[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PATCH v3 10/19] block-backend: Fix potential double blk_de
From: |
Kevin Wolf |
Subject: |
[Qemu-devel] [PATCH v3 10/19] block-backend: Fix potential double blk_delete() |
Date: |
Thu, 20 Sep 2018 18:19:49 +0200 |
blk_unref() first decreases the refcount of the BlockBackend and calls
blk_delete() if the refcount reaches zero. Requests can still be in
flight at this point, they are only drained during blk_delete():
At this point, arbitrary callbacks can run. If any callback takes a
temporary BlockBackend reference, it will first increase the refcount to
1 and then decrease it to 0 again, triggering another blk_delete(). This
will cause a use-after-free crash in the outer blk_delete().
Fix it by draining the BlockBackend before decreasing to refcount to 0.
Assert in blk_ref() that it never takes the first refcount (which would
mean that the BlockBackend is already being deleted).
Signed-off-by: Kevin Wolf <address@hidden>
Reviewed-by: Fam Zheng <address@hidden>
Reviewed-by: Max Reitz <address@hidden>
---
block/block-backend.c | 9 ++++++++-
1 file changed, 8 insertions(+), 1 deletion(-)
diff --git a/block/block-backend.c b/block/block-backend.c
index efa8d8011c..63eb2d10cd 100644
--- a/block/block-backend.c
+++ b/block/block-backend.c
@@ -435,6 +435,7 @@ int blk_get_refcnt(BlockBackend *blk)
*/
void blk_ref(BlockBackend *blk)
{
+ assert(blk->refcnt > 0);
blk->refcnt++;
}
@@ -447,7 +448,13 @@ void blk_unref(BlockBackend *blk)
{
if (blk) {
assert(blk->refcnt > 0);
- if (!--blk->refcnt) {
+ if (blk->refcnt > 1) {
+ blk->refcnt--;
+ } else {
+ blk_drain(blk);
+ /* blk_drain() cannot resurrect blk, nobody held a reference */
+ assert(blk->refcnt == 1);
+ blk->refcnt = 0;
blk_delete(blk);
}
}
--
2.13.6
- [Qemu-devel] [PATCH v3 03/19] aio-wait: Increase num_waiters even in home thread, (continued)
- [Qemu-devel] [PATCH v3 03/19] aio-wait: Increase num_waiters even in home thread, Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 05/19] test-blockjob: Acquire AioContext around job_cancel_sync(), Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 04/19] test-bdrv-drain: Drain with block jobs in an I/O thread, Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 06/19] job: Use AIO_WAIT_WHILE() in job_finish_sync(), Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 09/19] block-backend: Add .drained_poll callback, Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 11/19] block-backend: Decrease in_flight only after callback, Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 08/19] block: Add missing locking in bdrv_co_drain_bh_cb(), Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 12/19] blockjob: Lie better in child_job_drained_poll(), Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 13/19] block: Remove aio_poll() in bdrv_drain_poll variants, Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 10/19] block-backend: Fix potential double blk_delete(),
Kevin Wolf <=
- [Qemu-devel] [PATCH v3 14/19] test-bdrv-drain: Test nested poll in bdrv_drain_poll_top_level(), Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 07/19] test-bdrv-drain: Test AIO_WAIT_WHILE() in completion callback, Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 15/19] job: Avoid deadlocks in job_completed_txn_abort(), Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 16/19] test-bdrv-drain: AIO_WAIT_WHILE() in job .commit/.abort, Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 17/19] test-bdrv-drain: Fix outdated comments, Kevin Wolf, 2018/09/20
- [Qemu-devel] [PATCH v3 19/19] test-bdrv-drain: Test draining job source child and parent, Kevin Wolf, 2018/09/20