diff options
author | Isaac Huang <[email protected]> | 2015-02-26 22:46:45 -0700 |
---|---|---|
committer | Brian Behlendorf <[email protected]> | 2015-03-02 11:50:06 -0800 |
commit | d14cfd83dae0b1a261667acd416dba17a98d15fa (patch) | |
tree | b5052cb4d9fcd75ef873fbe767222774ae183b43 | |
parent | 87a63dd702370c00322e8d1d84102075c775324b (diff) |
Fix deadlock between zpool export and zfs list
Pool reference count is NOT checked in spa_export_common()
if the pool has been imported readonly=on, i.e. spa->spa_sync_on
is FALSE. Then zpool export and zfs list may deadlock:
1. Pool A is imported readonly.
2. zpool export A and zfs list are run concurrently.
3. zfs command gets reference on the spa, which holds a dbuf on
on the MOS meta dnode.
4. zpool command grabs spa_namespace_lock, and tries to evict dbufs
of the MOS meta dnode. The dbuf held by zfs command can't be
evicted as its reference count is not 0.
5. zpool command blocks in dnode_special_close() waiting for the
MOS meta dnode reference count to drop to 0, with
spa_namespace_lock held.
6. zfs command tries to get the spa_namespace_lock with a reference
on the spa held, which holds a dbuf on the MOS meta dnode.
7. Now zpool command and zfs command deadlock each other.
Also any further zfs/zpool command will block on spa_namespace_lock
forever.
The fix is to always check pool reference count in spa_export_common(),
no matter whether the pool was imported readonly or not.
Signed-off-by: Isaac Huang <[email protected]>
Signed-off-by: Brian Behlendorf <[email protected]>
Closes #2034
-rw-r--r-- | module/zfs/spa.c | 39 |
1 files changed, 20 insertions, 19 deletions
diff --git a/module/zfs/spa.c b/module/zfs/spa.c index 861a319af..81cef1952 100644 --- a/module/zfs/spa.c +++ b/module/zfs/spa.c @@ -4262,30 +4262,30 @@ spa_export_common(char *pool, int new_state, nvlist_t **oldconfig, mutex_enter(&spa_namespace_lock); spa_close(spa, FTAG); + if (spa->spa_state == POOL_STATE_UNINITIALIZED) + goto export_spa; /* - * The pool will be in core if it's openable, - * in which case we can modify its state. + * The pool will be in core if it's openable, in which case we can + * modify its state. Objsets may be open only because they're dirty, + * so we have to force it to sync before checking spa_refcnt. */ - if (spa->spa_state != POOL_STATE_UNINITIALIZED && spa->spa_sync_on) { - /* - * Objsets may be open only because they're dirty, so we - * have to force it to sync before checking spa_refcnt. - */ + if (spa->spa_sync_on) txg_wait_synced(spa->spa_dsl_pool, 0); - /* - * A pool cannot be exported or destroyed if there are active - * references. If we are resetting a pool, allow references by - * fault injection handlers. - */ - if (!spa_refcount_zero(spa) || - (spa->spa_inject_ref != 0 && - new_state != POOL_STATE_UNINITIALIZED)) { - spa_async_resume(spa); - mutex_exit(&spa_namespace_lock); - return (SET_ERROR(EBUSY)); - } + /* + * A pool cannot be exported or destroyed if there are active + * references. If we are resetting a pool, allow references by + * fault injection handlers. + */ + if (!spa_refcount_zero(spa) || + (spa->spa_inject_ref != 0 && + new_state != POOL_STATE_UNINITIALIZED)) { + spa_async_resume(spa); + mutex_exit(&spa_namespace_lock); + return (SET_ERROR(EBUSY)); + } + if (spa->spa_sync_on) { /* * A pool cannot be exported if it has an active shared spare. * This is to prevent other pools stealing the active spare @@ -4314,6 +4314,7 @@ spa_export_common(char *pool, int new_state, nvlist_t **oldconfig, } } +export_spa: spa_event_notify(spa, NULL, FM_EREPORT_ZFS_POOL_DESTROY); if (spa->spa_state != POOL_STATE_UNINITIALIZED) { |