diff options
author | David Chinner <dgc@sgi.com> | 2007-05-24 15:26:31 +1000 |
---|---|---|
committer | Tim Shimmin <tes@chook.melbourne.sgi.com> | 2007-07-14 15:28:50 +1000 |
commit | 92821e2ba4ae26887223326fb0b95cdab963b768 (patch) | |
tree | a40a2ef10e5b0791df3e522f3139193d39bf2454 /fs/xfs/xfs_trans.c | |
parent | 3260f78ad6d5b788e78ea709d377f58e569bee41 (diff) | |
download | kernel_samsung_espresso10-92821e2ba4ae26887223326fb0b95cdab963b768.zip kernel_samsung_espresso10-92821e2ba4ae26887223326fb0b95cdab963b768.tar.gz kernel_samsung_espresso10-92821e2ba4ae26887223326fb0b95cdab963b768.tar.bz2 |
[XFS] Lazy Superblock Counters
When we have a couple of hundred transactions on the fly at once, they all
typically modify the on disk superblock in some way.
create/unclink/mkdir/rmdir modify inode counts, allocation/freeing modify
free block counts.
When these counts are modified in a transaction, they must eventually lock
the superblock buffer and apply the mods. The buffer then remains locked
until the transaction is committed into the incore log buffer. The result
of this is that with enough transactions on the fly the incore superblock
buffer becomes a bottleneck.
The result of contention on the incore superblock buffer is that
transaction rates fall - the more pressure that is put on the superblock
buffer, the slower things go.
The key to removing the contention is to not require the superblock fields
in question to be locked. We do that by not marking the superblock dirty
in the transaction. IOWs, we modify the incore superblock but do not
modify the cached superblock buffer. In short, we do not log superblock
modifications to critical fields in the superblock on every transaction.
In fact we only do it just before we write the superblock to disk every
sync period or just before unmount.
This creates an interesting problem - if we don't log or write out the
fields in every transaction, then how do the values get recovered after a
crash? the answer is simple - we keep enough duplicate, logged information
in other structures that we can reconstruct the correct count after log
recovery has been performed.
It is the AGF and AGI structures that contain the duplicate information;
after recovery, we walk every AGI and AGF and sum their individual
counters to get the correct value, and we do a transaction into the log to
correct them. An optimisation of this is that if we have a clean unmount
record, we know the value in the superblock is correct, so we can avoid
the summation walk under normal conditions and so mount/recovery times do
not change under normal operation.
One wrinkle that was discovered during development was that the blocks
used in the freespace btrees are never accounted for in the AGF counters.
This was once a valid optimisation to make; when the filesystem is full,
the free space btrees are empty and consume no space. Hence when it
matters, the "accounting" is correct. But that means the when we do the
AGF summations, we would not have a correct count and xfs_check would
complain. Hence a new counter was added to track the number of blocks used
by the free space btrees. This is an *on-disk format change*.
As a result of this, lazy superblock counters are a mkfs option and at the
moment on linux there is no way to convert an old filesystem. This is
possible - xfs_db can be used to twiddle the right bits and then
xfs_repair will do the format conversion for you. Similarly, you can
convert backwards as well. At some point we'll add functionality to
xfs_admin to do the bit twiddling easily....
