summaryrefslogtreecommitdiffstats
diff options
context:
space:
mode:
authorYounger Liu <younger.liu@huawei.com>2013-09-11 14:19:44 -0700
committerLinus Torvalds <torvalds@linux-foundation.org>2013-09-11 15:56:28 -0700
commit2b1e55c389105b722cccadfa47f5615f57d8887f (patch)
treed5f6a8458a64e8a92f5df71822c75b25f6f0e065
parent5e42781caf6f5f1c77e842d6dcbbf5c51b8b2c29 (diff)
downloadtalos-obmc-linux-2b1e55c389105b722cccadfa47f5615f57d8887f.tar.gz
talos-obmc-linux-2b1e55c389105b722cccadfa47f5615f57d8887f.zip
ocfs2: lighten up allocate transaction
The issue scenario is as following: When fallocating a very large disk space for a small file, __ocfs2_extend_allocation attempts to get a very large transaction. For some journal sizes, there may be not enough room for this transaction, and the fallocate will fail. The patch below extends & restarts the transaction as necessary while allocating space, and should work with even the smallest journal. This patch refers ext4 resize. Test: # mkfs.ocfs2 -b 4K -C 32K -T datafiles /dev/sdc ...(jounral size is 32M) # mount.ocfs2 /dev/sdc /mnt/ocfs2/ # touch /mnt/ocfs2/1.log # fallocate -o 0 -l 400G /mnt/ocfs2/1.log fallocate: /mnt/ocfs2/1.log: fallocate failed: Cannot allocate memory # tail -f /var/log/messages [ 7372.278591] JBD: fallocate wants too many credits (2051 > 2048) [ 7372.278597] (fallocate,6438,0):__ocfs2_extend_allocation:709 ERROR: status = -12 [ 7372.278603] (fallocate,6438,0):ocfs2_allocate_unwritten_extents:1504 ERROR: status = -12 [ 7372.278607] (fallocate,6438,0):__ocfs2_change_file_space:1955 ERROR: status = -12 ^C With this patch, the test works well. Signed-off-by: Younger Liu <younger.liu@huawei.com> Cc: Jie Liu <jeff.liu@oracle.com> Cc: Joel Becker <jlbec@evilplan.org> Cc: Mark Fasheh <mfasheh@suse.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
-rw-r--r--fs/ocfs2/file.c6
-rw-r--r--fs/ocfs2/journal.c35
-rw-r--r--fs/ocfs2/journal.h11
-rw-r--r--fs/ocfs2/ocfs2_trace.h2
4 files changed, 49 insertions, 5 deletions
diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c
index 3261d71319ee..409c549ae02a 100644
--- a/fs/ocfs2/file.c
+++ b/fs/ocfs2/file.c
@@ -671,11 +671,7 @@ restarted_transaction:
} else {
BUG_ON(why != RESTART_TRANS);
- /* TODO: This can be more intelligent. */
- credits = ocfs2_calc_extend_credits(osb->sb,
- &fe->id2.i_list,
- clusters_to_add);
- status = ocfs2_extend_trans(handle, credits);
+ status = ocfs2_allocate_extend_trans(handle, 1);
if (status < 0) {
/* handle still has to be committed at
* this point. */
diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c
index 242170d83971..a126cb37ca4d 100644
--- a/fs/ocfs2/journal.c
+++ b/fs/ocfs2/journal.c
@@ -455,6 +455,41 @@ bail:
return status;
}
+/*
+ * If we have fewer than thresh credits, extend by OCFS2_MAX_TRANS_DATA.
+ * If that fails, restart the transaction & regain write access for the
+ * buffer head which is used for metadata modifications.
+ * Taken from Ext4: extend_or_restart_transaction()
+ */
+int ocfs2_allocate_extend_trans(handle_t *handle, int thresh)
+{
+ int status, old_nblks;
+
+ BUG_ON(!handle);
+
+ old_nblks = handle->h_buffer_credits;
+ trace_ocfs2_allocate_extend_trans(old_nblks, thresh);
+
+ if (old_nblks < thresh)
+ return 0;
+
+ status = jbd2_journal_extend(handle, OCFS2_MAX_TRANS_DATA);
+ if (status < 0) {
+ mlog_errno(status);
+ goto bail;
+ }
+
+ if (status > 0) {
+ status = jbd2_journal_restart(handle, OCFS2_MAX_TRANS_DATA);
+ if (status < 0)
+ mlog_errno(status);
+ }
+
+bail:
+ return status;
+}
+
+
struct ocfs2_triggers {
struct jbd2_buffer_trigger_type ot_triggers;
int ot_offset;
diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h
index 0a992737dcaf..0b479bab3671 100644
--- a/fs/ocfs2/journal.h
+++ b/fs/ocfs2/journal.h
@@ -258,6 +258,17 @@ handle_t *ocfs2_start_trans(struct ocfs2_super *osb,
int ocfs2_commit_trans(struct ocfs2_super *osb,
handle_t *handle);
int ocfs2_extend_trans(handle_t *handle, int nblocks);
+int ocfs2_allocate_extend_trans(handle_t *handle,
+ int thresh);
+
+/*
+ * Define an arbitrary limit for the amount of data we will anticipate
+ * writing to any given transaction. For unbounded transactions such as
+ * fallocate(2) we can write more than this, but we always
+ * start off at the maximum transaction size and grow the transaction
+ * optimistically as we go.
+ */
+#define OCFS2_MAX_TRANS_DATA 64U
/*
* Create access is for when we get a newly created buffer and we're
diff --git a/fs/ocfs2/ocfs2_trace.h b/fs/ocfs2/ocfs2_trace.h
index 3b481f490633..1b60c62aa9d6 100644
--- a/fs/ocfs2/ocfs2_trace.h
+++ b/fs/ocfs2/ocfs2_trace.h
@@ -2579,6 +2579,8 @@ DEFINE_OCFS2_INT_INT_EVENT(ocfs2_extend_trans);
DEFINE_OCFS2_INT_EVENT(ocfs2_extend_trans_restart);
+DEFINE_OCFS2_INT_INT_EVENT(ocfs2_allocate_extend_trans);
+
DEFINE_OCFS2_ULL_ULL_UINT_UINT_EVENT(ocfs2_journal_access);
DEFINE_OCFS2_ULL_EVENT(ocfs2_journal_dirty);
OpenPOWER on IntegriCloud