summaryrefslogtreecommitdiffstats
path: root/fs/xfs/xfs_alloc.c
diff options
context:
space:
mode:
authorDavid Chinner <dgc@sgi.com>2007-05-24 07:26:31 +0200
committerTim Shimmin <tes@chook.melbourne.sgi.com>2007-07-14 07:28:50 +0200
commit92821e2ba4ae26887223326fb0b95cdab963b768 (patch)
treea40a2ef10e5b0791df3e522f3139193d39bf2454 /fs/xfs/xfs_alloc.c
parent[XFS] Use generic shrinker interfaces in XFS. (diff)
downloadlinux-92821e2ba4ae26887223326fb0b95cdab963b768.tar.xz
linux-92821e2ba4ae26887223326fb0b95cdab963b768.zip
[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_alloc.c')
-rw-r--r--fs/xfs/xfs_alloc.c48
1 files changed, 38 insertions, 10 deletions
diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c
index 8e9a40aa0cd3..98f95d4c4bcc 100644
--- a/fs/xfs/xfs_alloc.c
+++ b/fs/xfs/xfs_alloc.c
@@ -1447,7 +1447,8 @@ xfs_alloc_ag_vextent_small(
else if (args->minlen == 1 && args->alignment == 1 && !args->isfl &&
(be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_flcount)
> args->minleft)) {
- if ((error = xfs_alloc_get_freelist(args->tp, args->agbp, &fbno)))
+ error = xfs_alloc_get_freelist(args->tp, args->agbp, &fbno, 0);
+ if (error)
goto error0;
if (fbno != NULLAGBLOCK) {
if (args->userdata) {
@@ -1923,7 +1924,8 @@ xfs_alloc_fix_freelist(
while (be32_to_cpu(agf->agf_flcount) > need) {
xfs_buf_t *bp;
- if ((error = xfs_alloc_get_freelist(tp, agbp, &bno)))
+ error = xfs_alloc_get_freelist(tp, agbp, &bno, 0);
+ if (error)
return error;
if ((error = xfs_free_ag_extent(tp, agbp, args->agno, bno, 1, 1)))
return error;
@@ -1973,8 +1975,9 @@ xfs_alloc_fix_freelist(
* Put each allocated block on the list.
*/
for (bno = targs.agbno; bno < targs.agbno + targs.len; bno++) {
- if ((error = xfs_alloc_put_freelist(tp, agbp, agflbp,
- bno)))
+ error = xfs_alloc_put_freelist(tp, agbp,
+ agflbp, bno, 0);
+ if (error)
return error;
}
}
@@ -1991,13 +1994,15 @@ int /* error */
xfs_alloc_get_freelist(
xfs_trans_t *tp, /* transaction pointer */
xfs_buf_t *agbp, /* buffer containing the agf structure */
- xfs_agblock_t *bnop) /* block address retrieved from freelist */
+ xfs_agblock_t *bnop, /* block address retrieved from freelist */
+ int btreeblk) /* destination is a AGF btree */
{
xfs_agf_t *agf; /* a.g. freespace structure */
xfs_agfl_t *agfl; /* a.g. freelist structure */
xfs_buf_t *agflbp;/* buffer for a.g. freelist structure */
xfs_agblock_t bno; /* block number returned */
int error;
+ int logflags;
#ifdef XFS_ALLOC_TRACE
static char fname[] = "xfs_alloc_get_freelist";
#endif
@@ -2032,8 +2037,16 @@ xfs_alloc_get_freelist(
be32_add(&agf->agf_flcount, -1);
xfs_trans_agflist_delta(tp, -1);
pag->pagf_flcount--;
- TRACE_MODAGF(NULL, agf, XFS_AGF_FLFIRST | XFS_AGF_FLCOUNT);
- xfs_alloc_log_agf(tp, agbp, XFS_AGF_FLFIRST | XFS_AGF_FLCOUNT);
+
+ logflags = XFS_AGF_FLFIRST | XFS_AGF_FLCOUNT;
+ if (btreeblk) {
+ be32_add(&agf->agf_btreeblks, 1);
+ pag->pagf_btreeblks++;
+ logflags |= XFS_AGF_BTREEBLKS;
+ }
+
+ TRACE_MODAGF(NULL, agf, logflags);
+ xfs_alloc_log_agf(tp, agbp, logflags);
*bnop = bno;
/*
@@ -2071,6 +2084,7 @@ xfs_alloc_log_agf(
offsetof(xfs_agf_t, agf_flcount),
offsetof(xfs_agf_t, agf_freeblks),
offsetof(xfs_agf_t, agf_longest),
+ offsetof(xfs_agf_t, agf_btreeblks),
sizeof(xfs_agf_t)
};
@@ -2106,12 +2120,14 @@ xfs_alloc_put_freelist(
xfs_trans_t *tp, /* transaction pointer */
xfs_buf_t *agbp, /* buffer for a.g. freelist header */
xfs_buf_t *agflbp,/* buffer for a.g. free block array */
- xfs_agblock_t bno) /* block being freed */
+ xfs_agblock_t bno, /* block being freed */
+ int btreeblk) /* block came from a AGF btree */
{
xfs_agf_t *agf; /* a.g. freespace structure */
xfs_agfl_t *agfl; /* a.g. free block array */
__be32 *blockp;/* pointer to array entry */
int error;
+ int logflags;
#ifdef XFS_ALLOC_TRACE
static char fname[] = "xfs_alloc_put_freelist";
#endif
@@ -2132,11 +2148,22 @@ xfs_alloc_put_freelist(
be32_add(&agf->agf_flcount, 1);
xfs_trans_agflist_delta(tp, 1);
pag->pagf_flcount++;
+
+ logflags = XFS_AGF_FLLAST | XFS_AGF_FLCOUNT;
+ if (btreeblk) {
+ be32_add(&agf->agf_btreeblks, -1);
+ pag->pagf_btreeblks--;
+ logflags |= XFS_AGF_BTREEBLKS;
+ }
+
+ TRACE_MODAGF(NULL, agf, logflags);
+ xfs_alloc_log_agf(tp, agbp, logflags);
+
ASSERT(be32_to_cpu(agf->agf_flcount) <= XFS_AGFL_SIZE(mp));
blockp = &agfl->agfl_bno[be32_to_cpu(agf->agf_fllast)];
*blockp = cpu_to_be32(bno);
- TRACE_MODAGF(NULL, agf, XFS_AGF_FLLAST | XFS_AGF_FLCOUNT);
- xfs_alloc_log_agf(tp, agbp, XFS_AGF_FLLAST | XFS_AGF_FLCOUNT);
+ TRACE_MODAGF(NULL, agf, logflags);
+ xfs_alloc_log_agf(tp, agbp, logflags);
xfs_trans_log_buf(tp, agflbp,
(int)((xfs_caddr_t)blockp - (xfs_caddr_t)agfl),
(int)((xfs_caddr_t)blockp - (xfs_caddr_t)agfl +
@@ -2196,6 +2223,7 @@ xfs_alloc_read_agf(
pag = &mp->m_perag[agno];
if (!pag->pagf_init) {
pag->pagf_freeblks = be32_to_cpu(agf->agf_freeblks);
+ pag->pagf_btreeblks = be32_to_cpu(agf->agf_btreeblks);
pag->pagf_flcount = be32_to_cpu(agf->agf_flcount);
pag->pagf_longest = be32_to_cpu(agf->agf_longest);
pag->pagf_levels[XFS_BTNUM_BNOi] =