lock); } static @@ -2461,10 +2447,13 @@ find_delegation_file(struct nfs4_file *fp, stateid_t *stid) { struct nfs4_delegation *dp; - list_for_each_entry(dp, &fp->fi_delegations, dl_perfile) { - if (dp->dl_stateid.si_stateownerid == stid->si_stateownerid) + spin_lock(&recall_lock); + list_for_each_entry(dp, &fp->fi_delegations, dl_perfile) + if (dp->dl_stateid.si_stateownerid == stid->si_stateownerid) { + spin_unlock(&recall_lock); return dp; - } + } + spin_unlock(&recall_lock); return NULL; } @@ -2641,6 +2630,66 @@ static bool nfsd4_cb_channel_good(struct nfs4_client *clp) return clp->cl_minorversion && clp->cl_cb_state == NFSD4_CB_UNKNOWN; } +static struct file_lock *nfs4_alloc_init_lease(struct nfs4_delegation *dp, int flag) +{ + struct file_lock *fl; + + fl = locks_alloc_lock(); + if (!fl) + return NULL; + locks_init_lock(fl); + fl->fl_lmops = &nfsd_lease_mng_ops; + fl->fl_flags = FL_LEASE; + fl->fl_type = flag == NFS4_OPEN_DELEGATE_READ? F_RDLCK: F_WRLCK; + fl->fl_end = OFFSET_MAX; + fl->fl_owner = (fl_owner_t)(dp->dl_file); + fl->fl_pid = current->tgid; + return fl; +} + +static int nfs4_setlease(struct nfs4_delegation *dp, int flag) +{ + struct nfs4_file *fp = dp->dl_file; + struct file_lock *fl; + int status; + + fl = nfs4_alloc_init_lease(dp, flag); + if (!fl) + return -ENOMEM; + fl->fl_file = find_readable_file(fp); + list_add(&dp->dl_perclnt, &dp->dl_client->cl_delegations); + status = vfs_setlease(fl->fl_file, fl->fl_type, &fl); + if (status) { + list_del_init(&dp->dl_perclnt); + locks_free_lock(fl); + return -ENOMEM; + } + fp->fi_lease = fl; + fp->fi_deleg_file = fl->fl_file; + get_file(fp->fi_deleg_file); + atomic_set(&fp->fi_delegees, 1); + list_add(&dp->dl_perfile, &fp->fi_delegations); + return 0; +} + +static int nfs4_set_delegation(struct nfs4_delegation *dp, int flag) +{ + struct nfs4_file *fp = dp->dl_file; + + if (!fp->fi_lease) + return nfs4_setlease(dp, flag); + spin_lock(&recall_lock); + if (fp->fi_had_conflict) { + spin_unlock(&recall_lock); + return -EAGAIN; + } + atomic_inc(&fp->fi_delegees); + list_add(&dp->dl_perfile, &fp->fi_delegations); + spin_unlock(&recall_lock); + list_add(&dp->dl_perclnt, &dp->dl_client->cl_delegations); + return 0; +} + /* * Attempt to hand out a delegation. */ @@ -2650,7 +2699,6 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, struct nfs4_sta struct nfs4_delegation *dp; struct nfs4_stateowner *sop = stp->st_stateowner; int cb_up; - struct file_lock *fl; int status, flag = 0; cb_up = nfsd4_cb_channel_good(sop->so_client); @@ -2681,36 +2729,11 @@ nfs4_open_delegation(struct svc_fh *fh, struct nfsd4_open *open, struct nfs4_sta } dp = alloc_init_deleg(sop->so_client, stp, fh, flag); - if (dp == NULL) { - flag = NFS4_OPEN_DELEGATE_NONE; - goto out; - } - status = -ENOMEM; - fl = locks_alloc_lock(); - if (!fl) - goto out; - locks_init_lock(fl); - fl->fl_lmops = &nfsd_lease_mng_ops; - fl->fl_flags = FL_LEASE; - fl->fl_type = flag == NFS4_OPEN_DELEGATE_READ? F_RDLCK: F_WRLCK; - fl->fl_end = OFFSET_MAX; - fl->fl_owner = (fl_owner_t)dp; - fl->fl_file = find_readable_file(stp->st_file); - BUG_ON(!fl->fl_file); - fl->fl_pid = current->tgid; - dp->dl_flock = fl; - - /* vfs_setlease checks to see if delegation should be handed out. - * the lock_manager callback fl_change is used - */ - if ((status = vfs_setlease(fl->fl_file, fl->fl_type, &fl))) { - dprintk("NFSD: setlease failed [%d], no delegation\n", status); - dp->dl_flock = NULL; - locks_free_lock(fl); - unhash_delegation(dp); - flag = NFS4_OPEN_DELEGATE_NONE; - goto out; - } + if (dp == NULL) + goto out_no_deleg; + status = nfs4_set_delegation(dp, flag); + if (status) + goto out_free; memcpy(&open->op_delegate_stateid, &dp->dl_stateid, sizeof(dp->dl_stateid)); @@ -2722,6 +2745,12 @@ out: && open->op_delegate_type != NFS4_OPEN_DELEGATE_NONE) dprintk("NFSD: WARNING: refusing delegation reclaim\n"); open->op_delegate_type = flag; + return; +out_free: + nfs4_put_delegation(dp); +out_no_deleg: + flag = NFS4_OPEN_DELEGATE_NONE; + goto out; } /* @@ -2916,8 +2945,6 @@ nfs4_laundromat(void) test_val = u; break; } - dprintk("NFSD: purging unused delegation dp %p, fp %p\n", - dp, dp->dl_flock); list_move(&dp->dl_recall_lru, &reaplist); } spin_unlock(&recall_lock); @@ -3128,7 +3155,7 @@ nfs4_preprocess_stateid_op(struct nfsd4_compound_state *cstate, goto out; renew_client(dp->dl_client); if (filpp) { - *filpp = find_readable_file(dp->dl_file); + *filpp = dp->dl_file->fi_deleg_file; BUG_ON(!*filpp); } } else { /* open or lock stateid */ diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index 956629b..615f0a9 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -317,8 +317,8 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, READ_BUF(dummy32); len += (XDR_QUADLEN(dummy32) << 2); READMEM(buf, dummy32); - if ((host_err = nfsd_map_name_to_uid(argp->rqstp, buf, dummy32, &iattr->ia_uid))) - goto out_nfserr; + if ((status = nfsd_map_name_to_uid(argp->rqstp, buf, dummy32, &iattr->ia_uid))) + return status; iattr->ia_valid |= ATTR_UID; } if (bmval[1] & FATTR4_WORD1_OWNER_GROUP) { @@ -328,8 +328,8 @@ nfsd4_decode_fattr(struct nfsd4_compoundargs *argp, u32 *bmval, READ_BUF(dummy32); len += (XDR_QUADLEN(dummy32) << 2); READMEM(buf, dummy32); - if ((host_err = nfsd_map_name_to_gid(argp->rqstp, buf, dummy32, &iattr->ia_gid))) - goto out_nfserr; + if ((status = nfsd_map_name_to_gid(argp->rqstp, buf, dummy32, &iattr->ia_gid))) + return status; iattr->ia_valid |= ATTR_GID; } if (bmval[1] & FATTR4_WORD1_TIME_ACCESS_SET) { @@ -1142,7 +1142,7 @@ nfsd4_decode_create_session(struct nfsd4_compoundargs *argp, u32 dummy; char *machine_name; - int i; + int i, j; int nr_secflavs; READ_BUF(16); @@ -1215,7 +1215,7 @@ nfsd4_decode_create_session(struct nfsd4_compoundargs *argp, READ_BUF(4); READ32(dummy); READ_BUF(dummy * 4); - for (i = 0; i < dummy; ++i) + for (j = 0; j < dummy; ++j) READ32(dummy); break; case RPC_AUTH_GSS: diff --git a/fs/nfsd/state.h b/fs/nfsd/state.h index 3074656..2d31224 100644 --- a/fs/nfsd/state.h +++ b/fs/nfsd/state.h @@ -83,8 +83,6 @@ struct nfs4_delegation { atomic_t dl_count; /* ref count */ struct nfs4_client *dl_client; struct nfs4_file *dl_file; - struct file *dl_vfs_file; - struct file_lock *dl_flock; u32 dl_type; time_t dl_time; /* For recall: */ @@ -379,6 +377,9 @@ struct nfs4_file { */ atomic_t fi_readers; atomic_t fi_writers; + struct file *fi_deleg_file; + struct file_lock *fi_lease; + atomic_t fi_delegees; struct inode *fi_inode; u32 fi_id; /* used with stateowner->so_id * for stateid_hashtbl hash */ diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 641117f..ff93025 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -87,7 +87,7 @@ nfsd_cross_mnt(struct svc_rqst *rqstp, struct dentry **dpp, .dentry = dget(dentry)}; int err = 0; - err = follow_down(&path, false); + err = follow_down(&path); if (err < 0) goto out; @@ -808,7 +808,7 @@ nfsd_get_raparms(dev_t dev, ino_t ino) if (ra->p_count == 0) frap = rap; } - depth = nfsdstats.ra_size*11/10; + depth = nfsdstats.ra_size; if (!frap) { spin_unlock(&rab->pb_lock); return NULL; @@ -1744,6 +1744,13 @@ nfsd_rename(struct svc_rqst *rqstp, struct svc_fh *ffhp, char *fname, int flen, host_err = nfsd_break_lease(odentry->d_inode); if (host_err) goto out_drop_write; + if (ndentry->d_inode) { + host_err = nfsd_break_lease(ndentry->d_inode); + if (host_err) + goto out_drop_write; + } + if (host_err) + goto out_drop_write; host_err = vfs_rename(fdir, odentry, tdir, ndentry); if (!host_err) { host_err = commit_metadata(tfhp); @@ -1812,22 +1819,22 @@ nfsd_unlink(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, host_err = mnt_want_write(fhp->fh_export->ex_path.mnt); if (host_err) - goto out_nfserr; + goto out_put; host_err = nfsd_break_lease(rdentry->d_inode); if (host_err) - goto out_put; + goto out_drop_write; if (type != S_IFDIR) host_err = vfs_unlink(dirp, rdentry); else host_err = vfs_rmdir(dirp, rdentry); -out_put: - dput(rdentry); - if (!host_err) host_err = commit_metadata(fhp); - +out_drop_write: mnt_drop_write(fhp->fh_export->ex_path.mnt); +out_put: + dput(rdentry); + out_nfserr: err = nfserrno(host_err); out: diff --git a/fs/nilfs2/alloc.c b/fs/nilfs2/alloc.c index d7fd696..0a0a66d 100644 --- a/fs/nilfs2/alloc.c +++ b/fs/nilfs2/alloc.c @@ -521,8 +521,8 @@ void nilfs_palloc_commit_free_entry(struct inode *inode, group_offset, bitmap)) printk(KERN_WARNING "%s: entry number %llu already freed\n", __func__, (unsigned long long)req->pr_entry_nr); - - nilfs_palloc_group_desc_add_entries(inode, group, desc, 1); + else + nilfs_palloc_group_desc_add_entries(inode, group, desc, 1); kunmap(req->pr_bitmap_bh->b_page); kunmap(req->pr_desc_bh->b_page); @@ -558,8 +558,8 @@ void nilfs_palloc_abort_alloc_entry(struct inode *inode, group_offset, bitmap)) printk(KERN_WARNING "%s: entry number %llu already freed\n", __func__, (unsigned long long)req->pr_entry_nr); - - nilfs_palloc_group_desc_add_entries(inode, group, desc, 1); + else + nilfs_palloc_group_desc_add_entries(inode, group, desc, 1); kunmap(req->pr_bitmap_bh->b_page); kunmap(req->pr_desc_bh->b_page); @@ -665,7 +665,7 @@ int nilfs_palloc_freev(struct inode *inode, __u64 *entry_nrs, size_t nitems) for (j = i, n = 0; (j < nitems) && nilfs_palloc_group_is_in(inode, group, entry_nrs[j]); - j++, n++) { + j++) { nilfs_palloc_group(inode, entry_nrs[j], &group_offset); if (!nilfs_clear_bit_atomic( nilfs_mdt_bgl_lock(inode, group), @@ -674,6 +674,8 @@ int nilfs_palloc_freev(struct inode *inode, __u64 *entry_nrs, size_t nitems) "%s: entry number %llu already freed\n", __func__, (unsigned long long)entry_nrs[j]); + } else { + n++; } } nilfs_palloc_group_desc_add_entries(inode, group, desc, n); diff --git a/fs/nilfs2/bmap.c b/fs/nilfs2/bmap.c index 3ee67c6..4723f04 100644 --- a/fs/nilfs2/bmap.c +++ b/fs/nilfs2/bmap.c @@ -25,7 +25,6 @@ #include <linux/errno.h> #include "nilfs.h" #include "bmap.h" -#include "sb.h" #include "btree.h" #include "direct.h" #include "btnode.h" @@ -425,17 +424,6 @@ int nilfs_bmap_test_and_clear_dirty(struct nilfs_bmap *bmap) /* * Internal use only */ - -void nilfs_bmap_add_blocks(const struct nilfs_bmap *bmap, int n) -{ - inode_add_bytes(bmap->b_inode, (1 << bmap->b_inode->i_blkbits) * n); -} - -void nilfs_bmap_sub_blocks(const struct nilfs_bmap *bmap, int n) -{ - inode_sub_bytes(bmap->b_inode, (1 << bmap->b_inode->i_blkbits) * n); -} - __u64 nilfs_bmap_data_get_key(const struct nilfs_bmap *bmap, const struct buffer_head *bh) { diff --git a/fs/nilfs2/bmap.h b/fs/nilfs2/bmap.h index bde1c0aa2..40d9f45 100644 --- a/fs/nilfs2/bmap.h +++ b/fs/nilfs2/bmap.h @@ -240,9 +240,6 @@ __u64 nilfs_bmap_data_get_key(const struct nilfs_bmap *, __u64 nilfs_bmap_find_target_seq(const struct nilfs_bmap *, __u64); __u64 nilfs_bmap_find_target_in_group(const struct nilfs_bmap *); -void nilfs_bmap_add_blocks(const struct nilfs_bmap *, int); -void nilfs_bmap_sub_blocks(const struct nilfs_bmap *, int); - /* Assume that bmap semaphore is locked. */ static inline int nilfs_bmap_dirty(const struct nilfs_bmap *bmap) diff --git a/fs/nilfs2/btnode.c b/fs/nilfs2/btnode.c index 388e9e8..85f7baa 100644 --- a/fs/nilfs2/btnode.c +++ b/fs/nilfs2/btnode.c @@ -35,11 +35,6 @@ #include "btnode.h" -void nilfs_btnode_cache_init_once(struct address_space *btnc) -{ - nilfs_mapping_init_once(btnc); -} - static const struct address_space_operations def_btnode_aops = { .sync_page = block_sync_page, }; diff --git a/fs/nilfs2/btnode.h b/fs/nilfs2/btnode.h index 7903749..1b8ebd8 100644 --- a/fs/nilfs2/btnode.h +++ b/fs/nilfs2/btnode.h @@ -37,7 +37,6 @@ struct nilfs_btnode_chkey_ctxt { struct buffer_head *newbh; }; -void nilfs_btnode_cache_init_once(struct address_space *); void nilfs_btnode_cache_init(struct address_space *, struct backing_dev_info *); void nilfs_btnode_cache_clear(struct address_space *); struct buffer_head *nilfs_btnode_create_block(struct address_space *btnc, diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c index 300c2bc..d451ae0 100644 --- a/fs/nilfs2/btree.c +++ b/fs/nilfs2/btree.c @@ -1174,7 +1174,7 @@ static int nilfs_btree_insert(struct nilfs_bmap *btree, __u64 key, __u64 ptr) if (ret < 0) goto out; nilfs_btree_commit_insert(btree, path, level, key, ptr); - nilfs_bmap_add_blocks(btree, stats.bs_nblocks); + nilfs_inode_add_blocks(btree->b_inode, stats.bs_nblocks); out: nilfs_btree_free_path(path); @@ -1511,7 +1511,7 @@ static int nilfs_btree_delete(struct nilfs_bmap *btree, __u64 key) if (ret < 0) goto out; nilfs_btree_commit_delete(btree, path, level, dat); - nilfs_bmap_sub_blocks(btree, stats.bs_nblocks); + nilfs_inode_sub_blocks(btree->b_inode, stats.bs_nblocks); out: nilfs_btree_free_path(path); @@ -1776,7 +1776,7 @@ int nilfs_btree_convert_and_insert(struct nilfs_bmap *btree, return ret; nilfs_btree_commit_convert_and_insert(btree, key, ptr, keys, ptrs, n, di, ni, bh); - nilfs_bmap_add_blocks(btree, stats.bs_nblocks); + nilfs_inode_add_blocks(btree->b_inode, stats.bs_nblocks); return 0; } diff --git a/fs/nilfs2/dir.c b/fs/nilfs2/dir.c index 9d45773..3a19239 100644 --- a/fs/nilfs2/dir.c +++ b/fs/nilfs2/dir.c @@ -440,7 +440,6 @@ void nilfs_set_link(struct inode *dir, struct nilfs_dir_entry *de, nilfs_commit_chunk(page, mapping, from, to); nilfs_put_page(page); dir->i_mtime = dir->i_ctime = CURRENT_TIME; -/* NILFS_I(dir)->i_flags &= ~NILFS_BTREE_FL; */ } /* @@ -531,7 +530,6 @@ got_it: nilfs_set_de_type(de, inode); nilfs_commit_chunk(page, page->mapping, from, to); dir->i_mtime = dir->i_ctime = CURRENT_TIME; -/* NILFS_I(dir)->i_flags &= ~NILFS_BTREE_FL; */ nilfs_mark_inode_dirty(dir); /* OFFSET_CACHE */ out_put: @@ -579,7 +577,6 @@ int nilfs_delete_entry(struct nilfs_dir_entry *dir, struct page *page) dir->inode = 0; nilfs_commit_chunk(page, mapping, from, to); inode->i_ctime = inode->i_mtime = CURRENT_TIME; -/* NILFS_I(inode)->i_flags &= ~NILFS_BTREE_FL; */ out: nilfs_put_page(page); return err; @@ -684,7 +681,7 @@ const struct file_operations nilfs_dir_operations = { .readdir = nilfs_readdir, .unlocked_ioctl = nilfs_ioctl, #ifdef CONFIG_COMPAT - .compat_ioctl = nilfs_ioctl, + .compat_ioctl = nilfs_compat_ioctl, #endif /* CONFIG_COMPAT */ .fsync = nilfs_sync_file, diff --git a/fs/nilfs2/direct.c b/fs/nilfs2/direct.c index 324d80c..82f4865 100644 --- a/fs/nilfs2/direct.c +++ b/fs/nilfs2/direct.c @@ -146,7 +146,7 @@ static int nilfs_direct_insert(struct nilfs_bmap *bmap, __u64 key, __u64 ptr) if (NILFS_BMAP_USE_VBN(bmap)) nilfs_bmap_set_target_v(bmap, key, req.bpr_ptr); - nilfs_bmap_add_blocks(bmap, 1); + nilfs_inode_add_blocks(bmap->b_inode, 1); } return ret; } @@ -168,7 +168,7 @@ static int nilfs_direct_delete(struct nilfs_bmap *bmap, __u64 key) if (!ret) { nilfs_bmap_commit_end_ptr(bmap, &req, dat); nilfs_direct_set_ptr(bmap, key, NILFS_BMAP_INVALID_PTR); - nilfs_bmap_sub_blocks(bmap, 1); + nilfs_inode_sub_blocks(bmap->b_inode, 1); } return ret; } diff --git a/fs/nilfs2/file.c b/fs/nilfs2/file.c index 2f560c9..93589fc 100644 --- a/fs/nilfs2/file.c +++ b/fs/nilfs2/file.c @@ -59,7 +59,7 @@ static int nilfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) struct nilfs_transaction_info ti; int ret; - if (unlikely(nilfs_near_disk_full(NILFS_SB(inode->i_sb)->s_nilfs))) + if (unlikely(nilfs_near_disk_full(inode->i_sb->s_fs_info))) return VM_FAULT_SIGBUS; /* -ENOSPC */ lock_page(page); @@ -142,7 +142,7 @@ const struct file_operations nilfs_file_operations = { .aio_write = generic_file_aio_write, .unlocked_ioctl = nilfs_ioctl, #ifdef CONFIG_COMPAT - .compat_ioctl = nilfs_ioctl, + .compat_ioctl = nilfs_compat_ioctl, #endif /* CONFIG_COMPAT */ .mmap = nilfs_file_mmap, .open = generic_file_open, diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index 2fd440d..d5625be 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -41,6 +41,24 @@ struct nilfs_iget_args { int for_gc; }; +void nilfs_inode_add_blocks(struct inode *inode, int n) +{ + struct nilfs_root *root = NILFS_I(inode)->i_root; + + inode_add_bytes(inode, (1 << inode->i_blkbits) * n); + if (root) + atomic_add(n, &root->blocks_count); +} + +void nilfs_inode_sub_blocks(struct inode *inode, int n) +{ + struct nilfs_root *root = NILFS_I(inode)->i_root; + + inode_sub_bytes(inode, (1 << inode->i_blkbits) * n); + if (root) + atomic_sub(n, &root->blocks_count); +} + /** * nilfs_get_block() - get a file block on the filesystem (callback function) * @inode - inode struct of the target file @@ -277,7 +295,7 @@ const struct address_space_operations nilfs_aops = { struct inode *nilfs_new_inode(struct inode *dir, int mode) { struct super_block *sb = dir->i_sb; - struct nilfs_sb_info *sbi = NILFS_SB(sb); + struct the_nilfs *nilfs = sb->s_fs_info; struct inode *inode; struct nilfs_inode_info *ii; struct nilfs_root *root; @@ -315,19 +333,16 @@ struct inode *nilfs_new_inode(struct inode *dir, int mode) /* No lock is needed; iget() ensures it. */ } - ii->i_flags = NILFS_I(dir)->i_flags; - if (S_ISLNK(mode)) - ii->i_flags &= ~(NILFS_IMMUTABLE_FL | NILFS_APPEND_FL); - if (!S_ISDIR(mode)) - ii->i_flags &= ~NILFS_DIRSYNC_FL; + ii->i_flags = nilfs_mask_flags( + mode, NILFS_I(dir)->i_flags & NILFS_FL_INHERITED); /* ii->i_file_acl = 0; */ /* ii->i_dir_acl = 0; */ ii->i_dir_start_lookup = 0; nilfs_set_inode_flags(inode); - spin_lock(&sbi->s_next_gen_lock); - inode->i_generation = sbi->s_next_generation++; - spin_unlock(&sbi->s_next_gen_lock); + spin_lock(&nilfs->ns_next_gen_lock); + inode->i_generation = nilfs->ns_next_generation++; + spin_unlock(&nilfs->ns_next_gen_lock); insert_inode_hash(inode); err = nilfs_init_acl(inode, dir); @@ -359,17 +374,15 @@ void nilfs_set_inode_flags(struct inode *inode) inode->i_flags &= ~(S_SYNC | S_APPEND | S_IMMUTABLE | S_NOATIME | S_DIRSYNC); - if (flags & NILFS_SYNC_FL) + if (flags & FS_SYNC_FL) inode->i_flags |= S_SYNC; - if (flags & NILFS_APPEND_FL) + if (flags & FS_APPEND_FL) inode->i_flags |= S_APPEND; - if (flags & NILFS_IMMUTABLE_FL) + if (flags & FS_IMMUTABLE_FL) inode->i_flags |= S_IMMUTABLE; -#ifndef NILFS_ATIME_DISABLE - if (flags & NILFS_NOATIME_FL) -#endif + if (flags & FS_NOATIME_FL) inode->i_flags |= S_NOATIME; - if (flags & NILFS_DIRSYNC_FL) + if (flags & FS_DIRSYNC_FL) inode->i_flags |= S_DIRSYNC; mapping_set_gfp_mask(inode->i_mapping, mapping_gfp_mask(inode->i_mapping) & ~__GFP_FS); @@ -420,7 +433,7 @@ static int __nilfs_read_inode(struct super_block *sb, struct nilfs_root *root, unsigned long ino, struct inode *inode) { - struct the_nilfs *nilfs = NILFS_SB(sb)->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; struct buffer_head *bh; struct nilfs_inode *raw_inode; int err; @@ -707,6 +720,7 @@ void nilfs_evict_inode(struct inode *inode) struct nilfs_transaction_info ti; struct super_block *sb = inode->i_sb; struct nilfs_inode_info *ii = NILFS_I(inode); + int ret; if (inode->i_nlink || !ii->i_root || unlikely(is_bad_inode(inode))) { if (inode->i_data.nrpages) @@ -725,8 +739,9 @@ void nilfs_evict_inode(struct inode *inode) nilfs_mark_inode_dirty(inode); end_writeback(inode); - nilfs_ifile_delete_inode(ii->i_root->ifile, inode->i_ino); - atomic_dec(&ii->i_root->inodes_count); + ret = nilfs_ifile_delete_inode(ii->i_root->ifile, inode->i_ino); + if (!ret) + atomic_dec(&ii->i_root->inodes_count); nilfs_clear_inode(inode); @@ -792,18 +807,18 @@ int nilfs_permission(struct inode *inode, int mask, unsigned int flags) int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh) { - struct nilfs_sb_info *sbi = NILFS_SB(inode->i_sb); + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; struct nilfs_inode_info *ii = NILFS_I(inode); int err; - spin_lock(&sbi->s_inode_lock); + spin_lock(&nilfs->ns_inode_lock); if (ii->i_bh == NULL) { - spin_unlock(&sbi->s_inode_lock); + spin_unlock(&nilfs->ns_inode_lock); err = nilfs_ifile_get_inode_block(ii->i_root->ifile, inode->i_ino, pbh); if (unlikely(err)) return err; - spin_lock(&sbi->s_inode_lock); + spin_lock(&nilfs->ns_inode_lock); if (ii->i_bh == NULL) ii->i_bh = *pbh; else { @@ -814,36 +829,36 @@ int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh) *pbh = ii->i_bh; get_bh(*pbh); - spin_unlock(&sbi->s_inode_lock); + spin_unlock(&nilfs->ns_inode_lock); return 0; } int nilfs_inode_dirty(struct inode *inode) { struct nilfs_inode_info *ii = NILFS_I(inode); - struct nilfs_sb_info *sbi = NILFS_SB(inode->i_sb); + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; int ret = 0; if (!list_empty(&ii->i_dirty)) { - spin_lock(&sbi->s_inode_lock); + spin_lock(&nilfs->ns_inode_lock); ret = test_bit(NILFS_I_DIRTY, &ii->i_state) || test_bit(NILFS_I_BUSY, &ii->i_state); - spin_unlock(&sbi->s_inode_lock); + spin_unlock(&nilfs->ns_inode_lock); } return ret; } int nilfs_set_file_dirty(struct inode *inode, unsigned nr_dirty) { - struct nilfs_sb_info *sbi = NILFS_SB(inode->i_sb); struct nilfs_inode_info *ii = NILFS_I(inode); + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; - atomic_add(nr_dirty, &sbi->s_nilfs->ns_ndirtyblks); + atomic_add(nr_dirty, &nilfs->ns_ndirtyblks); if (test_and_set_bit(NILFS_I_DIRTY, &ii->i_state)) return 0; - spin_lock(&sbi->s_inode_lock); + spin_lock(&nilfs->ns_inode_lock); if (!test_bit(NILFS_I_QUEUED, &ii->i_state) && !test_bit(NILFS_I_BUSY, &ii->i_state)) { /* Because this routine may race with nilfs_dispose_list(), @@ -851,18 +866,18 @@ int nilfs_set_file_dirty(struct inode *inode, unsigned nr_dirty) if (list_empty(&ii->i_dirty) && igrab(inode) == NULL) { /* This will happen when somebody is freeing this inode. */ - nilfs_warning(sbi->s_super, __func__, + nilfs_warning(inode->i_sb, __func__, "cannot get inode (ino=%lu)\n", inode->i_ino); - spin_unlock(&sbi->s_inode_lock); + spin_unlock(&nilfs->ns_inode_lock); return -EINVAL; /* NILFS_I_DIRTY may remain for freeing inode */ } list_del(&ii->i_dirty); - list_add_tail(&ii->i_dirty, &sbi->s_dirty_files); + list_add_tail(&ii->i_dirty, &nilfs->ns_dirty_files); set_bit(NILFS_I_QUEUED, &ii->i_state); } - spin_unlock(&sbi->s_inode_lock); + spin_unlock(&nilfs->ns_inode_lock); return 0; } diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c index 4967389..95c04c2 100644 --- a/fs/nilfs2/ioctl.c +++ b/fs/nilfs2/ioctl.c @@ -26,7 +26,9 @@ #include <linux/capability.h> /* capable() */ #include <linux/uaccess.h> /* copy_from_user(), copy_to_user() */ #include <linux/vmalloc.h> +#include <linux/compat.h> /* compat_ptr() */ #include <linux/mount.h> /* mnt_want_write(), mnt_drop_write() */ +#include <linux/buffer_head.h> #include <linux/nilfs2_fs.h> #include "nilfs.h" #include "segment.h" @@ -97,11 +99,74 @@ static int nilfs_ioctl_wrap_copy(struct the_nilfs *nilfs, return ret; } +static int nilfs_ioctl_getflags(struct inode *inode, void __user *argp) +{ + unsigned int flags = NILFS_I(inode)->i_flags & FS_FL_USER_VISIBLE; + + return put_user(flags, (int __user *)argp); +} + +static int nilfs_ioctl_setflags(struct inode *inode, struct file *filp, + void __user *argp) +{ + struct nilfs_transaction_info ti; + unsigned int flags, oldflags; + int ret; + + if (!is_owner_or_cap(inode)) + return -EACCES; + + if (get_user(flags, (int __user *)argp)) + return -EFAULT; + + ret = mnt_want_write(filp->f_path.mnt); + if (ret) + return ret; + + flags = nilfs_mask_flags(inode->i_mode, flags); + + mutex_lock(&inode->i_mutex); + + oldflags = NILFS_I(inode)->i_flags; + + /* + * The IMMUTABLE and APPEND_ONLY flags can only be changed by the + * relevant capability. + */ + ret = -EPERM; + if (((flags ^ oldflags) & (FS_APPEND_FL | FS_IMMUTABLE_FL)) && + !capable(CAP_LINUX_IMMUTABLE)) + goto out; + + ret = nilfs_transaction_begin(inode->i_sb, &ti, 0); + if (ret) + goto out; + + NILFS_I(inode)->i_flags = (oldflags & ~FS_FL_USER_MODIFIABLE) | + (flags & FS_FL_USER_MODIFIABLE); + + nilfs_set_inode_flags(inode); + inode->i_ctime = CURRENT_TIME; + if (IS_SYNC(inode)) + nilfs_set_transaction_flag(NILFS_TI_SYNC); + + nilfs_mark_inode_dirty(inode); + ret = nilfs_transaction_commit(inode->i_sb); +out: + mutex_unlock(&inode->i_mutex); + mnt_drop_write(filp->f_path.mnt); + return ret; +} + +static int nilfs_ioctl_getversion(struct inode *inode, void __user *argp) +{ + return put_user(inode->i_generation, (int __user *)argp); +} + static int nilfs_ioctl_change_cpmode(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) { - struct the_nilfs *nilfs = NILFS_SB(inode->i_sb)->s_nilfs; - struct inode *cpfile = nilfs->ns_cpfile; + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; struct nilfs_transaction_info ti; struct nilfs_cpmode cpmode; int ret; @@ -121,7 +186,7 @@ static int nilfs_ioctl_change_cpmode(struct inode *inode, struct file *filp, nilfs_transaction_begin(inode->i_sb, &ti, 0); ret = nilfs_cpfile_change_cpmode( - cpfile, cpmode.cm_cno, cpmode.cm_mode); + nilfs->ns_cpfile, cpmode.cm_cno, cpmode.cm_mode); if (unlikely(ret < 0)) nilfs_transaction_abort(inode->i_sb); else @@ -137,7 +202,7 @@ static int nilfs_ioctl_delete_checkpoint(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) { - struct inode *cpfile = NILFS_SB(inode->i_sb)->s_nilfs->ns_cpfile; + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; struct nilfs_transaction_info ti; __u64 cno; int ret; @@ -154,7 +219,7 @@ nilfs_ioctl_delete_checkpoint(struct inode *inode, struct file *filp, goto out; nilfs_transaction_begin(inode->i_sb, &ti, 0); - ret = nilfs_cpfile_delete_checkpoint(cpfile, cno); + ret = nilfs_cpfile_delete_checkpoint(nilfs->ns_cpfile, cno); if (unlikely(ret < 0)) nilfs_transaction_abort(inode->i_sb); else @@ -180,7 +245,7 @@ nilfs_ioctl_do_get_cpinfo(struct the_nilfs *nilfs, __u64 *posp, int flags, static int nilfs_ioctl_get_cpstat(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) { - struct the_nilfs *nilfs = NILFS_SB(inode->i_sb)->s_nilfs; + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; struct nilfs_cpstat cpstat; int ret; @@ -211,7 +276,7 @@ nilfs_ioctl_do_get_suinfo(struct the_nilfs *nilfs, __u64 *posp, int flags, static int nilfs_ioctl_get_sustat(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) { - struct the_nilfs *nilfs = NILFS_SB(inode->i_sb)->s_nilfs; + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; struct nilfs_sustat sustat; int ret; @@ -267,7 +332,7 @@ nilfs_ioctl_do_get_bdescs(struct the_nilfs *nilfs, __u64 *posp, int flags, static int nilfs_ioctl_get_bdescs(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp) { - struct the_nilfs *nilfs = NILFS_SB(inode->i_sb)->s_nilfs; + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; struct nilfs_argv argv; int ret; @@ -336,7 +401,7 @@ static int nilfs_ioctl_move_blocks(struct super_block *sb, struct nilfs_argv *argv, void *buf) { size_t nmembs = argv->v_nmembs; - struct the_nilfs *nilfs = NILFS_SB(sb)->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; struct inode *inode; struct nilfs_vdesc *vdesc; struct buffer_head *bh, *n; @@ -550,7 +615,7 @@ static int nilfs_ioctl_clean_segments(struct inode *inode, struct file *filp, ret = PTR_ERR(kbufs[4]); goto out; } - nilfs = NILFS_SB(inode->i_sb)->s_nilfs; + nilfs = inode->i_sb->s_fs_info; for (n = 0; n < 4; n++) { ret = -EINVAL; @@ -623,7 +688,7 @@ static int nilfs_ioctl_sync(struct inode *inode, struct file *filp, return ret; if (argp != NULL) { - nilfs = NILFS_SB(inode->i_sb)->s_nilfs; + nilfs = inode->i_sb->s_fs_info; down_read(&nilfs->ns_segctor_sem); cno = nilfs->ns_cno - 1; up_read(&nilfs->ns_segctor_sem); @@ -641,7 +706,7 @@ static int nilfs_ioctl_get_info(struct inode *inode, struct file *filp, void *, size_t, size_t)) { - struct the_nilfs *nilfs = NILFS_SB(inode->i_sb)->s_nilfs; + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; struct nilfs_argv argv; int ret; @@ -666,6 +731,12 @@ long nilfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) void __user *argp = (void __user *)arg; switch (cmd) { + case FS_IOC_GETFLAGS: + return nilfs_ioctl_getflags(inode, argp); + case FS_IOC_SETFLAGS: + return nilfs_ioctl_setflags(inode, filp, argp); + case FS_IOC_GETVERSION: + return nilfs_ioctl_getversion(inode, argp); case NILFS_IOCTL_CHANGE_CPMODE: return nilfs_ioctl_change_cpmode(inode, filp, cmd, argp); case NILFS_IOCTL_DELETE_CHECKPOINT: @@ -696,3 +767,23 @@ long nilfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return -ENOTTY; } } + +#ifdef CONFIG_COMPAT +long nilfs_compat_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) +{ + switch (cmd) { + case FS_IOC32_GETFLAGS: + cmd = FS_IOC_GETFLAGS; + break; + case FS_IOC32_SETFLAGS: + cmd = FS_IOC_SETFLAGS; + break; + case FS_IOC32_GETVERSION: + cmd = FS_IOC_GETVERSION; + break; + default: + return -ENOIOCTLCMD; + } + return nilfs_ioctl(filp, cmd, (unsigned long)compat_ptr(arg)); +} +#endif diff --git a/fs/nilfs2/mdt.c b/fs/nilfs2/mdt.c index 6a0e2a1..a0babd2 100644 --- a/fs/nilfs2/mdt.c +++ b/fs/nilfs2/mdt.c @@ -454,9 +454,9 @@ int nilfs_mdt_setup_shadow_map(struct inode *inode, struct backing_dev_info *bdi = inode->i_sb->s_bdi; INIT_LIST_HEAD(&shadow->frozen_buffers); - nilfs_mapping_init_once(&shadow->frozen_data); + address_space_init_once(&shadow->frozen_data); nilfs_mapping_init(&shadow->frozen_data, bdi, &shadow_map_aops); - nilfs_mapping_init_once(&shadow->frozen_btnodes); + address_space_init_once(&shadow->frozen_btnodes); nilfs_mapping_init(&shadow->frozen_btnodes, bdi, &shadow_map_aops); mi->mi_shadow = shadow; return 0; diff --git a/fs/nilfs2/mdt.h b/fs/nilfs2/mdt.h index b13734b..ed68563 100644 --- a/fs/nilfs2/mdt.h +++ b/fs/nilfs2/mdt.h @@ -66,7 +66,7 @@ static inline struct nilfs_mdt_info *NILFS_MDT(const struct inode *inode) static inline struct the_nilfs *NILFS_I_NILFS(struct inode *inode) { - return NILFS_SB(inode->i_sb)->s_nilfs; + return inode->i_sb->s_fs_info; } /* Default GFP flags using highmem */ diff --git a/fs/nilfs2/namei.c b/fs/nilfs2/namei.c index 9803427..546849b 100644 --- a/fs/nilfs2/namei.c +++ b/fs/nilfs2/namei.c @@ -397,7 +397,6 @@ static int nilfs_rename(struct inode *old_dir, struct dentry *old_dentry, new_de = nilfs_find_entry(new_dir, &new_dentry->d_name, &new_page); if (!new_de) goto out_dir; - inc_nlink(old_inode); nilfs_set_link(new_dir, new_de, new_page, old_inode); nilfs_mark_inode_dirty(new_dir); new_inode->i_ctime = CURRENT_TIME; @@ -411,13 +410,9 @@ static int nilfs_rename(struct inode *old_dir, struct dentry *old_dentry, if (new_dir->i_nlink >= NILFS_LINK_MAX) goto out_dir; } - inc_nlink(old_inode); err = nilfs_add_link(new_dentry, old_inode); - if (err) { - drop_nlink(old_inode); - nilfs_mark_inode_dirty(old_inode); + if (err) goto out_dir; - } if (dir_de) { inc_nlink(new_dir); nilfs_mark_inode_dirty(new_dir); @@ -431,7 +426,6 @@ static int nilfs_rename(struct inode *old_dir, struct dentry *old_dentry, old_inode->i_ctime = CURRENT_TIME; nilfs_delete_entry(old_de, old_page); - drop_nlink(old_inode); if (dir_de) { nilfs_set_link(old_inode, dir_de, dir_page, new_dir); @@ -488,7 +482,7 @@ static struct dentry *nilfs_get_dentry(struct super_block *sb, u64 cno, if (ino < NILFS_FIRST_INO(sb) && ino != NILFS_ROOT_INO) return ERR_PTR(-ESTALE); - root = nilfs_lookup_root(NILFS_SB(sb)->s_nilfs, cno); + root = nilfs_lookup_root(sb->s_fs_info, cno); if (!root) return ERR_PTR(-ESTALE); diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h index 777e8fd..856e8e4 100644 --- a/fs/nilfs2/nilfs.h +++ b/fs/nilfs2/nilfs.h @@ -30,7 +30,6 @@ #include <linux/blkdev.h> #include <linux/nilfs2_fs.h> #include "the_nilfs.h" -#include "sb.h" #include "bmap.h" /* @@ -122,7 +121,7 @@ enum { #define NILFS_SYS_INO_BITS \ ((unsigned int)(1 << NILFS_ROOT_INO) | NILFS_MDT_INO_BITS) -#define NILFS_FIRST_INO(sb) (NILFS_SB(sb)->s_nilfs->ns_first_ino) +#define NILFS_FIRST_INO(sb) (((struct the_nilfs *)sb->s_fs_info)->ns_first_ino) #define NILFS_MDT_INODE(sb, ino) \ ((ino) < NILFS_FIRST_INO(sb) && (NILFS_MDT_INO_BITS & (1 << (ino)))) @@ -212,6 +211,23 @@ static inline int nilfs_init_acl(struct inode *inode, struct inode *dir) #define NILFS_ATIME_DISABLE +/* Flags that should be inherited by new inodes from their parent. */ +#define NILFS_FL_INHERITED \ + (FS_SECRM_FL | FS_UNRM_FL | FS_COMPR_FL | FS_SYNC_FL | \ + FS_IMMUTABLE_FL | FS_APPEND_FL | FS_NODUMP_FL | FS_NOATIME_FL |\ + FS_COMPRBLK_FL | FS_NOCOMP_FL | FS_NOTAIL_FL | FS_DIRSYNC_FL) + +/* Mask out flags that are inappropriate for the given type of inode. */ +static inline __u32 nilfs_mask_flags(umode_t mode, __u32 flags) +{ + if (S_ISDIR(mode)) + return flags; + else if (S_ISREG(mode)) + return flags & ~(FS_DIRSYNC_FL | FS_TOPDIR_FL); + else + return flags & (FS_NODUMP_FL | FS_NOATIME_FL); +} + /* dir.c */ extern int nilfs_add_link(struct dentry *, struct inode *); extern ino_t nilfs_inode_by_name(struct inode *, const struct qstr *); @@ -229,10 +245,13 @@ extern int nilfs_sync_file(struct file *, int); /* ioctl.c */ long nilfs_ioctl(struct