Merge branch 'master' into export-slabh

11 files changed, 115 insertions, 96 deletions

diff --git a/fs/fat/namei_vfat.c b/fs/fat/namei_vfat.c
index c1ef501..6fcc7e7 100644
--- a/fs/fat/namei_vfat.c
+++ b/fs/fat/namei_vfat.c
@@ -309,7 +309,7 @@ static int vfat_create_shortname(struct inode *dir, struct nls_table *nls,
 {
 	struct fat_mount_options *opts = &MSDOS_SB(dir->i_sb)->options;
 	wchar_t *ip, *ext_start, *end, *name_start;
-	unsigned char base[9], ext[4], buf[8], *p;
+	unsigned char base[9], ext[4], buf[5], *p;
 	unsigned char charbuf[NLS_MAX_CHARSET_SIZE];
 	int chl, chi;
 	int sz = 0, extlen, baselen, i, numtail_baselen, numtail2_baselen;
@@ -467,7 +467,7 @@ static int vfat_create_shortname(struct inode *dir, struct nls_table *nls,
 			return 0;
 	}
 
-	i = jiffies & 0xffff;
+	i = jiffies;
 	sz = (jiffies >> 16) & 0x7;
 	if (baselen > 2) {
 		baselen = numtail2_baselen;
@@ -476,7 +476,7 @@ static int vfat_create_shortname(struct inode *dir, struct nls_table *nls,
 	name_res[baselen + 4] = '~';
 	name_res[baselen + 5] = '1' + sz;
 	while (1) {
-		sprintf(buf, "%04X", i);
+		snprintf(buf, sizeof(buf), "%04X", i & 0xffff);
 		memcpy(&name_res[baselen], buf, 4);
 		if (vfat_find_form(dir, name_res) < 0)
 			break;
diff --git a/fs/logfs/dev_bdev.c b/fs/logfs/dev_bdev.c
index 18e8c14..243c000 100644
--- a/fs/logfs/dev_bdev.c
+++ b/fs/logfs/dev_bdev.c
@@ -81,6 +81,7 @@ static void writeseg_end_io(struct bio *bio, int err)
 			prefetchw(&bvec->bv_page->flags);
 
 		end_page_writeback(page);
+		page_cache_release(page);
 	} while (bvec >= bio->bi_io_vec);
 	bio_put(bio);
 	if (atomic_dec_and_test(&super->s_pending_writes))
@@ -98,8 +99,10 @@ static int __bdev_writeseg(struct super_block *sb, u64 ofs, pgoff_t index,
 	unsigned int max_pages = queue_max_hw_sectors(q) >> (PAGE_SHIFT - 9);
 	int i;
 
+	if (max_pages > BIO_MAX_PAGES)
+		max_pages = BIO_MAX_PAGES;
 	bio = bio_alloc(GFP_NOFS, max_pages);
-	BUG_ON(!bio); /* FIXME: handle this */
+	BUG_ON(!bio);
 
 	for (i = 0; i < nr_pages; i++) {
 		if (i >= max_pages) {
@@ -192,8 +195,10 @@ static int do_erase(struct super_block *sb, u64 ofs, pgoff_t index,
 	unsigned int max_pages = queue_max_hw_sectors(q) >> (PAGE_SHIFT - 9);
 	int i;
 
+	if (max_pages > BIO_MAX_PAGES)
+		max_pages = BIO_MAX_PAGES;
 	bio = bio_alloc(GFP_NOFS, max_pages);
-	BUG_ON(!bio); /* FIXME: handle this */
+	BUG_ON(!bio);
 
 	for (i = 0; i < nr_pages; i++) {
 		if (i >= max_pages) {
diff --git a/fs/logfs/dir.c b/fs/logfs/dir.c
index e1cb995..2396a85 100644
--- a/fs/logfs/dir.c
+++ b/fs/logfs/dir.c
@@ -303,12 +303,12 @@ static int __logfs_readdir(struct file *file, void *buf, filldir_t filldir)
 				(filler_t *)logfs_readpage, NULL);
 		if (IS_ERR(page))
 			return PTR_ERR(page);
-		dd = kmap_atomic(page, KM_USER0);
+		dd = kmap(page);
 		BUG_ON(dd->namelen == 0);
 
 		full = filldir(buf, (char *)dd->name, be16_to_cpu(dd->namelen),
 				pos, be64_to_cpu(dd->ino), dd->type);
-		kunmap_atomic(dd, KM_USER0);
+		kunmap(page);
 		page_cache_release(page);
 		if (full)
 			break;
diff --git a/fs/logfs/journal.c b/fs/logfs/journal.c
index f186043..33bd260 100644
--- a/fs/logfs/journal.c
+++ b/fs/logfs/journal.c
@@ -801,6 +801,7 @@ void do_logfs_journal_wl_pass(struct super_block *sb)
 {
 	struct logfs_super *super = logfs_super(sb);
 	struct logfs_area *area = super->s_journal_area;
+	struct btree_head32 *head = &super->s_reserved_segments;
 	u32 segno, ec;
 	int i, err;
 