SGI-PV: 964999
SGI-Modid: xfs-linux-melb:xfs-kern:28652a
Signed-off-by: David Chinner <dgc@sgi.com>
Signed-off-by: Christoph Hellwig <hch@infradead.org>
Signed-off-by: Tim Shimmin <tes@sgi.com>
Diffstat (limited to 'fs/xfs/xfs_trans.c')
-rw-r--r-- | fs/xfs/xfs_trans.c | 58 |
1 files changed, 46 insertions, 12 deletions
diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index cc2d609..7133fd9 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -427,6 +427,14 @@ undo_blocks: * * Mark the transaction structure to indicate that the superblock * needs to be updated before committing. + * + * Because we may not be keeping track of allocated/free inodes and + * used filesystem blocks in the superblock, we do not mark the + * superblock dirty in this transaction if we modify these fields. + * We still need to update the transaction deltas so that they get + * applied to the incore superblock, but we don't want them to + * cause the superblock to get locked and logged if these are the + * only fields in the superblock that the transaction modifies. */ void xfs_trans_mod_sb( @@ -434,13 +442,19 @@ xfs_trans_mod_sb( uint field, int64_t delta) { + uint32_t flags = (XFS_TRANS_DIRTY|XFS_TRANS_SB_DIRTY); + xfs_mount_t *mp = tp->t_mountp; switch (field) { case XFS_TRANS_SB_ICOUNT: tp->t_icount_delta += delta; + if (xfs_sb_version_haslazysbcount(&mp->m_sb)) + flags &= ~XFS_TRANS_SB_DIRTY; break; case XFS_TRANS_SB_IFREE: tp->t_ifree_delta += delta; + if (xfs_sb_version_haslazysbcount(&mp->m_sb)) + flags &= ~XFS_TRANS_SB_DIRTY; break; case XFS_TRANS_SB_FDBLOCKS: /* @@ -453,6 +467,8 @@ xfs_trans_mod_sb( ASSERT(tp->t_blk_res_used <= tp->t_blk_res); } tp->t_fdblocks_delta += delta; + if (xfs_sb_version_haslazysbcount(&mp->m_sb)) + flags &= ~XFS_TRANS_SB_DIRTY; break; case XFS_TRANS_SB_RES_FDBLOCKS: /* @@ -462,6 +478,8 @@ xfs_trans_mod_sb( */ ASSERT(delta < 0); tp->t_res_fdblocks_delta += delta; + if (xfs_sb_version_haslazysbcount(&mp->m_sb)) + flags &= ~XFS_TRANS_SB_DIRTY; break; case XFS_TRANS_SB_FREXTENTS: /* @@ -544,18 +562,23 @@ xfs_trans_apply_sb_deltas( (tp->t_ag_freeblks_delta + tp->t_ag_flist_delta + tp->t_ag_btree_delta)); - if (tp->t_icount_delta != 0) { - INT_MOD(sbp->sb_icount, ARCH_CONVERT, tp->t_icount_delta); - } - if (tp->t_ifree_delta != 0) { - INT_MOD(sbp->sb_ifree, ARCH_CONVERT, tp->t_ifree_delta); - } + /* + * Only update the superblock counters if we are logging them + */ + if (!xfs_sb_version_haslazysbcount(&(tp->t_mountp->m_sb))) { + if (tp->t_icount_delta != 0) { + INT_MOD(sbp->sb_icount, ARCH_CONVERT, tp->t_icount_delta); + } + if (tp->t_ifree_delta != 0) { + INT_MOD(sbp->sb_ifree, ARCH_CONVERT, tp->t_ifree_delta); + } - if (tp->t_fdblocks_delta != 0) { - INT_MOD(sbp->sb_fdblocks, ARCH_CONVERT, tp->t_fdblocks_delta); - } - if (tp->t_res_fdblocks_delta != 0) { - INT_MOD(sbp->sb_fdblocks, ARCH_CONVERT, tp->t_res_fdblocks_delta); + if (tp->t_fdblocks_delta != 0) { + INT_MOD(sbp->sb_fdblocks, ARCH_CONVERT, tp->t_fdblocks_delta); + } + if (tp->t_res_fdblocks_delta != 0) { + INT_MOD(sbp->sb_fdblocks, ARCH_CONVERT, tp->t_res_fdblocks_delta); + } } if (tp->t_frextents_delta != 0) { @@ -627,6 +650,7 @@ xfs_trans_unreserve_and_mod_sb( { xfs_mod_sb_t msb[14]; /* If you add cases, add entries */ xfs_mod_sb_t *msbp; + xfs_mount_t *mp = tp->t_mountp; /* REFERENCED */ int error; int rsvd; @@ -659,8 +683,15 @@ xfs_trans_unreserve_and_mod_sb( * The t_res_fdblocks_delta and t_res_frextents_delta fields are * explicitly NOT applied to the in-core superblock. * The idea is that that has already been done. + * + * If we are not logging superblock counters, then the inode + * allocated/free and used block counts are not updated in the + * on disk superblock. In this case, XFS_TRANS_SB_DIRTY will + * not be set when the transaction is updated but we still need + * to update the incore superblock with the changes. */ - if (tp->t_flags & XFS_TRANS_SB_DIRTY) { + if (xfs_sb_version_haslazysbcount(&mp->m_sb) || + (tp->t_flags & XFS_TRANS_SB_DIRTY)) { if (tp->t_icount_delta != 0) { msbp->msb_field = XFS_SBS_ICOUNT; msbp->msb_delta = tp->t_icount_delta; @@ -676,6 +707,9 @@ xfs_trans_unreserve_and_mod_sb( msbp->msb_delta = tp->t_fdblocks_delta; msbp++; } + } + + if (tp->t_flags & XFS_TRANS_SB_DIRTY) { if (tp->t_frextents_delta != 0) { msbp->msb_field = XFS_SBS_FREXTENTS; msbp->msb_delta = tp->t_frextents_delta; |