[Qemu-devel] [PATCH] qcow2: Fix in-flight list after qcow2_cache_put failure

[Qemu-devel] [PATCH] qcow2: Fix in-flight list after qcow2_cache_put failure

[Kevin Wolf]

If qcow2_cache_put returns an error during cluster allocation and the
allocation fails, it must be removed from the list of in-flight allocations.
Otherwise we'd get a loop in the list when the ACB is used for the next
allocation.

Luckily, this qcow2_cache_put shouldn't fail anyway because the L2 table is
only read, so that qcow2_cache_put doesn't even involve I/O.

Signed-off-by: Kevin Wolf <kwolf@redhat.com>
---
 block/qcow2-cluster.c |    5 +++--
 1 files changed, 3 insertions(+), 2 deletions(-)

diff --git a/block/qcow2-cluster.c b/block/qcow2-cluster.c
index 750abe3..c56651c 100644
--- a/block/qcow2-cluster.c
+++ b/block/qcow2-cluster.c
@@ -812,7 +812,6 @@ int qcow2_alloc_cluster_offset(BlockDriverState *bs, uint64_t offset,
 
     cluster_offset = qcow2_alloc_clusters(bs, nb_clusters * s->cluster_size);
     if (cluster_offset < 0) {
-        QLIST_REMOVE(m, next_in_flight);
         ret = cluster_offset;
         goto fail;
     }
@@ -825,7 +824,7 @@ int qcow2_alloc_cluster_offset(BlockDriverState *bs, uint64_t offset,
 out:
     ret = qcow2_cache_put(bs, s->l2_table_cache, (void**) &l2_table);
     if (ret < 0) {
-        return ret;
+        goto fail_put;
     }
 
     m->nb_available = MIN(nb_clusters << (s->cluster_bits - 9), n_end);
@@ -837,6 +836,8 @@ out:
 
 fail:
     qcow2_cache_put(bs, s->l2_table_cache, (void**) &l2_table);
+fail_put:
+    QLIST_REMOVE(m, next_in_flight);
     return ret;
 }
 
-- 
1.7.2.3

Re: [Qemu-devel] [PATCH] qcow2: Fix in-flight list after qcow2_cache_put failure

[Christoph Hellwig]

On Tue, May 24, 2011 at 04:47:19PM +0200, Kevin Wolf wrote:
> If qcow2_cache_put returns an error during cluster allocation and the
> allocation fails, it must be removed from the list of in-flight allocations.
> Otherwise we'd get a loop in the list when the ACB is used for the next
> allocation.
> 
> Luckily, this qcow2_cache_put shouldn't fail anyway because the L2 table is
> only read, so that qcow2_cache_put doesn't even involve I/O.

Looks good,

Reviewed-by: Christoph Hellwig <hch@lst.de>