@@ -808,6 +809,7 @@ void do_logfs_journal_wl_pass(struct super_block *sb)
 	/* Drop old segments */
 	journal_for_each(i)
 		if (super->s_journal_seg[i]) {
+			btree_remove32(head, super->s_journal_seg[i]);
 			logfs_set_segment_unreserved(sb,
 					super->s_journal_seg[i],
 					super->s_journal_ec[i]);
@@ -820,8 +822,13 @@ void do_logfs_journal_wl_pass(struct super_block *sb)
 		super->s_journal_seg[i] = segno;
 		super->s_journal_ec[i] = ec;
 		logfs_set_segment_reserved(sb, segno);
+		err = btree_insert32(head, segno, (void *)1, GFP_KERNEL);
+		BUG_ON(err); /* mempool should prevent this */
+		err = logfs_erase_segment(sb, segno, 1);
+		BUG_ON(err); /* FIXME: remount-ro would be nicer */
 	}
 	/* Manually move journal_area */
+	freeseg(sb, area->a_segno);
 	area->a_segno = super->s_journal_seg[0];
 	area->a_is_open = 0;
 	area->a_used_bytes = 0;
diff --git a/fs/logfs/logfs.h b/fs/logfs/logfs.h
index 1297794..b84b0ee 100644
--- a/fs/logfs/logfs.h
+++ b/fs/logfs/logfs.h
@@ -587,6 +587,7 @@ void move_page_to_btree(struct page *page);
 int logfs_init_mapping(struct super_block *sb);
 void logfs_sync_area(struct logfs_area *area);
 void logfs_sync_segments(struct super_block *sb);
+void freeseg(struct super_block *sb, u32 segno);
 
 /* area handling */
 int logfs_init_areas(struct super_block *sb);
diff --git a/fs/logfs/readwrite.c b/fs/logfs/readwrite.c
index d5919af..bff4025 100644
--- a/fs/logfs/readwrite.c
+++ b/fs/logfs/readwrite.c
@@ -1595,7 +1595,6 @@ int logfs_delete(struct inode *inode, pgoff_t index,
 	return ret;
 }
 
-/* Rewrite cannot mark the inode dirty but has to write it immediatly. */
 int logfs_rewrite_block(struct inode *inode, u64 bix, u64 ofs,
 		gc_level_t gc_level, long flags)
 {
@@ -1612,6 +1611,18 @@ int logfs_rewrite_block(struct inode *inode, u64 bix, u64 ofs,
 		if (level != 0)
 			alloc_indirect_block(inode, page, 0);
 		err = logfs_write_buf(inode, page, flags);
+		if (!err && shrink_level(gc_level) == 0) {
+			/* Rewrite cannot mark the inode dirty but has to
+			 * write it immediatly.
+			 * Q: Can't we just create an alias for the inode
+			 * instead?  And if not, why not?
+			 */
+			if (inode->i_ino == LOGFS_INO_MASTER)
+				logfs_write_anchor(inode->i_sb);
+			else {
+				err = __logfs_write_inode(inode, flags);
+			}
+		}
 	}
 	logfs_put_write_page(page);
 	return err;
diff --git a/fs/logfs/segment.c b/fs/logfs/segment.c
index 614d7a6..801a3a1 100644
--- a/fs/logfs/segment.c
+++ b/fs/logfs/segment.c
@@ -94,50 +94,58 @@ void __logfs_buf_write(struct logfs_area *area, u64 ofs, void *buf, size_t len,
 	} while (len);
 }
 
