Permalink
Browse files

Fix issue with scanning dedup blocks as scan ends

This patch fixes an issue discovered by ztest where
dsl_scan_ddt_entry() could add I/Os to the dsl scan queues
between when the scan had finished all required work and
when the scan was marked as complete. This caused the scan
to spin indefinitely without ending.

Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov>
Reviewed-by: Serapheim Dimitropoulos <serapheim.dimitro@delphix.com>
Reviewed-by: Matthew Ahrens <mahrens@delphix.com>
Signed-off-by: Tom Caputi <tcaputi@datto.com>
Closes #8010
  • Loading branch information...
tcaputi authored and BrainSlayer committed Oct 18, 2018
1 parent f016fae commit fabc099a3331cb91126de29fc7adf9e54d174741
Showing with 16 additions and 0 deletions.
  1. +16 −0 module/zfs/dsl_scan.c
@@ -2386,6 +2386,20 @@ dsl_scan_ddt_entry(dsl_scan_t *scn, enum zio_checksum checksum,
if (!dsl_scan_is_running(scn))
return;
/*
* This function is special because it is the only thing
* that can add scan_io_t's to the vdev scan queues from
* outside dsl_scan_sync(). For the most part this is ok
* as long as it is called from within syncing context.
* However, dsl_scan_sync() expects that no new sio's will
* be added between when all the work for a scan is done
* and the next txg when the scan is actually marked as
* completed. This check ensures we do not issue new sio's
* during this period.
*/
if (scn->scn_done_txg != 0)
return;
for (p = 0; p < DDT_PHYS_TYPES; p++, ddp++) {
if (ddp->ddp_phys_birth == 0 ||
ddp->ddp_phys_birth > scn->scn_phys.scn_max_txg)
@@ -3468,6 +3482,8 @@ dsl_scan_sync(dsl_pool_t *dp, dmu_tx_t *tx)
(longlong_t)tx->tx_txg);
}
} else if (scn->scn_is_sorted && scn->scn_bytes_pending != 0) {
ASSERT(scn->scn_clearing);
/* need to issue scrubbing IOs from per-vdev queues */
scn->scn_zio_root = zio_root(dp->dp_spa, NULL,
NULL, ZIO_FLAG_CANFAIL);

0 comments on commit fabc099

Please sign in to comment.