Prevent race in blkptr_verify against device removal

When we check the vdev of the blkptr in zfs_blkptr_verify, we can run 
into a race condition where that vdev is temporarily unavailable. This 
happens when a device removal operation and the old vdev_t has been 
removed from the array, but the new indirect vdev has not yet been 
inserted.

We hold the spa_config_lock while doing our sensitive verification. 
To ensure that we don't deadlock, we only grab the lock if we don't 
have config_writer held. In addition, I had to const the tags of the 
refcounts and the spa_config_lock arguments.

Reviewed-by: Brian Behlendorf <[email protected]>
Reviewed-by: Serapheim Dimitropoulos <[email protected]>
Signed-off-by: Paul Dagnelie <[email protected]>
Closes #9112 

3 files changed, 23 insertions, 16 deletions

diff --git a/module/zfs/refcount.c b/module/zfs/refcount.c
index 89528e6d3..a612b2f40 100644
--- a/module/zfs/refcount.c
+++ b/module/zfs/refcount.c
@@ -121,7 +121,7 @@ zfs_refcount_count(zfs_refcount_t *rc)
 }
 
 int64_t
-zfs_refcount_add_many(zfs_refcount_t *rc, uint64_t number, void *holder)
+zfs_refcount_add_many(zfs_refcount_t *rc, uint64_t number, const void *holder)
 {
 	reference_t *ref = NULL;
 	int64_t count;
@@ -143,13 +143,14 @@ zfs_refcount_add_many(zfs_refcount_t *rc, uint64_t number, void *holder)
 }
 
 int64_t
-zfs_refcount_add(zfs_refcount_t *rc, void *holder)
+zfs_refcount_add(zfs_refcount_t *rc, const void *holder)
 {
 	return (zfs_refcount_add_many(rc, 1, holder));
 }
 
 int64_t
-zfs_refcount_remove_many(zfs_refcount_t *rc, uint64_t number, void *holder)
+zfs_refcount_remove_many(zfs_refcount_t *rc, uint64_t number,
+    const void *holder)
 {
 	reference_t *ref;
 	int64_t count;
@@ -197,7 +198,7 @@ zfs_refcount_remove_many(zfs_refcount_t *rc, uint64_t number, void *holder)
 }
 
 int64_t
-zfs_refcount_remove(zfs_refcount_t *rc, void *holder)
+zfs_refcount_remove(zfs_refcount_t *rc, const void *holder)
 {
 	return (zfs_refcount_remove_many(rc, 1, holder));
 }
@@ -235,7 +236,7 @@ zfs_refcount_transfer(zfs_refcount_t *dst, zfs_refcount_t *src)
 
 void
 zfs_refcount_transfer_ownership_many(zfs_refcount_t *rc, uint64_t number,
-    void *current_holder, void *new_holder)
+    const void *current_holder, const void *new_holder)
 {
 	reference_t *ref;
 	boolean_t found = B_FALSE;
@@ -260,8 +261,8 @@ zfs_refcount_transfer_ownership_many(zfs_refcount_t *rc, uint64_t number,
 }
 
 void
-zfs_refcount_transfer_ownership(zfs_refcount_t *rc, void *current_holder,
-    void *new_holder)
+zfs_refcount_transfer_ownership(zfs_refcount_t *rc, const void *current_holder,
+    const void *new_holder)
 {
 	return (zfs_refcount_transfer_ownership_many(rc, 1, current_holder,
 	    new_holder));
@@ -273,7 +274,7 @@ zfs_refcount_transfer_ownership(zfs_refcount_t *rc, void *current_holder,
  * might be held.
  */
 boolean_t
-zfs_refcount_held(zfs_refcount_t *rc, void *holder)
+zfs_refcount_held(zfs_refcount_t *rc, const void *holder)
 {
 	reference_t *ref;
 