file *, unsigned int, unsigned long); +long nilfs_compat_ioctl(struct file *file, unsigned int cmd, unsigned long arg); int nilfs_ioctl_prepare_clean_segments(struct the_nilfs *, struct nilfs_argv *, void **); /* inode.c */ +void nilfs_inode_add_blocks(struct inode *inode, int n); +void nilfs_inode_sub_blocks(struct inode *inode, int n); extern struct inode *nilfs_new_inode(struct inode *, int); extern void nilfs_free_inode(struct inode *); extern int nilfs_get_block(struct inode *, sector_t, struct buffer_head *, int); @@ -275,11 +294,11 @@ extern int nilfs_check_feature_compatibility(struct super_block *, struct nilfs_super_block *); extern void nilfs_set_log_cursor(struct nilfs_super_block *, struct the_nilfs *); -extern struct nilfs_super_block **nilfs_prepare_super(struct nilfs_sb_info *, - int flip); -extern int nilfs_commit_super(struct nilfs_sb_info *, int); -extern int nilfs_cleanup_super(struct nilfs_sb_info *); -int nilfs_attach_checkpoint(struct nilfs_sb_info *sbi, __u64 cno, int curr_mnt, +struct nilfs_super_block **nilfs_prepare_super(struct super_block *sb, + int flip); +int nilfs_commit_super(struct super_block *sb, int flag); +int nilfs_cleanup_super(struct super_block *sb); +int nilfs_attach_checkpoint(struct super_block *sb, __u64 cno, int curr_mnt, struct nilfs_root **root); int nilfs_checkpoint_is_mounted(struct super_block *sb, __u64 cno); diff --git a/fs/nilfs2/page.c b/fs/nilfs2/page.c index 0c43241..a585b35 100644 --- a/fs/nilfs2/page.c +++ b/fs/nilfs2/page.c @@ -492,19 +492,6 @@ unsigned nilfs_page_count_clean_buffers(struct page *page, return nc; } -void nilfs_mapping_init_once(struct address_space *mapping) -{ - memset(mapping, 0, sizeof(*mapping)); - INIT_RADIX_TREE(&mapping->page_tree, GFP_ATOMIC); - spin_lock_init(&mapping->tree_lock); - INIT_LIST_HEAD(&mapping->private_list); - spin_lock_init(&mapping->private_lock); - - spin_lock_init(&mapping->i_mmap_lock); - INIT_RAW_PRIO_TREE_ROOT(&mapping->i_mmap); - INIT_LIST_HEAD(&mapping->i_mmap_nonlinear); -} - void nilfs_mapping_init(struct address_space *mapping, struct backing_dev_info *bdi, const struct address_space_operations *aops) diff --git a/fs/nilfs2/page.h b/fs/nilfs2/page.h index 622df27..2a00953 100644 --- a/fs/nilfs2/page.h +++ b/fs/nilfs2/page.h @@ -61,7 +61,6 @@ void nilfs_free_private_page(struct page *); int nilfs_copy_dirty_pages(struct address_space *, struct address_space *); void nilfs_copy_back_pages(struct address_space *, struct address_space *); void nilfs_clear_dirty_pages(struct address_space *); -void nilfs_mapping_init_once(struct address_space *mapping); void nilfs_mapping_init(struct address_space *mapping, struct backing_dev_info *bdi, const struct address_space_operations *aops); diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c index 3dfcd3b..ba4a645 100644 --- a/fs/nilfs2/recovery.c +++ b/fs/nilfs2/recovery.c @@ -425,7 +425,7 @@ void nilfs_dispose_segment_list(struct list_head *head) } static int nilfs_prepare_segment_for_recovery(struct the_nilfs *nilfs, - struct nilfs_sb_info *sbi, + struct super_block *sb, struct nilfs_recovery_info *ri) { struct list_head *head = &ri->ri_used_segments; @@ -501,7 +501,7 @@ static int nilfs_recovery_copy_block(struct the_nilfs *nilfs, } static int nilfs_recover_dsync_blocks(struct the_nilfs *nilfs, - struct nilfs_sb_info *sbi, + struct super_block *sb, struct nilfs_root *root, struct list_head *head, unsigned long *nr_salvaged_blocks) @@ -514,7 +514,7 @@ static int nilfs_recover_dsync_blocks(struct the_nilfs *nilfs, int err = 0, err2 = 0; list_for_each_entry_safe(rb, n, head, list) { - inode = nilfs_iget(sbi->s_super, root, rb->ino); + inode = nilfs_iget(sb, root, rb->ino); if (IS_ERR(inode)) { err = PTR_ERR(inode); inode = NULL; @@ -572,11 +572,11 @@ static int nilfs_recover_dsync_blocks(struct the_nilfs *nilfs, * nilfs_do_roll_forward - salvage logical segments newer than the latest * checkpoint * @nilfs: nilfs object - * @sbi: nilfs_sb_info + * @sb: super block instance * @ri: pointer to a nilfs_recovery_info */ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, - struct nilfs_sb_info *sbi, + struct super_block *sb, struct nilfs_root *root, struct nilfs_recovery_info *ri) { @@ -648,7 +648,7 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, goto failed; if (flags & NILFS_SS_LOGEND) { err = nilfs_recover_dsync_blocks( - nilfs, sbi, root, &dsync_blocks, + nilfs, sb, root, &dsync_blocks, &nsalvaged_blocks); if (unlikely(err)) goto failed; @@ -681,7 +681,7 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, if (nsalvaged_blocks) { printk(KERN_INFO "NILFS (device %s): salvaged %lu blocks\n", - sbi->s_super->s_id, nsalvaged_blocks); + sb->s_id, nsalvaged_blocks); ri->ri_need_recovery = NILFS_RECOVERY_ROLLFORWARD_DONE; } out: @@ -695,7 +695,7 @@ static int nilfs_do_roll_forward(struct the_nilfs *nilfs, printk(KERN_ERR "NILFS (device %s): Error roll-forwarding " "(err=%d, pseg block=%llu). ", - sbi->s_super->s_id, err, (unsigned long long)pseg_start); + sb->s_id, err, (unsigned long long)pseg_start); goto out; } @@ -724,7 +724,7 @@ static void nilfs_finish_roll_forward(struct the_nilfs *nilfs, /** * nilfs_salvage_orphan_logs - salvage logs written after the latest checkpoint * @nilfs: nilfs object - * @sbi: nilfs_sb_info + * @sb: super block instance * @ri: pointer to a nilfs_recovery_info struct to store search results. * * Return Value: On success, 0 is returned. On error, one of the following @@ -741,7 +741,7 @@ static void nilfs_finish_roll_forward(struct the_nilfs *nilfs, * %-ENOMEM - Insufficient memory available. */ int nilfs_salvage_orphan_logs(struct the_nilfs *nilfs, - struct nilfs_sb_info *sbi, + struct super_block *sb, struct nilfs_recovery_info *ri) { struct nilfs_root *root; @@ -750,32 +750,32 @@ int nilfs_salvage_orphan_logs(struct the_nilfs *nilfs, if (ri->ri_lsegs_start == 0 || ri->ri_lsegs_end == 0) return 0; - err = nilfs_attach_checkpoint(sbi, ri->ri_cno, true, &root); + err = nilfs_attach_checkpoint(sb, ri->ri_cno, true, &root); if (unlikely(err)) { printk(KERN_ERR "NILFS: error loading the latest checkpoint.\n"); return err; } - err = nilfs_do_roll_forward(nilfs, sbi, root, ri); + err = nilfs_do_roll_forward(nilfs, sb, root, ri); if (unlikely(err)) goto failed; if (ri->ri_need_recovery == NILFS_RECOVERY_ROLLFORWARD_DONE) { - err = nilfs_prepare_segment_for_recovery(nilfs, sbi, ri); + err = nilfs_prepare_segment_for_recovery(nilfs, sb, ri); if (unlikely(err)) { printk(KERN_ERR "NILFS: Error preparing segments for " "recovery.\n"); goto failed; } - err = nilfs_attach_segment_constructor(sbi, root); + err = nilfs_attach_log_writer(sb, root); if (unlikely(err)) goto failed; set_nilfs_discontinued(nilfs); - err = nilfs_construct_segment(sbi->s_super); - nilfs_detach_segment_constructor(sbi); + err = nilfs_construct_segment(sb); + nilfs_detach_log_writer(sb); if (unlikely(err)) { printk(KERN_ERR "NILFS: Oops! recovery failed. " diff --git a/fs/nilfs2/sb.h b/fs/nilfs2/sb.h deleted file mode 100644 index 7a17715..0000000 --- a/fs/nilfs2/sb.h +++ /dev/null @@ -1,85 +0,0 @@ -/* - * sb.h - NILFS on-memory super block structure. - * - * Copyright (C) 2005-2008 Nippon Telegraph and Telephone Corporation. - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, write to the Free Software - * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA - * - * Written by Ryusuke Konishi <ryusuke@osrg.net> - * - */ - -#ifndef _NILFS_SB -#define _NILFS_SB - -#include <linux/types.h> -#include <linux/fs.h> - -struct the_nilfs; -struct nilfs_sc_info; - -/* - * NILFS super-block data in memory - */ -struct nilfs_sb_info { - /* Mount options */ - unsigned long s_mount_opt; - uid_t s_resuid; - gid_t s_resgid; - - unsigned long s_interval; /* construction interval */ - unsigned long s_watermark; /* threshold of data amount - for the segment construction */ - - /* Fundamental members */ - struct super_block *s_super; /* reverse pointer to super_block */ - struct the_nilfs *s_nilfs; - - /* Segment constructor */ - struct list_head s_dirty_files; /* dirty files list */ - struct nilfs_sc_info *s_sc_info; /* segment constructor info */ - spinlock_t s_inode_lock; /* Lock for the nilfs inode. - It covers s_dirty_files list */ - - /* Inode allocator */ - spinlock_t s_next_gen_lock; - u32 s_next_generation; -}; - -static inline struct nilfs_sb_info *NILFS_SB(struct super_block *sb) -{ - return sb->s_fs_info; -} - -static inline struct nilfs_sc_info *NILFS_SC(struct nilfs_sb_info *sbi) -{ - return sbi->s_sc_info; -} - -/* - * Bit operations for the mount option - */ -#define nilfs_clear_opt(sbi, opt) \ - do { (sbi)->s_mount_opt &= ~NILFS_MOUNT_##opt; } while (0) -#define nilfs_set_opt(sbi, opt) \ - do { (sbi)->s_mount_opt |= NILFS_MOUNT_##opt; } while (0) -#define nilfs_test_opt(sbi, opt) ((sbi)->s_mount_opt & NILFS_MOUNT_##opt) -#define nilfs_write_opt(sbi, mask, opt) \ - do { (sbi)->s_mount_opt = \ - (((sbi)->s_mount_opt & ~NILFS_MOUNT_##mask) | \ - NILFS_MOUNT_##opt); \ - } while (0) - -#endif /* _NILFS_SB */ diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index 55ebae5..afe4f21 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -104,8 +104,7 @@ struct nilfs_sc_operations { static void nilfs_segctor_start_timer(struct nilfs_sc_info *); static void nilfs_segctor_do_flush(struct nilfs_sc_info *, int); static void nilfs_segctor_do_immediate_flush(struct nilfs_sc_info *); -static void nilfs_dispose_list(struct nilfs_sb_info *, struct list_head *, - int); +static void nilfs_dispose_list(struct the_nilfs *, struct list_head *, int); #define nilfs_cnt32_gt(a, b) \ (typecheck(__u32, a) && typecheck(__u32, b) && \ @@ -182,7 +181,6 @@ int nilfs_transaction_begin(struct super_block *sb, struct nilfs_transaction_info *ti, int vacancy_check) { - struct nilfs_sb_info *sbi; struct the_nilfs *nilfs; int ret = nilfs_prepare_segment_lock(ti); @@ -193,8 +191,7 @@ int nilfs_transaction_begin(struct super_block *sb, vfs_check_frozen(sb, SB_FREEZE_WRITE); - sbi = NILFS_SB(sb); - nilfs = sbi->s_nilfs; + nilfs = sb->s_fs_info; down_read(&nilfs->ns_segctor_sem); if (vacancy_check && nilfs_near_disk_full(nilfs)) { up_read(&nilfs->ns_segctor_sem); @@ -225,8 +222,7 @@ int nilfs_transaction_begin(struct super_block *sb, int nilfs_transaction_commit(struct super_block *sb) { struct nilfs_transaction_info *ti = current->journal_info; - struct nilfs_sb_info *sbi; - struct nilfs_sc_info *sci; + struct the_nilfs *nilfs = sb->s_fs_info; int err = 0; BUG_ON(ti == NULL || ti->ti_magic != NILFS_TI_MAGIC); @@ -235,16 +231,15 @@ int nilfs_transaction_commit(struct super_block *sb) ti->ti_count--; return 0; } - sbi = NILFS_SB(sb); - sci = NILFS_SC(sbi); - if (sci != NULL) { + if (nilfs->ns_writer) { + struct nilfs_sc_info *sci = nilfs->ns_writer; + if (ti->ti_flags & NILFS_TI_COMMIT) nilfs_segctor_start_timer(sci); - if (atomic_read(&sbi->s_nilfs->ns_ndirtyblks) > - sci->sc_watermark) + if (atomic_read(&nilfs->ns_ndirtyblks) > sci->sc_watermark) nilfs_segctor_do_flush(sci, 0); } - up_read(&sbi->s_nilfs->ns_segctor_sem); + up_read(&nilfs->ns_segctor_sem); current->journal_info = ti->ti_save; if (ti->ti_flags & NILFS_TI_SYNC) @@ -257,13 +252,14 @@ int nilfs_transaction_commit(struct super_block *sb) void nilfs_transaction_abort(struct super_block *sb) { struct nilfs_transaction_info *ti = current->journal_info; + struct the_nilfs *nilfs = sb->s_fs_info; BUG_ON(ti == NULL || ti->ti_magic != NILFS_TI_MAGIC); if (ti->ti_count > 0) { ti->ti_count--; return; } - up_read(&NILFS_SB(sb)->s_nilfs->ns_segctor_sem); + up_read(&nilfs->ns_segctor_sem); current->journal_info = ti->ti_save; if (ti->ti_flags & NILFS_TI_DYNAMIC_ALLOC) @@ -272,9 +268,8 @@ void nilfs_transaction_abort(struct super_block *sb) void nilfs_relax_pressure_in_lock(struct super_block *sb) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); - struct nilfs_sc_info *sci = NILFS_SC(sbi); - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; + struct nilfs_sc_info *sci = nilfs->ns_writer; if (!sci || !sci->sc_flush_request) return; @@ -294,11 +289,13 @@ void nilfs_relax_pressure_in_lock(struct super_block *sb) downgrade_write(&nilfs->ns_segctor_sem); } -static void nilfs_transaction_lock(struct nilfs_sb_info *sbi, +static void nilfs_transaction_lock(struct super_block *sb, struct nilfs_transaction_info *ti, int gcflag) { struct nilfs_transaction_info *cur_ti = current->journal_info; + struct the_nilfs *nilfs = sb->s_fs_info; + struct nilfs_sc_info *sci = nilfs->ns_writer; WARN_ON(cur_ti); ti->ti_flags = NILFS_TI_WRITER; @@ -309,30 +306,31 @@ static void nilfs_transaction_lock(struct nilfs_sb_info *sbi, current->journal_info = ti; for (;;) { - down_write(&sbi->s_nilfs->ns_segctor_sem); - if (!test_bit(NILFS_SC_PRIOR_FLUSH, &NILFS_SC(sbi)->sc_flags)) + down_write(&nilfs->ns_segctor_sem); + if (!test_bit(NILFS_SC_PRIOR_FLUSH, &sci->sc_flags)) break; - nilfs_segctor_do_immediate_flush(NILFS_SC(sbi)); + nilfs_segctor_do_immediate_flush(sci); - up_write(&sbi->s_nilfs->ns_segctor_sem); + up_write(&nilfs->ns_segctor_sem); yield(); } if (gcflag) ti->ti_flags |= NILFS_TI_GC; } -static void nilfs_transaction_unlock(struct nilfs_sb_info *sbi) +static void nilfs_transaction_unlock(struct super_block *sb) { struct nilfs_transaction_info *ti = current->journal_info; + struct the_nilfs *nilfs = sb->s_fs_info; BUG_ON(ti == NULL || ti->ti_magic != NILFS_TI_MAGIC); BUG_ON(ti->ti_count > 0); - up_write(&sbi->s_nilfs->ns_segctor_sem); + up_write(&nilfs->ns_segctor_sem); current->journal_info = ti->ti_save; if (!list_empty(&ti->ti_garbage)) - nilfs_dispose_list(sbi, &ti->ti_garbage, 0); + nilfs_dispose_list(nilfs, &ti->ti_garbage, 0); } static void *nilfs_segctor_map_segsum_entry(struct nilfs_sc_info *sci, @@ -430,7 +428,8 @@ static void nilfs_segctor_begin_finfo(struct nilfs_sc_info *sci, nilfs_segctor_map_segsum_entry( sci, &sci->sc_binfo_ptr, sizeof(struct nilfs_finfo)); - if (inode->i_sb && !test_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags)) + if (NILFS_I(inode)->i_root && + !test_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags)) set_bit(NILFS_SC_HAVE_DELTA, &sci->sc_flags); /* skip finfo */ } @@ -713,7 +712,7 @@ static void nilfs_lookup_dirty_node_buffers(struct inode *inode, } } -static void nilfs_dispose_list(struct nilfs_sb_info *sbi, +static void nilfs_dispose_list(struct the_nilfs *nilfs, struct list_head *head, int force) { struct nilfs_inode_info *ii, *n; @@ -721,7 +720,7 @@ static void nilfs_dispose_list(struct nilfs_sb_info *sbi, unsigned nv = 0; while (!list_empty(head)) { - spin_lock(&sbi->s_inode_lock); + spin_lock(&nilfs->ns_inode_lock); list_for_each_entry_safe(ii, n, head, i_dirty) { list_del_init(&ii->i_dirty); if (force) { @@ -732,14 +731,14 @@ static void nilfs_dispose_list(struct nilfs_sb_info *sbi, } else if (test_bit(NILFS_I_DIRTY, &ii->i_state)) { set_bit(NILFS_I_QUEUED, &ii->i_state); list_add_tail(&ii->i_dirty, - &sbi->s_dirty_files); + &nilfs->ns_dirty_files); continue; } ivec[nv++] = ii; if (nv == SC_N_INODEVEC) break; } - spin_unlock(&sbi->s_inode_lock); + spin_unlock(&nilfs->ns_inode_lock); for (pii = ivec; nv > 0; pii++, nv--) iput(&(*pii)->vfs_inode); @@ -772,24 +771,23 @@ static int nilfs_segctor_clean(struct nilfs_sc_info *sci) static int nilfs_segctor_confirm(struct nilfs_sc_info *sci) { - struct nilfs_sb_info *sbi = sci->sc_sbi; + struct the_nilfs *nilfs = sci->sc_super->s_fs_info; int ret = 0; - if (nilfs_test_metadata_dirty(sbi->s_nilfs, sci->sc_root)) + if (nilfs_test_metadata_dirty(nilfs, sci->sc_root)) set_bit(NILFS_SC_DIRTY, &sci->sc_flags); - spin_lock(&sbi->s_inode_lock); - if (list_empty(&sbi->s_dirty_files) && nilfs_segctor_clean(sci)) + spin_lock(&nilfs->ns_inode_lock); + if (list_empty(&nilfs->ns_dirty_files) && nilfs_segctor_clean(sci)) ret++; - spin_unlock(&sbi->s_inode_lock); + spin_unlock(&nilfs->ns_inode_lock); return ret; } static void nilfs_segctor_clear_metadata_dirty(struct nilfs_sc_info *sci) { - struct nilfs_sb_info *sbi = sci->sc_sbi; - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sci->sc_super->s_fs_info; nilfs_mdt_clear_dirty(sci->sc_root->ifile); nilfs_mdt_clear_dirty(nilfs->ns_cpfile); @@ -799,7 +797,7 @@ static void nilfs_segctor_clear_metadata_dirty(struct nilfs_sc_info *sci) static int nilfs_segctor_create_checkpoint(struct nilfs_sc_info *sci) { - struct the_nilfs *nilfs = sci->sc_sbi->s_nilfs; + struct the_nilfs *nilfs = sci->sc_super->s_fs_info; struct buffer_head *bh_cp; struct nilfs_checkpoint *raw_cp; int err; @@ -823,8 +821,7 @@ static int nilfs_segctor_create_checkpoint(struct nilfs_sc_info *sci) static int nilfs_segctor_fill_in_checkpoint(struct nilfs_sc_info *sci) { - struct nilfs_sb_info *sbi = sci->sc_sbi; - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sci->sc_super->s_fs_info; struct buffer_head *bh_cp; struct nilfs_checkpoint *raw_cp; int err; @@ -1048,8 +1045,7 @@ static int nilfs_segctor_scan_file_dsync(struct nilfs_sc_info *sci, static int nilfs_segctor_collect_blocks(struct nilfs_sc_info *sci, int mode) { - struct nilfs_sb_info *sbi = sci->sc_sbi; - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sci->sc_super->s_fs_info; struct list_head *head; struct nilfs_inode_info *ii; size_t ndone; @@ -1858,7 +1854,7 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci) { struct nilfs_segment_buffer *segbuf; struct page *bd_page = NULL, *fs_page = NULL; - struct the_nilfs *nilfs = sci->sc_sbi->s_nilfs; + struct the_nilfs *nilfs = sci->sc_super->s_fs_info; int update_sr = false; list_for_each_entry(segbuf, &sci->sc_write_logs, sb_list) { @@ -1962,30 +1958,30 @@ static int nilfs_segctor_wait(struct nilfs_sc_info *sci) return ret; } -static int nilfs_segctor_check_in_files(struct nilfs_sc_info *sci, - struct nilfs_sb_info *sbi) +static int nilfs_segctor_collect_dirty_files(struct nilfs_sc_info *sci, + struct the_nilfs *nilfs) { struct nilfs_inode_info *ii, *n; struct inode *ifile = sci->sc_root->ifile; - spin_lock(&sbi->s_inode_lock); + spin_lock(&nilfs->ns_inode_lock); retry: - list_for_each_entry_safe(ii, n, &sbi->s_dirty_files, i_dirty) { + list_for_each_entry_safe(ii, n, &nilfs->ns_dirty_files, i_dirty) { if (!ii->i_bh) { struct buffer_head *ibh; int err; - spin_unlock(&sbi->s_inode_lock); + spin_unlock(&nilfs->ns_inode_lock); err = nilfs_ifile_get_inode_block( ifile, ii->vfs_inode.i_ino, &ibh); if (unlikely(err)) { - nilfs_warning(sbi->s_super, __func__, + nilfs_warning(sci->sc_super, __func__, "failed to get inode block.\n"); return err; } nilfs_mdt_mark_buffer_dirty(ibh); nilfs_mdt_mark_dirty(ifile); - spin_lock(&sbi->s_inode_lock); + spin_lock(&nilfs->ns_inode_lock); if (likely(!ii->i_bh)) ii->i_bh = ibh; else @@ -1998,18 +1994,18 @@ static int nilfs_segctor_check_in_files(struct nilfs_sc_info *sci, list_del(&ii->i_dirty); list_add_tail(&ii->i_dirty, &sci->sc_dirty_files); } - spin_unlock(&sbi->s_inode_lock); + spin_unlock(&nilfs->ns_inode_lock); return 0; } -static void nilfs_segctor_check_out_files(struct nilfs_sc_info *sci, - struct nilfs_sb_info *sbi) +static void nilfs_segctor_drop_written_files(struct nilfs_sc_info *sci, + struct the_nilfs *nilfs) { struct nilfs_transaction_info *ti = current->journal_info; struct nilfs_inode_info *ii, *n; - spin_lock(&sbi->s_inode_lock); + spin_lock(&nilfs->ns_inode_lock); list_for_each_entry_safe(ii, n, &sci->sc_dirty_files, i_dirty) { if (!test_and_clear_bit(NILFS_I_UPDATED, &ii->i_state) || test_bit(NILFS_I_DIRTY, &ii->i_state)) @@ -2021,7 +2017,7 @@ static void nilfs_segctor_check_out_files(struct nilfs_sc_info *sci, list_del(&ii->i_dirty); list_add_tail(&ii->i_dirty, &ti->ti_garbage); } - spin_unlock(&sbi->s_inode_lock); + spin_unlock(&nilfs->ns_inode_lock); } /* @@ -2029,15 +2025,14 @@ static void nilfs_segctor_check_out_files(struct nilfs_sc_info *sci, */ static int nilfs_segctor_do_construct(struct nilfs_sc_info *sci, int mode) { - struct nilfs_sb_info *sbi = sci->sc_sbi; - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sci->sc_super->s_fs_info; struct page *failed_page; int err; sci->sc_stage.scnt = NILFS_ST_INIT; sci->sc_cno = nilfs->ns_cno; - err = nilfs_segctor_check_in_files(sci, sbi); + err = nilfs_segctor_collect_dirty_files(sci, nilfs); if (unlikely(err)) goto out; @@ -2115,7 +2110,7 @@ static int nilfs_segctor_do_construct(struct nilfs_sc_info *sci, int mode) } while (sci->sc_stage.scnt != NILFS_ST_DONE); out: - nilfs_segctor_check_out_files(sci, sbi); + nilfs_segctor_drop_written_files(sci, nilfs); return err; failed_to_write: @@ -2168,8 +2163,8 @@ static void nilfs_segctor_do_flush(struct nilfs_sc_info *sci, int bn) */ void nilfs_flush_segment(struct super_block *sb, ino_t ino) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); - struct nilfs_sc_info *sci = NILFS_SC(sbi); + struct the_nilfs *nilfs = sb->s_fs_info; + struct nilfs_sc_info *sci = nilfs->ns_writer; if (!sci || nilfs_doing_construction()) return; @@ -2258,8 +2253,8 @@ static void nilfs_segctor_wakeup(struct nilfs_sc_info *sci, int err) */ int nilfs_construct_segment(struct super_block *sb) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); - struct nilfs_sc_info *sci = NILFS_SC(sbi); + struct the_nilfs *nilfs = sb->s_fs_info; + struct nilfs_sc_info *sci = nilfs->ns_writer; struct nilfs_transaction_info *ti; int err; @@ -2296,8 +2291,8 @@ int nilfs_construct_segment(struct super_block *sb) int nilfs_construct_dsync_segment(struct super_block *sb, struct inode *inode, loff_t start, loff_t end) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); - struct nilfs_sc_info *sci = NILFS_SC(sbi); + struct the_nilfs *nilfs = sb->s_fs_info; + struct nilfs_sc_info *sci = nilfs->ns_writer; struct nilfs_inode_info *ii; struct nilfs_transaction_info ti; int err = 0; @@ -2305,33 +2300,33 @@ int nilfs_construct_dsync_segment(struct super_block *sb, struct inode *inode, if (!sci) return -EROFS; - nilfs_transaction_lock(sbi, &ti, 0); + nilfs_transaction_lock(sb, &ti, 0); ii = NILFS_I(inode); if (test_bit(NILFS_I_INODE_DIRTY, &ii->i_state) || - nilfs_test_opt(sbi, STRICT_ORDER) || + nilfs_test_opt(nilfs, STRICT_ORDER) || test_bit(NILFS_SC_UNCLOSED, &sci->sc_flags) || - nilfs_discontinued(sbi->s_nilfs)) { - nilfs_transaction_unlock(sbi); + nilfs_discontinued(nilfs)) { + nilfs_transaction_unlock(sb); err = nilfs_segctor_sync(sci); return err; } - spin_lock(&sbi->s_inode_lock); + spin_lock(&nilfs->ns_inode_lock); if (!test_bit(NILFS_I_QUEUED, &ii->i_state) && !test_bit(NILFS_I_BUSY, &ii->i_state)) { - spin_unlock(&sbi->s_inode_lock); - nilfs_transaction_unlock(sbi); + spin_unlock(&nilfs->ns_inode_lock); + nilfs_transaction_unlock(sb); return 0; } - spin_unlock(&sbi->s_inode_lock); + spin_unlock(&nilfs->ns_inode_lock); sci->sc_dsync_inode = ii; sci->sc_dsync_start = start; sci->sc_dsync_end = end; err = nilfs_segctor_do_construct(sci, SC_LSEG_DSYNC); - nilfs_transaction_unlock(sbi); + nilfs_transaction_unlock(sb); return err; } @@ -2387,8 +2382,7 @@ static void nilfs_segctor_notify(struct nilfs_sc_info *sci, int mode, int err) */ static int nilfs_segctor_construct(struct nilfs_sc_info *sci, int mode) { - struct nilfs_sb_info *sbi = sci->sc_sbi; - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sci->sc_super->s_fs_info; struct nilfs_super_block **sbp; int err = 0; @@ -2406,11 +2400,12 @@ static int nilfs_segctor_construct(struct nilfs_sc_info *sci, int mode) nilfs_discontinued(nilfs)) { down_write(&nilfs->ns_sem); err = -EIO; - sbp = nilfs_prepare_super(sbi, + sbp = nilfs_prepare_super(sci->sc_super, nilfs_sb_will_flip(nilfs)); if (likely(sbp)) { nilfs_set_log_cursor(sbp[0], nilfs); - err = nilfs_commit_super(sbi, NILFS_SB_COMMIT); + err = nilfs_commit_super(sci->sc_super, + NILFS_SB_COMMIT); } up_write(&nilfs->ns_sem); } @@ -2442,16 +2437,15 @@ nilfs_remove_written_gcinodes(struct the_nilfs *nilfs, struct list_head *head) int nilfs_clean_segments(struct super_block *sb, struct nilfs_argv *argv, void **kbufs) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); - struct nilfs_sc_info *sci = NILFS_SC(sbi); - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; + struct nilfs_sc_info *sci = nilfs->ns_writer; struct nilfs_transaction_info ti; int err; if (unlikely(!sci)) return -EROFS; - nilfs_transaction_lock(sbi, &ti, 1); + nilfs_transaction_lock(sb, &ti, 1); err = nilfs_mdt_save_to_shadow_map(nilfs->ns_dat); if (unlikely(err)) @@ -2479,14 +2473,14 @@ int nilfs_clean_segments(struct super_block *sb, struct nilfs_argv *argv, set_current_state(TASK_INTERRUPTIBLE); schedule_timeout(sci->sc_interval); } - if (nilfs_test_opt(sbi, DISCARD)) { + if (nilfs_test_opt(nilfs, DISCARD)) { int ret = nilfs_discard_segments(nilfs, sci->sc_freesegs, sci->sc_nfreesegs); if (ret) { printk(KERN_WARNING "NILFS warning: error %d on discard request, " "turning discards off for the device\n", ret); - nilfs_clear_opt(sbi, DISCARD); + nilfs_clear_opt(nilfs, DISCARD); } } @@ -2494,16 +2488,15 @@ int nilfs_clean_segments(struct super_block *sb, struct nilfs_argv *argv, sci->sc_freesegs = NULL; sci->sc_nfreesegs = 0; nilfs_mdt_clear_shadow_map(nilfs->ns_dat); - nilfs_transaction_unlock(sbi); + nilfs_transaction_unlock(sb); return err; } static void nilfs_segctor_thread_construct(struct nilfs_sc_info *sci, int mode) { - struct nilfs_sb_info *sbi = sci->sc_sbi; struct nilfs_transaction_info ti; - nilfs_transaction_lock(sbi, &ti, 0); + nilfs_transaction_lock(sci->sc_super, &ti, 0); nilfs_segctor_construct(sci, mode); /* @@ -2514,7 +2507,7 @@ static void nilfs_segctor_thread_construct(struct nilfs_sc_info *sci, int mode) if (test_bit(NILFS_SC_UNCLOSED, &sci->sc_flags)) nilfs_segctor_start_timer(sci); - nilfs_transaction_unlock(sbi); + nilfs_transaction_unlock(sci->sc_super); } static void nilfs_segctor_do_immediate_flush(struct nilfs_sc_info *sci) @@ -2560,7 +2553,7 @@ static int nilfs_segctor_flush_mode(struct nilfs_sc_info *sci) static int nilfs_segctor_thread(void *arg) { struct nilfs_sc_info *sci = (struct nilfs_sc_info *)arg; - struct the_nilfs *nilfs = sci->sc_sbi->s_nilfs; + struct the_nilfs *nilfs = sci->sc_super->s_fs_info; int timeout = 0; sci->sc_timer.data = (unsigned long)current; @@ -2671,17 +2664,17 @@ static void nilfs_segctor_kill_thread(struct nilfs_sc_info *sci) /* * Setup & clean-up functions */ -static struct nilfs_sc_info *nilfs_segctor_new(struct nilfs_sb_info *sbi, +static struct nilfs_sc_info *nilfs_segctor_new(struct super_block *sb, struct nilfs_root *root) { + struct the_nilfs *nilfs = sb->s_fs_info; struct nilfs_sc_info *sci; sci = kzalloc(sizeof(*sci), GFP_KERNEL); if (!sci) return NULL; - sci->sc_sbi = sbi; - sci->sc_super = sbi->s_super; + sci->sc_super = sb; nilfs_get_root(root); sci->sc_root = root; @@ -2701,10 +2694,10 @@ static struct nilfs_sc_info *nilfs_segctor_new(struct nilfs_sb_info *sbi, sci->sc_mjcp_freq = HZ * NILFS_SC_DEFAULT_SR_FREQ; sci->sc_watermark = NILFS_SC_DEFAULT_WATERMARK; - if (sbi->s_interval) - sci->sc_interval = sbi->s_interval; - if (sbi->s_watermark) - sci->sc_watermark = sbi->s_watermark; + if (nilfs->ns_interval) + sci->sc_interval = nilfs->ns_interval; + if (nilfs->ns_watermark) + sci->sc_watermark = nilfs->ns_watermark; return sci; } @@ -2715,12 +2708,11 @@ static void nilfs_segctor_write_out(struct nilfs_sc_info *sci) /* The segctord thread was stopped and its timer was removed. But some tasks remain. */ do { - struct nilfs_sb_info *sbi = sci->sc_sbi; struct nilfs_transaction_info ti; - nilfs_transaction_lock(sbi, &ti, 0); + nilfs_transaction_lock(sci->sc_super, &ti, 0); ret = nilfs_segctor_construct(sci, SC_LSEG_SR); - nilfs_transaction_unlock(sbi); + nilfs_transaction_unlock(sci->sc_super); } while (ret && retrycount-- > 0); } @@ -2735,10 +2727,10 @@ static void nilfs_segctor_write_out(struct nilfs_sc_info *sci) */ static void nilfs_segctor_destroy(struct nilfs_sc_info *sci) { - struct nilfs_sb_info *sbi = sci->sc_sbi; + struct the_nilfs *nilfs = sci->sc_super->s_fs_info; int flag; - up_write(&sbi->s_nilfs->ns_segctor_sem); + up_write(&nilfs->ns_segctor_sem); spin_lock(&sci->sc_state_lock); nilfs_segctor_kill_thread(sci); @@ -2752,9 +2744,9 @@ static void nilfs_segctor_destroy(struct nilfs_sc_info *sci) WARN_ON(!list_empty(&sci->sc_copied_buffers)); if (!list_empty(&sci->sc_dirty_files)) { - nilfs_warning(sbi->s_super, __func__, + nilfs_warning(sci->sc_super, __func__, "dirty file(s) after the final construction\n"); - nilfs_dispose_list(sbi, &sci->sc_dirty_files, 1); + nilfs_dispose_list(nilfs, &sci->sc_dirty_files, 1); } WARN_ON(!list_empty(&sci->sc_segbufs)); @@ -2762,79 +2754,78 @@ static void nilfs_segctor_destroy(struct nilfs_sc_info *sci) nilfs_put_root(sci->sc_root); - down_write(&sbi->s_nilfs->ns_segctor_sem); + down_write(&nilfs->ns_segctor_sem); del_timer_sync(&sci->sc_timer); kfree(sci); } /** - * nilfs_attach_segment_constructor - attach a segment constructor - * @sbi: nilfs_sb_info + * nilfs_attach_log_writer - attach log writer + * @sb: super block instance * @root: root object of the current filesystem tree * - * nilfs_attach_segment_constructor() allocates a struct nilfs_sc_info, - * initializes it, and starts the segment constructor. + * This allocates a log writer object, initializes it, and starts the + * log writer. * * Return Value: On success, 0 is returned. On error, one of the following * negative error code is returned. * * %-ENOMEM - Insufficient memory available. */ -int nilfs_attach_segment_constructor(struct nilfs_sb_info *sbi, - struct nilfs_root *root) +int nilfs_attach_log_writer(struct super_block *sb, struct nilfs_root *root) { + struct the_nilfs *nilfs = sb->s_fs_info; int err; - if (NILFS_SC(sbi)) { + if (nilfs->ns_writer) { /* * This happens if the filesystem was remounted * read/write after nilfs_error degenerated it into a * read-only mount. */ - nilfs_detach_segment_constructor(sbi); + nilfs_detach_log_writer(sb); } - sbi->s_sc_info = nilfs_segctor_new(sbi, root); - if (!sbi->s_sc_info) + nilfs->ns_writer = nilfs_segctor_new(sb, root); + if (!nilfs->ns_writer) return -ENOMEM; - err = nilfs_segctor_start_thread(NILFS_SC(sbi)); + err = nilfs_segctor_start_thread(nilfs->ns_writer); if (err) { - kfree(sbi->s_sc_info); - sbi->s_sc_info = NULL; + kfree(nilfs->ns_writer); + nilfs->ns_writer = NULL; } return err; } /** - * nilfs_detach_segment_constructor - destroy the segment constructor - * @sbi: nilfs_sb_info + * nilfs_detach_log_writer - destroy log writer + * @sb: super block instance * - * nilfs_detach_segment_constructor() kills the segment constructor daemon, - * frees the struct nilfs_sc_info, and destroy the dirty file list. + * This kills log writer daemon, frees the log writer object, and + * destroys list of dirty files. */ -void nilfs_detach_segment_constructor(struct nilfs_sb_info *sbi) +void nilfs_detach_log_writer(struct super_block *sb) { - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; LIST_HEAD(garbage_list); down_write(&nilfs->ns_segctor_sem); - if (NILFS_SC(sbi)) { - nilfs_segctor_destroy(NILFS_SC(sbi)); - sbi->s_sc_info = NULL; + if (nilfs->ns_writer) { + nilfs_segctor_destroy(nilfs->ns_writer); + nilfs->ns_writer = NULL; } /* Force to free the list of dirty files */ - spin_lock(&sbi->s_inode_lock); - if (!list_empty(&sbi->s_dirty_files)) { - list_splice_init(&sbi->s_dirty_files, &garbage_list); - nilfs_warning(sbi->s_super, __func__, - "Non empty dirty list after the last " - "segment construction\n"); - } - spin_unlock(&sbi->s_inode_lock); + spin_lock(&nilfs->ns_inode_lock); + if (!list_empty(&nilfs->ns_dirty_files)) { + list_splice_init(&nilfs->ns_dirty_files, &garbage_list); + nilfs_warning(sb, __func__, + "Hit dirty file after stopped log writer\n"); + } + spin_unlock(&nilfs->ns_inode_lock); up_write(&nilfs->ns_segctor_sem); - nilfs_dispose_list(sbi, &garbage_list, 1); + nilfs_dispose_list(nilfs, &garbage_list, 1); } diff --git a/fs/nilfs2/segment.h b/fs/nilfs2/segment.h index cd8056e..6c02a86 100644 --- a/fs/nilfs2/segment.h +++ b/fs/nilfs2/segment.h @@ -27,7 +27,7 @@ #include <linux/fs.h> #include <linux/buffer_head.h> #include <linux/nilfs2_fs.h> -#include "sb.h" +#include "nilfs.h" struct nilfs_root; @@ -88,7 +88,6 @@ struct nilfs_segsum_pointer { /** * struct nilfs_sc_info - Segment constructor information * @sc_super: Back pointer to super_block struct - * @sc_sbi: Back pointer to nilfs_sb_info struct * @sc_root: root object of the current filesystem tree * @sc_nblk_inc: Block count of current generation * @sc_dirty_files: List of files to be written @@ -131,7 +130,6 @@ struct nilfs_segsum_pointer { */ struct nilfs_sc_info { struct super_block *sc_super; - struct nilfs_sb_info *sc_sbi; struct nilfs_root *sc_root; unsigned long sc_nblk_inc; @@ -235,18 +233,16 @@ extern void nilfs_flush_segment(struct super_block *, ino_t); extern int nilfs_clean_segments(struct super_block *, struct nilfs_argv *, void **); -int nilfs_attach_segment_constructor(struct nilfs_sb_info *sbi, - struct nilfs_root *root); -extern void nilfs_detach_segment_constructor(struct nilfs_sb_info *); +int nilfs_attach_log_writer(struct super_block *sb, struct nilfs_root *root); +void nilfs_detach_log_writer(struct super_block *sb); /* recovery.c */ extern int nilfs_read_super_root_block(struct the_nilfs *, sector_t, struct buffer_head **, int); extern int nilfs_search_super_root(struct the_nilfs *, struct nilfs_recovery_info *); -extern int nilfs_salvage_orphan_logs(struct the_nilfs *, - struct nilfs_sb_info *, - struct nilfs_recovery_info *); +int nilfs_salvage_orphan_logs(struct the_nilfs *nilfs, struct super_block *sb, + struct nilfs_recovery_info *ri); extern void nilfs_dispose_segment_list(struct list_head *); #endif /* _NILFS_SEGMENT_H */ diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 0994f6a7..062cca0 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -43,7 +43,6 @@ #include <linux/init.h> #include <linux/blkdev.h> #include <linux/parser.h> -#include <linux/random.h> #include <linux/crc32.h> #include <linux/vfs.h> #include <linux/writeback.h> @@ -72,23 +71,23 @@ struct kmem_cache *nilfs_transaction_cachep; struct kmem_cache *nilfs_segbuf_cachep; struct kmem_cache *nilfs_btree_path_cache; -static int nilfs_setup_super(struct nilfs_sb_info *sbi, int is_mount); +static int nilfs_setup_super(struct super_block *sb, int is_mount); static int nilfs_remount(struct super_block *sb, int *flags, char *data); -static void nilfs_set_error(struct nilfs_sb_info *sbi) +static void nilfs_set_error(struct super_block *sb) { - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; struct nilfs_super_block **sbp; down_write(&nilfs->ns_sem); if (!