Archive-Team
/
zfs
2
0
Fork 0
Code Issues Pull Requests Packages Projects Releases 103 Wiki Activity

zil_fail: handle the case where no itxs are found 

This is possible if spa_sync() completes before the ZIL write/flush
does, which then errors. At this point all itxs are in the past, leaving
us with nothing to wait for.

In a perfect world we would not fail the ZIL, but at this point we've
already locked out the itxgs early, so we have to see it through.

Signed-off-by: Rob Norris <rob.norris@klarasystems.com>
This commit is contained in:
Rob Norris 2023-08-25 15:02:11 +10:00 committed by Geoff Amey
parent e10da1db53
commit 27e39dd59a
1 changed files with 9 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
module/zfs/zil.c
View File

@ -1239,10 +1239,16 @@ zil_fail(zilog_t *zilog)
ASSERT(list_is_empty(&zilog->zl_itx_commit_list));
/*
* If this fails, then we didn't take any itxs at all. If that's true,
* how did we end up here?
* Its possible that we took no itxs if all the work was committed to
* the pool before the ZIL write errored; in that case there's only
* commit waiters left on the lists, and we just completed those.
*
* We've already flagged the itxs for failure though, so we still have
* to fail the whole ZIL. We'll just set its highest to be the last
* synced txg, so it can immediately reopen if the circumstances suit.
*/
ASSERT3U(fail_itxg->itxg_txg, <, UINT64_MAX);
if (fail_itxg->itxg_txg == UINT64_MAX)
fail_itxg->itxg_txg = highest_txg;
/* Prepare the live itxgs for failure */
for (int i = 0; i < TXG_SIZE; i++) {