diff options
| author | Tom Caputi <[email protected]> | 2018-04-10 14:15:05 -0400 |
|---|---|---|
| committer | Brian Behlendorf <[email protected]> | 2018-04-10 11:15:05 -0700 |
| commit | edc1e713c294d116702b034c59eed7b9a03fbc64 (patch) | |
| tree | 95423150271705e3015359791785f12221d896ff /module/zfs/dbuf.c | |
| parent | 10f88c5cd523eec8431941abf00017fc8fb67fd3 (diff) | |
Fix race in dnode_check_slots_free()
Currently, dnode_check_slots_free() works by checking dn->dn_type
in the dnode to determine if the dnode is reclaimable. However,
there is a small window of time between dnode_free_sync() in the
first call to dsl_dataset_sync() and when the useraccounting code
is run when the type is set DMU_OT_NONE, but the dnode is not yet
evictable, leading to crashes. This patch adds the ability for
dnodes to track which txg they were last dirtied in and adds a
check for this before performing the reclaim.
This patch also corrects several instances when dn_dirty_link was
treated as a list_node_t when it is technically a multilist_node_t.
Reviewed-by: Brian Behlendorf <[email protected]>
Signed-off-by: Tom Caputi <[email protected]>
Closes #7147
Closes #7388
Diffstat (limited to 'module/zfs/dbuf.c')
| -rw-r--r-- | module/zfs/dbuf.c | 3 |
1 files changed, 3 insertions, 0 deletions
diff --git a/module/zfs/dbuf.c b/module/zfs/dbuf.c index 18edd7834..1a6e560d2 100644 --- a/module/zfs/dbuf.c +++ b/module/zfs/dbuf.c @@ -1818,6 +1818,9 @@ dbuf_dirty(dmu_buf_impl_t *db, dmu_tx_t *tx) FTAG); } } + + if (tx->tx_txg > dn->dn_dirty_txg) + dn->dn_dirty_txg = tx->tx_txg; mutex_exit(&dn->dn_mtx); if (db->db_blkid == DMU_SPILL_BLKID) |