(nilfs->ns_mount_state & NILFS_ERROR_FS)) { nilfs->ns_mount_state |= NILFS_ERROR_FS; - sbp = nilfs_prepare_super(sbi, 0); + sbp = nilfs_prepare_super(sb, 0); if (likely(sbp)) { sbp[0]->s_state |= cpu_to_le16(NILFS_ERROR_FS); if (sbp[1]) sbp[1]->s_state |= cpu_to_le16(NILFS_ERROR_FS); - nilfs_commit_super(sbi, NILFS_SB_COMMIT_ALL); + nilfs_commit_super(sb, NILFS_SB_COMMIT_ALL); } } up_write(&nilfs->ns_sem); @@ -109,7 +108,7 @@ static void nilfs_set_error(struct nilfs_sb_info *sbi) void nilfs_error(struct super_block *sb, const char *function, const char *fmt, ...) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); + struct the_nilfs *nilfs = sb->s_fs_info; struct va_format vaf; va_list args; @@ -124,15 +123,15 @@ void nilfs_error(struct super_block *sb, const char *function, va_end(args); if (!(sb->s_flags & MS_RDONLY)) { - nilfs_set_error(sbi); + nilfs_set_error(sb); - if (nilfs_test_opt(sbi, ERRORS_RO)) { + if (nilfs_test_opt(nilfs, ERRORS_RO)) { printk(KERN_CRIT "Remounting filesystem read-only\n"); sb->s_flags |= MS_RDONLY; } } - if (nilfs_test_opt(sbi, ERRORS_PANIC)) + if (nilfs_test_opt(nilfs, ERRORS_PANIC)) panic("NILFS (device %s): panic forced after error\n", sb->s_id); } @@ -189,14 +188,14 @@ void nilfs_destroy_inode(struct inode *inode) call_rcu(&inode->i_rcu, nilfs_i_callback); } -static int nilfs_sync_super(struct nilfs_sb_info *sbi, int flag) +static int nilfs_sync_super(struct super_block *sb, int flag) { - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; int err; retry: set_buffer_dirty(nilfs->ns_sbh[0]); - if (nilfs_test_opt(sbi, BARRIER)) { + if (nilfs_test_opt(nilfs, BARRIER)) { err = __sync_dirty_buffer(nilfs->ns_sbh[0], WRITE_SYNC | WRITE_FLUSH_FUA); } else { @@ -263,10 +262,10 @@ void nilfs_set_log_cursor(struct nilfs_super_block *sbp, spin_unlock(&nilfs->ns_last_segment_lock); } -struct nilfs_super_block **nilfs_prepare_super(struct nilfs_sb_info *sbi, +struct nilfs_super_block **nilfs_prepare_super(struct super_block *sb, int flip) { - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; struct nilfs_super_block **sbp = nilfs->ns_sbp; /* nilfs->ns_sem must be locked by the caller. */ @@ -276,7 +275,7 @@ struct nilfs_super_block **nilfs_prepare_super(struct nilfs_sb_info *sbi, memcpy(sbp[0], sbp[1], nilfs->ns_sbsize); } else { printk(KERN_CRIT "NILFS: superblock broke on dev %s\n", - sbi->s_super->s_id); + sb->s_id); return NULL; } } else if (sbp[1] && @@ -290,9 +289,9 @@ struct nilfs_super_block **nilfs_prepare_super(struct nilfs_sb_info *sbi, return sbp; } -int nilfs_commit_super(struct nilfs_sb_info *sbi, int flag) +int nilfs_commit_super(struct super_block *sb, int flag) { - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; struct nilfs_super_block **sbp = nilfs->ns_sbp; time_t t; @@ -312,27 +311,28 @@ int nilfs_commit_super(struct nilfs_sb_info *sbi, int flag) nilfs->ns_sbsize)); } clear_nilfs_sb_dirty(nilfs); - return nilfs_sync_super(sbi, flag); + return nilfs_sync_super(sb, flag); } /** * nilfs_cleanup_super() - write filesystem state for cleanup - * @sbi: nilfs_sb_info to be unmounted or degraded to read-only + * @sb: super block instance to be unmounted or degraded to read-only * * This function restores state flags in the on-disk super block. * This will set "clean" flag (i.e. NILFS_VALID_FS) unless the * filesystem was not clean previously. */ -int nilfs_cleanup_super(struct nilfs_sb_info *sbi) +int nilfs_cleanup_super(struct super_block *sb) { + struct the_nilfs *nilfs = sb->s_fs_info; struct nilfs_super_block **sbp; int flag = NILFS_SB_COMMIT; int ret = -EIO; - sbp = nilfs_prepare_super(sbi, 0); + sbp = nilfs_prepare_super(sb, 0); if (sbp) { - sbp[0]->s_state = cpu_to_le16(sbi->s_nilfs->ns_mount_state); - nilfs_set_log_cursor(sbp[0], sbi->s_nilfs); + sbp[0]->s_state = cpu_to_le16(nilfs->ns_mount_state); + nilfs_set_log_cursor(sbp[0], nilfs); if (sbp[1] && sbp[0]->s_last_cno == sbp[1]->s_last_cno) { /* * make the "clean" flag also to the opposite @@ -342,21 +342,20 @@ int nilfs_cleanup_super(struct nilfs_sb_info *sbi) sbp[1]->s_state = sbp[0]->s_state; flag = NILFS_SB_COMMIT_ALL; } - ret = nilfs_commit_super(sbi, flag); + ret = nilfs_commit_super(sb, flag); } return ret; } static void nilfs_put_super(struct super_block *sb) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; - nilfs_detach_segment_constructor(sbi); + nilfs_detach_log_writer(sb); if (!(sb->s_flags & MS_RDONLY)) { down_write(&nilfs->ns_sem); - nilfs_cleanup_super(sbi); + nilfs_cleanup_super(sb); up_write(&nilfs->ns_sem); } @@ -365,15 +364,12 @@ static void nilfs_put_super(struct super_block *sb) iput(nilfs->ns_dat); destroy_nilfs(nilfs); - sbi->s_super = NULL; sb->s_fs_info = NULL; - kfree(sbi); } static int nilfs_sync_fs(struct super_block *sb, int wait) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; struct nilfs_super_block **sbp; int err = 0; @@ -383,10 +379,10 @@ static int nilfs_sync_fs(struct super_block *sb, int wait) down_write(&nilfs->ns_sem); if (nilfs_sb_dirty(nilfs)) { - sbp = nilfs_prepare_super(sbi, nilfs_sb_will_flip(nilfs)); + sbp = nilfs_prepare_super(sb, nilfs_sb_will_flip(nilfs)); if (likely(sbp)) { nilfs_set_log_cursor(sbp[0], nilfs); - nilfs_commit_super(sbi, NILFS_SB_COMMIT); + nilfs_commit_super(sb, NILFS_SB_COMMIT); } } up_write(&nilfs->ns_sem); @@ -394,10 +390,10 @@ static int nilfs_sync_fs(struct super_block *sb, int wait) return err; } -int nilfs_attach_checkpoint(struct nilfs_sb_info *sbi, __u64 cno, int curr_mnt, +int nilfs_attach_checkpoint(struct super_block *sb, __u64 cno, int curr_mnt, struct nilfs_root **rootp) { - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; struct nilfs_root *root; struct nilfs_checkpoint *raw_cp; struct buffer_head *bh_cp; @@ -426,7 +422,7 @@ int nilfs_attach_checkpoint(struct nilfs_sb_info *sbi, __u64 cno, int curr_mnt, goto failed; } - err = nilfs_ifile_read(sbi->s_super, root, nilfs->ns_inode_size, + err = nilfs_ifile_read(sb, root, nilfs->ns_inode_size, &raw_cp->cp_ifile_inode, &root->ifile); if (err) goto failed_bh; @@ -450,8 +446,7 @@ int nilfs_attach_checkpoint(struct nilfs_sb_info *sbi, __u64 cno, int curr_mnt, static int nilfs_freeze(struct super_block *sb) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; int err; if (sb->s_flags & MS_RDONLY) @@ -459,21 +454,20 @@ static int nilfs_freeze(struct super_block *sb) /* Mark super block clean */ down_write(&nilfs->ns_sem); - err = nilfs_cleanup_super(sbi); + err = nilfs_cleanup_super(sb); up_write(&nilfs->ns_sem); return err; } static int nilfs_unfreeze(struct super_block *sb) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; if (sb->s_flags & MS_RDONLY) return 0; down_write(&nilfs->ns_sem); - nilfs_setup_super(sbi, false); + nilfs_setup_super(sb, false); up_write(&nilfs->ns_sem); return 0; } @@ -530,22 +524,22 @@ static int nilfs_statfs(struct dentry *dentry, struct kstatfs *buf) static int nilfs_show_options(struct seq_file *seq, struct vfsmount *vfs) { struct super_block *sb = vfs->mnt_sb; - struct nilfs_sb_info *sbi = NILFS_SB(sb); + struct the_nilfs *nilfs = sb->s_fs_info; struct nilfs_root *root = NILFS_I(vfs->mnt_root->d_inode)->i_root; - if (!nilfs_test_opt(sbi, BARRIER)) + if (!nilfs_test_opt(nilfs, BARRIER)) seq_puts(seq, ",nobarrier"); if (root->cno != NILFS_CPTREE_CURRENT_CNO) seq_printf(seq, ",cp=%llu", (unsigned long long)root->cno); - if (nilfs_test_opt(sbi, ERRORS_PANIC)) + if (nilfs_test_opt(nilfs, ERRORS_PANIC)) seq_puts(seq, ",errors=panic"); - if (nilfs_test_opt(sbi, ERRORS_CONT)) + if (nilfs_test_opt(nilfs, ERRORS_CONT)) seq_puts(seq, ",errors=continue"); - if (nilfs_test_opt(sbi, STRICT_ORDER)) + if (nilfs_test_opt(nilfs, STRICT_ORDER)) seq_puts(seq, ",order=strict"); - if (nilfs_test_opt(sbi, NORECOVERY)) + if (nilfs_test_opt(nilfs, NORECOVERY)) seq_puts(seq, ",norecovery"); - if (nilfs_test_opt(sbi, DISCARD)) + if (nilfs_test_opt(nilfs, DISCARD)) seq_puts(seq, ",discard"); return 0; @@ -594,7 +588,7 @@ static match_table_t tokens = { static int parse_options(char *options, struct super_block *sb, int is_remount) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); + struct the_nilfs *nilfs = sb->s_fs_info; char *p; substring_t args[MAX_OPT_ARGS]; @@ -609,29 +603,29 @@ static int parse_options(char *options, struct super_block *sb, int is_remount) token = match_token(p, tokens, args); switch (token) { case Opt_barrier: - nilfs_set_opt(sbi, BARRIER); + nilfs_set_opt(nilfs, BARRIER); break; case Opt_nobarrier: - nilfs_clear_opt(sbi, BARRIER); + nilfs_clear_opt(nilfs, BARRIER); break; case Opt_order: if (strcmp(args[0].from, "relaxed") == 0) /* Ordered data semantics */ - nilfs_clear_opt(sbi, STRICT_ORDER); + nilfs_clear_opt(nilfs, STRICT_ORDER); else if (strcmp(args[0].from, "strict") == 0) /* Strict in-order semantics */ - nilfs_set_opt(sbi, STRICT_ORDER); + nilfs_set_opt(nilfs, STRICT_ORDER); else return 0; break; case Opt_err_panic: - nilfs_write_opt(sbi, ERROR_MODE, ERRORS_PANIC); + nilfs_write_opt(nilfs, ERROR_MODE, ERRORS_PANIC); break; case Opt_err_ro: - nilfs_write_opt(sbi, ERROR_MODE, ERRORS_RO); + nilfs_write_opt(nilfs, ERROR_MODE, ERRORS_RO); break; case Opt_err_cont: - nilfs_write_opt(sbi, ERROR_MODE, ERRORS_CONT); + nilfs_write_opt(nilfs, ERROR_MODE, ERRORS_CONT); break; case Opt_snapshot: if (is_remount) { @@ -642,13 +636,13 @@ static int parse_options(char *options, struct super_block *sb, int is_remount) } break; case Opt_norecovery: - nilfs_set_opt(sbi, NORECOVERY); + nilfs_set_opt(nilfs, NORECOVERY); break; case Opt_discard: - nilfs_set_opt(sbi, DISCARD); + nilfs_set_opt(nilfs, DISCARD); break; case Opt_nodiscard: - nilfs_clear_opt(sbi, DISCARD); + nilfs_clear_opt(nilfs, DISCARD); break; default: printk(KERN_ERR @@ -660,22 +654,24 @@ static int parse_options(char *options, struct super_block *sb, int is_remount) } static inline void -nilfs_set_default_options(struct nilfs_sb_info *sbi, +nilfs_set_default_options(struct super_block *sb, struct nilfs_super_block *sbp) { - sbi->s_mount_opt = + struct the_nilfs *nilfs = sb->s_fs_info; + + nilfs->ns_mount_opt = NILFS_MOUNT_ERRORS_RO | NILFS_MOUNT_BARRIER; } -static int nilfs_setup_super(struct nilfs_sb_info *sbi, int is_mount) +static int nilfs_setup_super(struct super_block *sb, int is_mount) { - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; struct nilfs_super_block **sbp; int max_mnt_count; int mnt_count; /* nilfs->ns_sem must be locked by the caller. */ - sbp = nilfs_prepare_super(sbi, 0); + sbp = nilfs_prepare_super(sb, 0); if (!sbp) return -EIO; @@ -704,8 +700,9 @@ skip_mount_setup: sbp[0]->s_state = cpu_to_le16(le16_to_cpu(sbp[0]->s_state) & ~NILFS_VALID_FS); /* synchronize sbp[1] with sbp[0] */ - memcpy(sbp[1], sbp[0], nilfs->ns_sbsize); - return nilfs_commit_super(sbi, NILFS_SB_COMMIT_ALL); + if (sbp[1]) + memcpy(sbp[1], sbp[0], nilfs->ns_sbsize); + return nilfs_commit_super(sb, NILFS_SB_COMMIT_ALL); } struct nilfs_super_block *nilfs_read_super_block(struct super_block *sb, @@ -726,7 +723,7 @@ int nilfs_store_magic_and_option(struct super_block *sb, struct nilfs_super_block *sbp, char *data) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); + struct the_nilfs *nilfs = sb->s_fs_info; sb->s_magic = le16_to_cpu(sbp->s_magic); @@ -735,12 +732,12 @@ int nilfs_store_magic_and_option(struct super_block *sb, sb->s_flags |= MS_NOATIME; #endif - nilfs_set_default_options(sbi, sbp); + nilfs_set_default_options(sb, sbp); - sbi->s_resuid = le16_to_cpu(sbp->s_def_resuid); - sbi->s_resgid = le16_to_cpu(sbp->s_def_resgid); - sbi->s_interval = le32_to_cpu(sbp->s_c_interval); - sbi->s_watermark = le32_to_cpu(sbp->s_c_block_max); + nilfs->ns_resuid = le16_to_cpu(sbp->s_def_resuid); + nilfs->ns_resgid = le16_to_cpu(sbp->s_def_resgid); + nilfs->ns_interval = le32_to_cpu(sbp->s_c_interval); + nilfs->ns_watermark = le32_to_cpu(sbp->s_c_block_max); return !parse_options(data, sb, 0) ? -EINVAL : 0 ; } @@ -821,7 +818,7 @@ static int nilfs_get_root_dentry(struct super_block *sb, static int nilfs_attach_snapshot(struct super_block *s, __u64 cno, struct dentry **root_dentry) { - struct the_nilfs *nilfs = NILFS_SB(s)->s_nilfs; + struct the_nilfs *nilfs = s->s_fs_info; struct nilfs_root *root; int ret; @@ -839,7 +836,7 @@ static int nilfs_attach_snapshot(struct super_block *s, __u64 cno, goto out; } - ret = nilfs_attach_checkpoint(NILFS_SB(s), cno, false, &root); + ret = nilfs_attach_checkpoint(s, cno, false, &root); if (ret) { printk(KERN_ERR "NILFS: error loading snapshot " "(checkpoint number=%llu).\n", @@ -873,7 +870,7 @@ static int nilfs_try_to_shrink_tree(struct dentry *root_dentry) int nilfs_checkpoint_is_mounted(struct super_block *sb, __u64 cno) { - struct the_nilfs *nilfs = NILFS_SB(sb)->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; struct nilfs_root *root; struct inode *inode; struct dentry *dentry; @@ -886,7 +883,7 @@ int nilfs_checkpoint_is_mounted(struct super_block *sb, __u64 cno) return true; /* protect recent checkpoints */ ret = false; - root = nilfs_lookup_root(NILFS_SB(sb)->s_nilfs, cno); + root = nilfs_lookup_root(nilfs, cno); if (root) { inode = nilfs_ilookup(sb, root, NILFS_ROOT_INO); if (inode) { @@ -916,43 +913,21 @@ static int nilfs_fill_super(struct super_block *sb, void *data, int silent) { struct the_nilfs *nilfs; - struct nilfs_sb_info *sbi; struct nilfs_root *fsroot; struct backing_dev_info *bdi; __u64 cno; int err; - sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); - if (!sbi) + nilfs = alloc_nilfs(sb->s_bdev); + if (!nilfs) return -ENOMEM; - sb->s_fs_info = sbi; - sbi->s_super = sb; - - nilfs = alloc_nilfs(sb->s_bdev); - if (!nilfs) { - err = -ENOMEM; - goto failed_sbi; - } - sbi->s_nilfs = nilfs; + sb->s_fs_info = nilfs; - err = init_nilfs(nilfs, sbi, (char *)data); + err = init_nilfs(nilfs, sb, (char *)data); if (err) goto failed_nilfs; - spin_lock_init(&sbi->s_inode_lock); - INIT_LIST_HEAD(&sbi->s_dirty_files); - - /* - * Following initialization is overlapped because - * nilfs_sb_info structure has been cleared at the beginning. - * But we reserve them to keep our interest and make ready - * for the future change. - */ - get_random_bytes(&sbi->s_next_generation, - sizeof(sbi->s_next_generation)); - spin_lock_init(&sbi->s_next_gen_lock); - sb->s_op = &nilfs_sops; sb->s_export_op = &nilfs_export_ops; sb->s_root = NULL; @@ -961,12 +936,12 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent) bdi = sb->s_bdev->bd_inode->i_mapping->backing_dev_info; sb->s_bdi = bdi ? : &default_backing_dev_info; - err = load_nilfs(nilfs, sbi); + err = load_nilfs(nilfs, sb); if (err) goto failed_nilfs; cno = nilfs_last_cno(nilfs); - err = nilfs_attach_checkpoint(sbi, cno, true, &fsroot); + err = nilfs_attach_checkpoint(sb, cno, true, &fsroot); if (err) { printk(KERN_ERR "NILFS: error loading last checkpoint " "(checkpoint number=%llu).\n", (unsigned long long)cno); @@ -974,7 +949,7 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent) } if (!(sb->s_flags & MS_RDONLY)) { - err = nilfs_attach_segment_constructor(sbi, fsroot); + err = nilfs_attach_log_writer(sb, fsroot); if (err) goto failed_checkpoint; } @@ -987,14 +962,14 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent) if (!(sb->s_flags & MS_RDONLY)) { down_write(&nilfs->ns_sem); - nilfs_setup_super(sbi, true); + nilfs_setup_super(sb, true); up_write(&nilfs->ns_sem); } return 0; failed_segctor: - nilfs_detach_segment_constructor(sbi); + nilfs_detach_log_writer(sb); failed_checkpoint: nilfs_put_root(fsroot); @@ -1006,23 +981,18 @@ nilfs_fill_super(struct super_block *sb, void *data, int silent) failed_nilfs: destroy_nilfs(nilfs); - - failed_sbi: - sb->s_fs_info = NULL; - kfree(sbi); return err; } static int nilfs_remount(struct super_block *sb, int *flags, char *data) { - struct nilfs_sb_info *sbi = NILFS_SB(sb); - struct the_nilfs *nilfs = sbi->s_nilfs; + struct the_nilfs *nilfs = sb->s_fs_info; unsigned long old_sb_flags; unsigned long old_mount_opt; int err; old_sb_flags = sb->s_flags; - old_mount_opt = sbi->s_mount_opt; + old_mount_opt = nilfs->ns_mount_opt; if (!parse_options(data, sb, 1)) { err = -EINVAL; @@ -1042,8 +1012,8 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data) if ((*flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY)) goto out; if (*flags & MS_RDONLY) { - /* Shutting down the segment constructor */ - nilfs_detach_segment_constructor(sbi); + /* Shutting down log writer */ + nilfs_detach_log_writer(sb); sb->s_flags |= MS_RDONLY; /* @@ -1051,7 +1021,7 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data) * the RDONLY flag and then mark the partition as valid again. */ down_write(&nilfs->ns_sem); - nilfs_cleanup_super(sbi); + nilfs_cleanup_super(sb); up_write(&nilfs->ns_sem); } else { __u64 features; @@ -1078,12 +1048,12 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data) sb->s_flags &= ~MS_RDONLY; root = NILFS_I(sb->s_root->d_inode)->i_root; - err = nilfs_attach_segment_constructor(sbi, root); + err = nilfs_attach_log_writer(sb, root); if (err) goto restore_opts; down_write(&nilfs->ns_sem); - nilfs_setup_super(sbi, true); + nilfs_setup_super(sb, true); up_write(&nilfs->ns_sem); } out: @@ -1091,13 +1061,12 @@ static int nilfs_remount(struct super_block *sb, int *flags, char *data) restore_opts: sb->s_flags = old_sb_flags; - sbi->s_mount_opt = old_mount_opt; + nilfs->ns_mount_opt = old_mount_opt; return err; } struct nilfs_super_data { struct block_device *bdev; - struct nilfs_sb_info *sbi; __u64 cno; int flags; }; @@ -1278,7 +1247,7 @@ static void nilfs_inode_init_once(void *obj) #ifdef CONFIG_NILFS_XATTR init_rwsem(&ii->xattr_sem); #endif - nilfs_btnode_cache_init_once(&ii->i_btnode_cache); + address_space_init_once(&ii->i_btnode_cache); ii->i_bmap = &ii->i_bmap_data; inode_init_once(&ii->vfs_inode); } diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c index ad4ac60..d2acd1a 100644 --- a/fs/nilfs2/the_nilfs.c +++ b/fs/nilfs2/the_nilfs.c @@ -25,6 +25,7 @@ #include <linux/slab.h> #include <linux/blkdev.h> #include <linux/backing-dev.h> +#include <linux/random.h> #include <linux/crc32.h> #include "nilfs.h" #include "segment.h" @@ -75,7 +76,10 @@ struct the_nilfs *alloc_nilfs(struct block_device *bdev) nilfs->ns_bdev = bdev; atomic_set(&nilfs->ns_ndirtyblks, 0); init_rwsem(&nilfs->ns_sem); + INIT_LIST_HEAD(&nilfs->ns_dirty_files); INIT_LIST_HEAD(&nilfs->ns_gc_inodes); + spin_lock_init(&nilfs->ns_inode_lock); + spin_lock_init(&nilfs->ns_next_gen_lock); spin_lock_init(&nilfs->ns_last_segment_lock); nilfs->ns_cptree = RB_ROOT; spin_lock_init(&nilfs->ns_cptree_lock); @@ -197,16 +201,16 @@ static int nilfs_store_log_cursor(struct the_nilfs *nilfs, /** * load_nilfs - load and recover the nilfs * @nilfs: the_nilfs structure to be released - * @sbi: nilfs_sb_info used to recover past segment + * @sb: super block isntance used to recover past segment * * load_nilfs() searches and load the latest super root, * attaches the last segment, and does recovery if needed. * The caller must call this exclusively for simultaneous mounts. */ -int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) +int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb) { struct nilfs_recovery_info ri; - unsigned int s_flags = sbi->s_super->s_flags; + unsigned int s_flags = sb->s_flags; int really_read_only = bdev_read_only(nilfs->ns_bdev); int valid_fs = nilfs_valid_fs(nilfs); int err; @@ -271,7 +275,7 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) goto scan_error; } - err = nilfs_load_super_root(nilfs, sbi->s_super, ri.ri_super_root); + err = nilfs_load_super_root(nilfs, sb, ri.ri_super_root); if (unlikely(err)) { printk(KERN_ERR "NILFS: error loading super root.\n"); goto failed; @@ -283,7 +287,7 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) if (s_flags & MS_RDONLY) { __u64 features; - if (nilfs_test_opt(sbi, NORECOVERY)) { + if (nilfs_test_opt(nilfs, NORECOVERY)) { printk(KERN_INFO "NILFS: norecovery option specified. " "skipping roll-forward recovery\n"); goto skip_recovery; @@ -304,21 +308,21 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) err = -EROFS; goto failed_unload; } - sbi->s_super->s_flags &= ~MS_RDONLY; - } else if (nilfs_test_opt(sbi, NORECOVERY)) { + sb->s_flags &= ~MS_RDONLY; + } else if (nilfs_test_opt(nilfs, NORECOVERY)) { printk(KERN_ERR "NILFS: recovery cancelled because norecovery " "option was specified for a read/write mount\n"); err = -EINVAL; goto failed_unload; } - err = nilfs_salvage_orphan_logs(nilfs, sbi, &ri); + err = nilfs_salvage_orphan_logs(nilfs, sb, &ri); if (err) goto failed_unload; down_write(&nilfs->ns_sem); nilfs->ns_mount_state |= NILFS_VALID_FS; /* set "clean" flag */ - err = nilfs_cleanup_super(sbi); + err = nilfs_cleanup_super(sb); up_write(&nilfs->ns_sem); if (err) { @@ -330,7 +334,7 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) skip_recovery: nilfs_clear_recovery_info(&ri); - sbi->s_super->s_flags = s_flags; + sb->s_flags = s_flags; return 0; scan_error: @@ -344,7 +348,7 @@ int load_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi) failed: nilfs_clear_recovery_info(&ri); - sbi->s_super->s_flags = s_flags; + sb->s_flags = s_flags; return err; } @@ -475,10 +479,13 @@ static int nilfs_load_super_block(struct the_nilfs *nilfs, return -EIO; } printk(KERN_WARNING - "NILFS warning: unable to read primary superblock\n"); - } else if (!sbp[1]) + "NILFS warning: unable to read primary superblock " + "(blocksize = %d)\n", blocksize); + } else if (!sbp[1]) { printk(KERN_WARNING - "NILFS warning: unable to read secondary superblock\n"); + "NILFS warning: unable to read secondary superblock " + "(blocksize = %d)\n", blocksize); + } /* * Compare two super blocks and set 1 in swp if the secondary @@ -505,7 +512,7 @@ static int nilfs_load_super_block(struct the_nilfs *nilfs, if (!valid[!swp]) printk(KERN_WARNING "NILFS warning: broken superblock. " - "using spare superblock.\n"); + "using spare superblock (blocksize = %d).\n", blocksize); if (swp) nilfs_swap_super_block(nilfs); @@ -519,7 +526,6 @@ static int nilfs_load_super_block(struct the_nilfs *nilfs, /** * init_nilfs - initialize a NILFS instance. * @nilfs: the_nilfs structure - * @sbi: nilfs_sb_info * @sb: super block * @data: mount options * @@ -530,9 +536,8 @@ static int nilfs_load_super_block(struct the_nilfs *nilfs, * Return Value: On success, 0 is returned. On error, a negative error * code is returned. */ -int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data) +int init_nilfs(struct the_nilfs *nilfs, struct super_block *sb, char *data) { - struct super_block *sb = sbi->s_super; struct nilfs_super_block *sbp; int blocksize; int err; @@ -588,6 +593,9 @@ int init_nilfs(struct the_nilfs *nilfs, struct nilfs_sb_info *sbi, char *data) nilfs->ns_blocksize_bits = sb->s_blocksize_bits; nilfs->ns_blocksize = blocksize; + get_random_bytes(&nilfs->ns_next_generation, + sizeof(nilfs->ns_next_generation)); + err = nilfs_store_disk_layout(nilfs, sbp); if (err) goto failed_sbh; diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h index fd85e4c..f496814 100644 --- a/fs/nilfs2/the_nilfs.h +++ b/fs/nilfs2/the_nilfs.h @@ -31,7 +31,8 @@ #include <linux/blkdev.h> #include <linux/backing-dev.h> #include <linux/slab.h> -#include "sb.h" + +struct nilfs_sc_info; /* the_nilfs struct */ enum { @@ -65,13 +66,23 @@ enum { * @ns_last_cno: checkpoint number of the latest segment * @ns_prot_seq: least sequence number of segments which must not be reclaimed * @ns_prev_seq: base sequence number used to decide if advance log cursor - * @ns_segctor_sem: segment constructor semaphore + * @ns_writer: log writer + * @ns_segctor_sem: semaphore protecting log write * @ns_dat: DAT file inode * @ns_cpfile: checkpoint file inode * @ns_sufile: segusage file inode * @ns_cptree: rb-tree of all mounted checkpoints (nilfs_root) * @ns_cptree_lock: lock protecting @ns_cptree + * @ns_dirty_files: list of dirty files + * @ns_inode_lock: lock protecting @ns_dirty_files * @ns_gc_inodes: dummy inodes to keep live blocks + * @ns_next_generation: next generation number for inodes + * @ns_next_gen_lock: lock protecting @ns_next_generation + * @ns_mount_opt: mount options + * @ns_resuid: uid for reserved blocks + * @ns_resgid: gid for reserved blocks + * @ns_interval: checkpoint creation interval + * @ns_watermark: watermark for the number of dirty buffers * @ns_blocksize_bits: bit length of block size * @ns_blocksize: block size * @ns_nsegments: number of segments in filesystem @@ -131,6 +142,7 @@ struct the_nilfs { u64 ns_prot_seq; u64 ns_prev_seq; + struct nilfs_sc_info *ns_writer; struct rw_semaphore ns_segctor_sem; /* @@ -145,9 +157,25 @@ struct the_nilfs { struct rb_root ns_cptree; spinlock_t ns_cptree_lock; + /* Dirty inode list */ + struct list_head ns_dirty_files; + spinlock_t ns_inode_lock; + /* GC inode list */ struct list_head ns_gc_inodes; + /* Inode allocator */ + u32 ns_next_generation; + spinlock_t ns_next_gen_lock; + + /* Mount options */ + unsigned long ns_mount_opt; + + uid_t ns_resuid; + gid_t ns_resgid; + unsigned long ns_interval; + unsigned long ns_watermark; + /* Disk layout information (static) */ unsigned int ns_blocksize_bits; unsigned int ns_blocksize; @@ -180,6 +208,20 @@ THE_NILFS_FNS(DISCONTINUED, discontinued) THE_NILFS_FNS(GC_RUNNING, gc_running) THE_NILFS_FNS(SB_DIRTY, sb_dirty) +/* + * Mount option operations + */ +#define nilfs_clear_opt(nilfs, opt) \ + do { (nilfs)->ns_mount_opt &= ~NILFS_MOUNT_##opt; } while (0) +#define nilfs_set_opt(nilfs, opt) \ + do { (nilfs)->ns_mount_opt |= NILFS_MOUNT_##opt; } while (0) +#define nilfs_test_opt(nilfs, opt) ((nilfs)->ns_mount_opt & NILFS_MOUNT_##opt) +#define nilfs_write_opt(nilfs, mask, opt) \ + do { (nilfs)->ns_mount_opt = \ + (((nilfs)->ns_mount_opt & ~NILFS_MOUNT_##mask) | \ + NILFS_MOUNT_##opt); \ + } while (0) + /** * struct nilfs_root - nilfs root object * @cno: checkpoint number @@ -224,15 +266,14 @@ static inline int nilfs_sb_will_flip(struct the_nilfs *nilfs) void nilfs_set_last_segment(struct the_nilfs *, sector_t, u64, __u64); struct the_nilfs *alloc_nilfs(struct block_device *bdev); void destroy_nilfs(struct the_nilfs *nilfs); -int init_nilfs(struct the_nilfs *, struct nilfs_sb_info *, char *); -int load_nilfs(struct the_nilfs *, struct nilfs_sb_info *); +int init_nilfs(struct the_nilfs *nilfs, struct super_block *sb, char *data); +int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb); int nilfs_discard_segments(struct the_nilfs *, __u64 *, size_t); int nilfs_count_free_blocks(struct the_nilfs *, sector_t *); struct nilfs_root *nilfs_lookup_root(struct the_nilfs *nilfs, __u64 cno); struct nilfs_root *nilfs_find_or_create_root(struct the_nilfs *nilfs, __u64 cno); void nilfs_put_root(struct nilfs_root *root); -struct nilfs_sb_info *nilfs_find_sbinfo(struct the_nilfs *, int, __u64); int nilfs_near_disk_full(struct the_nilfs *); void nilfs_fall_back_super_block(struct the_nilfs *); void nilfs_swap_super_block(struct the_nilfs *); diff --git a/fs/notify/fanotify/fanotify_user.c b/fs/notify/fanotify/fanotify_user.c index 8b61220..6b1305d 100644 --- a/fs/notify/fanotify/fanotify_user.c +++ b/fs/notify/fanotify/fanotify_user.c @@ -876,7 +876,7 @@ SYSCALL_ALIAS(sys_fanotify_mark, SyS_fanotify_mark); #endif /* - * fanotify_user_setup - Our initialization function. Note that we cannnot return + * fanotify_user_setup - Our initialization function. Note that we cannot return * error because we have compiled-in VFS hooks. So an (unlikely) failure here * must result in panic(). */ diff --git a/fs/notify/inotify/inotify_user.c b/fs/notify/inotify/inotify_user.c index 4cd5d5d..bd46e7c 100644 --- a/fs/notify/inotify/inotify_user.c +++ b/fs/notify/inotify/inotify_user.c @@ -841,7 +841,7 @@ out: } /* - * inotify_user_setup - Our initialization function. Note that we cannnot return + * inotify_user_setup - Our initialization function. Note that we cannot return * error because we have compiled-in VFS hooks. So an (unlikely) failure here * must result in panic(). */ diff --git a/fs/ntfs/Makefile b/fs/ntfs/Makefile index 4ff028f..30206b2 100644 --- a/fs/ntfs/Makefile +++ b/fs/ntfs/Makefile @@ -2,18 +2,13 @@ obj-$(CONFIG_NTFS_FS) += ntfs.o -ntfs-objs := aops.o attrib.o collate.o compress.o debug.o dir.o file.o \ - index.o inode.o mft.o mst.o namei.o runlist.o super.o sysctl.o \ - unistr.o upcase.o +ntfs-y := aops.o attrib.o collate.o compress.o debug.o dir.o file.o \ + index.o inode.o mft.o mst.o namei.o runlist.o super.o sysctl.o \ + unistr.o upcase.o -EXTRA_CFLAGS = -DNTFS_VERSION=\"2.1.30\" +ntfs-$(CONFIG_NTFS_RW) += bitmap.o lcnalloc.o logfile.o quota.o usnjrnl.o -ifeq ($(CONFIG_NTFS_DEBUG),y) -EXTRA_CFLAGS += -DDEBUG -endif +ccflags-y := -DNTFS_VERSION=\"2.1.30\" +ccflags-$(CONFIG_NTFS_DEBUG) += -DDEBUG +ccflags-$(CONFIG_NTFS_RW) += -DNTFS_RW -ifeq ($(CONFIG_NTFS_RW),y) -EXTRA_CFLAGS += -DNTFS_RW - -ntfs-objs += bitmap.o lcnalloc.o logfile.o quota.o usnjrnl.o -endif diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c index b572b67..326e747 100644 --- a/fs/ntfs/mft.c +++ b/fs/ntfs/mft.c @@ -1,7 +1,7 @@ /** * mft.c - NTFS kernel mft record operations. Part of the Linux-NTFS project. * - * Copyright (c) 2001-2006 Anton Altaparmakov + * Copyright (c) 2001-2011 Anton Altaparmakov and Tuxera Inc. * Copyright (c) 2002 Richard Russon * * This program/include file is free software; you can redistribute it and/or @@ -2576,6 +2576,8 @@ mft_rec_already_initialized: flush_dcache_page(page); SetPageUptodate(page); if (base_ni) { + MFT_RECORD *m_tmp; + /* * Setup the base mft record in the extent mft record. This * completes initialization of the allocated extent mft record @@ -2588,11 +2590,11 @@ mft_rec_already_initialized: * attach it to the base inode @base_ni and map, pin, and lock * its, i.e. the allocated, mft record. */ - m = map_extent_mft_record(base_ni, bit, &ni); - if (IS_ERR(m)) { + m_tmp = map_extent_mft_record(base_ni, bit, &ni); + if (IS_ERR(m_tmp)) { ntfs_error(vol->sb, "Failed to map allocated extent " "mft record 0x%llx.", (long long)bit); - err = PTR_ERR(m); + err = PTR_ERR(m_tmp); /* Set the mft record itself not in use. */ m->flags &= cpu_to_le16( ~le16_to_cpu(MFT_RECORD_IN_USE)); @@ -2603,6 +2605,7 @@ mft_rec_already_initialized: ntfs_unmap_page(page); goto undo_mftbmp_alloc; } + BUG_ON(m != m_tmp); /* * Make sure the allocated mft record is written out to disk. * No need to set the inode dirty because the caller is going diff --git a/fs/ocfs2/Makefile b/fs/ocfs2/Makefile index 07d9fd8..d8a0313 100644 --- a/fs/ocfs2/Makefile +++ b/fs/ocfs2/Makefile @@ -1,6 +1,6 @@ -EXTRA_CFLAGS += -Ifs/ocfs2 +ccflags-y := -Ifs/ocfs2 -EXTRA_CFLAGS += -DCATCH_BH_JBD_RACES +ccflags-y += -DCATCH_BH_JBD_RACES obj-$(CONFIG_OCFS2_FS) += \ ocfs2.o \ diff --git a/fs/ocfs2/dcache.c b/fs/ocfs2/dcache.c index 6d80ecc..7eb9040 100644 --- a/fs/ocfs2/dcache.c +++ b/fs/ocfs2/dcache.c @@ -56,7 +56,7 @@ static int ocfs2_dentry_revalidate(struct dentry *dentry, int ret = 0; /* if all else fails, just return false */ struct ocfs2_super *osb; - if (nd->flags & LOOKUP_RCU) + if (nd && nd->flags & LOOKUP_RCU) return -ECHILD; inode = dentry->d_inode; diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index d417b3f..f97b6f1 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -354,7 +354,7 @@ static inline int ocfs2_match(int len, /* * Returns 0 if not found, -1 on failure, and 1 on success */ -static int inline ocfs2_search_dirblock(struct buffer_head *bh, +static inline int ocfs2_search_dirblock(struct buffer_head *bh, struct inode *dir, const char *name, int namelen, unsigned long offset, diff --git a/fs/ocfs2/dlm/Makefile b/fs/ocfs2/dlm/Makefile index dcebf0d..c8a044e 100644 --- a/fs/ocfs2/dlm/Makefile +++ b/fs/ocfs2/dlm/Makefile @@ -1,4 +1,4 @@ -EXTRA_CFLAGS += -Ifs/ocfs2 +ccflags-y := -Ifs/ocfs2 obj-$(CONFIG_OCFS2_FS_O2CB) += ocfs2_dlm.o diff --git a/fs/ocfs2/dlmfs/Makefile b/fs/ocfs2/dlmfs/Makefile index df69b48..f14be89 100644 --- a/fs/ocfs2/dlmfs/Makefile +++ b/fs/ocfs2/dlmfs/Makefile @@ -1,4 +1,4 @@ -EXTRA_CFLAGS += -Ifs/ocfs2 +ccflags-y := -Ifs/ocfs2 obj-$(CONFIG_OCFS2_FS) += ocfs2_dlmfs.o diff --git a/fs/ocfs2/export.c b/fs/ocfs2/export.c index 5dbc306..254652a 100644 --- a/fs/ocfs2/export.c +++ b/fs/ocfs2/export.c @@ -197,8 +197,12 @@ static int ocfs2_encode_fh(struct dentry *dentry, u32 *fh_in, int *max_len, dentry->d_name.len, dentry->d_name.name, fh, len, connectable); - if (len < 3 || (connectable && len < 6)) { - mlog(ML_ERROR, "fh buffer is too small for encoding\n"); + if (connectable && (len < 6)) { + *max_len = 6; + type = 255; + goto bail; + } else if (len < 3) { + *max_len = 3; type = 255; goto bail; } diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h index 43e56b9..6180da1 100644 --- a/fs/ocfs2/journal.h +++ b/fs/ocfs2/journal.h @@ -405,9 +405,9 @@ static inline int ocfs2_remove_extent_credits(struct super_block *sb) ocfs2_quota_trans_credits(sb); } -/* data block for new dir/symlink, 2 for bitmap updates (bitmap fe + - * bitmap block for the new bit) dx_root update for free list */ -#define OCFS2_DIR_LINK_ADDITIONAL_CREDITS (1 + 2 + 1) +/* data block for new dir/symlink, allocation of directory block, dx_root + * update for free list */ +#define OCFS2_DIR_LINK_ADDITIONAL_CREDITS (1 + OCFS2_SUBALLOC_ALLOC + 1) static inline int ocfs2_add_dir_index_credits(struct super_block *sb) { diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index 849fb4a..d6c25d7 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -293,7 +293,7 @@ static int ocfs2_mknod(struct inode *dir, } /* get security xattr */ - status = ocfs2_init_security_get(inode, dir, &si); + status = ocfs2_init_security_get(inode, dir, &dentry->d_name, &si); if (status) { if (status == -EOPNOTSUPP) si.enable = 0; @@ -1665,7 +1665,7 @@ static int ocfs2_symlink(struct inode *dir, } /* get security xattr */ - status = ocfs2_init_security_get(inode, dir, &si); + status = ocfs2_init_security_get(inode, dir, &dentry->d_name, &si); if (status) { if (status == -EOPNOTSUPP) si.enable = 0; diff --git a/fs/ocfs2/quota.h b/fs/ocfs2/quota.h index 196fcb5..d5ab56c 100644 --- a/fs/ocfs2/quota.h +++ b/fs/ocfs2/quota.h @@ -114,7 +114,4 @@ int ocfs2_local_write_dquot(struct dquot *dquot); extern const struct dquot_operations ocfs2_quota_operations; extern struct quota_format_type ocfs2_quota_format; -int ocfs2_quota_setup(void); -void ocfs2_quota_shutdown(void); - #endif /* _OCFS2_QUOTA_H */ diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index 4607923..a73f641 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -63,8 +63,6 @@ * write to gf */ -static struct workqueue_struct *ocfs2_quota_wq = NULL; - static void qsync_work_fn(struct work_struct *work); static void ocfs2_global_disk2memdqb(struct dquot *dquot, void *dp) @@ -400,8 +398,8 @@ int ocfs2_global_read_info(struct super_block *sb, int type) OCFS2_QBLK_RESERVED_SPACE; oinfo->dqi_gi.dqi_qtree_depth = qtree_depth(&oinfo->dqi_gi); INIT_DELAYED_WORK(&oinfo->dqi_sync_work, qsync_work_fn); - queue_delayed_work(ocfs2_quota_wq, &oinfo->dqi_sync_work, - msecs_to_jiffies(oinfo->dqi_syncms)); + schedule_delayed_work(&oinfo->dqi_sync_work, + msecs_to_jiffies(oinfo->dqi_syncms)); out_err: mlog_exit(status); @@ -635,8 +633,8 @@ static void qsync_work_fn(struct work_struct *work) struct super_block *sb = oinfo->dqi_gqinode->i_sb; dquot_scan_active(sb, ocfs2_sync_dquot_helper, oinfo->dqi_type); - queue_delayed_work(ocfs2_quota_wq, &oinfo->dqi_sync_work, - msecs_to_jiffies(oinfo->dqi_syncms)); + schedule_delayed_work(&oinfo->dqi_sync_work, + msecs_to_jiffies(oinfo->dqi_syncms)); } /* @@ -923,20 +921,3 @@ const struct dquot_operations ocfs2_quota_operations = { .alloc_dquot = ocfs2_alloc_dquot, .destroy_dquot = ocfs2_destroy_dquot, }; - -int ocfs2_quota_setup(void) -{ - ocfs2_quota_wq = create_workqueue("o2quot"); - if (!ocfs2_quota_wq) - return -ENOMEM; - return 0; -} - -void ocfs2_quota_shutdown(void) -{ - if (ocfs2_quota_wq) { - flush_workqueue(ocfs2_quota_wq); - destroy_workqueue(ocfs2_quota_wq); - ocfs2_quota_wq = NULL; - } -} diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c index b5f9160..c384d63 100644 --- a/fs/ocfs2/refcounttree.c +++ b/fs/ocfs2/refcounttree.c @@ -3228,7 +3228,7 @@ static int ocfs2_make_clusters_writable(struct super_block *sb, u32 num_clusters, unsigned int e_flags) { int ret, delete, index, credits = 0; - u32 new_bit, new_len; + u32 new_bit, new_len, orig_num_clusters; unsigned int set_len; struct ocfs2_super *osb = OCFS2_SB(sb); handle_t *handle; @@ -3261,6 +3261,8 @@ static int ocfs2_make_clusters_writable(struct super_block *sb, goto out; } + orig_num_clusters = num_clusters; + while (num_clusters) { ret = ocfs2_get_refcount_rec(ref_ci, context->ref_root_bh, p_cluster, num_clusters, @@ -3348,7 +3350,8 @@ static int ocfs2_make_clusters_writable(struct super_block *sb, * in write-back mode. */ if (context->get_clusters == ocfs2_di_get_clusters) { - ret = ocfs2_cow_sync_writeback(sb, context, cpos, num_clusters); + ret = ocfs2_cow_sync_writeback(sb, context, cpos, + orig_num_clusters); if (ret) mlog_errno(ret); } @@ -4325,7 +4328,8 @@ static int ocfs2_reflink(struct dentry *old_dentry, struct inode *dir, /* If the security isn't preserved, we need to re-initialize them. */ if (!preserve) { - error = ocfs2_init_security_and_acl(dir, new_orphan_inode); + error = ocfs2_init_security_and_acl(dir, new_orphan_inode, + &new_dentry->d_name); if (error) mlog_errno(error); } @@ -4376,7 +4380,7 @@ static int ocfs2_user_path_parent(const char __user *path, if (IS_ERR(s)) return PTR_ERR(s); - error = path_lookup(s, LOOKUP_PARENT, nd); + error = kern_path_parent(s, nd); if (error) putname(s); else diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 06d1f74..236ed1b 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -993,8 +993,7 @@ static void ocfs2_disable_quotas(struct ocfs2_super *osb) } /* Handle quota on quotactl */ -static int ocfs2_quota_on(struct super_block *sb, int type, int format_id, - char *path) +static int ocfs2_quota_on(struct super_block *sb, int type, int format_id) { unsigned int feature[MAXQUOTAS] = { OCFS2_FEATURE_RO_COMPAT_USRQUOTA, OCFS2_FEATURE_RO_COMPAT_GRPQUOTA}; @@ -1013,7 +1012,7 @@ static int ocfs2_quota_off(struct super_block *sb, int type) } static const struct quotactl_ops ocfs2_quotactl_ops = { - .quota_on = ocfs2_quota_on, + .quota_on_meta = ocfs2_quota_on, .quota_off = ocfs2_quota_off, .quota_sync = dquot_quota_sync, .get_info = dquot_get_dqinfo, @@ -1317,7 +1316,7 @@ static int ocfs2_parse_options(struct super_block *sb, struct mount_options *mopt, int is_remount) { - int status; + int status, user_stack = 0; char *p; u32 tmp; @@ -1460,6 +1459,15 @@ static int ocfs2_parse_options(struct super_block *sb, memcpy(mopt->cluster_stack, args[0].from, OCFS2_STACK_LABEL_LEN); mopt->cluster_stack[OCFS2_STACK_LABEL_LEN] = '\0'; + /* + * Open code the memcmp here as we don't have + * an osb to pass to + * ocfs2_userspace_stack(). + */ + if (memcmp(mopt->cluster_stack, + OCFS2_CLASSIC_CLUSTER_STACK, + OCFS2_STACK_LABEL_LEN)) + user_stack = 1; break; case Opt_inode64: mopt->mount_opt |= OCFS2_MOUNT_INODE64; @@ -1515,13 +1523,16 @@ static int ocfs2_parse_options(struct super_block *sb, } } - /* Ensure only one heartbeat mode */ - tmp = mopt->mount_opt & (OCFS2_MOUNT_HB_LOCAL | OCFS2_MOUNT_HB_GLOBAL | - OCFS2_MOUNT_HB_NONE); - if (hweight32(tmp) != 1) { - mlog(ML_ERROR, "Invalid heartbeat mount options\n"); - status = 0; - goto bail; + if (user_stack == 0) { + /* Ensure only one heartbeat mode */ + tmp = mopt->mount_opt & (OCFS2_MOUNT_HB_LOCAL | + OCFS2_MOUNT_HB_GLOBAL | + OCFS2_MOUNT_HB_NONE); + if (hweight32(tmp) != 1) { + mlog(ML_ERROR, "Invalid heartbeat mount options\n"); + status = 0; + goto bail; + } } status = 1; @@ -1646,16 +1657,11 @@ static int __init ocfs2_init(void) mlog(ML_ERROR, "Unable to create ocfs2 debugfs root.\n"); } - status = ocfs2_quota_setup(); - if (status) - goto leave; - ocfs2_set_locking_protocol(); status = register_quota_format(&ocfs2_quota_format); leave: if (status < 0) { - ocfs2_quota_shutdown(); ocfs2_free_mem_caches(); exit_ocfs2_uptodate_cache(); } @@ -1672,8 +1678,6 @@ static void __exit ocfs2_exit(void) { mlog_entry_void(); - ocfs2_quota_shutdown(); - if (ocfs2_wq) { flush_workqueue(ocfs2_wq); destroy_workqueue(ocfs2_wq); diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index 67cd439..6bb6024 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -7185,7 +7185,8 @@ out: * must not hold any lock expect i_mutex. */ int ocfs2_init_security_and_acl(struct inode *dir, - struct inode *inode) + struct inode *inode, + const struct qstr *qstr) { int ret = 0; struct buffer_head *dir_bh = NULL; @@ -7193,7 +7194,7 @@ int ocfs2_init_security_and_acl(struct inode *dir, .enable = 1, }; - ret = ocfs2_init_security_get(inode, dir, &si); + ret = ocfs2_init_security_get(inode, dir, qstr, &si); if (!ret) { ret = ocfs2_xattr_set(inode, OCFS2_XATTR_INDEX_SECURITY, si.name, si.value, si.value_len, @@ -7261,13 +7262,14 @@ static int ocfs2_xattr_security_set(struct dentry *dentry, const char *name, int ocfs2_init_security_get(struct inode *inode, struct inode *dir, + const struct qstr *qstr, struct ocfs2_security_xattr_info *si) { /* check whether ocfs2 support feature xattr */ if (!ocfs2_supports_xattr(OCFS2_SB(dir->i_sb))) return -EOPNOTSUPP; - return security_inode_init_security(inode, dir, &si->name, &si->value, - &si->value_len); + return security_inode_init_security(inode, dir, qstr, &si->name, + &si->value, &si->value_len); } int ocfs2_init_security_set(handle_t *handle, diff --git a/fs/ocfs2/xattr.h b/fs/ocfs2/xattr.h index aa64bb3..d63cfb7 100644 --- a/fs/ocfs2/xattr.h +++ b/fs/ocfs2/xattr.h @@ -57,6 +57,7 @@ int ocfs2_has_inline_xattr_value_outside(struct inode *inode, struct ocfs2_dinode *di); int ocfs2_xattr_remove(struct inode *, struct buffer_head *); int ocfs2_init_security_get(struct inode *, struct inode *, + const struct qstr *, struct ocfs2_security_xattr_info *); int ocfs2_init_security_set(handle_t *, struct inode *, struct buffer_head *, @@ -94,5 +95,6 @@ int ocfs2_reflink_xattrs(struct inode *old_inode, struct buffer_head *new_bh, bool preserve_security); int ocfs2_init_security_and_acl(struct inode *dir, - struct inode *inode); + struct inode *inode, + const struct qstr *qstr); #endif /* OCFS2_XATTR_H */ diff --git a/fs/omfs/dir.c b/fs/omfs/dir.c index 393f3f6..de4ff29 100644 --- a/fs/omfs/dir.c +++ b/fs/omfs/dir.c @@ -235,33 +235,22 @@ static int omfs_dir_is_empty(struct inode *inode) return *ptr != ~0; } -static int omfs_unlink(struct inode *dir, struct dentry *dentry) +static int omfs_remove(struct inode *dir, struct dentry *dentry) { - int ret; struct inode *inode = dentry->d_inode; + int ret; + + if (S_ISDIR(inode->i_mode) && !omfs_dir_is_empty(inode)) + return -ENOTEMPTY; ret = omfs_delete_entry(dentry); if (ret) - goto end_unlink; - - inode_dec_link_count(inode); + return ret; + + clear_nlink(inode); + mark_inode_dirty(inode); mark_inode_dirty(dir); - -end_unlink: - return ret; -} - -static int omfs_rmdir(struct inode *dir, struct dentry *dentry) -{ - int err = -ENOTEMPTY; - struct inode *inode = dentry->d_inode; - - if (omfs_dir_is_empty(inode)) { - err = omfs_unlink(dir, dentry); - if (!err) - inode_dec_link_count(inode); - } - return err; + return 0; } static int omfs_add_node(struct inode *dir, struct dentry *dentry, int mode) @@ -372,9 +361,10 @@ static int omfs_fill_chain(struct file *filp, void *dirent, filldir_t filldir, res = filldir(dirent, oi->i_name, strnlen(oi->i_name, OMFS_NAMELEN), filp->f_pos, self, d_type); - if (res == 0) - filp->f_pos++; brelse(bh); + if (res < 0) + break; + filp->f_pos++; } out: return res; @@ -385,44 +375,28 @@ static int omfs_rename(struct inode *old_dir, struct dentry *old_dentry, { struct inode *new_inode = new_dentry->d_inode; struct inode *old_inode = old_dentry->d_inode; - struct buffer_head *bh; - int is_dir; int err; - is_dir = S_ISDIR(old_inode->i_mode); - if (new_inode) { /* overwriting existing file/dir */ - err = -ENOTEMPTY; - if (is_dir && !omfs_dir_is_empty(new_inode)) - goto out; - - err = -ENOENT; - bh = omfs_find_entry(new_dir, new_dentry->d_name.name, - new_dentry->d_name.len); - if (IS_ERR(bh)) - goto out; - brelse(bh); - - err = omfs_unlink(new_dir, new_dentry); + err = omfs_remove(new_dir, new_dentry); if (err) goto out; } /* since omfs locates files by name, we need to unlink _before_ * adding the new link or we won't find the old one */ - inode_inc_link_count(old_inode); - err = omfs_unlink(old_dir, old_dentry); - if (err) { - inode_dec_link_count(old_inode); + err = omfs_delete_entry(old_dentry); + if (err) goto out; - } + mark_inode_dirty(old_dir); err = omfs_add_link(new_dentry, old_inode); if (err) goto out; old_inode->i_ctime = CURRENT_TIME_SEC; + mark_inode_dirty(old_inode); out: return err; } @@ -488,8 +462,8 @@ const struct inode_operations omfs_dir_inops = { .mkdir = omfs_mkdir, .rename = omfs_rename, .create = omfs_create, - .unlink = omfs_unlink, - .rmdir = omfs_rmdir, + .unlink = omfs_remove, + .rmdir = omfs_remove, }; const struct file_operations omfs_dir_operations = { @@ -233,6 +233,14 @@ int do_fallocate(struct file *file, int mode, loff_t offset, loff_t len) if (!(file->f_mode & FMODE_WRITE)) return -EBADF; + + /* It's not possible punch hole on append only file */ + if (mode & FALLOC_FL_PUNCH_HOLE && IS_APPEND(inode)) + return -EPERM; + + if (IS_IMMUTABLE(inode)) + return -EPERM; + /* * Revalidate the write permissions, in case security policy has * changed since the files were opened. @@ -565,13 +573,15 @@ SYSCALL_DEFINE5(fchownat, int, dfd, const char __user *, filename, uid_t, user, { struct path path; int error = -EINVAL; - int follow; + int lookup_flags; - if ((flag & ~AT_SYMLINK_NOFOLLOW) != 0) + if ((flag & ~(AT_SYMLINK_NOFOLLOW | AT_EMPTY_PATH)) != 0) goto out; - follow = (flag & AT_SYMLINK_NOFOLLOW) ? 0 : LOOKUP_FOLLOW; - error = user_path_at(dfd, filename, follow, &path); + lookup_flags = (flag & AT_SYMLINK_NOFOLLOW) ? 0 : LOOKUP_FOLLOW; + if (flag & AT_EMPTY_PATH) + lookup_flags |= LOOKUP_EMPTY; + error = user_path_at(dfd, filename, lookup_flags, &path); if (error) goto out; error = mnt_want_write(path.mnt); @@ -661,11 +671,16 @@ static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt, int (*open)(struct inode *, struct file *), const struct cred *cred) { + static const struct file_operations empty_fops = {}; struct inode *inode; int error; f->f_mode = OPEN_FMODE(f->f_flags) | FMODE_LSEEK | FMODE_PREAD | FMODE_PWRITE; + + if (unlikely(f->f_flags & O_PATH)) + f->f_mode = FMODE_PATH; + inode = dentry->d_inode; if (f->f_mode & FMODE_WRITE) { error = __get_file_write_access(inode, mnt); @@ -679,9 +694,15 @@ static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt, f->f_path.dentry = dentry; f->f_path.mnt = mnt; f->f_pos = 0; - f->f_op = fops_get(inode->i_fop); file_sb_list_add(f, inode->i_sb); + if (unlikely(f->f_mode & FMODE_PATH)) { + f->f_op = &empty_fops; + return f; + } + + f->f_op = fops_get(inode->i_fop); + error = security_dentry_open(f, cred); if (error) goto cleanup_all; @@ -693,7 +714,8 @@ static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt, if (error) goto cleanup_all; } - ima_counts_get(f); + if ((f->f_mode & (FMODE_READ | FMODE_WRITE)) == FMODE_READ) + i_readcount_inc(inode); f->f_flags &= ~(O_CREAT | O_EXCL | O_NOCTTY | O_TRUNC); @@ -790,6 +812,8 @@ struct file *nameidata_to_filp(struct nameidata *nd) /* Pick up the filp from the open intent */ filp = nd->intent.open.file; + nd->intent.open.file = NULL; + /* Has the filesystem initialised the file for us? */ if (filp->f_path.dentry == NULL) { path_get(&nd->path); @@ -811,17 +835,8 @@ struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags, validate_creds(cred); - /* - * We must always pass in a valid mount pointer. Historically - * callers got away with not passing it, but we must enforce this at - * the earliest possible point now to avoid strange problems deep in the - * filesystem stack. - */ - if (!mnt) { - printk(KERN_WARNING "%s called with NULL vfsmount\n", __func__); - dump_stack(); - return ERR_PTR(-EINVAL); - } + /* We must always pass in a valid mount pointer. */ + BUG_ON(!mnt); error = -ENFILE; f = get_empty_filp(); @@ -880,15 +895,110 @@ void fd_install(unsigned int fd, struct file *file) EXPORT_SYMBOL(fd_install); +static inline int build_open_flags(int flags, int mode, struct open_flags *op) +{ + int lookup_flags = 0; + int acc_mode; + + if (!(flags & O_CREAT)) + mode = 0; + op->mode = mode; + + /* Must never be set by userspace */ + flags &= ~FMODE_NONOTIFY; + + /* + * O_SYNC is implemented as __O_SYNC|O_DSYNC. As many places only + * check for O_DSYNC if the need any syncing at all we enforce it's + * always set instead of having to deal with possibly weird behaviour + * for malicious applications setting only __O_SYNC. + */ + if (flags & __O_SYNC) + flags |= O_DSYNC; + + /* + * If we have O_PATH in the open flag. Then we + * cannot have anything other than the below set of flags + */ + if (flags & O_PATH) { + flags &= O_DIRECTORY | O_NOFOLLOW | O_PATH; + acc_mode = 0; + } else { + acc_mode = MAY_OPEN | ACC_MODE(flags); + } + + op->open_flag = flags; + + /* O_TRUNC implies we need access checks for write permissions */ + if (flags & O_TRUNC) + acc_mode |= MAY_WRITE; + + /* Allow the LSM permission hook to distinguish append + access from general write access. */ + if (flags & O_APPEND) + acc_mode |= MAY_APPEND; + + op->acc_mode = acc_mode; + + op->intent = flags & O_PATH ? 0 : LOOKUP_OPEN; + + if (flags & O_CREAT) { + op->intent |= LOOKUP_CREATE; + if (flags & O_EXCL) + op->intent |= LOOKUP_EXCL; + } + + if (flags & O_DIRECTORY) + lookup_flags |= LOOKUP_DIRECTORY; + if (!(flags & O_NOFOLLOW)) + lookup_flags |= LOOKUP_FOLLOW; + return lookup_flags; +} + +/** + * filp_open - open file and return file pointer + * + * @filename: path to open + * @flags: open flags as per the open(2) second argument + * @mode: mode for the new file if O_CREAT is set, else ignored + * + * This is the helper to open a file from kernelspace if you really + * have to. But in generally you should not do this, so please move + * along, nothing to see here.. + */ +struct file *filp_open(const char *filename, int flags, int mode) +{ + struct open_flags op; + int lookup = build_open_flags(flags, mode, &op); + return do_filp_open(AT_FDCWD, filename, &op, lookup); +} +EXPORT_SYMBOL(filp_open); + +struct file *file_open_root(struct dentry *dentry, struct vfsmount *mnt, + const char *filename, int flags) +{ + struct open_flags op; + int lookup = build_open_flags(flags, 0, &op); + if (flags & O_CREAT) + return ERR_PTR(-EINVAL); + if (!filename && (flags & O_DIRECTORY)) + if (!dentry->d_inode->i_op->lookup) + return ERR_PTR(-ENOTDIR); + return do_file_open_root(dentry, mnt, filename, &op, lookup); +} +EXPORT_SYMBOL(file_open_root); + long do_sys_open(int dfd, const char __user *filename, int flags, int mode) { + struct open_flags op; + int lookup = build_open_flags(flags, mode, &op); char *tmp = getname(filename); int fd = PTR_ERR(tmp); if (!IS_ERR(tmp)) { fd = get_unused_fd_flags(flags); if (fd >= 0) { - struct file *f = do_filp_open(dfd, tmp, flags, mode, 0); + struct file *f = do_filp_open(dfd, tmp, &op, lookup); if (IS_ERR(f)) { put_unused_fd(fd); fd = PTR_ERR(f); @@ -958,8 +1068,10 @@ int filp_close(struct file *filp, fl_owner_t id) if (filp->f_op && filp->f_op->flush) retval = filp->f_op->flush(filp, id); - dnotify_flush(filp, id); - locks_remove_posix(filp, id); + if (likely(!(filp->f_mode & FMODE_PATH))) { + dnotify_flush(filp, id); + locks_remove_posix(filp, id); + } fput(filp); return retval; } diff --git a/fs/partitions/ldm.c b/fs/partitions/ldm.c index 789c625..b10e354 100644 --- a/fs/partitions/ldm.c +++ b/fs/partitions/ldm.c @@ -251,6 +251,11 @@ static bool ldm_parse_vmdb (const u8 *data, struct vmdb *vm) } vm->vblk_size = get_unaligned_be32(data + 0x08); + if (vm->vblk_size == 0) { + ldm_error ("Illegal VBLK size"); + return false; + } + vm->vblk_offset = get_unaligned_be32(data + 0x0C); vm->last_vblk_seq = get_unaligned_be32(data + 0x04); diff --git a/fs/partitions/mac.c b/fs/partitions/mac.c index 68d6a21..11f688b 100644 --- a/fs/partitions/mac.c +++ b/fs/partitions/mac.c @@ -29,10 +29,9 @@ static inline void mac_fix_string(char *stg, int len) int mac_partition(struct parsed_partitions *state) { - int slot = 1; Sector sect; unsigned char *data; - int blk, blocks_in_map; + int slot, blocks_in_map; unsigned secsize; #ifdef CONFIG_PPC_PMAC int found_root = 0; @@ -59,10 +58,14 @@ int mac_partition(struct parsed_partitions *state) put_dev_sector(sect); return 0; /* not a MacOS disk */ } - strlcat(state->pp_buf, " [mac]", PAGE_SIZE); blocks_in_map = be32_to_cpu(part->map_count); - for (blk = 1; blk <= blocks_in_map; ++blk) { - int pos = blk * secsize; + if (blocks_in_map < 0 || blocks_in_map >= DISK_MAX_PARTS) { + put_dev_sector(sect); + return 0; + } + strlcat(state->pp_buf, " [mac]", PAGE_SIZE); + for (slot = 1; slot <= blocks_in_map; ++slot) { + int pos = slot * secsize; put_dev_sector(sect); data = read_part_sector(state, pos/512, §); if (!data) @@ -113,13 +116,11 @@ int mac_partition(struct parsed_partitions *state) } if (goodness > found_root_goodness) { - found_root = blk; + found_root = slot; found_root_goodness = goodness; } } #endif /* CONFIG_PPC_PMAC */ - - ++slot; } #ifdef CONFIG_PPC_PMAC if (found_root_goodness) diff --git a/fs/partitions/osf.c b/fs/partitions/osf.c index 48cec7c..764b86a 100644 --- a/fs/partitions/osf.c +++ b/fs/partitions/osf.c @@ -10,10 +10,13 @@ #include "check.h" #include "osf.h" +#define MAX_OSF_PARTITIONS 18 + int osf_partition(struct parsed_partitions *state) { int i; int slot = 1; + unsigned int npartitions; Sector sect; unsigned char *data; struct disklabel { @@ -45,7 +48,7 @@ int osf_partition(struct parsed_partitions *state) u8 p_fstype; u8 p_frag; __le16 p_cpg; - } d_partitions[8]; + } d_partitions[MAX_OSF_PARTITIONS]; } * label; struct d_partition * partition; @@ -63,7 +66,12 @@ int osf_partition(struct parsed_partitions *state) put_dev_sector(sect); return 0; } - for (i = 0 ; i < le16_to_cpu(label->d_npartitions); i++, partition++) { + npartitions = le16_to_cpu(label->d_npartitions); + if (npartitions > MAX_OSF_PARTITIONS) { + put_dev_sector(sect); + return 0; + } + for (i = 0 ; i < npartitions; i++, partition++) { if (slot == state->limit) break; if (le32_to_cpu(partition->p_size)) @@ -441,7 +441,7 @@ redo: break; } if (do_wakeup) { - wake_up_interruptible_sync_poll(&pipe->wait, POLLOUT); + wake_up_interruptible_sync_poll(&pipe->wait, POLLOUT | POLLWRNORM); kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); } pipe_wait(pipe); @@ -450,7 +450,7 @@ redo: /* Signal writers asynchronously that there is more room. */ if (do_wakeup) { - wake_up_interruptible_sync_poll(&pipe->wait, POLLOUT); + wake_up_interruptible_sync_poll(&pipe->wait, POLLOUT | POLLWRNORM); kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); } if (ret > 0) @@ -612,7 +612,7 @@ redo2: break; } if (do_wakeup) { - wake_up_interruptible_sync_poll(&pipe->wait, POLLIN); + wake_up_interruptible_sync_poll(&pipe->wait, POLLIN | POLLRDNORM); kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); do_wakeup = 0; } @@ -623,7 +623,7 @@ redo2: out: mutex_unlock(&inode->i_mutex); if (do_wakeup) { - wake_up_interruptible_sync_poll(&pipe->wait, POLLIN); + wake_up_interruptible_sync_poll(&pipe->wait, POLLIN | POLLRDNORM); kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); } if (ret > 0) @@ -715,7 +715,7 @@ pipe_release(struct inode *inode, int decr, int decw) if (!pipe->readers && !pipe->writers) { free_pipe_info(inode); } else { - wake_up_interruptible_sync_poll(&pipe->wait, POLLIN | POLLOUT); + wake_up_interruptible_sync_poll(&pipe->wait, POLLIN | POLLOUT | POLLRDNORM | POLLWRNORM | POLLERR | POLLHUP); kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); } diff --git a/fs/posix_acl.c b/fs/posix_acl.c index 39df95a..b1cf6bf 100644 --- a/fs/posix_acl.c +++ b/fs/posix_acl.c @@ -22,6 +22,7 @@ #include <linux/errno.h> +EXPORT_SYMBOL(posix_acl_init); EXPORT_SYMBOL(posix_acl_alloc); EXPORT_SYMBOL(posix_acl_clone); EXPORT_SYMBOL(posix_acl_valid); @@ -32,6 +33,16 @@ EXPORT_SYMBOL(posix_acl_chmod_masq); EXPORT_SYMBOL(posix_acl_permission); /* + * Init a fresh posix_acl + */ +void +posix_acl_init(struct posix_acl *acl, int count) +{ + atomic_set(&acl->a_refcount, 1); + acl->a_count = count; +} + +/* * Allocate a new ACL with the specified number of entries. */ struct posix_acl * @@ -40,10 +51,8 @@ posix_acl_alloc(int count, gfp_t flags) const size_t size = sizeof(struct posix_acl) + count * sizeof(struct posix_acl_entry); struct posix_acl *acl = kmalloc(size, flags); - if (acl) { - atomic_set(&acl->a_refcount, 1); - acl->a_count = count; - } + if (acl) + posix_acl_init(acl, count); return acl; } diff --git a/fs/proc/Kconfig b/fs/proc/Kconfig index 6a00688..15af622 100644 --- a/fs/proc/Kconfig +++ b/fs/proc/Kconfig @@ -1,5 +1,5 @@ config PROC_FS - bool "/proc file system support" if EMBEDDED + bool "/proc file system support" if EXPERT default y help This is a virtual file system providing information about the status @@ -40,7 +40,7 @@ config PROC_VMCORE Exports the dump image of crashed kernel in ELF format. config PROC_SYSCTL - bool "Sysctl support (/proc/sys)" if EMBEDDED + bool "Sysctl support (/proc/sys)" if EXPERT depends on PROC_FS select SYSCTL default y @@ -61,7 +61,7 @@ config PROC_SYSCTL config PROC_PAGE_MONITOR default y depends on PROC_FS && MMU - bool "Enable /proc page monitoring" if EMBEDDED + bool "Enable /proc page monitoring" if EXPERT help Various /proc files exist to monitor process memory utilization: /proc/pid/smaps, /proc/pid/clear_refs, /proc/pid/pagemap, diff --git a/fs/proc/array.c b/fs/proc/array.c index df2b703..7c99c1c 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -353,9 +353,6 @@ int proc_pid_status(struct seq_file *m, struct pid_namespace *ns, task_cap(m, task); task_cpus_allowed(m, task); cpuset_task_status_allowed(m, task); -#if defined(CONFIG_S390) - task_show_regs(m, task); -#endif task_context_switch_counts(m, task); return 0; } diff --git a/fs/proc/base.c b/fs/proc/base.c index 9d096e8..d49c4b5 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -2620,35 +2620,6 @@ static const struct pid_entry proc_base_stuff[] = { &proc_self_inode_operations, NULL, {}), }; -/* - * Exceptional case: normally we are not allowed to unhash a busy - * directory. In this case, however, we can do it - no aliasing problems - * due to the way we treat inodes. - */ -static int proc_base_revalidate(struct dentry *dentry, struct nameidata *nd) -{ - struct inode *inode; - struct task_struct *task; - - if (nd->flags & LOOKUP_RCU) - return -ECHILD; - - inode = dentry->d_inode; - task = get_proc_task(inode); - if (task) { - put_task_struct(task); - return 1; - } - d_drop(dentry); - return 0; -} - -static const struct dentry_operations proc_base_dentry_operations = -{ - .d_revalidate = proc_base_revalidate, - .d_delete = pid_delete_dentry, -}; - static struct dentry *proc_base_instantiate(struct inode *dir, struct dentry *dentry, struct task_struct *task, const void *ptr) { @@ -2685,7 +2656,6 @@ static struct dentry *proc_base_instantiate(struct inode *dir, if (p->fop) inode->i_fop = p->fop; ei->op = p->op; - d_set_d_op(dentry, &proc_base_dentry_operations); d_add(dentry, inode); error = NULL; out: diff --git a/fs/proc/consoles.c b/fs/proc/consoles.c index eafc22a..b701eaa 100644 --- a/fs/proc/consoles.c +++ b/fs/proc/consoles.c @@ -67,7 +67,7 @@ static void *c_start(struct seq_file *m, loff_t *pos) struct console *con; loff_t off = 0; - acquire_console_sem(); + console_lock(); for_each_console(con) if (off++ == *pos) break; @@ -84,7 +84,7 @@ static void *c_next(struct seq_file *m, void *v, loff_t *pos) static void c_stop(struct seq_file *m, void *v) { - release_console_sem(); + console_unlock(); } static const struct seq_operations consoles_op = { diff --git a/fs/proc/inode.c b/fs/proc/inode.c index 176ce4c..d6a7ca1 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -27,6 +27,7 @@ static void proc_evict_inode(struct inode *inode) { struct proc_dir_entry *de; + struct ctl_table_header *head; truncate_inode_pages(&inode->i_data, 0); end_writeback(inode); @@ -38,8 +39,11 @@ static void proc_evict_inode(struct inode *inode) de = PROC_I(inode)->pde; if (de) pde_put(de); - if (PROC_I(inode)->sysctl) - sysctl_head_put(PROC_I(inode)->sysctl); + head = PROC_I(inode)->sysctl; + if (head) { + rcu_assign_pointer(PROC_I(inode)->sysctl, NULL); + sysctl_head_put(head); + } } struct vfsmount *proc_mnt; diff --git a/fs/proc/proc_devtree.c b/fs/proc/proc_devtree.c index d9396a4..927cbd1 100644 --- a/fs/proc/proc_devtree.c +++ b/fs/proc/proc_devtree.c @@ -233,7 +233,7 @@ void __init proc_device_tree_init(void) return; root = of_find_node_by_path("/"); if (root == NULL) { - printk(KERN_ERR "/proc/device-tree: can't find root\n"); + pr_debug("/proc/device-tree: can't find root\n"); return; } proc_device_tree_add_node(root, proc_device_tree); diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index 09a1f92..f50133c 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -32,7 +32,6 @@ static struct inode *proc_sys_make_inode(struct super_block *sb, ei->sysctl_entry = table; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; - inode->i_flags |= S_PRIVATE; /* tell selinux to ignore this inode */ inode->i_mode = table->mode; if (!table->child) { inode->i_mode |= S_IFREG; @@ -408,15 +407,18 @@ static int proc_sys_compare(const struct dentry *parent, const struct dentry *dentry, const struct inode *inode, unsigned int len, const char *str, const struct qstr *name) { + struct ctl_table_header *head; /* Although proc doesn't have negative dentries, rcu-walk means * that inode here can be NULL */ + /* AV: can it, indeed? */ if (!inode) - return 0; + return 1; if (name->len != len) return 1; if (memcmp(name->name, str, len)) return 1; - return !sysctl_is_seen(PROC_I(inode)->sysctl); + head = rcu_dereference(PROC_I(inode)->sysctl); + return !head || !sysctl_is_seen(head); } static const struct dentry_operations proc_sys_dentry_operations = { diff --git a/fs/pstore/Kconfig b/fs/pstore/Kconfig new file mode 100644 index 0000000..867d0ac --- /dev/null +++ b/fs/pstore/Kconfig @@ -0,0 +1,13 @@ +config PSTORE + bool "Persistant store support" + default n + help + This option enables generic access to platform level + persistent storage via "pstore" filesystem that can + be mounted as /dev/pstore. Only useful if you have + a platform level driver that registers with pstore to + provide the data, so you probably should just go say "Y" + (or "M") to a platform specific persistent store driver + (e.g. ACPI_APEI on X86) which will select this for you. + If you don't have a platform persistent store driver, + say N. diff --git a/fs/pstore/Makefile b/fs/pstore/Makefile new file mode 100644 index 0000000..760f4bc --- /dev/null +++ b/fs/pstore/Makefile @@ -0,0 +1,7 @@ +# +# Makefile for the linux pstorefs routines. +# + +obj-y += pstore.o + +pstore-objs += inode.o platform.o diff --git a/fs/pstore/inode.c b/fs/pstore/inode.c new file mode 100644 index 0000000..977ed27 --- /dev/null +++ b/fs/pstore/inode.c @@ -0,0 +1,311 @@ +/* + * Persistent Storage - ramfs parts. + * + * Copyright (C) 2010 Intel Corporation <tony.luck@intel.