-/*
- * bdev_writeseg will write full pages.  Memset the tail to prevent data leaks.
- */
-static void pad_wbuf(struct logfs_area *area, int final)
+static void pad_partial_page(struct logfs_area *area)
 {
 	struct super_block *sb = area->a_sb;
-	struct logfs_super *super = logfs_super(sb);
 	struct page *page;
 	u64 ofs = dev_ofs(sb, area->a_segno, area->a_used_bytes);
 	pgoff_t index = ofs >> PAGE_SHIFT;
 	long offset = ofs & (PAGE_SIZE-1);
 	u32 len = PAGE_SIZE - offset;
 
-	if (len == PAGE_SIZE) {
-		/* The math in this function can surely use some love */
-		len = 0;
-	}
-	if (len) {
-		BUG_ON(area->a_used_bytes >= super->s_segsize);
-
-		page = get_mapping_page(area->a_sb, index, 0);
+	if (len % PAGE_SIZE) {
+		page = get_mapping_page(sb, index, 0);
 		BUG_ON(!page); /* FIXME: reserve a pool */
 		memset(page_address(page) + offset, 0xff, len);
 		SetPagePrivate(page);
 		page_cache_release(page);
 	}
+}
 
-	if (!final)
-		return;
+static void pad_full_pages(struct logfs_area *area)
+{
+	struct super_block *sb = area->a_sb;
+	struct logfs_super *super = logfs_super(sb);
+	u64 ofs = dev_ofs(sb, area->a_segno, area->a_used_bytes);
+	u32 len = super->s_segsize - area->a_used_bytes;
+	pgoff_t index = PAGE_CACHE_ALIGN(ofs) >> PAGE_CACHE_SHIFT;
+	pgoff_t no_indizes = len >> PAGE_CACHE_SHIFT;
+	struct page *page;
 
-	area->a_used_bytes += len;
-	for ( ; area->a_used_bytes < super->s_segsize;
-			area->a_used_bytes += PAGE_SIZE) {
-		/* Memset another page */
-		index++;
-		page = get_mapping_page(area->a_sb, index, 0);
+	while (no_indizes) {
+		page = get_mapping_page(sb, index, 0);
 		BUG_ON(!page); /* FIXME: reserve a pool */
-		memset(page_address(page), 0xff, PAGE_SIZE);
+		SetPageUptodate(page);
+		memset(page_address(page), 0xff, PAGE_CACHE_SIZE);
 		SetPagePrivate(page);
 		page_cache_release(page);
+		index++;
+		no_indizes--;
 	}
 }
 
 /*
+ * bdev_writeseg will write full pages.  Memset the tail to prevent data leaks.
+ * Also make sure we allocate (and memset) all pages for final writeout.
+ */
+static void pad_wbuf(struct logfs_area *area, int final)
+{
+	pad_partial_page(area);
+	if (final)
+		pad_full_pages(area);
+}
+
+/*
  * We have to be careful with the alias tree.  Since lookup is done by bix,
  * it needs to be normalized, so 14, 15, 16, etc. all match when dealing with
  * indirect blocks.  So always use it through accessor functions.
@@ -684,7 +692,7 @@ int logfs_segment_delete(struct inode *inode, struct logfs_shadow *shadow)
 	return 0;
 }
 
-static void freeseg(struct super_block *sb, u32 segno)
+void freeseg(struct super_block *sb, u32 segno)
 {
 	struct logfs_super *super = logfs_super(sb);
 	struct address_space *mapping = super->s_mapping_inode->i_mapping;
diff --git a/fs/logfs/super.c b/fs/logfs/super.c
index 46990ea..b60bfac 100644
--- a/fs/logfs/super.c
+++ b/fs/logfs/super.c
@@ -278,7 +278,7 @@ static int logfs_recover_sb(struct super_block *sb)
 	}
 	if (valid0 && valid1 && ds_cmp(ds0, ds1)) {
 		printk(KERN_INFO"Superblocks don't match - fixing.\n");
-		return write_one_sb(sb, super->s_devops->find_last_sb);
+		return logfs_write_sb(sb);
 	}
 	/* If neither is valid now, something's wrong.  Didn't we properly
 	 * check them before?!? */
@@ -290,6 +290,10 @@ static int logfs_make_writeable(struct super_block *sb)
 {
 	int err;
 
+	err = logfs_open_segfile(sb);
+	if (err)
+		return err;
+
 	/* Repair any broken superblock copies */
 	err = logfs_recover_sb(sb);
 	if (err)
@@ -300,10 +304,6 @@ static int logfs_make_writeable(struct super_block *sb)
 	if (err)
 		return err;
 
