diff options
author | Dave Chinner <dchinner@redhat.com> | 2015-01-22 09:10:26 +1100 |
---|---|---|
committer | Dave Chinner <david@fromorbit.com> | 2015-01-22 09:10:26 +1100 |
commit | 4d11a40239405e531fc0e9dcd07921f00b965931 (patch) | |
tree | f8a6b43d76e4d698f7cb1a5daf9027d2a2f22ca7 /fs/xfs/xfs_qm_syscalls.c | |
parent | 97bf6af1f928216fd6c5a66e8a57bfa95a659672 (diff) | |
download | linux-stable-4d11a40239405e531fc0e9dcd07921f00b965931.tar.gz linux-stable-4d11a40239405e531fc0e9dcd07921f00b965931.tar.bz2 linux-stable-4d11a40239405e531fc0e9dcd07921f00b965931.zip |
xfs: remove bitfield based superblock updates
When we log changes to the superblock, we first have to write them
to the on-disk buffer, and then log that. Right now we have a
complex bitfield based arrangement to only write the modified field
to the buffer before we log it.
This used to be necessary as a performance optimisation because we
logged the superblock buffer in every extent or inode allocation or
freeing, and so performance was extremely important. We haven't done
this for years, however, ever since the lazy superblock counters
pulled the superblock logging out of the transaction commit
fast path.
Hence we have a bunch of complexity that is not necessary that makes
writing the in-core superblock to disk much more complex than it
needs to be. We only need to log the superblock now during
management operations (e.g. during mount, unmount or quota control
operations) so it is not a performance critical path anymore.
As such, remove the complex field based logging mechanism and
replace it with a simple conversion function similar to what we use
for all other on-disk structures.
This means we always log the entirity of the superblock, but again
because we rarely modify the superblock this is not an issue for log
bandwidth or CPU time. Indeed, if we do log the superblock
frequently, delayed logging will minimise the impact of this
overhead.
[Fixed gquota/pquota inode sharing regression noticed by bfoster.]
Signed-off-by: Dave Chinner <dchinner@redhat.com>
Reviewed-by: Brian Foster <bfoster@redhat.com>
Signed-off-by: Dave Chinner <david@fromorbit.com>
Diffstat (limited to 'fs/xfs/xfs_qm_syscalls.c')
-rw-r--r-- | fs/xfs/xfs_qm_syscalls.c | 13 |
1 files changed, 4 insertions, 9 deletions
diff --git a/fs/xfs/xfs_qm_syscalls.c b/fs/xfs/xfs_qm_syscalls.c index 74fca68e43b6..8d7e5f068803 100644 --- a/fs/xfs/xfs_qm_syscalls.c +++ b/fs/xfs/xfs_qm_syscalls.c @@ -92,8 +92,7 @@ xfs_qm_scall_quotaoff( mutex_unlock(&q->qi_quotaofflock); /* XXX what to do if error ? Revert back to old vals incore ? */ - error = xfs_qm_write_sb_changes(mp, XFS_SB_QFLAGS); - return error; + return xfs_qm_write_sb_changes(mp); } dqtype = 0; @@ -314,7 +313,6 @@ xfs_qm_scall_quotaon( { int error; uint qf; - __int64_t sbflags; flags &= (XFS_ALL_QUOTA_ACCT | XFS_ALL_QUOTA_ENFD); /* @@ -322,8 +320,6 @@ xfs_qm_scall_quotaon( */ flags &= ~(XFS_ALL_QUOTA_ACCT); - sbflags = 0; - if (flags == 0) { xfs_debug(mp, "%s: zero flags, m_qflags=%x", __func__, mp->m_qflags); @@ -370,11 +366,10 @@ xfs_qm_scall_quotaon( /* * There's nothing to change if it's the same. */ - if ((qf & flags) == flags && sbflags == 0) + if ((qf & flags) == flags) return -EEXIST; - sbflags |= XFS_SB_QFLAGS; - if ((error = xfs_qm_write_sb_changes(mp, sbflags))) + if ((error = xfs_qm_write_sb_changes(mp))) return error; /* * If we aren't trying to switch on quota enforcement, we are done. @@ -801,7 +796,7 @@ xfs_qm_log_quotaoff( mp->m_sb.sb_qflags = (mp->m_qflags & ~(flags)) & XFS_MOUNT_QUOTA_ALL; spin_unlock(&mp->m_sb_lock); - xfs_mod_sb(tp, XFS_SB_QFLAGS); + xfs_mod_sb(tp); /* * We have to make sure that the transaction is secure on disk before we |