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include <linux/module.h> +#include <linux/fs.h> +#include <linux/fsnotify.h> +#include <linux/pagemap.h> +#include <linux/highmem.h> +#include <linux/time.h> +#include <linux/init.h> +#include <linux/string.h> +#include <linux/mount.h> +#include <linux/ramfs.h> +#include <linux/parser.h> +#include <linux/sched.h> +#include <linux/magic.h> +#include <linux/pstore.h> +#include <linux/slab.h> +#include <linux/uaccess.h> + +#include "internal.h" + +#define PSTORE_NAMELEN 64 + +struct pstore_private { + u64 id; + int (*erase)(u64); + ssize_t size; + char data[]; +}; + +static int pstore_file_open(struct inode *inode, struct file *file) +{ + file->private_data = inode->i_private; + return 0; +} + +static ssize_t pstore_file_read(struct file *file, char __user *userbuf, + size_t count, loff_t *ppos) +{ + struct pstore_private *ps = file->private_data; + + return simple_read_from_buffer(userbuf, count, ppos, ps->data, ps->size); +} + +static const struct file_operations pstore_file_operations = { + .open = pstore_file_open, + .read = pstore_file_read, + .llseek = default_llseek, +}; + +/* + * When a file is unlinked from our file system we call the + * platform driver to erase the record from persistent store. + */ +static int pstore_unlink(struct inode *dir, struct dentry *dentry) +{ + struct pstore_private *p = dentry->d_inode->i_private; + + p->erase(p->id); + + return simple_unlink(dir, dentry); +} + +static void pstore_evict_inode(struct inode *inode) +{ + end_writeback(inode); + kfree(inode->i_private); +} + +static const struct inode_operations pstore_dir_inode_operations = { + .lookup = simple_lookup, + .unlink = pstore_unlink, +}; + +static struct inode *pstore_get_inode(struct super_block *sb, + const struct inode *dir, int mode, dev_t dev) +{ + struct inode *inode = new_inode(sb); + + if (inode) { + inode->i_ino = get_next_ino(); + inode->i_uid = inode->i_gid = 0; + inode->i_mode = mode; + inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; + switch (mode & S_IFMT) { + case S_IFREG: + inode->i_fop = &pstore_file_operations; + break; + case S_IFDIR: + inode->i_op = &pstore_dir_inode_operations; + inode->i_fop = &simple_dir_operations; + inc_nlink(inode); + break; + } + } + return inode; +} + +enum { + Opt_kmsg_bytes, Opt_err +}; + +static const match_table_t tokens = { + {Opt_kmsg_bytes, "kmsg_bytes=%u"}, + {Opt_err, NULL} +}; + +static void parse_options(char *options) +{ + char *p; + substring_t args[MAX_OPT_ARGS]; + int option; + + if (!options) + return; + + while ((p = strsep(&options, ",")) != NULL) { + int token; + + if (!*p) + continue; + + token = match_token(p, tokens, args); + switch (token) { + case Opt_kmsg_bytes: + if (!match_int(&args[0], &option)) + pstore_set_kmsg_bytes(option); + break; + } + } +} + +static int pstore_remount(struct super_block *sb, int *flags, char *data) +{ + parse_options(data); + + return 0; +} + +static const struct super_operations pstore_ops = { + .statfs = simple_statfs, + .drop_inode = generic_delete_inode, + .evict_inode = pstore_evict_inode, + .remount_fs = pstore_remount, + .show_options = generic_show_options, +}; + +static struct super_block *pstore_sb; + +int pstore_is_mounted(void) +{ + return pstore_sb != NULL; +} + +/* + * Make a regular file in the root directory of our file system. + * Load it up with "size" bytes of data from "buf". + * Set the mtime & ctime to the date that this record was originally stored. + */ +int pstore_mkfile(enum pstore_type_id type, char *psname, u64 id, + char *data, size_t size, + struct timespec time, int (*erase)(u64)) +{ + struct dentry *root = pstore_sb->s_root; + struct dentry *dentry; + struct inode *inode; + int rc; + char name[PSTORE_NAMELEN]; + struct pstore_private *private; + + rc = -ENOMEM; + inode = pstore_get_inode(pstore_sb, root->d_inode, S_IFREG | 0444, 0); + if (!inode) + goto fail; + private = kmalloc(sizeof *private + size, GFP_KERNEL); + if (!private) + goto fail_alloc; + private->id = id; + private->erase = erase; + + switch (type) { + case PSTORE_TYPE_DMESG: + sprintf(name, "dmesg-%s-%lld", psname, id); + break; + case PSTORE_TYPE_MCE: + sprintf(name, "mce-%s-%lld", psname, id); + break; + case PSTORE_TYPE_UNKNOWN: + sprintf(name, "unknown-%s-%lld", psname, id); + break; + default: + sprintf(name, "type%d-%s-%lld", type, psname, id); + break; + } + + mutex_lock(&root->d_inode->i_mutex); + + rc = -ENOSPC; + dentry = d_alloc_name(root, name); + if (IS_ERR(dentry)) + goto fail_lockedalloc; + + memcpy(private->data, data, size); + inode->i_size = private->size = size; + + inode->i_private = private; + + if (time.tv_sec) + inode->i_mtime = inode->i_ctime = time; + + d_add(dentry, inode); + + mutex_unlock(&root->d_inode->i_mutex); + + return 0; + +fail_lockedalloc: + mutex_unlock(&root->d_inode->i_mutex); + kfree(private); +fail_alloc: + iput(inode); + +fail: + return rc; +} + +int pstore_fill_super(struct super_block *sb, void *data, int silent) +{ + struct inode *inode = NULL; + struct dentry *root; + int err; + + save_mount_options(sb, data); + + pstore_sb = sb; + + sb->s_maxbytes = MAX_LFS_FILESIZE; + sb->s_blocksize = PAGE_CACHE_SIZE; + sb->s_blocksize_bits = PAGE_CACHE_SHIFT; + sb->s_magic = PSTOREFS_MAGIC; + sb->s_op = &pstore_ops; + sb->s_time_gran = 1; + + parse_options(data); + + inode = pstore_get_inode(sb, NULL, S_IFDIR | 0755, 0); + if (!inode) { + err = -ENOMEM; + goto fail; + } + /* override ramfs "dir" options so we catch unlink(2) */ + inode->i_op = &pstore_dir_inode_operations; + + root = d_alloc_root(inode); + sb->s_root = root; + if (!root) { + err = -ENOMEM; + goto fail; + } + + pstore_get_records(); + + return 0; +fail: + iput(inode); + return err; +} + +static struct dentry *pstore_mount(struct file_system_type *fs_type, + int flags, const char *dev_name, void *data) +{ + return mount_single(fs_type, flags, data, pstore_fill_super); +} + +static void pstore_kill_sb(struct super_block *sb) +{ + kill_litter_super(sb); + pstore_sb = NULL; +} + +static struct file_system_type pstore_fs_type = { + .name = "pstore", + .mount = pstore_mount, + .kill_sb = pstore_kill_sb, +}; + +static int __init init_pstore_fs(void) +{ + return register_filesystem(&pstore_fs_type); +} +module_init(init_pstore_fs) + +MODULE_AUTHOR("Tony Luck <tony.luck@intel.com>"); +MODULE_LICENSE("GPL"); diff --git a/fs/pstore/internal.h b/fs/pstore/internal.h new file mode 100644 index 0000000..8c9f23e --- /dev/null +++ b/fs/pstore/internal.h @@ -0,0 +1,6 @@ +extern void pstore_set_kmsg_bytes(int); +extern void pstore_get_records(void); +extern int pstore_mkfile(enum pstore_type_id, char *psname, u64 id, + char *data, size_t size, + struct timespec time, int (*erase)(u64)); +extern int pstore_is_mounted(void); diff --git a/fs/pstore/platform.c b/fs/pstore/platform.c new file mode 100644 index 0000000..ce9ad84 --- /dev/null +++ b/fs/pstore/platform.c @@ -0,0 +1,192 @@ +/* + * Persistent Storage - platform driver interface parts. + * + * Copyright (C) 2010 Intel Corporation <tony.luck@intel.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + */ + +#include <linux/atomic.h> +#include <linux/types.h> +#include <linux/errno.h> +#include <linux/init.h> +#include <linux/kmsg_dump.h> +#include <linux/module.h> +#include <linux/pstore.h> +#include <linux/string.h> +#include <linux/slab.h> +#include <linux/uaccess.h> + +#include "internal.h" + +/* + * pstore_lock just protects "psinfo" during + * calls to pstore_register() + */ +static DEFINE_SPINLOCK(pstore_lock); +static struct pstore_info *psinfo; + +/* How much of the console log to snapshot */ +static unsigned long kmsg_bytes = 10240; + +void pstore_set_kmsg_bytes(int bytes) +{ + kmsg_bytes = bytes; +} + +/* Tag each group of saved records with a sequence number */ +static int oopscount; + +/* + * callback from kmsg_dump. (s2,l2) has the most recently + * written bytes, older bytes are in (s1,l1). Save as much + * as we can from the end of the buffer. + */ +static void pstore_dump(struct kmsg_dumper *dumper, + enum kmsg_dump_reason reason, + const char *s1, unsigned long l1, + const char *s2, unsigned long l2) +{ + unsigned long s1_start, s2_start; + unsigned long l1_cpy, l2_cpy; + unsigned long size, total = 0; + char *dst; + u64 id; + int hsize, part = 1; + + mutex_lock(&psinfo->buf_mutex); + oopscount++; + while (total < kmsg_bytes) { + dst = psinfo->buf; + hsize = sprintf(dst, "Oops#%d Part%d\n", oopscount, part++); + size = psinfo->bufsize - hsize; + dst += hsize; + + l2_cpy = min(l2, size); + l1_cpy = min(l1, size - l2_cpy); + + if (l1_cpy + l2_cpy == 0) + break; + + s2_start = l2 - l2_cpy; + s1_start = l1 - l1_cpy; + + memcpy(dst, s1 + s1_start, l1_cpy); + memcpy(dst + l1_cpy, s2 + s2_start, l2_cpy); + + id = psinfo->write(PSTORE_TYPE_DMESG, hsize + l1_cpy + l2_cpy); + if (pstore_is_mounted()) + pstore_mkfile(PSTORE_TYPE_DMESG, psinfo->name, id, + psinfo->buf, hsize + l1_cpy + l2_cpy, + CURRENT_TIME, psinfo->erase); + l1 -= l1_cpy; + l2 -= l2_cpy; + total += l1_cpy + l2_cpy; + } + mutex_unlock(&psinfo->buf_mutex); +} + +static struct kmsg_dumper pstore_dumper = { + .dump = pstore_dump, +}; + +/* + * platform specific persistent storage driver registers with + * us here. If pstore is already mounted, call the platform + * read function right away to populate the file system. If not + * then the pstore mount code will call us later to fill out + * the file system. + * + * Register with kmsg_dump to save last part of console log on panic. + */ +int pstore_register(struct pstore_info *psi) +{ + struct module *owner = psi->owner; + + spin_lock(&pstore_lock); + if (psinfo) { + spin_unlock(&pstore_lock); + return -EBUSY; + } + psinfo = psi; + spin_unlock(&pstore_lock); + + if (owner && !try_module_get(owner)) { + psinfo = NULL; + return -EINVAL; + } + + if (pstore_is_mounted()) + pstore_get_records(); + + kmsg_dump_register(&pstore_dumper); + + return 0; +} +EXPORT_SYMBOL_GPL(pstore_register); + +/* + * Read all the records from the persistent store. Create and + * file files in our filesystem. + */ +void pstore_get_records(void) +{ + struct pstore_info *psi = psinfo; + size_t size; + u64 id; + enum pstore_type_id type; + struct timespec time; + int failed = 0; + + if (!psi) + return; + + mutex_lock(&psinfo->buf_mutex); + while ((size = psi->read(&id, &type, &time)) > 0) { + if (pstore_mkfile(type, psi->name, id, psi->buf, size, + time, psi->erase)) + failed++; + } + mutex_unlock(&psinfo->buf_mutex); + + if (failed) + printk(KERN_WARNING "pstore: failed to load %d record(s) from '%s'\n", + failed, psi->name); +} + +/* + * Call platform driver to write a record to the + * persistent store. + */ +int pstore_write(enum pstore_type_id type, char *buf, size_t size) +{ + u64 id; + + if (!psinfo) + return -ENODEV; + + if (size > psinfo->bufsize) + return -EFBIG; + + mutex_lock(&psinfo->buf_mutex); + memcpy(psinfo->buf, buf, size); + id = psinfo->write(type, size); + if (pstore_is_mounted()) + pstore_mkfile(PSTORE_TYPE_DMESG, psinfo->name, id, psinfo->buf, + size, CURRENT_TIME, psinfo->erase); + mutex_unlock(&psinfo->buf_mutex); + + return 0; +} +EXPORT_SYMBOL_GPL(pstore_write); diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index 84becd3..a2a622e 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -2189,8 +2189,8 @@ int dquot_resume(struct super_block *sb, int type) } EXPORT_SYMBOL(dquot_resume); -int dquot_quota_on_path(struct super_block *sb, int type, int format_id, - struct path *path) +int dquot_quota_on(struct super_block *sb, int type, int format_id, + struct path *path) { int error = security_quota_on(path->dentry); if (error) @@ -2204,20 +2204,6 @@ int dquot_quota_on_path(struct super_block *sb, int type, int format_id, DQUOT_LIMITS_ENABLED); return error; } -EXPORT_SYMBOL(dquot_quota_on_path); - -int dquot_quota_on(struct super_block *sb, int type, int format_id, char *name) -{ - struct path path; - int error; - - error = kern_path(name, LOOKUP_FOLLOW, &path); - if (!error) { - error = dquot_quota_on_path(sb, type, format_id, &path); - path_put(&path); - } - return error; -} EXPORT_SYMBOL(dquot_quota_on); /* diff --git a/fs/quota/quota.c b/fs/quota/quota.c index b299961..b34bdb2 100644 --- a/fs/quota/quota.c +++ b/fs/quota/quota.c @@ -64,18 +64,15 @@ static int quota_sync_all(int type) } static int quota_quotaon(struct super_block *sb, int type, int cmd, qid_t id, - void __user *addr) + struct path *path) { - char *pathname; - int ret = -ENOSYS; - - pathname = getname(addr); - if (IS_ERR(pathname)) - return PTR_ERR(pathname); - if (sb->s_qcop->quota_on) - ret = sb->s_qcop->quota_on(sb, type, id, pathname); - putname(pathname); - return ret; + if (!sb->s_qcop->quota_on && !sb->s_qcop->quota_on_meta) + return -ENOSYS; + if (sb->s_qcop->quota_on_meta) + return sb->s_qcop->quota_on_meta(sb, type, id); + if (IS_ERR(path)) + return PTR_ERR(path); + return sb->s_qcop->quota_on(sb, type, id, path); } static int quota_getfmt(struct super_block *sb, int type, void __user *addr) @@ -241,7 +238,7 @@ static int quota_getxquota(struct super_block *sb, int type, qid_t id, /* Copy parameters and call proper function */ static int do_quotactl(struct super_block *sb, int type, int cmd, qid_t id, - void __user *addr) + void __user *addr, struct path *path) { int ret; @@ -256,7 +253,7 @@ static int do_quotactl(struct super_block *sb, int type, int cmd, qid_t id, switch (cmd) { case Q_QUOTAON: - return quota_quotaon(sb, type, cmd, id, addr); + return quota_quotaon(sb, type, cmd, id, path); case Q_QUOTAOFF: if (!sb->s_qcop->quota_off) return -ENOSYS; @@ -335,6 +332,7 @@ SYSCALL_DEFINE4(quotactl, unsigned int, cmd, const char __user *, special, { uint cmds, type; struct super_block *sb = NULL; + struct path path, *pathp = NULL; int ret; cmds = cmd >> SUBCMDSHIFT; @@ -351,12 +349,27 @@ SYSCALL_DEFINE4(quotactl, unsigned int, cmd, const char __user *, special, return -ENODEV; } + /* + * Path for quotaon has to be resolved before grabbing superblock + * because that gets s_umount sem which is also possibly needed by path + * resolution (think about autofs) and thus deadlocks could arise. + */ + if (cmds == Q_QUOTAON) { + ret = user_path_at(AT_FDCWD, addr, LOOKUP_FOLLOW, &path); + if (ret) + pathp = ERR_PTR(ret); + else + pathp = &path; + } + sb = quotactl_block(special); if (IS_ERR(sb)) return PTR_ERR(sb); - ret = do_quotactl(sb, type, cmds, id, addr); + ret = do_quotactl(sb, type, cmds, id, addr, pathp); drop_super(sb); + if (pathp && !IS_ERR(pathp)) + path_put(pathp); return ret; } diff --git a/fs/quota/quota_v2.c b/fs/quota/quota_v2.c index 65444d2..f1ab360 100644 --- a/fs/quota/quota_v2.c +++ b/fs/quota/quota_v2.c @@ -112,7 +112,7 @@ static int v2_read_file_info(struct super_block *sb, int type) if (!info->dqi_priv) { printk(KERN_WARNING "Not enough memory for quota information structure.\n"); - return -1; + return -ENOMEM; } qinfo = info->dqi_priv; if (version == 0) { diff --git a/fs/reiserfs/Makefile b/fs/reiserfs/Makefile index 792b3cb..3c3b001 100644 --- a/fs/reiserfs/Makefile +++ b/fs/reiserfs/Makefile @@ -31,9 +31,7 @@ endif # and causing a panic. Since this behavior only affects ppc32, this ifeq # will work around it. If any other architecture displays this behavior, # add it here. -ifeq ($(CONFIG_PPC32),y) -EXTRA_CFLAGS := $(call cc-ifversion, -lt, 0400, -O1) -endif +ccflags-$(CONFIG_PPC32) := $(call cc-ifversion, -lt, 0400, -O1) TAGS: etags *.c diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index 0bae036..1bba24b 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -1593,8 +1593,13 @@ int reiserfs_encode_fh(struct dentry *dentry, __u32 * data, int *lenp, struct inode *inode = dentry->d_inode; int maxlen = *lenp; - if (maxlen < 3) + if (need_parent && (maxlen < 5)) { + *lenp = 5; return 255; + } else if (maxlen < 3) { + *lenp = 3; + return 255; + } data[0] = inode->i_ino; data[1] = le32_to_cpu(INODE_PKEY(inode)->k_dir_id); diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c index 3eea859..c77514b 100644 --- a/fs/reiserfs/journal.c +++ b/fs/reiserfs/journal.c @@ -2876,7 +2876,7 @@ int journal_init(struct super_block *sb, const char *j_dev_name, reiserfs_mounted_fs_count++; if (reiserfs_mounted_fs_count <= 1) { reiserfs_write_unlock(sb); - commit_wq = create_workqueue("reiserfs"); + commit_wq = alloc_workqueue("reiserfs", WQ_MEM_RECLAIM, 0); reiserfs_write_lock(sb); } diff --git a/fs/reiserfs/namei.c b/fs/reiserfs/namei.c index ba5f51e..1186626 100644 --- a/fs/reiserfs/namei.c +++ b/fs/reiserfs/namei.c @@ -593,7 +593,7 @@ static int reiserfs_create(struct inode *dir, struct dentry *dentry, int mode, new_inode_init(inode, dir, mode); jbegin_count += reiserfs_cache_default_acl(dir); - retval = reiserfs_security_init(dir, inode, &security); + retval = reiserfs_security_init(dir, inode, &dentry->d_name, &security); if (retval < 0) { drop_new_inode(inode); return retval; @@ -667,7 +667,7 @@ static int reiserfs_mknod(struct inode *dir, struct dentry *dentry, int mode, new_inode_init(inode, dir, mode); jbegin_count += reiserfs_cache_default_acl(dir); - retval = reiserfs_security_init(dir, inode, &security); + retval = reiserfs_security_init(dir, inode, &dentry->d_name, &security); if (retval < 0) { drop_new_inode(inode); return retval; @@ -747,7 +747,7 @@ static int reiserfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) new_inode_init(inode, dir, mode); jbegin_count += reiserfs_cache_default_acl(dir); - retval = reiserfs_security_init(dir, inode, &security); + retval = reiserfs_security_init(dir, inode, &dentry->d_name, &security); if (retval < 0) { drop_new_inode(inode); return retval; @@ -771,7 +771,7 @@ static int reiserfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) EMPTY_DIR_SIZE_V1 : EMPTY_DIR_SIZE, dentry, inode, &security); if (retval) { - dir->i_nlink--; + DEC_DIR_INODE_NLINK(dir) goto out_failed; } @@ -1032,7 +1032,8 @@ static int reiserfs_symlink(struct inode *parent_dir, } new_inode_init(inode, parent_dir, mode); - retval = reiserfs_security_init(parent_dir, inode, &security); + retval = reiserfs_security_init(parent_dir, inode, &dentry->d_name, + &security); if (retval < 0) { drop_new_inode(inode); return retval; @@ -1122,10 +1123,6 @@ static int reiserfs_link(struct dentry *old_dentry, struct inode *dir, reiserfs_write_unlock(dir->i_sb); return -EMLINK; } - if (inode->i_nlink == 0) { - reiserfs_write_unlock(dir->i_sb); - return -ENOENT; - } /* inc before scheduling so reiserfs_unlink knows we are here */ inc_nlink(inode); diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index 2575682..0aab04f 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -632,7 +632,7 @@ static int reiserfs_acquire_dquot(struct dquot *); static int reiserfs_release_dquot(struct dquot *); static int reiserfs_mark_dquot_dirty(struct dquot *); static int reiserfs_write_info(struct super_block *, int); -static int reiserfs_quota_on(struct super_block *, int, int, char *); +static int reiserfs_quota_on(struct super_block *, int, int, struct path *); static const struct dquot_operations reiserfs_quota_operations = { .write_dquot = reiserfs_write_dquot, @@ -2048,25 +2048,21 @@ static int reiserfs_quota_on_mount(struct super_block *sb, int type) * Standard function to be called on quota_on */ static int reiserfs_quota_on(struct super_block *sb, int type, int format_id, - char *name) + struct path *path) { int err; - struct path path; struct inode *inode; struct reiserfs_transaction_handle th; if (!(REISERFS_SB(sb)->s_mount_opt & (1 << REISERFS_QUOTA))) return -EINVAL; - err = kern_path(name, LOOKUP_FOLLOW, &path); - if (err) - return err; /* Quotafile not on the same filesystem? */ - if (path.mnt->mnt_sb != sb) { + if (path->mnt->mnt_sb != sb) { err = -EXDEV; goto out; } - inode = path.dentry->d_inode; + inode = path->dentry->d_inode; /* We must not pack tails for quota files on reiserfs for quota IO to work */ if (!(REISERFS_I(inode)->i_flags & i_nopack_mask)) { err = reiserfs_unpack(inode, NULL); @@ -2082,7 +2078,7 @@ static int reiserfs_quota_on(struct super_block *sb, int type, int format_id, /* Journaling quota? */ if (REISERFS_SB(sb)->s_qf_names[type]) { /* Quotafile not of fs root? */ - if (path.dentry->d_parent != sb->s_root) + if (path->dentry->d_parent != sb->s_root) reiserfs_warning(sb, "super-6521", "Quota file not on filesystem root. " "Journalled quota will not work."); @@ -2101,9 +2097,8 @@ static int reiserfs_quota_on(struct super_block *sb, int type, int format_id, if (err) goto out; } - err = dquot_quota_on_path(sb, type, format_id, &path); + err = dquot_quota_on(sb, type, format_id, path); out: - path_put(&path); return err; } diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c index 3cfb2e9..5c11ca8 100644 --- a/fs/reiserfs/xattr.c +++ b/fs/reiserfs/xattr.c @@ -978,8 +978,6 @@ int reiserfs_permission(struct inode *inode, int mask, unsigned int flags) static int xattr_hide_revalidate(struct dentry *dentry, struct nameidata *nd) { - if (nd->flags & LOOKUP_RCU) - return -ECHILD; return -EPERM; } diff --git a/fs/reiserfs/xattr_security.c b/fs/reiserfs/xattr_security.c index 237c692..ef66c18 100644 --- a/fs/reiserfs/xattr_security.c +++ b/fs/reiserfs/xattr_security.c @@ -54,6 +54,7 @@ static size_t security_list(struct dentry *dentry, char *list, size_t list_len, * of blocks needed for the transaction. If successful, reiserfs_security * must be released using reiserfs_security_free when the caller is done. */ int reiserfs_security_init(struct inode *dir, struct inode *inode, + const struct qstr *qstr, struct reiserfs_security_handle *sec) { int blocks = 0; @@ -65,7 +66,7 @@ int reiserfs_security_init(struct inode *dir, struct inode *inode, if (IS_PRIVATE(dir)) return 0; - error = security_inode_init_security(inode, dir, &sec->name, + error = security_inode_init_security(inode, dir, qstr, &sec->name, &sec->value, &sec->length); if (error) { if (error == -EOPNOTSUPP) diff --git a/fs/select.c b/fs/select.c index e56560d..d33418f 100644 --- a/fs/select.c +++ b/fs/select.c @@ -517,9 +517,6 @@ int do_select(int n, fd_set_bits *fds, struct timespec *end_time) * Update: ERESTARTSYS breaks at least the xview clock binary, so * I'm trying ERESTARTNOHAND which restart only when you want to. */ -#define MAX_SELECT_SECONDS \ - ((unsigned long) (MAX_SCHEDULE_TIMEOUT / HZ)-1) - int core_sys_select(int n, fd_set __user *inp, fd_set __user *outp, fd_set __user *exp, struct timespec *end_time) { diff --git a/fs/squashfs/block.c b/fs/squashfs/block.c index 2fb2882..8ab48bc 100644 --- a/fs/squashfs/block.c +++ b/fs/squashfs/block.c @@ -63,6 +63,14 @@ static struct buffer_head *get_block_length(struct super_block *sb, *length = (unsigned char) bh->b_data[*offset] | (unsigned char) bh->b_data[*offset + 1] << 8; *offset += 2; + + if (*offset == msblk->devblksize) { + put_bh(bh); + bh = sb_bread(sb, ++(*cur_index)); + if (bh == NULL) + return NULL; + *offset = 0; + } } return bh; diff --git a/fs/squashfs/xz_wrapper.c b/fs/squashfs/xz_wrapper.c index 856756c..c4eb400 100644 --- a/fs/squashfs/xz_wrapper.c +++ b/fs/squashfs/xz_wrapper.c @@ -95,12 +95,6 @@ static int squashfs_xz_uncompress(struct squashfs_sb_info *msblk, void **buffer, if (!buffer_uptodate(bh[k])) goto release_mutex; - if (avail == 0) { - offset = 0; - put_bh(bh[k++]); - continue; - } - stream->buf.in = bh[k]->b_data + offset; stream->buf.in_size = avail; stream->buf.in_pos = 0; diff --git a/fs/squashfs/zlib_wrapper.c b/fs/squashfs/zlib_wrapper.c index 818a5e0..4661ae2 100644 --- a/fs/squashfs/zlib_wrapper.c +++ b/fs/squashfs/zlib_wrapper.c @@ -82,12 +82,6 @@ static int zlib_uncompress(struct squashfs_sb_info *msblk, void **buffer, if (!buffer_uptodate(bh[k])) goto release_mutex; - if (avail == 0) { - offset = 0; - put_bh(bh[k++]); - continue; - } - stream->next_in = bh[k]->b_data + offset; stream->avail_in = avail; offset = 0; @@ -75,13 +75,16 @@ int vfs_fstatat(int dfd, const char __user *filename, struct kstat *stat, int error = -EINVAL; int lookup_flags = 0; - if ((flag & ~(AT_SYMLINK_NOFOLLOW | AT_NO_AUTOMOUNT)) != 0) + if ((flag & ~(AT_SYMLINK_NOFOLLOW | AT_NO_AUTOMOUNT | + AT_EMPTY_PATH)) != 0) goto out; if (!(flag & AT_SYMLINK_NOFOLLOW)) lookup_flags |= LOOKUP_FOLLOW; if (flag & AT_NO_AUTOMOUNT) lookup_flags |= LOOKUP_NO_AUTOMOUNT; + if (flag & AT_EMPTY_PATH) + lookup_flags |= LOOKUP_EMPTY; error = user_path_at(dfd, filename, lookup_flags, &path); if (error) @@ -297,7 +300,7 @@ SYSCALL_DEFINE4(readlinkat, int, dfd, const char __user *, pathname, if (bufsiz <= 0) return -EINVAL; - error = user_path_at(dfd, pathname, 0, &path); + error = user_path_at(dfd, pathname, LOOKUP_EMPTY, &path); if (!error) { struct inode *inode = path.dentry->d_inode; diff --git a/fs/statfs.c b/fs/statfs.c index 30ea8c8..8244924 100644 --- a/fs/statfs.c +++ b/fs/statfs.c @@ -73,149 +73,135 @@ int vfs_statfs(struct path *path, struct kstatfs *buf) } EXPORT_SYMBOL(vfs_statfs); -static int do_statfs_native(struct path *path, struct statfs *buf) +int user_statfs(const char __user *pathname, struct kstatfs *st) { - struct kstatfs st; - int retval; + struct path path; + int error = user_path(pathname, &path); + if (!error) { + error = vfs_statfs(&path, st); + path_put(&path); + } + return error; +} - retval = vfs_statfs(path, &st); - if (retval) - return retval; +int fd_statfs(int fd, struct kstatfs *st) +{ + struct file *file = fget(fd); + int error = -EBADF; + if (file) { + error = vfs_statfs(&file->f_path, st); + fput(file); + } + return error; +} - if (sizeof(*buf) == sizeof(st)) - memcpy(buf, &st, sizeof(st)); +static int do_statfs_native(struct kstatfs *st, struct statfs __user *p) +{ + struct statfs buf; + + if (sizeof(buf) == sizeof(*st)) + memcpy(&buf, st, sizeof(*st)); else { - if (sizeof buf->f_blocks == 4) { - if ((st.f_blocks | st.f_bfree | st.f_bavail | - st.f_bsize | st.f_frsize) & + if (sizeof buf.f_blocks == 4) { + if ((st->f_blocks | st->f_bfree | st->f_bavail | + st->f_bsize | st->f_frsize) & 0xffffffff00000000ULL) return -EOVERFLOW; /* * f_files and f_ffree may be -1; it's okay to stuff * that into 32 bits */ - if (st.f_files != -1 && - (st.f_files & 0xffffffff00000000ULL)) + if (st->f_files != -1 && + (st->f_files & 0xffffffff00000000ULL)) return -EOVERFLOW; - if (st.f_ffree != -1 && - (st.f_ffree & 0xffffffff00000000ULL)) + if (st->f_ffree != -1 && + (st->f_ffree & 0xffffffff00000000ULL)) return -EOVERFLOW; } - buf->f_type = st.f_type; - buf->f_bsize = st.f_bsize; - buf->f_blocks = st.f_blocks; - buf->f_bfree = st.f_bfree; - buf->f_bavail = st.f_bavail; - buf->f_files = st.f_files; - buf->f_ffree = st.f_ffree; - buf->f_fsid = st.f_fsid; - buf->f_namelen = st.f_namelen; - buf->f_frsize = st.f_frsize; - buf->f_flags = st.f_flags; - memset(buf->f_spare, 0, sizeof(buf->f_spare)); + buf.f_type = st->f_type; + buf.f_bsize = st->f_bsize; + buf.f_blocks = st->f_blocks; + buf.f_bfree = st->f_bfree; + buf.f_bavail = st->f_bavail; + buf.f_files = st->f_files; + buf.f_ffree = st->f_ffree; + buf.f_fsid = st->f_fsid; + buf.f_namelen = st->f_namelen; + buf.f_frsize = st->f_frsize; + buf.f_flags = st->f_flags; + memset(buf.f_spare, 0, sizeof(buf.f_spare)); } + if (copy_to_user(p, &buf, sizeof(buf))) + return -EFAULT; return 0; } -static int do_statfs64(struct path *path, struct statfs64 *buf) +static int do_statfs64(struct kstatfs *st, struct statfs64 __user *p) { - struct kstatfs st; - int retval; - - retval = vfs_statfs(path, &st); - if (retval) - return retval; - - if (sizeof(*buf) == sizeof(st)) - memcpy(buf, &st, sizeof(st)); + struct statfs64 buf; + if (sizeof(buf) == sizeof(*st)) + memcpy(&buf, st, sizeof(*st)); else { - buf->f_type = st.f_type; - buf->f_bsize = st.f_bsize; - buf->f_blocks = st.f_blocks; - buf->f_bfree = st.f_bfree; - buf->f_bavail = st.f_bavail; - buf->f_files = st.f_files; - buf->f_ffree = st.f_ffree; - buf->f_fsid = st.f_fsid; - buf->f_namelen = st.f_namelen; - buf->f_frsize = st.f_frsize; - buf->f_flags = st.f_flags; - memset(buf->f_spare, 0, sizeof(buf->f_spare)); + buf.f_type = st->f_type; + buf.f_bsize = st->f_bsize; + buf.f_blocks = st->f_blocks; + buf.f_bfree = st->f_bfree; + buf.f_bavail = st->f_bavail; + buf.f_files = st->f_files; + buf.f_ffree = st->f_ffree; + buf.f_fsid = st->f_fsid; + buf.f_namelen = st->f_namelen; + buf.f_frsize = st->f_frsize; + buf.f_flags = st->f_flags; + memset(buf.f_spare, 0, sizeof(buf.f_spare)); } + if (copy_to_user(p, &buf, sizeof(buf))) + return -EFAULT; return 0; } SYSCALL_DEFINE2(statfs, const char __user *, pathname, struct statfs __user *, buf) { - struct path path; - int error; - - error = user_path(pathname, &path); - if (!error) { - struct statfs tmp; - error = do_statfs_native(&path, &tmp); - if (!error && copy_to_user(buf, &tmp, sizeof(tmp))) - error = -EFAULT; - path_put(&path); - } + struct kstatfs st; + int error = user_statfs(pathname, &st); + if (!error) + error = do_statfs_native(&st, buf); return error; } SYSCALL_DEFINE3(statfs64, const char __user *, pathname, size_t, sz, struct statfs64 __user *, buf) { - struct path path; - long error; - + struct kstatfs st; + int error; if (sz != sizeof(*buf)) return -EINVAL; - error = user_path(pathname, &path); - if (!error) { - struct statfs64 tmp; - error = do_statfs64(&path, &tmp); - if (!error && copy_to_user(buf, &tmp, sizeof(tmp))) - error = -EFAULT; - path_put(&path); - } + error = user_statfs(pathname, &st); + if (!error) + error = do_statfs64(&st, buf); return error; } SYSCALL_DEFINE2(fstatfs, unsigned int, fd, struct statfs __user *, buf) { - struct file *file; - struct statfs tmp; - int error; - - error = -EBADF; - file = fget(fd); - if (!file) - goto out; - error = do_statfs_native(&file->f_path, &tmp); - if (!error && copy_to_user(buf, &tmp, sizeof(tmp))) - error = -EFAULT; - fput(file); -out: + struct kstatfs st; + int error = fd_statfs(fd, &st); + if (!error) + error = do_statfs_native(&st, buf); return error; } SYSCALL_DEFINE3(fstatfs64, unsigned int, fd, size_t, sz, struct statfs64 __user *, buf) { - struct file *file; - struct statfs64 tmp; + struct kstatfs st; int error; if (sz != sizeof(*buf)) return -EINVAL; - error = -EBADF; - file = fget(fd); - if (!file) - goto out; - error = do_statfs64(&file->f_path, &tmp); - if (!error && copy_to_user(buf, &tmp, sizeof(tmp))) - error = -EFAULT; - fput(file); -out: + error = fd_statfs(fd, &st); + if (!error) + error = do_statfs64(&st, buf); return error; } @@ -177,6 +177,11 @@ void deactivate_locked_super(struct super_block *s) struct file_system_type *fs = s->s_type; if (atomic_dec_and_test(&s->s_active)) { fs->kill_sb(s); + /* + * We need to call rcu_barrier so all the delayed rcu free + * inodes are flushed before we release the fs module. + */ + rcu_barrier(); put_filesystem(fs); put_super(s); } else { @@ -838,23 +843,6 @@ error: } EXPORT_SYMBOL(mount_bdev); -int get_sb_bdev(struct file_system_type *fs_type, - int flags, const char *dev_name, void *data, - int (*fill_super)(struct super_block *, void *, int), - struct vfsmount *mnt) -{ - struct dentry *root; - - root = mount_bdev(fs_type, flags, dev_name, data, fill_super); - if (IS_ERR(root)) - return PTR_ERR(root); - mnt->mnt_root = root; - mnt->mnt_sb = root->d_sb; - return 0; -} - -EXPORT_SYMBOL(get_sb_bdev); - void kill_block_super(struct super_block *sb) { struct block_device *bdev = sb->s_bdev; @@ -892,22 +880,6 @@ struct dentry *mount_nodev(struct file_system_type *fs_type, } EXPORT_SYMBOL(mount_nodev); -int get_sb_nodev(struct file_system_type *fs_type, - int flags, void *data, - int (*fill_super)(struct super_block *, void *, int), - struct vfsmount *mnt) -{ - struct dentry *root; - - root = mount_nodev(fs_type, flags, data, fill_super); - if (IS_ERR(root)) - return PTR_ERR(root); - mnt->mnt_root = root; - mnt->mnt_sb = root->d_sb; - return 0; -} -EXPORT_SYMBOL(get_sb_nodev); - static int compare_single(struct super_block *s, void *p) { return 1; @@ -938,69 +910,35 @@ struct dentry *mount_single(struct file_system_type *fs_type, } EXPORT_SYMBOL(mount_single); -int get_sb_single(struct file_system_type *fs_type, - int flags, void *data, - int (*fill_super)(struct super_block *, void *, int), - struct vfsmount *mnt) -{ - struct dentry *root; - root = mount_single(fs_type, flags, data, fill_super); - if (IS_ERR(root)) - return PTR_ERR(root); - mnt->mnt_root = root; - mnt->mnt_sb = root->d_sb; - return 0; -} - -EXPORT_SYMBOL(get_sb_single); - -struct vfsmount * -vfs_kern_mount(struct file_system_type *type, int flags, const char *name, void *data) +struct dentry * +mount_fs(struct file_system_type *type, int flags, const char *name, void *data) { - struct vfsmount *mnt; struct dentry *root; + struct super_block *sb; char *secdata = NULL; - int error; - - if (!type) - return ERR_PTR(-ENODEV); - - error = -ENOMEM; - mnt = alloc_vfsmnt(name); - if (!mnt) - goto out; - - if (flags & MS_KERNMOUNT) - mnt->mnt_flags = MNT_INTERNAL; + int error = -ENOMEM; if (data && !