diff options
author | Jeff Mahoney <jeffm@suse.com> | 2013-08-08 17:34:46 -0400 |
---|---|---|
committer | Jeff Mahoney <jeffm@suse.de> | 2013-08-08 17:34:46 -0400 |
commit | 278f6679f454bf185a07d9a4ca355b153482d17a (patch) | |
tree | ffead073e67cfdc1ddfc3949ebc93c06dcaaab8f /fs/reiserfs/dir.c | |
parent | 4c05141df57f4ffc1a9a28f1925434924179bfe4 (diff) | |
download | op-kernel-dev-278f6679f454bf185a07d9a4ca355b153482d17a.zip op-kernel-dev-278f6679f454bf185a07d9a4ca355b153482d17a.tar.gz |
reiserfs: locking, handle nested locks properly
The reiserfs write lock replaced the BKL and uses similar semantics.
Frederic's locking code makes a distinction between when the lock is nested
and when it's being acquired/released, but I don't think that's the right
distinction to make.
The right distinction is between the lock being released at end-of-use and
the lock being released for a schedule. The unlock should return the depth
and the lock should restore it, rather than the other way around as it is now.
This patch implements that and adds a number of places where the lock
should be dropped.
Signed-off-by: Jeff Mahoney <jeffm@suse.com>
Diffstat (limited to 'fs/reiserfs/dir.c')
-rw-r--r-- | fs/reiserfs/dir.c | 7 |
1 files changed, 4 insertions, 3 deletions
diff --git a/fs/reiserfs/dir.c b/fs/reiserfs/dir.c index 03e4ca5..1fd2051 100644 --- a/fs/reiserfs/dir.c +++ b/fs/reiserfs/dir.c @@ -71,6 +71,7 @@ int reiserfs_readdir_inode(struct inode *inode, struct dir_context *ctx) char small_buf[32]; /* avoid kmalloc if we can */ struct reiserfs_dir_entry de; int ret = 0; + int depth; reiserfs_write_lock(inode->i_sb); @@ -181,17 +182,17 @@ int reiserfs_readdir_inode(struct inode *inode, struct dir_context *ctx) * Since filldir might sleep, we can release * the write lock here for other waiters */ - reiserfs_write_unlock(inode->i_sb); + depth = reiserfs_write_unlock_nested(inode->i_sb); if (!dir_emit (ctx, local_buf, d_reclen, d_ino, DT_UNKNOWN)) { - reiserfs_write_lock(inode->i_sb); + reiserfs_write_lock_nested(inode->i_sb, depth); if (local_buf != small_buf) { kfree(local_buf); } goto end; } - reiserfs_write_lock(inode->i_sb); + reiserfs_write_lock_nested(inode->i_sb, depth); if (local_buf != small_buf) { kfree(local_buf); } |