-	err = logfs_open_segfile(sb);
-	if (err)
-		return err;
-
 	/* Do one GC pass before any data gets dirtied */
 	logfs_gc_pass(sb);
 
@@ -329,7 +329,7 @@ static int logfs_get_sb_final(struct super_block *sb, struct vfsmount *mnt)
 
 	sb->s_root = d_alloc_root(rootdir);
 	if (!sb->s_root)
-		goto fail;
+		goto fail2;
 
 	super->s_erase_page = alloc_pages(GFP_KERNEL, 0);
 	if (!super->s_erase_page)
@@ -573,8 +573,7 @@ int logfs_get_sb_device(struct file_system_type *type, int flags,
 	return 0;
 
 err1:
-	up_write(&sb->s_umount);
-	deactivate_super(sb);
+	deactivate_locked_super(sb);
 	return err;
 err0:
 	kfree(super);
diff --git a/fs/proc/base.c b/fs/proc/base.c
index 9e82adc..7621db8 100644
--- a/fs/proc/base.c
+++ b/fs/proc/base.c
@@ -443,12 +443,13 @@ static const struct file_operations proc_lstats_operations = {
 unsigned long badness(struct task_struct *p, unsigned long uptime);
 static int proc_oom_score(struct task_struct *task, char *buffer)
 {
-	unsigned long points;
+	unsigned long points = 0;
 	struct timespec uptime;
 
 	do_posix_clock_monotonic_gettime(&uptime);
 	read_lock(&tasklist_lock);
-	points = badness(task->group_leader, uptime.tv_sec);
+	if (pid_alive(task))
+		points = badness(task, uptime.tv_sec);
 	read_unlock(&tasklist_lock);
 	return sprintf(buffer, "%lu\n", points);
 }
diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c
index 2d45889..caf0337 100644
--- a/fs/proc/task_mmu.c
+++ b/fs/proc/task_mmu.c
@@ -407,6 +407,7 @@ static int show_smap(struct seq_file *m, void *v)
 
 	memset(&mss, 0, sizeof mss);
 	mss.vma = vma;
+	/* mmap_sem is held in m_start */
 	if (vma->vm_mm && !is_vm_hugetlb_page(vma))
 		walk_page_range(vma->vm_start, vma->vm_end, &smaps_walk);
 
@@ -553,7 +554,8 @@ const struct file_operations proc_clear_refs_operations = {
 };
 
 struct pagemapread {
-	u64 __user *out, *end;
+	int pos, len;
+	u64 *buffer;
 };
 
 #define PM_ENTRY_BYTES      sizeof(u64)
@@ -576,10 +578,8 @@ struct pagemapread {
 static int add_to_pagemap(unsigned long addr, u64 pfn,
 			  struct pagemapread *pm)
 {
-	if (put_user(pfn, pm->out))
-		return -EFAULT;
-	pm->out++;
-	if (pm->out >= pm->end)
+	pm->buffer[pm->pos++] = pfn;
+	if (pm->pos >= pm->len)
 		return PM_END_OF_BUFFER;
 	return 0;
 }
@@ -721,21 +721,20 @@ static int pagemap_hugetlb_range(pte_t *pte, unsigned long addr,
  * determine which areas of memory are actually mapped and llseek to
  * skip over unmapped regions.
  */
+#define PAGEMAP_WALK_SIZE	(PMD_SIZE)
 static ssize_t pagemap_read(struct file *file, char __user *buf,
 			    size_t count, loff_t *ppos)
 {
 	struct task_struct *task = get_proc_task(file->f_path.dentry->d_inode);
-	struct page **pages, *page;
-	unsigned long uaddr, uend;
 	struct mm_struct *mm;
 	struct pagemapread pm;
-	int pagecount;
 	int ret = -ESRCH;
 	struct mm_walk pagemap_walk = {};
 	unsigned long src;
 	unsigned long svpfn;
 	unsigned long start_vaddr;
 	unsigned long end_vaddr;
+	int copied = 0;
 
 	if (!task)
 		goto out;
@@ -758,35 +757,12 @@ static ssize_t pagemap_read(struct file *file, char __user *buf,
 	if (!mm)
 		goto out_task;
 