(type->fs_flags & FS_BINARY_MOUNTDATA)) { secdata = alloc_secdata(); if (!secdata) - goto out_mnt; + goto out; error = security_sb_copy_data(data, secdata); if (error) goto out_free_secdata; } - if (type->mount) { - root = type->mount(type, flags, name, data); - if (IS_ERR(root)) { - error = PTR_ERR(root); - goto out_free_secdata; - } - mnt->mnt_root = root; - mnt->mnt_sb = root->d_sb; - } else { - error = type->get_sb(type, flags, name, data, mnt); - if (error < 0) - goto out_free_secdata; + root = type->mount(type, flags, name, data); + if (IS_ERR(root)) { + error = PTR_ERR(root); + goto out_free_secdata; } - BUG_ON(!mnt->mnt_sb); - WARN_ON(!mnt->mnt_sb->s_bdi); - mnt->mnt_sb->s_flags |= MS_BORN; + sb = root->d_sb; + BUG_ON(!sb); + WARN_ON(!sb->s_bdi); + sb->s_flags |= MS_BORN; - error = security_sb_kern_mount(mnt->mnt_sb, flags, secdata); + error = security_sb_kern_mount(sb, flags, secdata); if (error) goto out_sb; @@ -1011,27 +949,21 @@ vfs_kern_mount(struct file_system_type *type, int flags, const char *name, void * violate this rule. This warning should be either removed or * converted to a BUG() in 2.6.34. */ - WARN((mnt->mnt_sb->s_maxbytes < 0), "%s set sb->s_maxbytes to " - "negative value (%lld)\n", type->name, mnt->mnt_sb->s_maxbytes); + WARN((sb->s_maxbytes < 0), "%s set sb->s_maxbytes to " + "negative value (%lld)\n", type->name, sb->s_maxbytes); - mnt->mnt_mountpoint = mnt->mnt_root; - mnt->mnt_parent = mnt; - up_write(&mnt->mnt_sb->s_umount); + up_write(&sb->s_umount); free_secdata(secdata); - return mnt; + return root; out_sb: - dput(mnt->mnt_root); - deactivate_locked_super(mnt->mnt_sb); + dput(root); + deactivate_locked_super(sb); out_free_secdata: free_secdata(secdata); -out_mnt: - free_vfsmnt(mnt); out: return ERR_PTR(error); } -EXPORT_SYMBOL_GPL(vfs_kern_mount); - /** * freeze_super - lock the filesystem and force it into a consistent state * @sb: the super to lock @@ -1121,49 +1053,3 @@ out: return 0; } EXPORT_SYMBOL(thaw_super); - -static struct vfsmount *fs_set_subtype(struct vfsmount *mnt, const char *fstype) -{ - int err; - const char *subtype = strchr(fstype, '.'); - if (subtype) { - subtype++; - err = -EINVAL; - if (!subtype[0]) - goto err; - } else - subtype = ""; - - mnt->mnt_sb->s_subtype = kstrdup(subtype, GFP_KERNEL); - err = -ENOMEM; - if (!mnt->mnt_sb->s_subtype) - goto err; - return mnt; - - err: - mntput(mnt); - return ERR_PTR(err); -} - -struct vfsmount * -do_kern_mount(const char *fstype, int flags, const char *name, void *data) -{ - struct file_system_type *type = get_fs_type(fstype); - struct vfsmount *mnt; - if (!type) - return ERR_PTR(-ENODEV); - mnt = vfs_kern_mount(type, flags, name, data); - if (!IS_ERR(mnt) && (type->fs_flags & FS_HAS_SUBTYPE) && - !mnt->mnt_sb->s_subtype) - mnt = fs_set_subtype(mnt, fstype); - put_filesystem(type); - return mnt; -} -EXPORT_SYMBOL_GPL(do_kern_mount); - -struct vfsmount *kern_mount_data(struct file_system_type *type, void *data) -{ - return vfs_kern_mount(type, MS_KERNMOUNT, type->name, data); -} - -EXPORT_SYMBOL_GPL(kern_mount_data); @@ -7,6 +7,7 @@ #include <linux/fs.h> #include <linux/slab.h> #include <linux/module.h> +#include <linux/namei.h> #include <linux/sched.h> #include <linux/writeback.h> #include <linux/syscalls.h> @@ -128,6 +129,29 @@ void emergency_sync(void) } } +/* + * sync a single super + */ +SYSCALL_DEFINE1(syncfs, int, fd) +{ + struct file *file; + struct super_block *sb; + int ret; + int fput_needed; + + file = fget_light(fd, &fput_needed); + if (!file) + return -EBADF; + sb = file->f_dentry->d_sb; + + down_read(&sb->s_umount); + ret = sync_filesystem(sb); + up_read(&sb->s_umount); + + fput_light(file, fput_needed); + return ret; +} + /** * vfs_fsync_range - helper to sync a range of data & metadata to disk * @file: file to sync diff --git a/fs/sysfs/Kconfig b/fs/sysfs/Kconfig index f4b6758..8c41fea 100644 --- a/fs/sysfs/Kconfig +++ b/fs/sysfs/Kconfig @@ -1,5 +1,5 @@ config SYSFS - bool "sysfs file system support" if EMBEDDED + bool "sysfs file system support" if EXPERT default y help The sysfs filesystem is a virtual filesystem that the kernel uses to diff --git a/fs/sysv/namei.c b/fs/sysv/namei.c index b427b12..e474fbc 100644 --- a/fs/sysv/namei.c +++ b/fs/sysv/namei.c @@ -245,7 +245,6 @@ static int sysv_rename(struct inode * old_dir, struct dentry * old_dentry, new_de = sysv_find_entry(new_dentry, &new_page); if (!new_de) goto out_dir; - inode_inc_link_count(old_inode); sysv_set_link(new_de, new_page, old_inode); new_inode->i_ctime = CURRENT_TIME_SEC; if (dir_de) @@ -257,18 +256,15 @@ static int sysv_rename(struct inode * old_dir, struct dentry * old_dentry, if (new_dir->i_nlink >= SYSV_SB(new_dir->i_sb)->s_link_max) goto out_dir; } - inode_inc_link_count(old_inode); err = sysv_add_link(new_dentry, old_inode); - if (err) { - inode_dec_link_count(old_inode); + if (err) goto out_dir; - } if (dir_de) inode_inc_link_count(new_dir); } sysv_delete_entry(old_de, old_page); - inode_dec_link_count(old_inode); + mark_inode_dirty(old_inode); if (dir_de) { sysv_set_link(dir_de, dir_page, new_dir); diff --git a/fs/ubifs/Kconfig b/fs/ubifs/Kconfig index 830e3f7..1d1859d 100644 --- a/fs/ubifs/Kconfig +++ b/fs/ubifs/Kconfig @@ -44,23 +44,20 @@ config UBIFS_FS_ZLIB # Debugging-related stuff config UBIFS_FS_DEBUG - bool "Enable debugging" + bool "Enable debugging support" depends on UBIFS_FS select DEBUG_FS select KALLSYMS_ALL help - This option enables UBIFS debugging. - -config UBIFS_FS_DEBUG_MSG_LVL - int "Default message level (0 = no extra messages, 3 = lots)" - depends on UBIFS_FS_DEBUG - default "0" - help - This controls the amount of debugging messages produced by UBIFS. - If reporting bugs, please try to have available a full dump of the - messages at level 1 while the misbehaviour was occurring. Level 2 - may become necessary if level 1 messages were not enough to find the - bug. Generally Level 3 should be avoided. + This option enables UBIFS debugging support. It makes sure various + assertions, self-checks, debugging messages and test modes are compiled + in (this all is compiled out otherwise). Assertions are light-weight + and this option also enables them. Self-checks, debugging messages and + test modes are switched off by default. Thus, it is safe and actually + recommended to have debugging support enabled, and it should not slow + down UBIFS. You can then further enable / disable individual debugging + features using UBIFS module parameters and the corresponding sysfs + interfaces. config UBIFS_FS_DEBUG_CHKS bool "Enable extra checks" diff --git a/fs/ubifs/commit.c b/fs/ubifs/commit.c index 02429d8..b148fbc 100644 --- a/fs/ubifs/commit.c +++ b/fs/ubifs/commit.c @@ -48,6 +48,56 @@ #include <linux/slab.h> #include "ubifs.h" +/* + * nothing_to_commit - check if there is nothing to commit. + * @c: UBIFS file-system description object + * + * This is a helper function which checks if there is anything to commit. It is + * used as an optimization to avoid starting the commit if it is not really + * necessary. Indeed, the commit operation always assumes flash I/O (e.g., + * writing the commit start node to the log), and it is better to avoid doing + * this unnecessarily. E.g., 'ubifs_sync_fs()' runs the commit, but if there is + * nothing to commit, it is more optimal to avoid any flash I/O. + * + * This function has to be called with @c->commit_sem locked for writing - + * this function does not take LPT/TNC locks because the @c->commit_sem + * guarantees that we have exclusive access to the TNC and LPT data structures. + * + * This function returns %1 if there is nothing to commit and %0 otherwise. + */ +static int nothing_to_commit(struct ubifs_info *c) +{ + /* + * During mounting or remounting from R/O mode to R/W mode we may + * commit for various recovery-related reasons. + */ + if (c->mounting || c->remounting_rw) + return 0; + + /* + * If the root TNC node is dirty, we definitely have something to + * commit. + */ + if (c->zroot.znode && test_bit(DIRTY_ZNODE, &c->zroot.znode->flags)) + return 0; + + /* + * Even though the TNC is clean, the LPT tree may have dirty nodes. For + * example, this may happen if the budgeting subsystem invoked GC to + * make some free space, and the GC found an LEB with only dirty and + * free space. In this case GC would just change the lprops of this + * LEB (by turning all space into free space) and unmap it. + */ + if (c->nroot && test_bit(DIRTY_CNODE, &c->nroot->flags)) + return 0; + + ubifs_assert(atomic_long_read(&c->dirty_zn_cnt) == 0); + ubifs_assert(c->dirty_pn_cnt == 0); + ubifs_assert(c->dirty_nn_cnt == 0); + + return 1; +} + /** * do_commit - commit the journal. * @c: UBIFS file-system description object @@ -70,6 +120,12 @@ static int do_commit(struct ubifs_info *c) goto out_up; } + if (nothing_to_commit(c)) { + up_write(&c->commit_sem); + err = 0; + goto out_cancel; + } + /* Sync all write buffers (necessary for recovery) */ for (i = 0; i < c->jhead_cnt; i++) { err = ubifs_wbuf_sync(&c->jheads[i].wbuf); @@ -162,12 +218,12 @@ static int do_commit(struct ubifs_info *c) if (err) goto out; +out_cancel: spin_lock(&c->cs_lock); c->cmt_state = COMMIT_RESTING; wake_up(&c->cmt_wq); dbg_cmt("commit end"); spin_unlock(&c->cs_lock); - return 0; out_up: diff --git a/fs/ubifs/debug.c b/fs/ubifs/debug.c index 0bee4db..01c2b02 100644 --- a/fs/ubifs/debug.c +++ b/fs/ubifs/debug.c @@ -43,8 +43,8 @@ DEFINE_SPINLOCK(dbg_lock); static char dbg_key_buf0[128]; static char dbg_key_buf1[128]; -unsigned int ubifs_msg_flags = UBIFS_MSG_FLAGS_DEFAULT; -unsigned int ubifs_chk_flags = UBIFS_CHK_FLAGS_DEFAULT; +unsigned int ubifs_msg_flags; +unsigned int ubifs_chk_flags; unsigned int ubifs_tst_flags; module_param_named(debug_msgs, ubifs_msg_flags, uint, S_IRUGO | S_IWUSR); @@ -810,16 +810,24 @@ void dbg_dump_leb(const struct ubifs_info *c, int lnum) { struct ubifs_scan_leb *sleb; struct ubifs_scan_node *snod; + void *buf; if (dbg_failure_mode) return; printk(KERN_DEBUG "(pid %d) start dumping LEB %d\n", current->pid, lnum); - sleb = ubifs_scan(c, lnum, 0, c->dbg->buf, 0); + + buf = __vmalloc(c->leb_size, GFP_KERNEL | GFP_NOFS, PAGE_KERNEL); + if (!buf) { + ubifs_err("cannot allocate memory for dumping LEB %d", lnum); + return; + } + + sleb = ubifs_scan(c, lnum, 0, buf, 0); if (IS_ERR(sleb)) { ubifs_err("scan error %d", (int)PTR_ERR(sleb)); - return; + goto out; } printk(KERN_DEBUG "LEB %d has %d nodes ending at %d\n", lnum, @@ -835,6 +843,9 @@ void dbg_dump_leb(const struct ubifs_info *c, int lnum) printk(KERN_DEBUG "(pid %d) finish dumping LEB %d\n", current->pid, lnum); ubifs_scan_destroy(sleb); + +out: + vfree(buf); return; } @@ -2690,16 +2701,8 @@ int ubifs_debugging_init(struct ubifs_info *c) if (!c->dbg) return -ENOMEM; - c->dbg->buf = vmalloc(c->leb_size); - if (!c->dbg->buf) - goto out; - failure_mode_init(c); return 0; - -out: - kfree(c->dbg); - return -ENOMEM; } /** @@ -2709,7 +2712,6 @@ out: void ubifs_debugging_exit(struct ubifs_info *c) { failure_mode_exit(c); - vfree(c->dbg->buf); kfree(c->dbg); } @@ -2813,19 +2815,19 @@ int dbg_debugfs_init_fs(struct ubifs_info *c) } fname = "dump_lprops"; - dent = debugfs_create_file(fname, S_IWUGO, d->dfs_dir, c, &dfs_fops); + dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, c, &dfs_fops); if (IS_ERR(dent)) goto out_remove; d->dfs_dump_lprops = dent; fname = "dump_budg"; - dent = debugfs_create_file(fname, S_IWUGO, d->dfs_dir, c, &dfs_fops); + dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, c, &dfs_fops); if (IS_ERR(dent)) goto out_remove; d->dfs_dump_budg = dent; fname = "dump_tnc"; - dent = debugfs_create_file(fname, S_IWUGO, d->dfs_dir, c, &dfs_fops); + dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, c, &dfs_fops); if (IS_ERR(dent)) goto out_remove; d->dfs_dump_tnc = dent; diff --git a/fs/ubifs/debug.h b/fs/ubifs/debug.h index 69ebe47..919f0de 100644 --- a/fs/ubifs/debug.h +++ b/fs/ubifs/debug.h @@ -27,7 +27,6 @@ /** * ubifs_debug_info - per-FS debugging information. - * @buf: a buffer of LEB size, used for various purposes * @old_zroot: old index root - used by 'dbg_check_old_index()' * @old_zroot_level: old index root level - used by 'dbg_check_old_index()' * @old_zroot_sqnum: old index root sqnum - used by 'dbg_check_old_index()' @@ -54,7 +53,6 @@ * dfs_dump_tnc: "dump TNC" debugfs knob */ struct ubifs_debug_info { - void *buf; struct ubifs_zbranch old_zroot; int old_zroot_level; unsigned long long old_zroot_sqnum; @@ -173,7 +171,7 @@ const char *dbg_key_str1(const struct ubifs_info *c, #define dbg_rcvry(fmt, ...) dbg_do_msg(UBIFS_MSG_RCVRY, fmt, ##__VA_ARGS__) /* - * Debugging message type flags (must match msg_type_names in debug.c). + * Debugging message type flags. * * UBIFS_MSG_GEN: general messages * UBIFS_MSG_JNL: journal messages @@ -205,14 +203,8 @@ enum { UBIFS_MSG_RCVRY = 0x1000, }; -/* Debugging message type flags for each default debug message level */ -#define UBIFS_MSG_LVL_0 0 -#define UBIFS_MSG_LVL_1 0x1 -#define UBIFS_MSG_LVL_2 0x7f -#define UBIFS_MSG_LVL_3 0xffff - /* - * Debugging check flags (must match chk_names in debug.c). + * Debugging check flags. * * UBIFS_CHK_GEN: general checks * UBIFS_CHK_TNC: check TNC @@ -233,7 +225,7 @@ enum { }; /* - * Special testing flags (must match tst_names in debug.c). + * Special testing flags. * * UBIFS_TST_FORCE_IN_THE_GAPS: force the use of in-the-gaps method * UBIFS_TST_RCVRY: failure mode for recovery testing @@ -243,22 +235,6 @@ enum { UBIFS_TST_RCVRY = 0x4, }; -#if CONFIG_UBIFS_FS_DEBUG_MSG_LVL == 1 -#define UBIFS_MSG_FLAGS_DEFAULT UBIFS_MSG_LVL_1 -#elif CONFIG_UBIFS_FS_DEBUG_MSG_LVL == 2 -#define UBIFS_MSG_FLAGS_DEFAULT UBIFS_MSG_LVL_2 -#elif CONFIG_UBIFS_FS_DEBUG_MSG_LVL == 3 -#define UBIFS_MSG_FLAGS_DEFAULT UBIFS_MSG_LVL_3 -#else -#define UBIFS_MSG_FLAGS_DEFAULT UBIFS_MSG_LVL_0 -#endif - -#ifdef CONFIG_UBIFS_FS_DEBUG_CHKS -#define UBIFS_CHK_FLAGS_DEFAULT 0xffffffff -#else -#define UBIFS_CHK_FLAGS_DEFAULT 0 -#endif - extern spinlock_t dbg_lock; extern unsigned int ubifs_msg_flags; diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c index 14f64b6..7217d67 100644 --- a/fs/ubifs/dir.c +++ b/fs/ubifs/dir.c @@ -522,24 +522,6 @@ static int ubifs_link(struct dentry *old_dentry, struct inode *dir, ubifs_assert(mutex_is_locked(&dir->i_mutex)); ubifs_assert(mutex_is_locked(&inode->i_mutex)); - /* - * Return -ENOENT if we've raced with unlink and i_nlink is 0. Doing - * otherwise has the potential to corrupt the orphan inode list. - * - * Indeed, consider a scenario when 'vfs_link(dirA/fileA)' and - * 'vfs_unlink(dirA/fileA, dirB/fileB)' race. 'vfs_link()' does not - * lock 'dirA->i_mutex', so this is possible. Both of the functions - * lock 'fileA->i_mutex' though. Suppose 'vfs_unlink()' wins, and takes - * 'fileA->i_mutex' mutex first. Suppose 'fileA->i_nlink' is 1. In this - * case 'ubifs_unlink()' will drop the last reference, and put 'inodeA' - * to the list of orphans. After this, 'vfs_link()' will link - * 'dirB/fileB' to 'inodeA'. This is a problem because, for example, - * the subsequent 'vfs_unlink(dirB/fileB)' will add the same inode - * to the list of orphans. - */ - if (inode->i_nlink == 0) - return -ENOENT; - err = dbg_check_synced_i_size(inode); if (err) return err; diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c index d821731..dfd168b 100644 --- a/fs/ubifs/io.c +++ b/fs/ubifs/io.c @@ -31,6 +31,26 @@ * buffer is full or when it is not used for some time (by timer). This is * similar to the mechanism is used by JFFS2. * + * UBIFS distinguishes between minimum write size (@c->min_io_size) and maximum + * write size (@c->max_write_size). The latter is the maximum amount of bytes + * the underlying flash is able to program at a time, and writing in + * @c->max_write_size units should presumably be faster. Obviously, + * @c->min_io_size <= @c->max_write_size. Write-buffers are of + * @c->max_write_size bytes in size for maximum performance. However, when a + * write-buffer is flushed, only the portion of it (aligned to @c->min_io_size + * boundary) which contains data is written, not the whole write-buffer, + * because this is more space-efficient. + * + * This optimization adds few complications to the code. Indeed, on the one + * hand, we want to write in optimal @c->max_write_size bytes chunks, which + * also means aligning writes at the @c->max_write_size bytes offsets. On the + * other hand, we do not want to waste space when synchronizing the write + * buffer, so during synchronization we writes in smaller chunks. And this makes + * the next write offset to be not aligned to @c->max_write_size bytes. So the + * have to make sure that the write-buffer offset (@wbuf->offs) becomes aligned + * to @c->max_write_size bytes again. We do this by temporarily shrinking + * write-buffer size (@wbuf->size). + * * Write-buffers are defined by 'struct ubifs_wbuf' objects and protected by * mutexes defined inside these objects. Since sometimes upper-level code * has to lock the write-buffer (e.g. journal space reservation code), many @@ -46,8 +66,8 @@ * UBIFS uses padding when it pads to the next min. I/O unit. In this case it * uses padding nodes or padding bytes, if the padding node does not fit. * - * All UBIFS nodes are protected by CRC checksums and UBIFS checks all nodes - * every time they are read from the flash media. + * All UBIFS nodes are protected by CRC checksums and UBIFS checks CRC when + * they are read from the flash media. */ #include <linux/crc32.h> @@ -88,8 +108,12 @@ void ubifs_ro_mode(struct ubifs_info *c, int err) * This function may skip data nodes CRC checking if @c->no_chk_data_crc is * true, which is controlled by corresponding UBIFS mount option. However, if * @must_chk_crc is true, then @c->no_chk_data_crc is ignored and CRC is - * checked. Similarly, if @c->always_chk_crc is true, @c->no_chk_data_crc is - * ignored and CRC is checked. + * checked. Similarly, if @c->mounting or @c->remounting_rw is true (we are + * mounting or re-mounting to R/W mode), @c->no_chk_data_crc is ignored and CRC + * is checked. This is because during mounting or re-mounting from R/O mode to + * R/W mode we may read journal nodes (when replying the journal or doing the + * recovery) and the journal nodes may potentially be corrupted, so checking is + * required. * * This function returns zero in case of success and %-EUCLEAN in case of bad * CRC or magic. @@ -131,8 +155,8 @@ int ubifs_check_node(const struct ubifs_info *c, const void *buf, int lnum, node_len > c->ranges[type].max_len) goto out_len; - if (!must_chk_crc && type == UBIFS_DATA_NODE && !c->always_chk_crc && - c->no_chk_data_crc) + if (!must_chk_crc && type == UBIFS_DATA_NODE && !c->mounting && + !c->remounting_rw && c->no_chk_data_crc) return 0; crc = crc32(UBIFS_CRC32_INIT, buf + 8, node_len - 8); @@ -343,11 +367,17 @@ static void cancel_wbuf_timer_nolock(struct ubifs_wbuf *wbuf) * * This function synchronizes write-buffer @buf and returns zero in case of * success or a negative error code in case of failure. + * + * Note, although write-buffers are of @c->max_write_size, this function does + * not necessarily writes all @c->max_write_size bytes to the flash. Instead, + * if the write-buffer is only partially filled with data, only the used part + * of the write-buffer (aligned on @c->min_io_size boundary) is synchronized. + * This way we waste less space. */ int ubifs_wbuf_sync_nolock(struct ubifs_wbuf *wbuf) { struct ubifs_info *c = wbuf->c; - int err, dirt; + int err, dirt, sync_len; cancel_wbuf_timer_nolock(wbuf); if (!wbuf->used || wbuf->lnum == -1) @@ -357,27 +387,53 @@ int ubifs_wbuf_sync_nolock(struct ubifs_wbuf *wbuf) dbg_io("LEB %d:%d, %d bytes, jhead %s", wbuf->lnum, wbuf->offs, wbuf->used, dbg_jhead(wbuf->jhead)); ubifs_assert(!(wbuf->avail & 7)); - ubifs_assert(wbuf->offs + c->min_io_size <= c->leb_size); + ubifs_assert(wbuf->offs + wbuf->size <= c->leb_size); + ubifs_assert(wbuf->size >= c->min_io_size); + ubifs_assert(wbuf->size <= c->max_write_size); + ubifs_assert(wbuf->size % c->min_io_size == 0); ubifs_assert(!c->ro_media && !c->ro_mount); + if (c->leb_size - wbuf->offs >= c->max_write_size) + ubifs_assert(!((wbuf->offs + wbuf->size) % c->max_write_size )); if (c->ro_error) return -EROFS; - ubifs_pad(c, wbuf->buf + wbuf->used, wbuf->avail); + /* + * Do not write whole write buffer but write only the minimum necessary + * amount of min. I/O units. + */ + sync_len = ALIGN(wbuf->used, c->min_io_size); + dirt = sync_len - wbuf->used; + if (dirt) + ubifs_pad(c, wbuf->buf + wbuf->used, dirt); err = ubi_leb_write(c->ubi, wbuf->lnum, wbuf->buf, wbuf->offs, - c->min_io_size, wbuf->dtype); + sync_len, wbuf->dtype); if (err) { ubifs_err("cannot write %d bytes to LEB %d:%d", - c->min_io_size, wbuf->lnum, wbuf->offs); + sync_len, wbuf->lnum, wbuf->offs); dbg_dump_stack(); return err; } - dirt = wbuf->avail; - spin_lock(&wbuf->lock); - wbuf->offs += c->min_io_size; - wbuf->avail = c->min_io_size; + wbuf->offs += sync_len; + /* + * Now @wbuf->offs is not necessarily aligned to @c->max_write_size. + * But our goal is to optimize writes and make sure we write in + * @c->max_write_size chunks and to @c->max_write_size-aligned offset. + * Thus, if @wbuf->offs is not aligned to @c->max_write_size now, make + * sure that @wbuf->offs + @wbuf->size is aligned to + * @c->max_write_size. This way we make sure that after next + * write-buffer flush we are again at the optimal offset (aligned to + * @c->max_write_size). + */ + if (c->leb_size - wbuf->offs < c->max_write_size) + wbuf->size = c->leb_size - wbuf->offs; + else if (wbuf->offs & (c->max_write_size - 1)) + wbuf->size = ALIGN(wbuf->offs, c->max_write_size) - wbuf->offs; + else + wbuf->size = c->max_write_size; + wbuf->avail = wbuf->size; wbuf->used = 0; wbuf->next_ino = 0; spin_unlock(&wbuf->lock); @@ -420,7 +476,13 @@ int ubifs_wbuf_seek_nolock(struct ubifs_wbuf *wbuf, int lnum, int offs, spin_lock(&wbuf->lock); wbuf->lnum = lnum; wbuf->offs = offs; - wbuf->avail = c->min_io_size; + if (c->leb_size - wbuf->offs < c->max_write_size) + wbuf->size = c->leb_size - wbuf->offs; + else if (wbuf->offs & (c->max_write_size - 1)) + wbuf->size = ALIGN(wbuf->offs, c->max_write_size) - wbuf->offs; + else + wbuf->size = c->max_write_size; + wbuf->avail = wbuf->size; wbuf->used = 0; spin_unlock(&wbuf->lock); wbuf->dtype = dtype; @@ -500,8 +562,9 @@ out_timers: * * This function writes data to flash via write-buffer @wbuf. This means that * the last piece of the node won't reach the flash media immediately if it - * does not take whole minimal I/O unit. Instead, the node will sit in RAM - * until the write-buffer is synchronized (e.g., by timer). + * does not take whole max. write unit (@c->max_write_size). Instead, the node + * will sit in RAM until the write-buffer is synchronized (e.g., by timer, or + * because more data are appended to the write-buffer). * * This function returns zero in case of success and a negative error code in * case of failure. If the node cannot be written because there is no more @@ -518,9 +581,14 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) ubifs_assert(len > 0 && wbuf->lnum >= 0 && wbuf->lnum < c->leb_cnt); ubifs_assert(wbuf->offs >= 0 && wbuf->offs % c->min_io_size == 0); ubifs_assert(!(wbuf->offs & 7) && wbuf->offs <= c->leb_size); - ubifs_assert(wbuf->avail > 0 && wbuf->avail <= c->min_io_size); + ubifs_assert(wbuf->avail > 0 && wbuf->avail <= wbuf->size); + ubifs_assert(wbuf->size >= c->min_io_size); + ubifs_assert(wbuf->size <= c->max_write_size); + ubifs_assert(wbuf->size % c->min_io_size == 0); ubifs_assert(mutex_is_locked(&wbuf->io_mutex)); ubifs_assert(!c->ro_media && !c->ro_mount); + if (c->leb_size - wbuf->offs >= c->max_write_size) + ubifs_assert(!((wbuf->offs + wbuf->size) % c->max_write_size )); if (c->leb_size - wbuf->offs - wbuf->used < aligned_len) { err = -ENOSPC; @@ -543,14 +611,18 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) dbg_io("flush jhead %s wbuf to LEB %d:%d", dbg_jhead(wbuf->jhead), wbuf->lnum, wbuf->offs); err = ubi_leb_write(c->ubi, wbuf->lnum, wbuf->buf, - wbuf->offs, c->min_io_size, + wbuf->offs, wbuf->size, wbuf->dtype); if (err) goto out; spin_lock(&wbuf->lock); - wbuf->offs += c->min_io_size; - wbuf->avail = c->min_io_size; + wbuf->offs += wbuf->size; + if (c->leb_size - wbuf->offs >= c->max_write_size) + wbuf->size = c->max_write_size; + else + wbuf->size = c->leb_size - wbuf->offs; + wbuf->avail = wbuf->size; wbuf->used = 0; wbuf->next_ino = 0; spin_unlock(&wbuf->lock); @@ -564,33 +636,57 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) goto exit; } - /* - * The node is large enough and does not fit entirely within current - * minimal I/O unit. We have to fill and flush write-buffer and switch - * to the next min. I/O unit. - */ - dbg_io("flush jhead %s wbuf to LEB %d:%d", - dbg_jhead(wbuf->jhead), wbuf->lnum, wbuf->offs); - memcpy(wbuf->buf + wbuf->used, buf, wbuf->avail); - err = ubi_leb_write(c->ubi, wbuf->lnum, wbuf->buf, wbuf->offs, - c->min_io_size, wbuf->dtype); - if (err) - goto out; + offs = wbuf->offs; + written = 0; - offs = wbuf->offs + c->min_io_size; - len -= wbuf->avail; - aligned_len -= wbuf->avail; - written = wbuf->avail; + if (wbuf->used) { + /* + * The node is large enough and does not fit entirely within + * current available space. We have to fill and flush + * write-buffer and switch to the next max. write unit. + */ + dbg_io("flush jhead %s wbuf to LEB %d:%d", + dbg_jhead(wbuf->jhead), wbuf->lnum, wbuf->offs); + memcpy(wbuf->buf + wbuf->used, buf, wbuf->avail); + err = ubi_leb_write(c->ubi, wbuf->lnum, wbuf->buf, wbuf->offs, + wbuf->size, wbuf->dtype); + if (err) + goto out; + + offs += wbuf->size; + len -= wbuf->avail; + aligned_len -= wbuf->avail; + written += wbuf->avail; + } else if (wbuf->offs & (c->max_write_size - 1)) { + /* + * The write-buffer offset is not aligned to + * @c->max_write_size and @wbuf->size is less than + * @c->max_write_size. Write @wbuf->size bytes to make sure the + * following writes are done in optimal @c->max_write_size + * chunks. + */ + dbg_io("write %d bytes to LEB %d:%d", + wbuf->size, wbuf->lnum, wbuf->offs); + err = ubi_leb_write(c->ubi, wbuf->lnum, buf, wbuf->offs, + wbuf->size, wbuf->dtype); + if (err) + goto out; + + offs += wbuf->size; + len -= wbuf->size; + aligned_len -= wbuf->size; + written += wbuf->size; + } /* - * The remaining data may take more whole min. I/O units, so write the - * remains multiple to min. I/O unit size directly to the flash media. + * The remaining data may take more whole max. write units, so write the + * remains multiple to max. write unit size directly to the flash media. * We align node length to 8-byte boundary because we anyway flash wbuf * if the remaining space is less than 8 bytes. */ - n = aligned_len >> c->min_io_shift; + n = aligned_len >> c->max_write_shift; if (n) { - n <<= c->min_io_shift; + n <<= c->max_write_shift; dbg_io("write %d bytes to LEB %d:%d", n, wbuf->lnum, offs); err = ubi_leb_write(c->ubi, wbuf->lnum, buf + written, offs, n, wbuf->dtype); @@ -606,14 +702,18 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) if (aligned_len) /* * And now we have what's left and what does not take whole - * min. I/O unit, so write it to the write-buffer and we are + * max. write unit, so write it to the write-buffer and we are * done. */ memcpy(wbuf->buf, buf + written, len); wbuf->offs = offs; + if (c->leb_size - wbuf->offs >= c->max_write_size) + wbuf->size = c->max_write_size; + else + wbuf->size = c->leb_size - wbuf->offs; + wbuf->avail = wbuf->size - aligned_len; wbuf->used = aligned_len; - wbuf->avail = c->min_io_size - aligned_len; wbuf->next_ino = 0; spin_unlock(&wbuf->lock); @@ -837,11 +937,11 @@ int ubifs_wbuf_init(struct ubifs_info *c, struct ubifs_wbuf *wbuf) { size_t size; - wbuf->buf = kmalloc(c->min_io_size, GFP_KERNEL); + wbuf->buf = kmalloc(c->max_write_size, GFP_KERNEL); if (!wbuf->buf) return -ENOMEM; - size = (c->min_io_size / UBIFS_CH_SZ + 1) * sizeof(ino_t); + size = (c->max_write_size / UBIFS_CH_SZ + 1) * sizeof(ino_t); wbuf->inodes = kmalloc(size, GFP_KERNEL); if (!wbuf->inodes) { kfree(wbuf->buf); @@ -851,7 +951,14 @@ int ubifs_wbuf_init(struct ubifs_info *c, struct ubifs_wbuf *wbuf) wbuf->used = 0; wbuf->lnum = wbuf->offs = -1; - wbuf->avail = c->min_io_size; + /* + * If the LEB starts at the max. write size aligned address, then + * write-buffer size has to be set to @c->max_write_size. Otherwise, + * set it to something smaller so that it ends at the closest max. + * write size boundary. + */ + size = c->max_write_size - (c->leb_start % c->max_write_size); + wbuf->avail = wbuf->size = size; wbuf->dtype = UBI_UNKNOWN; wbuf->sync_callback = NULL; mutex_init(&wbuf->io_mutex); diff --git a/fs/ubifs/journal.c b/fs/ubifs/journal.c index 914f1bd..aed25e8 100644 --- a/fs/ubifs/journal.c +++ b/fs/ubifs/journal.c @@ -690,7 +690,7 @@ int ubifs_jnl_write_data(struct ubifs_info *c, const struct inode *inode, { struct ubifs_data_node *data; int err, lnum, offs, compr_type, out_len; - int dlen = UBIFS_DATA_NODE_SZ + UBIFS_BLOCK_SIZE * WORST_COMPR_FACTOR; + int dlen = COMPRESSED_DATA_NODE_BUF_SZ, allocated = 1; struct ubifs_inode *ui = ubifs_inode(inode); dbg_jnl("ino %lu, blk %u, len %d, key %s", @@ -698,9 +698,19 @@ int ubifs_jnl_write_data(struct ubifs_info *c, const struct inode *inode, DBGKEY(key)); ubifs_assert(len <= UBIFS_BLOCK_SIZE); - data = kmalloc(dlen, GFP_NOFS); - if (!data) - return -ENOMEM; + data = kmalloc(dlen, GFP_NOFS | __GFP_NOWARN); + if (!data) { + /* + * Fall-back to the write reserve buffer. Note, we might be + * currently on the memory reclaim path, when the kernel is + * trying to free some memory by writing out dirty pages. The + * write reserve buffer helps us to guarantee that we are + * always able to write the data. + */ + allocated = 0; + mutex_lock(&c->write_reserve_mutex); + data = c->write_reserve_buf; + } data->ch.node_type = UBIFS_DATA_NODE; key_write(c, key, &data->key); @@ -736,7 +746,10 @@ int ubifs_jnl_write_data(struct ubifs_info *c, const struct inode *inode, goto out_ro; finish_reservation(c); - kfree(data); + if (!allocated) + mutex_unlock(&c->write_reserve_mutex); + else + kfree(data); return 0; out_release: @@ -745,7 +758,10 @@ out_ro: ubifs_ro_mode(c, err); finish_reservation(c); out_free: - kfree(data); + if (!allocated) + mutex_unlock(&c->write_reserve_mutex); + else + kfree(data); return err; } diff --git a/fs/ubifs/lprops.c b/fs/ubifs/lprops.c index 4d4ca38..c7b25e2 100644 --- a/fs/ubifs/lprops.c +++ b/fs/ubifs/lprops.c @@ -1035,7 +1035,8 @@ static int scan_check_cb(struct ubifs_info *c, struct ubifs_scan_leb *sleb; struct ubifs_scan_node *snod; struct ubifs_lp_stats *lst = &data->lst; - int cat, lnum = lp->lnum, is_idx = 0, used = 0, free, dirty; + int cat, lnum = lp->lnum, is_idx = 0, used = 0, free, dirty, ret; + void *buf = NULL; cat = lp->flags & LPROPS_CAT_MASK; if (cat != LPROPS_UNCAT) { @@ -1093,7 +1094,13 @@ static int scan_check_cb(struct ubifs_info *c, } } - sleb = ubifs_scan(c, lnum, 0, c->dbg->buf, 0); + buf = __vmalloc(c->leb_size, GFP_KERNEL | GFP_NOFS, PAGE_KERNEL); + if (!buf) { + ubifs_err("cannot allocate memory to scan LEB %d", lnum); + goto out; + } + + sleb = ubifs_scan(c, lnum, 0, buf, 0); if (IS_ERR(sleb)) { /* * After an unclean unmount, empty and freeable LEBs @@ -1105,7 +1112,8 @@ static int scan_check_cb(struct ubifs_info *c, lst->empty_lebs += 1; lst->total_free += c->leb_size; lst->total_dark += ubifs_calc_dark(c, c->leb_size); - return LPT_SCAN_CONTINUE; + ret = LPT_SCAN_CONTINUE; + goto exit; } if (lp->free + lp->dirty == c->leb_size && @@ -1115,10 +1123,12 @@ static int scan_check_cb(struct ubifs_info *c, lst->total_free += lp->free; lst->total_dirty += lp->dirty; lst->total_dark += ubifs_calc_dark(c, c->leb_size); - return LPT_SCAN_CONTINUE; + ret = LPT_SCAN_CONTINUE; + goto exit; } data->err = PTR_ERR(sleb); - return LPT_SCAN_STOP; + ret = LPT_SCAN_STOP; + goto exit; } is_idx = -1; @@ -1236,7 +1246,10 @@ static int scan_check_cb(struct ubifs_info *c, } ubifs_scan_destroy(sleb); - return LPT_SCAN_CONTINUE; + ret = LPT_SCAN_CONTINUE; +exit: + vfree(buf); + return ret; out_print: ubifs_err("bad accounting of LEB %d: free %d, dirty %d flags %#x, " @@ -1246,6 +1259,7 @@ out_print: out_destroy: ubifs_scan_destroy(sleb); out: + vfree(buf); data->err = -EINVAL; return LPT_SCAN_STOP; } diff --git a/fs/ubifs/lpt_commit.c b/fs/ubifs/lpt_commit.c index 5c90dec..0a3c2c3 100644 --- a/fs/ubifs/lpt_commit.c +++ b/fs/ubifs/lpt_commit.c @@ -1628,29 +1628,35 @@ static int dbg_check_ltab_lnum(struct ubifs_info *c, int lnum) { int err, len = c->leb_size, dirty = 0, node_type, node_num, node_len; int ret; - void *buf = c->dbg->buf; + void *buf, *p; if (!