Prevent deadlock in arc_read in Linux memory reclaim callback

Using zfs with Lustre, an arc_read can trigger kernel memory allocation
that in turn leads to a memory reclaim callback and a deadlock within a
single zfs process. This change uses spl_fstrans_mark and
spl_trans_unmark to prevent the reclaim attempt and the deadlock
(https://zfsonlinux.topicbox.com/groups/zfs-devel/T4db2c705ec1804ba).
The stack trace observed is:

    __schedule at ffffffff81610f2e
    schedule at ffffffff81611558
    schedule_preempt_disabled at ffffffff8161184a
    __mutex_lock at ffffffff816131e8
    arc_buf_destroy at ffffffffa0bf37d7 [zfs]
    dbuf_destroy at ffffffffa0bfa6fe [zfs]
    dbuf_evict_one at ffffffffa0bfaa96 [zfs]
    dbuf_rele_and_unlock at ffffffffa0bfa561 [zfs]
    dbuf_rele_and_unlock at ffffffffa0bfa32b [zfs]
    osd_object_delete at ffffffffa0b64ecc [osd_zfs]
    lu_object_free at ffffffffa06d6a74 [obdclass]
    lu_site_purge_objects at ffffffffa06d7fc1 [obdclass]
    lu_cache_shrink_scan at ffffffffa06d81b8 [obdclass]
    shrink_slab at ffffffff811ca9d8
    shrink_node at ffffffff811cfd94
    do_try_to_free_pages at ffffffff811cfe63
    try_to_free_pages at ffffffff811d01c4
    __alloc_pages_slowpath at ffffffff811be7f2
    __alloc_pages_nodemask at ffffffff811bf3ed
    new_slab at ffffffff81226304
    ___slab_alloc at ffffffff812272ab
    __slab_alloc at ffffffff8122740c
    kmem_cache_alloc at ffffffff81227578
    spl_kmem_cache_alloc at ffffffffa048a1fd [spl]
    arc_buf_alloc_impl at ffffffffa0befba2 [zfs]
    arc_read at ffffffffa0bf0924 [zfs]
    dbuf_read at ffffffffa0bf9083 [zfs]
    dmu_buf_hold_by_dnode at ffffffffa0c04869 [zfs]

Reviewed-by: Brian Behlendorf <[email protected]>
Signed-off-by: Mark Roper <[email protected]>
Closes #9987 

1 files changed, 12 insertions, 0 deletions

diff --git a/module/zfs/arc.c b/module/zfs/arc.c
index d49d85db0..6c9164f76 100644
--- a/module/zfs/arc.c
+++ b/module/zfs/arc.c
@@ -5540,6 +5540,17 @@ arc_read(zio_t *pio, spa_t *spa, const blkptr_t *bp,
 	ASSERT(!BP_IS_HOLE(bp));
 	ASSERT(!BP_IS_REDACTED(bp));
 
+	/*
+	 * Normally SPL_FSTRANS will already be set since kernel threads which
+	 * expect to call the DMU interfaces will set it when created.  System
+	 * calls are similarly handled by setting/cleaning the bit in the
+	 * registered callback (module/os/.../zfs/zpl_*).
+	 *
+	 * External consumers such as Lustre which call the exported DMU
+	 * interfaces may not have set SPL_FSTRANS.  To avoid a deadlock
+	 * on the hash_lock always set and clear the bit.
+	 */
+	fstrans_cookie_t cookie = spl_fstrans_mark();
 top:
 	if (!embedded_bp) {
 		/*
@@ -6014,6 +6025,7 @@ out:
 	/* embedded bps don't actually go to disk */
 	if (!embedded_bp)
 		spa_read_history_add(spa, zb, *arc_flags);
+	spl_fstrans_unmark(cookie);
 	return (rc);
 }