-
-	uaddr = (unsigned long)buf & PAGE_MASK;
-	uend = (unsigned long)(buf + count);
-	pagecount = (PAGE_ALIGN(uend) - uaddr) / PAGE_SIZE;
-	ret = 0;
-	if (pagecount == 0)
-		goto out_mm;
-	pages = kcalloc(pagecount, sizeof(struct page *), GFP_KERNEL);
+	pm.len = PM_ENTRY_BYTES * (PAGEMAP_WALK_SIZE >> PAGE_SHIFT);
+	pm.buffer = kmalloc(pm.len, GFP_TEMPORARY);
 	ret = -ENOMEM;
-	if (!pages)
+	if (!pm.buffer)
 		goto out_mm;
 
-	down_read(&current->mm->mmap_sem);
-	ret = get_user_pages(current, current->mm, uaddr, pagecount,
-			     1, 0, pages, NULL);
-	up_read(&current->mm->mmap_sem);
-
-	if (ret < 0)
-		goto out_free;
-
-	if (ret != pagecount) {
-		pagecount = ret;
-		ret = -EFAULT;
-		goto out_pages;
-	}
-
-	pm.out = (u64 __user *)buf;
-	pm.end = (u64 __user *)(buf + count);
-
 	pagemap_walk.pmd_entry = pagemap_pte_range;
 	pagemap_walk.pte_hole = pagemap_pte_hole;
 	pagemap_walk.hugetlb_entry = pagemap_hugetlb_range;
@@ -808,23 +784,36 @@ static ssize_t pagemap_read(struct file *file, char __user *buf,
 	 * user buffer is tracked in "pm", and the walk
 	 * will stop when we hit the end of the buffer.
 	 */
-	ret = walk_page_range(start_vaddr, end_vaddr, &pagemap_walk);
-	if (ret == PM_END_OF_BUFFER)
-		ret = 0;
-	/* don't need mmap_sem for these, but this looks cleaner */
-	*ppos += (char __user *)pm.out - buf;
-	if (!ret)
-		ret = (char __user *)pm.out - buf;
-
-out_pages:
-	for (; pagecount; pagecount--) {
-		page = pages[pagecount-1];
-		if (!PageReserved(page))
-			SetPageDirty(page);
-		page_cache_release(page);
+	ret = 0;
+	while (count && (start_vaddr < end_vaddr)) {
+		int len;
+		unsigned long end;
+
+		pm.pos = 0;
+		end = start_vaddr + PAGEMAP_WALK_SIZE;
+		/* overflow ? */
+		if (end < start_vaddr || end > end_vaddr)
+			end = end_vaddr;
+		down_read(&mm->mmap_sem);
+		ret = walk_page_range(start_vaddr, end, &pagemap_walk);
+		up_read(&mm->mmap_sem);
+		start_vaddr = end;
+
+		len = min(count, PM_ENTRY_BYTES * pm.pos);
+		if (copy_to_user(buf, pm.buffer, len) < 0) {
+			ret = -EFAULT;
+			goto out_free;
+		}
+		copied += len;
+		buf += len;
+		count -= len;
 	}
+	*ppos += copied;
+	if (!ret || ret == PM_END_OF_BUFFER)
+		ret = copied;
+
 out_free:
-	kfree(pages);
+	kfree(pm.buffer);
 out_mm:
 	mmput(mm);
 out_task:
diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c
index d8fd90d..59125fb 100644
--- a/fs/reiserfs/super.c
+++ b/fs/reiserfs/super.c
@@ -1619,10 +1619,8 @@ static int reiserfs_fill_super(struct super_block *s, void *data, int silent)
 	save_mount_options(s, data);
 
 	sbi = kzalloc(sizeof(struct reiserfs_sb_info), GFP_KERNEL);
-	if (!sbi) {
-		errval = -ENOMEM;
-		goto error_alloc;
-	}
+	if (!sbi)
+		return -ENOMEM;
 	s->s_fs_info = sbi;
 	/* Set default values for options: non-aggressive tails, RO on errors */
 	REISERFS_SB(s)->s_mount_opt |= (1 << REISERFS_SMALLTAIL);
@@ -1879,12 +1877,12 @@ static int reiserfs_fill_super(struct super_block *s, void *data, int silent)
 	return (0);
 
 error:
-	reiserfs_write_unlock(s);
-error_alloc:
 	if (jinit_done) {	/* kill the commit thread, free journal ram */
 		journal_release_error(NULL, s);
 	}
 
+	reiserfs_write_unlock(s);
+
 	reiserfs_free_bitmap_cache(s);
 	if (SB_BUFFER_WITH_SB(s))
 		brelse(SB_BUFFER_WITH_SB(s));