(ubifs_chk_flags & UBIFS_CHK_LPROPS)) return 0; + buf = p = __vmalloc(c->leb_size, GFP_KERNEL | GFP_NOFS, PAGE_KERNEL); + if (!buf) { + ubifs_err("cannot allocate memory for ltab checking"); + return 0; + } + dbg_lp("LEB %d", lnum); err = ubi_read(c->ubi, lnum, buf, 0, c->leb_size); if (err) { dbg_msg("ubi_read failed, LEB %d, error %d", lnum, err); - return err; + goto out; } while (1) { - if (!is_a_node(c, buf, len)) { + if (!is_a_node(c, p, len)) { int i, pad_len; - pad_len = get_pad_len(c, buf, len); + pad_len = get_pad_len(c, p, len); if (pad_len) { - buf += pad_len; + p += pad_len; len -= pad_len; dirty += pad_len; continue; } - if (!dbg_is_all_ff(buf, len)) { + if (!dbg_is_all_ff(p, len)) { dbg_msg("invalid empty space in LEB %d at %d", lnum, c->leb_size - len); err = -EINVAL; @@ -1668,16 +1674,21 @@ static int dbg_check_ltab_lnum(struct ubifs_info *c, int lnum) lnum, dirty, c->ltab[i].dirty); err = -EINVAL; } - return err; + goto out; } - node_type = get_lpt_node_type(c, buf, &node_num); + node_type = get_lpt_node_type(c, p, &node_num); node_len = get_lpt_node_len(c, node_type); ret = dbg_is_node_dirty(c, node_type, lnum, c->leb_size - len); if (ret == 1) dirty += node_len; - buf += node_len; + p += node_len; len -= node_len; } + + err = 0; +out: + vfree(buf); + return err; } /** @@ -1870,25 +1881,31 @@ int dbg_chk_lpt_sz(struct ubifs_info *c, int action, int len) static void dump_lpt_leb(const struct ubifs_info *c, int lnum) { int err, len = c->leb_size, node_type, node_num, node_len, offs; - void *buf = c->dbg->buf; + void *buf, *p; printk(KERN_DEBUG "(pid %d) start dumping LEB %d\n", current->pid, lnum); + buf = p = __vmalloc(c->leb_size, GFP_KERNEL | GFP_NOFS, PAGE_KERNEL); + if (!buf) { + ubifs_err("cannot allocate memory to dump LPT"); + return; + } + err = ubi_read(c->ubi, lnum, buf, 0, c->leb_size); if (err) { ubifs_err("cannot read LEB %d, error %d", lnum, err); - return; + goto out; } while (1) { offs = c->leb_size - len; - if (!is_a_node(c, buf, len)) { + if (!is_a_node(c, p, len)) { int pad_len; - pad_len = get_pad_len(c, buf, len); + pad_len = get_pad_len(c, p, len); if (pad_len) { printk(KERN_DEBUG "LEB %d:%d, pad %d bytes\n", lnum, offs, pad_len); - buf += pad_len; + p += pad_len; len -= pad_len; continue; } @@ -1898,7 +1915,7 @@ static void dump_lpt_leb(const struct ubifs_info *c, int lnum) break; } - node_type = get_lpt_node_type(c, buf, &node_num); + node_type = get_lpt_node_type(c, p, &node_num); switch (node_type) { case UBIFS_LPT_PNODE: { @@ -1923,7 +1940,7 @@ static void dump_lpt_leb(const struct ubifs_info *c, int lnum) else printk(KERN_DEBUG "LEB %d:%d, nnode, ", lnum, offs); - err = ubifs_unpack_nnode(c, buf, &nnode); + err = ubifs_unpack_nnode(c, p, &nnode); for (i = 0; i < UBIFS_LPT_FANOUT; i++) { printk(KERN_CONT "%d:%d", nnode.nbranch[i].lnum, nnode.nbranch[i].offs); @@ -1944,15 +1961,18 @@ static void dump_lpt_leb(const struct ubifs_info *c, int lnum) break; default: ubifs_err("LPT node type %d not recognized", node_type); - return; + goto out; } - buf += node_len; + p += node_len; len -= node_len; } printk(KERN_DEBUG "(pid %d) finish dumping LEB %d\n", current->pid, lnum); +out: + vfree(buf); + return; } /** diff --git a/fs/ubifs/orphan.c b/fs/ubifs/orphan.c index 82009c7..2cdbd31 100644 --- a/fs/ubifs/orphan.c +++ b/fs/ubifs/orphan.c @@ -892,15 +892,22 @@ static int dbg_read_orphans(struct check_info *ci, struct ubifs_scan_leb *sleb) static int dbg_scan_orphans(struct ubifs_info *c, struct check_info *ci) { int lnum, err = 0; + void *buf; /* Check no-orphans flag and skip this if no orphans */ if (c->no_orphs) return 0; + buf = __vmalloc(c->leb_size, GFP_KERNEL | GFP_NOFS, PAGE_KERNEL); + if (!buf) { + ubifs_err("cannot allocate memory to check orphans"); + return 0; + } + for (lnum = c->orph_first; lnum <= c->orph_last; lnum++) { struct ubifs_scan_leb *sleb; - sleb = ubifs_scan(c, lnum, 0, c->dbg->buf, 0); + sleb = ubifs_scan(c, lnum, 0, buf, 0); if (IS_ERR(sleb)) { err = PTR_ERR(sleb); break; @@ -912,6 +919,7 @@ static int dbg_scan_orphans(struct ubifs_info *c, struct check_info *ci) break; } + vfree(buf); return err; } diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c index 77e9b87..936f2cb 100644 --- a/fs/ubifs/recovery.c +++ b/fs/ubifs/recovery.c @@ -28,6 +28,23 @@ * UBIFS always cleans away all remnants of an unclean un-mount, so that * errors do not accumulate. However UBIFS defers recovery if it is mounted * read-only, and the flash is not modified in that case. + * + * The general UBIFS approach to the recovery is that it recovers from + * corruptions which could be caused by power cuts, but it refuses to recover + * from corruption caused by other reasons. And UBIFS tries to distinguish + * between these 2 reasons of corruptions and silently recover in the former + * case and loudly complain in the latter case. + * + * UBIFS writes only to erased LEBs, so it writes only to the flash space + * containing only 0xFFs. UBIFS also always writes strictly from the beginning + * of the LEB to the end. And UBIFS assumes that the underlying flash media + * writes in @c->max_write_size bytes at a time. + * + * Hence, if UBIFS finds a corrupted node at offset X, it expects only the min. + * I/O unit corresponding to offset X to contain corrupted data, all the + * following min. I/O units have to contain empty space (all 0xFFs). If this is + * not true, the corruption cannot be the result of a power cut, and UBIFS + * refuses to mount. */ #include <linux/crc32.h> @@ -362,8 +379,9 @@ int ubifs_write_rcvrd_mst_node(struct ubifs_info *c) * @offs: offset to check * * This function returns %1 if @offs was in the last write to the LEB whose data - * is in @buf, otherwise %0 is returned. The determination is made by checking - * for subsequent empty space starting from the next @c->min_io_size boundary. + * is in @buf, otherwise %0 is returned. The determination is made by checking + * for subsequent empty space starting from the next @c->max_write_size + * boundary. */ static int is_last_write(const struct ubifs_info *c, void *buf, int offs) { @@ -371,10 +389,10 @@ static int is_last_write(const struct ubifs_info *c, void *buf, int offs) uint8_t *p; /* - * Round up to the next @c->min_io_size boundary i.e. @offs is in the - * last wbuf written. After that should be empty space. + * Round up to the next @c->max_write_size boundary i.e. @offs is in + * the last wbuf written. After that should be empty space. */ - empty_offs = ALIGN(offs + 1, c->min_io_size); + empty_offs = ALIGN(offs + 1, c->max_write_size); check_len = c->leb_size - empty_offs; p = buf + empty_offs - offs; return is_empty(p, check_len); @@ -429,7 +447,7 @@ static int no_more_nodes(const struct ubifs_info *c, void *buf, int len, int skip, dlen = le32_to_cpu(ch->len); /* Check for empty space after the corrupt node's common header */ - skip = ALIGN(offs + UBIFS_CH_SZ, c->min_io_size) - offs; + skip = ALIGN(offs + UBIFS_CH_SZ, c->max_write_size) - offs; if (is_empty(buf + skip, len - skip)) return 1; /* @@ -441,7 +459,7 @@ static int no_more_nodes(const struct ubifs_info *c, void *buf, int len, return 0; } /* Now we know the corrupt node's length we can skip over it */ - skip = ALIGN(offs + dlen, c->min_io_size) - offs; + skip = ALIGN(offs + dlen, c->max_write_size) - offs; /* After which there should be empty space */ if (is_empty(buf + skip, len - skip)) return 1; @@ -671,10 +689,14 @@ struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum, } else { int corruption = first_non_ff(buf, len); + /* + * See header comment for this file for more + * explanations about the reasons we have this check. + */ ubifs_err("corrupt empty space LEB %d:%d, corruption " "starts at %d", lnum, offs, corruption); /* Make sure we dump interesting non-0xFF data */ - offs = corruption; + offs += corruption; buf += corruption; goto corrupted; } @@ -836,12 +858,8 @@ struct ubifs_scan_leb *ubifs_recover_log_leb(struct ubifs_info *c, int lnum, static int recover_head(const struct ubifs_info *c, int lnum, int offs, void *sbuf) { - int len, err; + int len = c->max_write_size, err; - if (c->min_io_size > 1) - len = c->min_io_size; - else - len = 512; if (offs + len > c->leb_size) len = c->leb_size - offs; diff --git a/fs/ubifs/scan.c b/fs/ubifs/scan.c index 3e1ee57..36216b4 100644 --- a/fs/ubifs/scan.c +++ b/fs/ubifs/scan.c @@ -328,7 +328,7 @@ struct ubifs_scan_leb *ubifs_scan(const struct ubifs_info *c, int lnum, if (!quiet) ubifs_err("empty space starts at non-aligned offset %d", offs); - goto corrupted;; + goto corrupted; } ubifs_end_scan(c, sleb, lnum, offs); diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 6e11c29..e5dc1e1 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -512,9 +512,12 @@ static int init_constants_early(struct ubifs_info *c) c->leb_cnt = c->vi.size; c->leb_size = c->vi.usable_leb_size; + c->leb_start = c->di.leb_start; c->half_leb_size = c->leb_size / 2; c->min_io_size = c->di.min_io_size; c->min_io_shift = fls(c->min_io_size) - 1; + c->max_write_size = c->di.max_write_size; + c->max_write_shift = fls(c->max_write_size) - 1; if (c->leb_size < UBIFS_MIN_LEB_SZ) { ubifs_err("too small LEBs (%d bytes), min. is %d bytes", @@ -534,6 +537,18 @@ static int init_constants_early(struct ubifs_info *c) } /* + * Maximum write size has to be greater or equivalent to min. I/O + * size, and be multiple of min. I/O size. + */ + if (c->max_write_size < c->min_io_size || + c->max_write_size % c->min_io_size || + !is_power_of_2(c->max_write_size)) { + ubifs_err("bad write buffer size %d for %d min. I/O unit", + c->max_write_size, c->min_io_size); + return -EINVAL; + } + + /* * UBIFS aligns all node to 8-byte boundary, so to make function in * io.c simpler, assume minimum I/O unit size to be 8 bytes if it is * less than 8. @@ -541,6 +556,10 @@ static int init_constants_early(struct ubifs_info *c) if (c->min_io_size < 8) { c->min_io_size = 8; c->min_io_shift = 3; + if (c->max_write_size < c->min_io_size) { + c->max_write_size = c->min_io_size; + c->max_write_shift = c->min_io_shift; + } } c->ref_node_alsz = ALIGN(UBIFS_REF_NODE_SZ, c->min_io_size); @@ -1202,11 +1221,14 @@ static int mount_ubifs(struct ubifs_info *c) if (c->bulk_read == 1) bu_init(c); - /* - * We have to check all CRCs, even for data nodes, when we mount the FS - * (specifically, when we are replaying). - */ - c->always_chk_crc = 1; + if (!c->ro_mount) { + c->write_reserve_buf = kmalloc(COMPRESSED_DATA_NODE_BUF_SZ, + GFP_KERNEL); + if (!c->write_reserve_buf) + goto out_free; + } + + c->mounting = 1; err = ubifs_read_superblock(c); if (err) @@ -1382,7 +1404,7 @@ static int mount_ubifs(struct ubifs_info *c) if (err) goto out_infos; - c->always_chk_crc = 0; + c->mounting = 0; ubifs_msg("mounted UBI device %d, volume %d, name \"%s\"", c->vi.ubi_num, c->vi.vol_id, c->vi.name); @@ -1403,6 +1425,7 @@ static int mount_ubifs(struct ubifs_info *c) dbg_msg("compiled on: " __DATE__ " at " __TIME__); dbg_msg("min. I/O unit size: %d bytes", c->min_io_size); + dbg_msg("max. write size: %d bytes", c->max_write_size); dbg_msg("LEB size: %d bytes (%d KiB)", c->leb_size, c->leb_size >> 10); dbg_msg("data journal heads: %d", @@ -1432,9 +1455,9 @@ static int mount_ubifs(struct ubifs_info *c) UBIFS_TRUN_NODE_SZ, UBIFS_SB_NODE_SZ, UBIFS_MST_NODE_SZ); dbg_msg("node sizes: ref %zu, cmt. start %zu, orph %zu", UBIFS_REF_NODE_SZ, UBIFS_CS_NODE_SZ, UBIFS_ORPH_NODE_SZ); - dbg_msg("max. node sizes: data %zu, inode %zu dentry %zu", + dbg_msg("max. node sizes: data %zu, inode %zu dentry %zu, idx %d", UBIFS_MAX_DATA_NODE_SZ, UBIFS_MAX_INO_NODE_SZ, - UBIFS_MAX_DENT_NODE_SZ); + UBIFS_MAX_DENT_NODE_SZ, ubifs_idx_node_sz(c, c->fanout)); dbg_msg("dead watermark: %d", c->dead_wm); dbg_msg("dark watermark: %d", c->dark_wm); dbg_msg("LEB overhead: %d", c->leb_overhead); @@ -1474,6 +1497,7 @@ out_wbufs: out_cbuf: kfree(c->cbuf); out_free: + kfree(c->write_reserve_buf); kfree(c->bu.buf); vfree(c->ileb_buf); vfree(c->sbuf); @@ -1512,6 +1536,7 @@ static void ubifs_umount(struct ubifs_info *c) kfree(c->cbuf); kfree(c->rcvrd_mst_node); kfree(c->mst_node); + kfree(c->write_reserve_buf); kfree(c->bu.buf); vfree(c->ileb_buf); vfree(c->sbuf); @@ -1543,7 +1568,6 @@ static int ubifs_remount_rw(struct ubifs_info *c) mutex_lock(&c->umount_mutex); dbg_save_space_info(c); c->remounting_rw = 1; - c->always_chk_crc = 1; err = check_free_space(c); if (err) @@ -1598,6 +1622,10 @@ static int ubifs_remount_rw(struct ubifs_info *c) goto out; } + c->write_reserve_buf = kmalloc(COMPRESSED_DATA_NODE_BUF_SZ, GFP_KERNEL); + if (!c->write_reserve_buf) + goto out; + err = ubifs_lpt_init(c, 0, 1); if (err) goto out; @@ -1650,7 +1678,6 @@ static int ubifs_remount_rw(struct ubifs_info *c) dbg_gen("re-mounted read-write"); c->ro_mount = 0; c->remounting_rw = 0; - c->always_chk_crc = 0; err = dbg_check_space_info(c); mutex_unlock(&c->umount_mutex); return err; @@ -1663,11 +1690,12 @@ out: c->bgt = NULL; } free_wbufs(c); + kfree(c->write_reserve_buf); + c->write_reserve_buf = NULL; vfree(c->ileb_buf); c->ileb_buf = NULL; ubifs_lpt_free(c, 1); c->remounting_rw = 0; - c->always_chk_crc = 0; mutex_unlock(&c->umount_mutex); return err; } @@ -1707,6 +1735,8 @@ static void ubifs_remount_ro(struct ubifs_info *c) free_wbufs(c); vfree(c->orph_buf); c->orph_buf = NULL; + kfree(c->write_reserve_buf); + c->write_reserve_buf = NULL; vfree(c->ileb_buf); c->ileb_buf = NULL; ubifs_lpt_free(c, 1); @@ -1937,6 +1967,7 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent) mutex_init(&c->mst_mutex); mutex_init(&c->umount_mutex); mutex_init(&c->bu_mutex); + mutex_init(&c->write_reserve_mutex); init_waitqueue_head(&c->cmt_wq); c->buds = RB_ROOT; c->old_idx = RB_ROOT; @@ -1954,6 +1985,7 @@ static int ubifs_fill_super(struct super_block *sb, void *data, int silent) INIT_LIST_HEAD(&c->old_buds); INIT_LIST_HEAD(&c->orph_list); INIT_LIST_HEAD(&c->orph_new); + c->no_chk_data_crc = 1; c->vfs_sb = sb; c->highest_inum = UBIFS_FIRST_INO; diff --git a/fs/ubifs/tnc.c b/fs/ubifs/tnc.c index ad9cf01..de48597 100644 --- a/fs/ubifs/tnc.c +++ b/fs/ubifs/tnc.c @@ -447,8 +447,11 @@ static int tnc_read_node_nm(struct ubifs_info *c, struct ubifs_zbranch *zbr, * * Note, this function does not check CRC of data nodes if @c->no_chk_data_crc * is true (it is controlled by corresponding mount option). However, if - * @c->always_chk_crc is true, @c->no_chk_data_crc is ignored and CRC is always - * checked. + * @c->mounting or @c->remounting_rw is true (we are mounting or re-mounting to + * R/W mode), @c->no_chk_data_crc is ignored and CRC is checked. This is + * because during mounting or re-mounting from R/O mode to R/W mode we may read + * journal nodes (when replying the journal or doing the recovery) and the + * journal nodes may potentially be corrupted, so checking is required. */ static int try_read_node(const struct ubifs_info *c, void *buf, int type, int len, int lnum, int offs) @@ -476,7 +479,8 @@ static int try_read_node(const struct ubifs_info *c, void *buf, int type, if (node_len != len) return 0; - if (type == UBIFS_DATA_NODE && !c->always_chk_crc && c->no_chk_data_crc) + if (type == UBIFS_DATA_NODE && c->no_chk_data_crc && !c->mounting && + !c->remounting_rw) return 1; crc = crc32(UBIFS_CRC32_INIT, buf + 8, node_len - 8); diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index 381d6b2..8c40ad3 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -151,6 +151,12 @@ */ #define WORST_COMPR_FACTOR 2 +/* + * How much memory is needed for a buffer where we comress a data node. + */ +#define COMPRESSED_DATA_NODE_BUF_SZ \ + (UBIFS_DATA_NODE_SZ + UBIFS_BLOCK_SIZE * WORST_COMPR_FACTOR) + /* Maximum expected tree height for use by bottom_up_buf */ #define BOTTOM_UP_HEIGHT 64 @@ -646,6 +652,7 @@ typedef int (*ubifs_lpt_scan_callback)(struct ubifs_info *c, * @offs: write-buffer offset in this logical eraseblock * @avail: number of bytes available in the write-buffer * @used: number of used bytes in the write-buffer + * @size: write-buffer size (in [@c->min_io_size, @c->max_write_size] range) * @dtype: type of data stored in this LEB (%UBI_LONGTERM, %UBI_SHORTTERM, * %UBI_UNKNOWN) * @jhead: journal head the mutex belongs to (note, needed only to shut lockdep @@ -680,6 +687,7 @@ struct ubifs_wbuf { int offs; int avail; int used; + int size; int dtype; int jhead; int (*sync_callback)(struct ubifs_info *c, int lnum, int free, int pad); @@ -1003,6 +1011,11 @@ struct ubifs_debug_info; * @bu_mutex: protects the pre-allocated bulk-read buffer and @c->bu * @bu: pre-allocated bulk-read information * + * @write_reserve_mutex: protects @write_reserve_buf + * @write_reserve_buf: on the write path we allocate memory, which might + * sometimes be unavailable, in which case we use this + * write reserve buffer + * * @log_lebs: number of logical eraseblocks in the log * @log_bytes: log size in bytes * @log_last: last LEB of the log @@ -1024,7 +1037,12 @@ struct ubifs_debug_info; * * @min_io_size: minimal input/output unit size * @min_io_shift: number of bits in @min_io_size minus one + * @max_write_size: maximum amount of bytes the underlying flash can write at a + * time (MTD write buffer size) + * @max_write_shift: number of bits in @max_write_size minus one * @leb_size: logical eraseblock size in bytes + * @leb_start: starting offset of logical eraseblocks within physical + * eraseblocks * @half_leb_size: half LEB size * @idx_leb_size: how many bytes of an LEB are effectively available when it is * used to store indexing nodes (@leb_size - @max_idx_node_sz) @@ -1166,22 +1184,21 @@ struct ubifs_debug_info; * @rp_uid: reserved pool user ID * @rp_gid: reserved pool group ID * - * @empty: if the UBI device is empty + * @empty: %1 if the UBI device is empty + * @need_recovery: %1 if the file-system needs recovery + * @replaying: %1 during journal replay + * @mounting: %1 while mounting + * @remounting_rw: %1 while re-mounting from R/O mode to R/W mode * @replay_tree: temporary tree used during journal replay * @replay_list: temporary list used during journal replay * @replay_buds: list of buds to replay * @cs_sqnum: sequence number of first node in the log (commit start node) * @replay_sqnum: sequence number of node currently being replayed - * @need_recovery: file-system needs recovery - * @replaying: set to %1 during journal replay * @unclean_leb_list: LEBs to recover when re-mounting R/O mounted FS to R/W * mode * @rcvrd_mst_node: recovered master node to write when re-mounting R/O mounted * FS to R/W mode * @size_tree: inode size information for recovery - * @remounting_rw: set while re-mounting from R/O mode to R/W mode - * @always_chk_crc: always check CRCs (while mounting and remounting to R/W - * mode) * @mount_opts: UBIFS-specific mount options * * @dbg: debugging-related information @@ -1250,6 +1267,9 @@ struct ubifs_info { struct mutex bu_mutex; struct bu_info bu; + struct mutex write_reserve_mutex; + void *write_reserve_buf; + int log_lebs; long long log_bytes; int log_last; @@ -1271,7 +1291,10 @@ struct ubifs_info { int min_io_size; int min_io_shift; + int max_write_size; + int max_write_shift; int leb_size; + int leb_start; int half_leb_size; int idx_leb_size; int leb_cnt; @@ -1402,19 +1425,19 @@ struct ubifs_info { gid_t rp_gid; /* The below fields are used only during mounting and re-mounting */ - int empty; + unsigned int empty:1; + unsigned int need_recovery:1; + unsigned int replaying:1; + unsigned int mounting:1; + unsigned int remounting_rw:1; struct rb_root replay_tree; struct list_head replay_list; struct list_head replay_buds; unsigned long long cs_sqnum; unsigned long long replay_sqnum; - int need_recovery; - int replaying; struct list_head unclean_leb_list; struct ubifs_mst_node *rcvrd_mst_node; struct rb_root size_tree; - int remounting_rw; - int always_chk_crc; struct ubifs_mount_opts mount_opts; #ifdef CONFIG_UBIFS_FS_DEBUG diff --git a/fs/udf/balloc.c b/fs/udf/balloc.c index 306ee39..8994dd0 100644 --- a/fs/udf/balloc.c +++ b/fs/udf/balloc.c @@ -31,7 +31,7 @@ #define udf_set_bit(nr, addr) ext2_set_bit(nr, addr) #define udf_test_bit(nr, addr) ext2_test_bit(nr, addr) #define udf_find_next_one_bit(addr, size, offset) \ - ext2_find_next_bit(addr, size, offset) + ext2_find_next_bit((unsigned long *)(addr), size, offset) static int read_block_bitmap(struct super_block *sb, struct udf_bitmap *bitmap, unsigned int block, @@ -297,7 +297,7 @@ repeat: break; } } else { - bit = udf_find_next_one_bit((char *)bh->b_data, + bit = udf_find_next_one_bit(bh->b_data, sb->s_blocksize << 3, group_start << 3); if (bit < sb->s_blocksize << 3) diff --git a/fs/udf/file.c b/fs/udf/file.c index 89c7848..f391a2a 100644 --- a/fs/udf/file.c +++ b/fs/udf/file.c @@ -123,8 +123,8 @@ static ssize_t udf_file_aio_write(struct kiocb *iocb, const struct iovec *iov, if (inode->i_sb->s_blocksize < (udf_file_entry_alloc_offset(inode) + pos + count)) { - udf_expand_file_adinicb(inode, pos + count, &err); - if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB) { + err = udf_expand_file_adinicb(inode); + if (err) { udf_debug("udf_expand_adinicb: err=%d\n", err); up_write(&iinfo->i_data_sem); return err; @@ -237,7 +237,7 @@ static int udf_setattr(struct dentry *dentry, struct iattr *attr) if ((attr->ia_valid & ATTR_SIZE) && attr->ia_size != i_size_read(inode)) { - error = vmtruncate(inode, attr->ia_size); + error = udf_setsize(inode, attr->ia_size); if (error) return error; } @@ -249,5 +249,4 @@ static int udf_setattr(struct dentry *dentry, struct iattr *attr) const struct inode_operations udf_file_inode_operations = { .setattr = udf_setattr, - .truncate = udf_truncate, }; diff --git a/fs/udf/inode.c b/fs/udf/inode.c index c6a2e78..ccc8143 100644 --- a/fs/udf/inode.c +++ b/fs/udf/inode.c @@ -73,14 +73,12 @@ void udf_evict_inode(struct inode *inode) struct udf_inode_info *iinfo = UDF_I(inode); int want_delete = 0; - truncate_inode_pages(&inode->i_data, 0); - if (!inode->i_nlink && !is_bad_inode(inode)) { want_delete = 1; - inode->i_size = 0; - udf_truncate(inode); + udf_setsize(inode, 0); udf_update_inode(inode, IS_SYNC(inode)); - } + } else + truncate_inode_pages(&inode->i_data, 0); invalidate_inode_buffers(inode); end_writeback(inode); if (iinfo->i_alloc_type != ICBTAG_FLAG_AD_IN_ICB && @@ -117,9 +115,18 @@ static int udf_write_begin(struct file *file, struct address_space *mapping, ret = block_write_begin(mapping, pos, len, flags, pagep, udf_get_block); if (unlikely(ret)) { - loff_t isize = mapping->host->i_size; - if (pos + len > isize) - vmtruncate(mapping->host, isize); + struct inode *inode = mapping->host; + struct udf_inode_info *iinfo = UDF_I(inode); + loff_t isize = inode->i_size; + + if (pos + len > isize) { + truncate_pagecache(inode, pos + len, isize); + if (iinfo->i_alloc_type != ICBTAG_FLAG_AD_IN_ICB) { + down_write(&iinfo->i_data_sem); + udf_truncate_extents(inode); + up_write(&iinfo->i_data_sem); + } + } } return ret; @@ -139,30 +146,31 @@ const struct address_space_operations udf_aops = { .bmap = udf_bmap, }; -void udf_expand_file_adinicb(struct inode *inode, int newsize, int *err) +int udf_expand_file_adinicb(struct inode *inode) { struct page *page; char *kaddr; struct udf_inode_info *iinfo = UDF_I(inode); + int err; struct writeback_control udf_wbc = { .sync_mode = WB_SYNC_NONE, .nr_to_write = 1, }; - /* from now on we have normal address_space methods */ - inode->i_data.a_ops = &udf_aops; - if (!iinfo->i_lenAlloc) { if (UDF_QUERY_FLAG(inode->i_sb, UDF_FLAG_USE_SHORT_AD)) iinfo->i_alloc_type = ICBTAG_FLAG_AD_SHORT; else iinfo->i_alloc_type = ICBTAG_FLAG_AD_LONG; + /* from now on we have normal address_space methods */ + inode->i_data.a_ops = &udf_aops; mark_inode_dirty(inode); - return; + return 0; } - page = grab_cache_page(inode->i_mapping, 0); - BUG_ON(!PageLocked(page)); + page = find_or_create_page(inode->i_mapping, 0, GFP_NOFS); + if (!page) + return -ENOMEM; if (!PageUptodate(page)) { kaddr = kmap(page); @@ -181,11 +189,24 @@ void udf_expand_file_adinicb(struct inode *inode, int newsize, int *err) iinfo->i_alloc_type = ICBTAG_FLAG_AD_SHORT; else iinfo->i_alloc_type = ICBTAG_FLAG_AD_LONG; - - inode->i_data.a_ops->writepage(page, &udf_wbc); + /* from now on we have normal address_space methods */ + inode->i_data.a_ops = &udf_aops; + err = inode->i_data.a_ops->writepage(page, &udf_wbc); + if (err) { + /* Restore everything back so that we don't lose data... */ + lock_page(page); + kaddr = kmap(page); + memcpy(iinfo->i_ext.i_data + iinfo->i_lenEAttr, kaddr, + inode->i_size); + kunmap(page); + unlock_page(page); + iinfo->i_alloc_type = ICBTAG_FLAG_AD_IN_ICB; + inode->i_data.a_ops = &udf_adinicb_aops; + } page_cache_release(page); - mark_inode_dirty(inode); + + return err; } struct buffer_head *udf_expand_dir_adinicb(struct inode *inode, int *block, @@ -348,8 +369,10 @@ static struct buffer_head *udf_getblk(struct inode *inode, long block, } /* Extend the file by 'blocks' blocks, return the number of extents added */ -int udf_extend_file(struct inode *inode, struct extent_position *last_pos, - struct kernel_long_ad *last_ext, sector_t blocks) +static int udf_do_extend_file(struct inode *inode, + struct extent_position *last_pos, + struct kernel_long_ad *last_ext, + sector_t blocks) { sector_t add; int count = 0, fake = !(last_ext->extLength & UDF_EXTENT_LENGTH_MASK); @@ -357,6 +380,7 @@ int udf_extend_file(struct inode *inode, struct extent_position *last_pos, struct kernel_lb_addr prealloc_loc = {}; int prealloc_len = 0; struct udf_inode_info *iinfo; + int err; /* The previous extent is fake and we should not extend by anything * - there's nothing to do... */ @@ -422,26 +446,29 @@ int udf_extend_file(struct inode *inode, struct extent_position *last_pos, /* Create enough extents to cover the whole hole */ while (blocks > add) { blocks -= add; - if (udf_add_aext(inode, last_pos, &last_ext->extLocation, - last_ext->extLength, 1) == -1) - return -1; + err = udf_add_aext(inode, last_pos, &last_ext->extLocation, + last_ext->extLength, 1); + if (err) + return err; count++; } if (blocks) { last_ext->extLength = EXT_NOT_RECORDED_NOT_ALLOCATED | (blocks << sb->s_blocksize_bits); - if (udf_add_aext(inode, last_pos, &last_ext->extLocation, - last_ext->extLength, 1) == -1) - return -1; + err = udf_add_aext(inode, last_pos, &last_ext->extLocation, + last_ext->extLength, 1); + if (err) + return err; count++; } out: /* Do we have some preallocated blocks saved? */ if (prealloc_len) { - if (udf_add_aext(inode, last_pos, &prealloc_loc, - prealloc_len, 1) == -1) - return -1; + err = udf_add_aext(inode, last_pos, &prealloc_loc, + prealloc_len, 1); + if (err) + return err; last_ext->extLocation = prealloc_loc; last_ext->extLength = prealloc_len; count++; @@ -453,11 +480,68 @@ out: else if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_LONG) last_pos->offset -= sizeof(struct long_ad); else - return -1; + return -EIO; return count; } +static int udf_extend_file(struct inode *inode, loff_t newsize) +{ + + struct extent_position epos; + struct kernel_lb_addr eloc; + uint32_t elen; + int8_t etype; + struct super_block *sb = inode->i_sb; + sector_t first_block = newsize >> sb->s_blocksize_bits, offset; + int adsize; + struct udf_inode_info *iinfo = UDF_I(inode); + struct kernel_long_ad extent; + int err; + + if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_SHORT) + adsize = sizeof(struct short_ad); + else if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_LONG) + adsize = sizeof(struct long_ad); + else + BUG(); + + etype = inode_bmap(inode, first_block, &epos, &eloc, &elen, &offset); + + /* File has extent covering the new size (could happen when extending + * inside a block)? */ + if (etype != -1) + return 0; + if (newsize & (sb->s_blocksize - 1)) + offset++; + /* Extended file just to the boundary of the last file block? */ + if (offset == 0) + return 0; + + /* Truncate is extending the file by 'offset' blocks */ + if ((!epos.bh && epos.offset == udf_file_entry_alloc_offset(inode)) || + (epos.bh && epos.offset == sizeof(struct allocExtDesc))) { + /* File has no extents at all or has empty last + * indirect extent! Create a fake extent... */ + extent.extLocation.logicalBlockNum = 0; + extent.extLocation.partitionReferenceNum = 0; + extent.extLength = EXT_NOT_RECORDED_NOT_ALLOCATED; + } else { + epos.offset -= adsize; + etype = udf_next_aext(inode, &epos, &extent.extLocation, + &extent.extLength, 0); + extent.extLength |= etype << 30; + } + err = udf_do_extend_file(inode, &epos, &extent, offset); + if (err < 0) + goto out; + err = 0; + iinfo->i_lenExtents = newsize; +out: + brelse(epos.bh); + return err; +} + static struct buffer_head *inode_getblk(struct inode *inode, sector_t block, int *err, sector_t *phys, int *new) { @@ -540,7 +624,7 @@ static struct buffer_head *inode_getblk(struct inode *inode, sector_t block, elen = EXT_RECORDED_ALLOCATED | ((elen + inode->i_sb->s_blocksize - 1) & ~(inode->i_sb->s_blocksize - 1)); - etype = udf_write_aext(inode, &cur_epos, &eloc, elen, 1); + udf_write_aext(inode, &cur_epos, &eloc, elen, 1); } brelse(prev_epos.bh); brelse(cur_epos.bh); @@ -564,19 +648,17 @@ static struct buffer_head *inode_getblk(struct inode *inode, sector_t block, memset(&laarr[0].extLocation, 0x00, sizeof(struct kernel_lb_addr)); laarr[0].extLength = EXT_NOT_RECORDED_NOT_ALLOCATED; - /* Will udf_extend_file() create real extent from + /* Will udf_do_extend_file() create real extent from a fake one? */ startnum = (offset > 0); } /* Create extents for the hole between EOF and offset */ - ret = udf_extend_file(inode, &prev_epos, laarr, offset); - if (ret == -1) { + ret = udf_do_extend_file(inode, &prev_epos, laarr, offset); + if (ret < 0) { brelse(prev_epos.bh); brelse(cur_epos.bh); brelse(next_epos.bh); - /* We don't really know the error here so we just make - * something up */ - *err = -ENOSPC; + *err = ret; return NULL; } c = 0; @@ -1005,52 +1087,66 @@ struct buffer_head *udf_bread(struct inode *inode, int block, return NULL; } -void udf_truncate(struct inode *inode) +int udf_setsize(struct inode *inode, loff_t newsize) { - int offset; int err; struct udf_inode_info *iinfo; + int bsize = 1 << inode->i_blkbits; if (!