Skip to contentSkip to conversation

blk-mq: Fix a use-after-free

PATCH New Bart Van Assche 1 participant 1 message Oct 30, 2014 Block layer

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 blk-mq: Fix a use-after-free

Since the tag set structure can be embedded in a structure that is freed before the last queue put occurs (e.g. the SCSI host structure), the block layer must not access the tag set after the SCSI host structure has been freed. This patch avoids that the following kernel oops is triggered with memory poisoning enabled:

general protection fault: 0000 [#1] SMP
RAX: 0000000000000000 RBX: 6b6b6b6b6b6b6b6b RCX: 0000000000000000
Call Trace:
 [<ffffffff8109a7c4>] lock_acquire+0xc4/0x270
 [<ffffffff814ce111>] mutex_lock_nested+0x61/0x380
 [<ffffffff812575f0>] blk_mq_free_queue+0x30/0x180
 [<ffffffff8124d654>] blk_release_queue+0x84/0xd0
 [<ffffffff8126c29b>] kobject_cleanup+0x7b/0x1a0
 [<ffffffff8126c140>] kobject_put+0x30/0x70
 [<ffffffff81245895>] blk_put_queue+0x15/0x20
 [<ffffffff8125c409>] disk_release+0x99/0xd0
 [<ffffffff8133d056>] device_release+0x36/0xb0
 [<ffffffff8126c29b>] kobject_cleanup+0x7b/0x1a0
 [<ffffffff8126c140>] kobject_put+0x30/0x70
 [<ffffffff8125a78a>] put_disk+0x1a/0x20
 [<ffffffff811d4cb5>] __blkdev_put+0x135/0x1b0
 [<ffffffff811d56a0>] blkdev_put+0x50/0x160
 [<ffffffff81199eb4>] kill_block_super+0x44/0x70
 [<ffffffff8119a2a4>] deactivate_locked_super+0x44/0x60
 [<ffffffff8119a87e>] deactivate_super+0x4e/0x70
 [<ffffffff811b9833>] cleanup_mnt+0x43/0x90
 [<ffffffff811b98d2>] __cleanup_mnt+0x12/0x20
 [<ffffffff8107252c>] task_work_run+0xac/0xe0
 [<ffffffff81002c01>] do_notify_resume+0x61/0xa0
 [<ffffffff814d2c58>] int_signal+0x12/0x17
Cc Christoph Hellwig <[email protected]>
Cc Robert Elliott <[email protected]>
Cc Ming Lei <[email protected]>
Cc <[email protected]> # v3.12+
---
 block/blk-core.c  | 3 +++
 block/blk-sysfs.c | 4 +---
 2 files changed, 4 insertions(+), 3 deletions(-)
block/blk-core.c +3 -0
index 0421b53..93f9152 100644
@@ -525,6 +525,9 @@ void blk_cleanup_queue(struct request_queue *q)
 	del_timer_sync(&q->backing_dev_info.laptop_mode_wb_timer);
 	blk_sync_queue(q);
 
+	if (q->mq_ops)
+		blk_mq_free_queue(q);
+
 	spin_lock_irq(lock);
 	if (q->queue_lock != &q->__queue_lock)
 		q->queue_lock = &q->__queue_lock;
block/blk-sysfs.c +1 -3
index 1fac434..3ce0da1 100644
@@ -517,9 +517,7 @@ static void blk_release_queue(struct kobject *kobj)
 	if (q->queue_tags)
 		__blk_queue_free_tags(q);
 
-	if (q->mq_ops)
-		blk_mq_free_queue(q);
-	else
+	if (!q->mq_ops)
 		blk_free_flush_queue(q->fq);
 
 	blk_trace_shutdown(q);
Signature · 1 line
1.8.4.5