@@ -301,7 +302,7 @@ zfs_refcount_held(zfs_refcount_t *rc, void *holder)
  * since the reference might not be held.
  */
 boolean_t
-zfs_refcount_not_held(zfs_refcount_t *rc, void *holder)
+zfs_refcount_not_held(zfs_refcount_t *rc, const void *holder)
 {
 	reference_t *ref;
 
diff --git a/module/zfs/spa_misc.c b/module/zfs/spa_misc.c
index 3d6c0ee3e..d998fe225 100644
--- a/module/zfs/spa_misc.c
+++ b/module/zfs/spa_misc.c
@@ -484,7 +484,7 @@ spa_config_tryenter(spa_t *spa, int locks, void *tag, krw_t rw)
 }
 
 void
-spa_config_enter(spa_t *spa, int locks, void *tag, krw_t rw)
+spa_config_enter(spa_t *spa, int locks, const void *tag, krw_t rw)
 {
 	int wlocks_held = 0;
 
@@ -517,7 +517,7 @@ spa_config_enter(spa_t *spa, int locks, void *tag, krw_t rw)
 }
 
 void
-spa_config_exit(spa_t *spa, int locks, void *tag)
+spa_config_exit(spa_t *spa, int locks, const void *tag)
 {
 	for (int i = SCL_LOCKS - 1; i >= 0; i--) {
 		spa_config_lock_t *scl = &spa->spa_config_lock[i];
diff --git a/module/zfs/zio.c b/module/zfs/zio.c
index b740afde6..7b4ab341f 100644
--- a/module/zfs/zio.c
+++ b/module/zfs/zio.c
@@ -881,8 +881,8 @@ zio_root(spa_t *spa, zio_done_func_t *done, void *private, enum zio_flag flags)
 	return (zio_null(NULL, spa, NULL, done, private, flags));
 }
 
-void
-zfs_blkptr_verify(spa_t *spa, const blkptr_t *bp)
+static void
+zfs_blkptr_verify(spa_t *spa, const blkptr_t *bp, boolean_t config_held)
 {
 	if (!DMU_OT_IS_VALID(BP_GET_TYPE(bp))) {
 		zfs_panic_recover("blkptr at %p has invalid TYPE %llu",
@@ -921,6 +921,10 @@ zfs_blkptr_verify(spa_t *spa, const blkptr_t *bp)
 	if (!spa->spa_trust_config)
 		return;
 
+	if (!config_held)
+		spa_config_enter(spa, SCL_VDEV, bp, RW_READER);
+	else
+		ASSERT(spa_config_held(spa, SCL_VDEV, RW_WRITER));
 	/*
 	 * Pool-specific checks.
 	 *
@@ -969,6 +973,8 @@ zfs_blkptr_verify(spa_t *spa, const blkptr_t *bp)
 			    bp, i, (longlong_t)offset);
 		}
 	}
+	if (!config_held)
+		spa_config_exit(spa, SCL_VDEV, bp);
 }
 
 boolean_t
@@ -1008,7 +1014,7 @@ zio_read(zio_t *pio, spa_t *spa, const blkptr_t *bp,
 {
 	zio_t *zio;
 
-	zfs_blkptr_verify(spa, bp);
+	zfs_blkptr_verify(spa, bp, flags & ZIO_FLAG_CONFIG_WRITER);
 
 	zio = zio_create(pio, spa, BP_PHYSICAL_BIRTH(bp), bp,
 	    data, size, size, done, private,
@@ -1101,7 +1107,7 @@ void
 zio_free(spa_t *spa, uint64_t txg, const blkptr_t *bp)
 {
 
-	zfs_blkptr_verify(spa, bp);
+	zfs_blkptr_verify(spa, bp, B_FALSE);
 
 	/*
 	 * The check for EMBEDDED is a performance optimization.  We
@@ -1171,7 +1177,7 @@ zio_claim(zio_t *pio, spa_t *spa, uint64_t txg, const blkptr_t *bp,
 {
 	zio_t *zio;
 
-	zfs_blkptr_verify(spa, bp);
+	zfs_blkptr_verify(spa, bp, flags & ZIO_FLAG_CONFIG_WRITER);
 
 	if (BP_IS_EMBEDDED(bp))
 		return (zio_null(pio, spa, NULL, NULL, NULL, 0));