(S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || S_ISLNK(inode->i_mode))) - return; + return -EINVAL; if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) - return; + return -EPERM; iinfo = UDF_I(inode); - if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB) { + if (newsize > inode->i_size) { down_write(&iinfo->i_data_sem); - if (inode->i_sb->s_blocksize < - (udf_file_entry_alloc_offset(inode) + - inode->i_size)) { - udf_expand_file_adinicb(inode, inode->i_size, &err); - if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB) { - inode->i_size = iinfo->i_lenAlloc; - up_write(&iinfo->i_data_sem); - return; + if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB) { + if (bsize < + (udf_file_entry_alloc_offset(inode) + newsize)) { + err = udf_expand_file_adinicb(inode); + if (err) { + up_write(&iinfo->i_data_sem); + return err; + } } else - udf_truncate_extents(inode); - } else { - offset = inode->i_size & (inode->i_sb->s_blocksize - 1); - memset(iinfo->i_ext.i_data + iinfo->i_lenEAttr + offset, - 0x00, inode->i_sb->s_blocksize - - offset - udf_file_entry_alloc_offset(inode)); - iinfo->i_lenAlloc = inode->i_size; + iinfo->i_lenAlloc = newsize; + } + err = udf_extend_file(inode, newsize); + if (err) { + up_write(&iinfo->i_data_sem); + return err; } + truncate_setsize(inode, newsize); up_write(&iinfo->i_data_sem); } else { - block_truncate_page(inode->i_mapping, inode->i_size, - udf_get_block); + if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_IN_ICB) { + down_write(&iinfo->i_data_sem); + memset(iinfo->i_ext.i_data + iinfo->i_lenEAttr + newsize, + 0x00, bsize - newsize - + udf_file_entry_alloc_offset(inode)); + iinfo->i_lenAlloc = newsize; + truncate_setsize(inode, newsize); + up_write(&iinfo->i_data_sem); + goto update_time; + } + err = block_truncate_page(inode->i_mapping, newsize, + udf_get_block); + if (err) + return err; down_write(&iinfo->i_data_sem); + truncate_setsize(inode, newsize); udf_truncate_extents(inode); up_write(&iinfo->i_data_sem); } - +update_time: inode->i_mtime = inode->i_ctime = current_fs_time(inode->i_sb); if (IS_SYNC(inode)) udf_sync_inode(inode); else mark_inode_dirty(inode); + return 0; } static void __udf_read_inode(struct inode *inode) @@ -1637,14 +1733,13 @@ struct inode *udf_iget(struct super_block *sb, struct kernel_lb_addr *ino) return NULL; } -int8_t udf_add_aext(struct inode *inode, struct extent_position *epos, - struct kernel_lb_addr *eloc, uint32_t elen, int inc) +int udf_add_aext(struct inode *inode, struct extent_position *epos, + struct kernel_lb_addr *eloc, uint32_t elen, int inc) { int adsize; struct short_ad *sad = NULL; struct long_ad *lad = NULL; struct allocExtDesc *aed; - int8_t etype; uint8_t *ptr; struct udf_inode_info *iinfo = UDF_I(inode); @@ -1660,7 +1755,7 @@ int8_t udf_add_aext(struct inode *inode, struct extent_position *epos, else if (iinfo->i_alloc_type == ICBTAG_FLAG_AD_LONG) adsize = sizeof(struct long_ad); else - return -1; + return -EIO; if (epos->offset + (2 * adsize) > inode->i_sb->s_blocksize) { unsigned char *sptr, *dptr; @@ -1672,12 +1767,12 @@ int8_t udf_add_aext(struct inode *inode, struct extent_position *epos, obloc.partitionReferenceNum, obloc.logicalBlockNum, &err); if (!epos->block.logicalBlockNum) - return -1; + return -ENOSPC; nbh = udf_tgetblk(inode->i_sb, udf_get_lb_pblock(inode->i_sb, &epos->block, 0)); if (!nbh) - return -1; + return -EIO; lock_buffer(nbh); memset(nbh->b_data, 0x00, inode->i_sb->s_blocksize); set_buffer_uptodate(nbh); @@ -1746,7 +1841,7 @@ int8_t udf_add_aext(struct inode *inode, struct extent_position *epos, epos->bh = nbh; } - etype = udf_write_aext(inode, epos, eloc, elen, inc); + udf_write_aext(inode, epos, eloc, elen, inc); if (!epos->bh) { iinfo->i_lenAlloc += adsize; @@ -1764,11 +1859,11 @@ int8_t udf_add_aext(struct inode *inode, struct extent_position *epos, mark_buffer_dirty_inode(epos->bh, inode); } - return etype; + return 0; } -int8_t udf_write_aext(struct inode *inode, struct extent_position *epos, - struct kernel_lb_addr *eloc, uint32_t elen, int inc) +void udf_write_aext(struct inode *inode, struct extent_position *epos, + struct kernel_lb_addr *eloc, uint32_t elen, int inc) { int adsize; uint8_t *ptr; @@ -1798,7 +1893,7 @@ int8_t udf_write_aext(struct inode *inode, struct extent_position *epos, adsize = sizeof(struct long_ad); break; default: - return -1; + return; } if (epos->bh) { @@ -1817,8 +1912,6 @@ int8_t udf_write_aext(struct inode *inode, struct extent_position *epos, if (inc) epos->offset += adsize; - - return (elen >> 30); } int8_t udf_next_aext(struct inode *inode, struct extent_position *epos, diff --git a/fs/udf/namei.c b/fs/udf/namei.c index 2be0f9e..f1dce84 100644 --- a/fs/udf/namei.c +++ b/fs/udf/namei.c @@ -32,6 +32,8 @@ #include <linux/crc-itu-t.h> #include <linux/exportfs.h> +enum { UDF_MAX_LINKS = 0xffff }; + static inline int udf_match(int len1, const unsigned char *name1, int len2, const unsigned char *name2) { @@ -650,7 +652,7 @@ static int udf_mkdir(struct inode *dir, struct dentry *dentry, int mode) struct udf_inode_info *iinfo; err = -EMLINK; - if (dir->i_nlink >= (256 << sizeof(dir->i_nlink)) - 1) + if (dir->i_nlink >= UDF_MAX_LINKS) goto out; err = -EIO; @@ -1034,9 +1036,8 @@ static int udf_link(struct dentry *old_dentry, struct inode *dir, struct fileIdentDesc cfi, *fi; int err; - if (inode->i_nlink >= (256 << sizeof(inode->i_nlink)) - 1) { + if (inode->i_nlink >= UDF_MAX_LINKS) return -EMLINK; - } fi = udf_add_entry(dir, dentry, &fibh, &cfi, &err); if (!fi) { @@ -1131,9 +1132,7 @@ static int udf_rename(struct inode *old_dir, struct dentry *old_dentry, goto end_rename; retval = -EMLINK; - if (!new_inode && - new_dir->i_nlink >= - (256 << sizeof(new_dir->i_nlink)) - 1) + if (!new_inode && new_dir->i_nlink >= UDF_MAX_LINKS) goto end_rename; } if (!nfi) { @@ -1287,8 +1286,13 @@ static int udf_encode_fh(struct dentry *de, __u32 *fh, int *lenp, struct fid *fid = (struct fid *)fh; int type = FILEID_UDF_WITHOUT_PARENT; - if (len < 3 || (connectable && len < 5)) + if (connectable && (len < 5)) { + *lenp = 5; return 255; + } else if (len < 3) { + *lenp = 3; + return 255; + } *lenp = 3; fid->udf.block = location.logicalBlockNum; diff --git a/fs/udf/truncate.c b/fs/udf/truncate.c index 225527c..8424308 100644 --- a/fs/udf/truncate.c +++ b/fs/udf/truncate.c @@ -197,6 +197,11 @@ static void udf_update_alloc_ext_desc(struct inode *inode, mark_buffer_dirty_inode(epos->bh, inode); } +/* + * Truncate extents of inode to inode->i_size. This function can be used only + * for making file shorter. For making file longer, udf_extend_file() has to + * be used. + */ void udf_truncate_extents(struct inode *inode) { struct extent_position epos; @@ -219,96 +224,65 @@ void udf_truncate_extents(struct inode *inode) etype = inode_bmap(inode, first_block, &epos, &eloc, &elen, &offset); byte_offset = (offset << sb->s_blocksize_bits) + (inode->i_size & (sb->s_blocksize - 1)); - if (etype != -1) { - epos.offset -= adsize; - extent_trunc(inode, &epos, &eloc, etype, elen, byte_offset); - epos.offset += adsize; - if (byte_offset) - lenalloc = epos.offset; - else - lenalloc = epos.offset - adsize; - - if (!epos.bh) - lenalloc -= udf_file_entry_alloc_offset(inode); - else - lenalloc -= sizeof(struct allocExtDesc); - - while ((etype = udf_current_aext(inode, &epos, &eloc, - &elen, 0)) != -1) { - if (etype == (EXT_NEXT_EXTENT_ALLOCDECS >> 30)) { - udf_write_aext(inode, &epos, &neloc, nelen, 0); - if (indirect_ext_len) { - /* We managed to free all extents in the - * indirect extent - free it too */ - BUG_ON(!epos.bh); - udf_free_blocks(sb, inode, &epos.block, - 0, indirect_ext_len); - } else if (!epos.bh) { - iinfo->i_lenAlloc = lenalloc; - mark_inode_dirty(inode); - } else - udf_update_alloc_ext_desc(inode, - &epos, lenalloc); - brelse(epos.bh); - epos.offset = sizeof(struct allocExtDesc); - epos.block = eloc; - epos.bh = udf_tread(sb, - udf_get_lb_pblock(sb, &eloc, 0)); - if (elen) - indirect_ext_len = - (elen + sb->s_blocksize - 1) >> - sb->s_blocksize_bits; - else - indirect_ext_len = 1; - } else { - extent_trunc(inode, &epos, &eloc, etype, - elen, 0); - epos.offset += adsize; - } - } + if (etype == -1) { + /* We should extend the file? */ + WARN_ON(byte_offset); + return; + } + epos.offset -= adsize; + extent_trunc(inode, &epos, &eloc, etype, elen, byte_offset); + epos.offset += adsize; + if (byte_offset) + lenalloc = epos.offset; + else + lenalloc = epos.offset - adsize; - if (indirect_ext_len) { - BUG_ON(!epos.bh); - udf_free_blocks(sb, inode, &epos.block, 0, - indirect_ext_len); - } else if (!epos.bh) { - iinfo->i_lenAlloc = lenalloc; - mark_inode_dirty(inode); - } else - udf_update_alloc_ext_desc(inode, &epos, lenalloc); - } else if (inode->i_size) { - if (byte_offset) { - struct kernel_long_ad extent; + if (!epos.bh) + lenalloc -= udf_file_entry_alloc_offset(inode); + else + lenalloc -= sizeof(struct allocExtDesc); - /* - * OK, there is not extent covering inode->i_size and - * no extent above inode->i_size => truncate is - * extending the file by 'offset' blocks. - */ - if ((!epos.bh && - epos.offset == - udf_file_entry_alloc_offset(inode)) || - (epos.bh && epos.offset == - sizeof(struct allocExtDesc))) { - /* File has no extents at all or has empty last - * indirect extent! Create a fake extent... */ - extent.extLocation.logicalBlockNum = 0; - extent.extLocation.partitionReferenceNum = 0; - extent.extLength = - EXT_NOT_RECORDED_NOT_ALLOCATED; - } else { - epos.offset -= adsize; - etype = udf_next_aext(inode, &epos, - &extent.extLocation, - &extent.extLength, 0); - extent.extLength |= etype << 30; - } - udf_extend_file(inode, &epos, &extent, - offset + - ((inode->i_size & - (sb->s_blocksize - 1)) != 0)); + while ((etype = udf_current_aext(inode, &epos, &eloc, + &elen, 0)) != -1) { + if (etype == (EXT_NEXT_EXTENT_ALLOCDECS >> 30)) { + udf_write_aext(inode, &epos, &neloc, nelen, 0); + if (indirect_ext_len) { + /* We managed to free all extents in the + * indirect extent - free it too */ + BUG_ON(!epos.bh); + udf_free_blocks(sb, inode, &epos.block, + 0, indirect_ext_len); + } else if (!epos.bh) { + iinfo->i_lenAlloc = lenalloc; + mark_inode_dirty(inode); + } else + udf_update_alloc_ext_desc(inode, + &epos, lenalloc); + brelse(epos.bh); + epos.offset = sizeof(struct allocExtDesc); + epos.block = eloc; + epos.bh = udf_tread(sb, + udf_get_lb_pblock(sb, &eloc, 0)); + if (elen) + indirect_ext_len = + (elen + sb->s_blocksize - 1) >> + sb->s_blocksize_bits; + else + indirect_ext_len = 1; + } else { + extent_trunc(inode, &epos, &eloc, etype, elen, 0); + epos.offset += adsize; } } + + if (indirect_ext_len) { + BUG_ON(!epos.bh); + udf_free_blocks(sb, inode, &epos.block, 0, indirect_ext_len); + } else if (!epos.bh) { + iinfo->i_lenAlloc = lenalloc; + mark_inode_dirty(inode); + } else + udf_update_alloc_ext_desc(inode, &epos, lenalloc); iinfo->i_lenExtents = inode->i_size; brelse(epos.bh); diff --git a/fs/udf/udfdecl.h b/fs/udf/udfdecl.h index eba4820..dbd52d4b 100644 --- a/fs/udf/udfdecl.h +++ b/fs/udf/udfdecl.h @@ -136,22 +136,20 @@ extern int udf_write_fi(struct inode *inode, struct fileIdentDesc *, extern long udf_ioctl(struct file *, unsigned int, unsigned long); /* inode.c */ extern struct inode *udf_iget(struct super_block *, struct kernel_lb_addr *); -extern void udf_expand_file_adinicb(struct inode *, int, int *); +extern int udf_expand_file_adinicb(struct inode *); extern struct buffer_head *udf_expand_dir_adinicb(struct inode *, int *, int *); extern struct buffer_head *udf_bread(struct inode *, int, int, int *); -extern void udf_truncate(struct inode *); +extern int udf_setsize(struct inode *, loff_t); extern void udf_read_inode(struct inode *); extern void udf_evict_inode(struct inode *); extern int udf_write_inode(struct inode *, struct writeback_control *wbc); extern long udf_block_map(struct inode *, sector_t); -extern int udf_extend_file(struct inode *, struct extent_position *, - struct kernel_long_ad *, sector_t); extern int8_t inode_bmap(struct inode *, sector_t, struct extent_position *, struct kernel_lb_addr *, uint32_t *, sector_t *); -extern int8_t udf_add_aext(struct inode *, struct extent_position *, +extern int udf_add_aext(struct inode *, struct extent_position *, + struct kernel_lb_addr *, uint32_t, int); +extern void udf_write_aext(struct inode *, struct extent_position *, struct kernel_lb_addr *, uint32_t, int); -extern int8_t udf_write_aext(struct inode *, struct extent_position *, - struct kernel_lb_addr *, uint32_t, int); extern int8_t udf_delete_aext(struct inode *, struct extent_position, struct kernel_lb_addr, uint32_t); extern int8_t udf_next_aext(struct inode *, struct extent_position *, diff --git a/fs/ufs/Kconfig b/fs/ufs/Kconfig index 30c8f22..e4f10a4 100644 --- a/fs/ufs/Kconfig +++ b/fs/ufs/Kconfig @@ -1,7 +1,6 @@ config UFS_FS tristate "UFS file system support (read only)" depends on BLOCK - depends on BKL # probably fixable help BSD and derivate versions of Unix (such as SunOS, FreeBSD, NetBSD, OpenBSD and NeXTstep) use a file system called UFS. Some System V diff --git a/fs/ufs/inode.c b/fs/ufs/inode.c index 2b251f2..03c255f 100644 --- a/fs/ufs/inode.c +++ b/fs/ufs/inode.c @@ -34,7 +34,6 @@ #include <linux/stat.h> #include <linux/string.h> #include <linux/mm.h> -#include <linux/smp_lock.h> #include <linux/buffer_head.h> #include <linux/writeback.h> @@ -43,7 +42,7 @@ #include "swab.h" #include "util.h" -static u64 ufs_frag_map(struct inode *inode, sector_t frag); +static u64 ufs_frag_map(struct inode *inode, sector_t frag, bool needs_lock); static int ufs_block_to_path(struct inode *inode, sector_t i_block, sector_t offsets[4]) { @@ -82,7 +81,7 @@ static int ufs_block_to_path(struct inode *inode, sector_t i_block, sector_t off * the begining of the filesystem. */ -static u64 ufs_frag_map(struct inode *inode, sector_t frag) +static u64 ufs_frag_map(struct inode *inode, sector_t frag, bool needs_lock) { struct ufs_inode_info *ufsi = UFS_I(inode); struct super_block *sb = inode->i_sb; @@ -107,7 +106,8 @@ static u64 ufs_frag_map(struct inode *inode, sector_t frag) p = offsets; - lock_kernel(); + if (needs_lock) + lock_ufs(sb); if ((flags & UFS_TYPE_MASK) == UFS_TYPE_UFS2) goto ufs2; @@ -152,7 +152,8 @@ ufs2: ret = temp + (u64) (frag & uspi->s_fpbmask); out: - unlock_kernel(); + if (needs_lock) + unlock_ufs(sb); return ret; } @@ -415,14 +416,16 @@ out: int ufs_getfrag_block(struct inode *inode, sector_t fragment, struct buffer_head *bh_result, int create) { struct super_block * sb = inode->i_sb; - struct ufs_sb_private_info * uspi = UFS_SB(sb)->s_uspi; + struct ufs_sb_info * sbi = UFS_SB(sb); + struct ufs_sb_private_info * uspi = sbi->s_uspi; struct buffer_head * bh; int ret, err, new; unsigned long ptr,phys; u64 phys64 = 0; + bool needs_lock = (sbi->mutex_owner != current); if (!create) { - phys64 = ufs_frag_map(inode, fragment); + phys64 = ufs_frag_map(inode, fragment, needs_lock); UFSD("phys64 = %llu\n", (unsigned long long)phys64); if (phys64) map_bh(bh_result, sb, phys64); @@ -436,7 +439,8 @@ int ufs_getfrag_block(struct inode *inode, sector_t fragment, struct buffer_head ret = 0; bh = NULL; - lock_kernel(); + if (needs_lock) + lock_ufs(sb); UFSD("ENTER, ino %lu, fragment %llu\n", inode->i_ino, (unsigned long long)fragment); if (fragment > @@ -498,7 +502,9 @@ out: set_buffer_new(bh_result); map_bh(bh_result, sb, phys); abort: - unlock_kernel(); + if (needs_lock) + unlock_ufs(sb); + return err; abort_too_big: @@ -506,48 +512,6 @@ abort_too_big: goto abort; } -static struct buffer_head *ufs_getfrag(struct inode *inode, - unsigned int fragment, - int create, int *err) -{ - struct buffer_head dummy; - int error; - - dummy.b_state = 0; - dummy.b_blocknr = -1000; - error = ufs_getfrag_block(inode, fragment, &dummy, create); - *err = error; - if (!error && buffer_mapped(&dummy)) { - struct buffer_head *bh; - bh = sb_getblk(inode->i_sb, dummy.b_blocknr); - if (buffer_new(&dummy)) { - memset(bh->b_data, 0, inode->i_sb->s_blocksize); - set_buffer_uptodate(bh); - mark_buffer_dirty(bh); - } - return bh; - } - return NULL; -} - -struct buffer_head * ufs_bread (struct inode * inode, unsigned fragment, - int create, int * err) -{ - struct buffer_head * bh; - - UFSD("ENTER, ino %lu, fragment %u\n", inode->i_ino, fragment); - bh = ufs_getfrag (inode, fragment, create, err); - if (!bh || buffer_uptodate(bh)) - return bh; - ll_rw_block (READ, 1, &bh); - wait_on_buffer (bh); - if (buffer_uptodate(bh)) - return bh; - brelse (bh); - *err = -EIO; - return NULL; -} - static int ufs_writepage(struct page *page, struct writeback_control *wbc) { return block_write_full_page(page,ufs_getfrag_block,wbc); @@ -900,9 +864,9 @@ static int ufs_update_inode(struct inode * inode, int do_sync) int ufs_write_inode(struct inode *inode, struct writeback_control *wbc) { int ret; - lock_kernel(); + lock_ufs(inode->i_sb); ret = ufs_update_inode(inode, wbc->sync_mode == WB_SYNC_ALL); - unlock_kernel(); + unlock_ufs(inode->i_sb); return ret; } @@ -922,22 +886,22 @@ void ufs_evict_inode(struct inode * inode) if (want_delete) { loff_t old_i_size; /*UFS_I(inode)->i_dtime = CURRENT_TIME;*/ - lock_kernel(); + lock_ufs(inode->i_sb); mark_inode_dirty(inode); ufs_update_inode(inode, IS_SYNC(inode)); old_i_size = inode->i_size; inode->i_size = 0; if (inode->i_blocks && ufs_truncate(inode, old_i_size)) ufs_warning(inode->i_sb, __func__, "ufs_truncate failed\n"); - unlock_kernel(); + unlock_ufs(inode->i_sb); } invalidate_inode_buffers(inode); end_writeback(inode); if (want_delete) { - lock_kernel(); + lock_ufs(inode->i_sb); ufs_free_inode (inode); - unlock_kernel(); + unlock_ufs(inode->i_sb); } } diff --git a/fs/ufs/namei.c b/fs/ufs/namei.c index 12f39b9..29309e2 100644 --- a/fs/ufs/namei.c +++ b/fs/ufs/namei.c @@ -29,7 +29,6 @@ #include <linux/time.h> #include <linux/fs.h> -#include <linux/smp_lock.h> #include "ufs_fs.h" #include "ufs.h" @@ -55,16 +54,16 @@ static struct dentry *ufs_lookup(struct inode * dir, struct dentry *dentry, stru if (dentry->d_name.len > UFS_MAXNAMLEN) return ERR_PTR(-ENAMETOOLONG); - lock_kernel(); + lock_ufs(dir->i_sb); ino = ufs_inode_by_name(dir, &dentry->d_name); if (ino) { inode = ufs_iget(dir->i_sb, ino); if (IS_ERR(inode)) { - unlock_kernel(); + unlock_ufs(dir->i_sb); return ERR_CAST(inode); } } - unlock_kernel(); + unlock_ufs(dir->i_sb); d_add(dentry, inode); return NULL; } @@ -93,9 +92,9 @@ static int ufs_create (struct inode * dir, struct dentry * dentry, int mode, inode->i_fop = &ufs_file_operations; inode->i_mapping->a_ops = &ufs_aops; mark_inode_dirty(inode); - lock_kernel(); + lock_ufs(dir->i_sb); err = ufs_add_nondir(dentry, inode); - unlock_kernel(); + unlock_ufs(dir->i_sb); } UFSD("END: err=%d\n", err); return err; @@ -115,9 +114,9 @@ static int ufs_mknod (struct inode * dir, struct dentry *dentry, int mode, dev_t init_special_inode(inode, mode, rdev); ufs_set_inode_dev(inode->i_sb, UFS_I(inode), rdev); mark_inode_dirty(inode); - lock_kernel(); + lock_ufs(dir->i_sb); err = ufs_add_nondir(dentry, inode); - unlock_kernel(); + unlock_ufs(dir->i_sb); } return err; } @@ -133,7 +132,7 @@ static int ufs_symlink (struct inode * dir, struct dentry * dentry, if (l > sb->s_blocksize) goto out_notlocked; - lock_kernel(); + lock_ufs(dir->i_sb); inode = ufs_new_inode(dir, S_IFLNK | S_IRWXUGO); err = PTR_ERR(inode); if (IS_ERR(inode)) @@ -156,7 +155,7 @@ static int ufs_symlink (struct inode * dir, struct dentry * dentry, err = ufs_add_nondir(dentry, inode); out: - unlock_kernel(); + unlock_ufs(dir->i_sb); out_notlocked: return err; @@ -172,9 +171,9 @@ static int ufs_link (struct dentry * old_dentry, struct inode * dir, struct inode *inode = old_dentry->d_inode; int error; - lock_kernel(); + lock_ufs(dir->i_sb); if (inode->i_nlink >= UFS_LINK_MAX) { - unlock_kernel(); + unlock_ufs(dir->i_sb); return -EMLINK; } @@ -183,7 +182,7 @@ static int ufs_link (struct dentry * old_dentry, struct inode * dir, ihold(inode); error = ufs_add_nondir(dentry, inode); - unlock_kernel(); + unlock_ufs(dir->i_sb); return error; } @@ -195,7 +194,7 @@ static int ufs_mkdir(struct inode * dir, struct dentry * dentry, int mode) if (dir->i_nlink >= UFS_LINK_MAX) goto out; - lock_kernel(); + lock_ufs(dir->i_sb); inode_inc_link_count(dir); inode = ufs_new_inode(dir, S_IFDIR|mode); @@ -216,7 +215,7 @@ static int ufs_mkdir(struct inode * dir, struct dentry * dentry, int mode) err = ufs_add_link(dentry, inode); if (err) goto out_fail; - unlock_kernel(); + unlock_ufs(dir->i_sb); d_instantiate(dentry, inode); out: @@ -228,7 +227,7 @@ out_fail: iput (inode); out_dir: inode_dec_link_count(dir); - unlock_kernel(); + unlock_ufs(dir->i_sb); goto out; } @@ -259,7 +258,7 @@ static int ufs_rmdir (struct inode * dir, struct dentry *dentry) struct inode * inode = dentry->d_inode; int err= -ENOTEMPTY; - lock_kernel(); + lock_ufs(dir->i_sb); if (ufs_empty_dir (inode)) { err = ufs_unlink(dir, dentry); if (!err) { @@ -268,7 +267,7 @@ static int ufs_rmdir (struct inode * dir, struct dentry *dentry) inode_dec_link_count(dir); } } - unlock_kernel(); + unlock_ufs(dir->i_sb); return err; } @@ -306,7 +305,6 @@ static int ufs_rename(struct inode *old_dir, struct dentry *old_dentry, new_de = ufs_find_entry(new_dir, &new_dentry->d_name, &new_page); if (!new_de) goto out_dir; - inode_inc_link_count(old_inode); ufs_set_link(new_dir, new_de, new_page, old_inode); new_inode->i_ctime = CURRENT_TIME_SEC; if (dir_de) @@ -318,12 +316,9 @@ static int ufs_rename(struct inode *old_dir, struct dentry *old_dentry, if (new_dir->i_nlink >= UFS_LINK_MAX) goto out_dir; } - inode_inc_link_count(old_inode); err = ufs_add_link(new_dentry, old_inode); - if (err) { - inode_dec_link_count(old_inode); + if (err) goto out_dir; - } if (dir_de) inode_inc_link_count(new_dir); } @@ -331,12 +326,11 @@ static int ufs_rename(struct inode *old_dir, struct dentry *old_dentry, /* * Like most other Unix systems, set the ctime for inodes on a * rename. - * inode_dec_link_count() will mark the inode dirty. */ old_inode->i_ctime = CURRENT_TIME_SEC; ufs_delete_entry(old_dir, old_de, old_page); - inode_dec_link_count(old_inode); + mark_inode_dirty(old_inode); if (dir_de) { ufs_set_link(old_inode, dir_de, dir_page, new_dir); diff --git a/fs/ufs/super.c b/fs/ufs/super.c index 2c61ac5..7693d62 100644 --- a/fs/ufs/super.c +++ b/fs/ufs/super.c @@ -84,7 +84,6 @@ #include <linux/blkdev.h> #include <linux/init.h> #include <linux/parser.h> -#include <linux/smp_lock.h> #include <linux/buffer_head.h> #include <linux/vfs.h> #include <linux/log2.h> @@ -96,6 +95,26 @@ #include "swab.h" #include "util.h" +void lock_ufs(struct super_block *sb) +{ +#if defined(CONFIG_SMP) || defined (CONFIG_PREEMPT) + struct ufs_sb_info *sbi = UFS_SB(sb); + + mutex_lock(&sbi->mutex); + sbi->mutex_owner = current; +#endif +} + +void unlock_ufs(struct super_block *sb) +{ +#if defined(CONFIG_SMP) || defined (CONFIG_PREEMPT) + struct ufs_sb_info *sbi = UFS_SB(sb); + + sbi->mutex_owner = NULL; + mutex_unlock(&sbi->mutex); +#endif +} + static struct inode *ufs_nfs_get_inode(struct super_block *sb, u64 ino, u32 generation) { struct ufs_sb_private_info *uspi = UFS_SB(sb)->s_uspi; @@ -313,7 +332,6 @@ void ufs_panic (struct super_block * sb, const char * function, struct ufs_super_block_first * usb1; va_list args; - lock_kernel(); uspi = UFS_SB(sb)->s_uspi; usb1 = ubh_get_usb_first(uspi); @@ -521,7 +539,7 @@ static int ufs_read_cylinder_structures(struct super_block *sb) */ size = uspi->s_cssize; blks = (size + uspi->s_fsize - 1) >> uspi->s_fshift; - base = space = kmalloc(size, GFP_KERNEL); + base = space = kmalloc(size, GFP_NOFS); if (!base) goto failed; sbi->s_csp = (struct ufs_csum *)space; @@ -546,7 +564,7 @@ static int ufs_read_cylinder_structures(struct super_block *sb) * Read cylinder group (we read only first fragment from block * at this time) and prepare internal data structures for cg caching. */ - if (!(sbi->s_ucg = kmalloc (sizeof(struct buffer_head *) * uspi->s_ncg, GFP_KERNEL))) + if (!(sbi->s_ucg = kmalloc (sizeof(struct buffer_head *) * uspi->s_ncg, GFP_NOFS))) goto failed; for (i = 0; i < uspi->s_ncg; i++) sbi->s_ucg[i] = NULL; @@ -564,7 +582,7 @@ static int ufs_read_cylinder_structures(struct super_block *sb) ufs_print_cylinder_stuff(sb, (struct ufs_cylinder_group *) sbi->s_ucg[i]->b_data); } for (i = 0; i < UFS_MAX_GROUP_LOADED; i++) { - if (!(sbi->s_ucpi[i] = kmalloc (sizeof(struct ufs_cg_private_info), GFP_KERNEL))) + if (!(sbi->s_ucpi[i] = kmalloc (sizeof(struct ufs_cg_private_info), GFP_NOFS))) goto failed; sbi->s_cgno[i] = UFS_CGNO_EMPTY; } @@ -646,8 +664,6 @@ static void ufs_put_super_internal(struct super_block *sb) UFSD("ENTER\n"); - lock_kernel(); - ufs_put_cstotal(sb); size = uspi->s_cssize; blks = (size + uspi->s_fsize - 1) >> uspi->s_fshift; @@ -676,8 +692,6 @@ static void ufs_put_super_internal(struct super_block *sb) kfree (sbi->s_ucg); kfree (base); - unlock_kernel(); - UFSD("EXIT\n"); } @@ -696,8 +710,6 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) unsigned maxsymlen; int ret = -EINVAL; - lock_kernel(); - uspi = NULL; ubh = NULL; flags = 0; @@ -718,6 +730,7 @@ static int ufs_fill_super(struct super_block *sb, void *data, int silent) goto failed; } #endif + mutex_init(&sbi->mutex); /* * Set default mount options * Parse mount options @@ -1165,7 +1178,6 @@ magic_found: goto failed; UFSD("EXIT\n"); - unlock_kernel(); return 0; dalloc_failed: @@ -1177,12 +1189,10 @@ failed: kfree(sbi); sb->s_fs_info = NULL; UFSD("EXIT (FAILED)\n"); - unlock_kernel(); return ret; failed_nomem: UFSD("EXIT (NOMEM)\n"); - unlock_kernel(); return -ENOMEM; } @@ -1193,8 +1203,8 @@ static int ufs_sync_fs(struct super_block *sb, int wait) struct ufs_super_block_third * usb3; unsigned flags; + lock_ufs(sb); lock_super(sb); - lock_kernel(); UFSD("ENTER\n"); @@ -1213,8 +1223,8 @@ static int ufs_sync_fs(struct super_block *sb, int wait) sb->s_dirt = 0; UFSD("EXIT\n"); - unlock_kernel(); unlock_super(sb); + unlock_ufs(sb); return 0; } @@ -1256,7 +1266,7 @@ static int ufs_remount (struct super_block *sb, int *mount_flags, char *data) unsigned new_mount_opt, ufstype; unsigned flags; - lock_kernel(); + lock_ufs(sb); lock_super(sb); uspi = UFS_SB(sb)->s_uspi; flags = UFS_SB(sb)->s_flags; @@ -1272,7 +1282,7 @@ static int ufs_remount (struct super_block *sb, int *mount_flags, char *data) ufs_set_opt (new_mount_opt, ONERROR_LOCK); if (!ufs_parse_options (data, &new_mount_opt)) { unlock_super(sb); - unlock_kernel(); + unlock_ufs(sb); return -EINVAL; } if (!(new_mount_opt & UFS_MOUNT_UFSTYPE)) { @@ -1280,14 +1290,14 @@ static int ufs_remount (struct super_block *sb, int *mount_flags, char *data) } else if ((new_mount_opt & UFS_MOUNT_UFSTYPE) != ufstype) { printk("ufstype can't be changed during remount\n"); unlock_super(sb); - unlock_kernel(); + unlock_ufs(sb); return -EINVAL; } if ((*mount_flags & MS_RDONLY) == (sb->s_flags & MS_RDONLY)) { UFS_SB(sb)->s_mount_opt = new_mount_opt; unlock_super(sb); - unlock_kernel(); + unlock_ufs(sb); return 0; } @@ -1313,7 +1323,7 @@ static int ufs_remount (struct super_block *sb, int *mount_flags, char *data) printk("ufs was compiled with read-only support, " "can't be mounted as read-write\n"); unlock_super(sb); - unlock_kernel(); + unlock_ufs(sb); return -EINVAL; #else if (ufstype != UFS_MOUNT_UFSTYPE_SUN && @@ -1323,13 +1333,13 @@ static int ufs_remount (struct super_block *sb, int *mount_flags, char *data) ufstype != UFS_MOUNT_UFSTYPE_UFS2) { printk("this ufstype is read-only supported\n"); unlock_super(sb); - unlock_kernel(); + unlock_ufs(sb); return -EINVAL; } if (!ufs_read_cylinder_structures(sb)) { printk("failed during remounting\n"); unlock_super(sb); - unlock_kernel(); + unlock_ufs(sb); return -EPERM; } sb->s_flags &= ~MS_RDONLY; @@ -1337,7 +1347,7 @@ static int ufs_remount (struct super_block *sb, int *mount_flags, char *data) } UFS_SB(sb)->s_mount_opt = new_mount_opt; unlock_super(sb); - unlock_kernel(); + unlock_ufs(sb); return 0; } @@ -1371,7 +1381,7 @@ static int ufs_statfs(struct dentry *dentry, struct kstatfs *buf) struct ufs_super_block_third *usb3; u64 id = huge_encode_dev(sb->s_bdev->bd_dev); - lock_kernel(); + lock_ufs(sb); usb1 = ubh_get_usb_first(uspi); usb2 = ubh_get_usb_second(uspi); @@ -1395,7 +1405,7 @@ static int ufs_statfs(struct dentry *dentry, struct kstatfs *buf) buf->f_fsid.val[0] = (u32)id; buf->f_fsid.val[1] = (u32)(id >> 32); - unlock_kernel(); + unlock_ufs(sb); return 0; } @@ -1405,7 +1415,7 @@ static struct kmem_cache * ufs_inode_cachep; static struct inode *ufs_alloc_inode(struct super_block *sb) { struct ufs_inode_info *ei; - ei = (struct ufs_inode_info *)kmem_cache_alloc(ufs_inode_cachep, GFP_KERNEL); + ei = (struct ufs_inode_info *)kmem_cache_alloc(ufs_inode_cachep, GFP_NOFS); if (!ei) return NULL; ei->vfs_inode.i_version = 1; diff --git a/fs/ufs/truncate.c b/fs/ufs/truncate.c index a58f915..e56a4f5 100644 --- a/fs/ufs/truncate.c +++ b/fs/ufs/truncate.c @@ -40,7 +40,6 @@ #include <linux/time.h> #include <linux/stat.h> #include <linux/string.h> -#include <linux/smp_lock.h> #include <linux/buffer_head.h> #include <linux/blkdev.h> #include <linux/sched.h> @@ -467,7 +466,6 @@ int ufs_truncate(struct inode *inode, loff_t old_i_size) block_truncate_page(inode->i_mapping, inode->i_size, ufs_getfrag_block); - lock_kernel(); while (1) { retry = ufs_trunc_direct(inode); retry |= ufs_trunc_indirect(inode, UFS_IND_BLOCK, @@ -487,7 +485,6 @@ int ufs_truncate(struct inode *inode, loff_t old_i_size) inode->i_mtime = inode->i_ctime = CURRENT_TIME_SEC; ufsi->i_lastfrag = DIRECT_FRAGMENT; - unlock_kernel(); mark_inode_dirty(inode); out: UFSD("EXIT: err %d\n", err); @@ -510,7 +507,9 @@ int ufs_setattr(struct dentry *dentry, struct iattr *attr) /* XXX(truncate): truncate_setsize should be called last */ truncate_setsize(inode, attr->ia_size); + lock_ufs(inode->i_sb); error = ufs_truncate(inode, old_i_size); + unlock_ufs(inode->i_sb); if (error) return error; } diff --git a/fs/ufs/ufs.h b/fs/ufs/ufs.h index c08782e..5be2755 100644 --- a/fs/ufs/ufs.h +++ b/fs/ufs/ufs.h @@ -18,6 +18,8 @@ struct ufs_sb_info { unsigned s_cgno[UFS_MAX_GROUP_LOADED]; unsigned short s_cg_loaded; unsigned s_mount_opt; + struct mutex mutex; + struct task_struct *mutex_owner; }; struct ufs_inode_info { @@ -109,7 +111,6 @@ extern struct inode *ufs_iget(struct super_block *, unsigned long); extern int ufs_write_inode (struct inode *, struct writeback_control *); extern int ufs_sync_inode (struct inode *); extern void ufs_evict_inode (struct inode *); -extern struct buffer_head * ufs_bread (struct inode *, unsigned, int, int *); extern int ufs_getfrag_block (struct inode *inode, sector_t fragment, struct buffer_head *bh_result, int create); /* namei.c */ @@ -154,4 +155,7 @@ static inline u32 ufs_dtogd(struct ufs_sb_private_info * uspi, u64 b) return do_div(b, uspi->s_fpg); } +extern void lock_ufs(struct super_block *sb); +extern void unlock_ufs(struct super_block *sb); + #endif /* _UFS_UFS_H */ diff --git a/fs/ufs/util.c b/fs/ufs/util.c index d2c36d5..95425b5 100644 --- a/fs/ufs/util.c +++ b/fs/ufs/util.c @@ -27,7 +27,7 @@ struct ufs_buffer_head * _ubh_bread_ (struct ufs_sb_private_info * uspi, if (count > UFS_MAXFRAG) return NULL; ubh = (struct ufs_buffer_head *) - kmalloc (sizeof (struct ufs_buffer_head), GFP_KERNEL); + kmalloc (sizeof (struct ufs_buffer_head), GFP_NOFS); if (!ubh) return NULL; ubh->fragment = fragment; diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile index fee7901..284a7c8 100644 --- a/fs/xfs/Makefile +++ b/fs/xfs/Makefile @@ -16,14 +16,11 @@ # Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA # -EXTRA_CFLAGS += -I$(src) -I$(src)/linux-2.6 +ccflags-y := -I$(src) -I$(src)/linux-2.6 +ccflags-$(CONFIG_XFS_DEBUG) += -g XFS_LINUX := linux-2.6 -ifeq ($(CONFIG_XFS_DEBUG),y) - EXTRA_CFLAGS += -g -endif - obj-$(CONFIG_XFS_FS) += xfs.o xfs-y += linux-2.6/xfs_trace.o diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 3cc671c..5cb230f 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -2021,11 +2021,12 @@ xfs_buf_init(void) if (!xfslogd_workqueue) goto out_free_buf_zone; - xfsdatad_workqueue = create_workqueue("xfsdatad"); + xfsdatad_workqueue = alloc_workqueue("xfsdatad", WQ_MEM_RECLAIM, 1); if (!xfsdatad_workqueue) goto out_destroy_xfslogd_workqueue; - xfsconvertd_workqueue = create_workqueue("xfsconvertd"); + xfsconvertd_workqueue = alloc_workqueue("xfsconvertd", + WQ_MEM_RECLAIM, 1); if (!xfsconvertd_workqueue) goto out_destroy_xfsdatad_workqueue; diff --git a/fs/xfs/linux-2.6/xfs_export.c b/fs/xfs/linux-2.6/xfs_export.c index fc0114d..f4f878f 100644 --- a/fs/xfs/linux-2.6/xfs_export.c +++ b/fs/xfs/linux-2.6/xfs_export.c @@ -89,8 +89,10 @@ xfs_fs_encode_fh( * seven combinations work. The real answer is "don't use v2". */ len = xfs_fileid_length(fileid_type); - if (*max_len < len) + if (*max_len < len) { + *max_len = len; return 255; + } *max_len = len; switch (fileid_type) { diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c index bd57278..9ff7fc6 100644 --- a/fs/xfs/linux-2.6/xfs_iops.c +++ b/fs/xfs/linux-2.6/xfs_iops.c @@ -102,7 +102,8 @@ xfs_mark_inode_dirty( STATIC int xfs_init_security( struct inode *inode, - struct inode *dir) + struct inode *dir, + const struct qstr *qstr) { struct xfs_inode *ip = XFS_I(inode); size_t length; @@ -110,7 +111,7 @@ xfs_init_security( unsigned char *name; int error; - error = security_inode_init_security(inode, dir, (char **)&name, + error = security_inode_init_security(inode, dir, qstr, (char **)&name, &value, &length); if (error) { if (error == -EOPNOTSUPP) @@ -194,7 +195,7 @@ xfs_vn_mknod( inode = VFS_I(ip); - error = xfs_init_security(inode, dir); + error = xfs_init_security(inode, dir, &dentry->d_name); if (unlikely(error)) goto out_cleanup_inode; @@ -367,7 +368,7 @@ xfs_vn_symlink( inode = VFS_I(cip); - error = xfs_init_security(inode, dir); + error = xfs_init_security(inode, dir, &dentry->d_name); if (unlikely(error)) goto out_cleanup_inode; diff --git a/fs/xfs/xfs_mru_cache.c b/fs/xfs/xfs_mru_cache.c index edfa178..4aff563 100644 --- a/fs/xfs/xfs_mru_cache.c +++ b/fs/xfs/xfs_mru_cache.c @@ -309,7 +309,7 @@ xfs_mru_cache_init(void) if (!xfs_mru_elem_zone) goto out; - xfs_mru_reap_wq = create_singlethread_workqueue("xfs_mru_cache"); + xfs_mru_reap_wq = alloc_workqueue("xfs_mru_cache", WQ_MEM_RECLAIM, 1); if (!xfs_mru_reap_wq) goto out_destroy_mru_elem_zone; |