In case of -EAGAIN returned by update_refcount(), we should discard the
cluster offset we were trying to allocate and request a new one, because
in theory that old offset might now be taken by a refcount block.
In practice, this was not the case due to update_refcount() generally
returning strictly monotonic increasing cluster offsets. However, this
behavior is not set in stone, and it is also not obvious when looking at
qcow2_alloc_bytes() alone, so we should not rely on it.
Reported-by: Kevin Wolf <kw...@redhat.com>
Signed-off-by: Max Reitz <mre...@redhat.com>
---
block/qcow2-refcount.c | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/block/qcow2-refcount.c b/block/qcow2-refcount.c
index e8430ec..c30bb14 100644
--- a/block/qcow2-refcount.c
+++ b/block/qcow2-refcount.c
@@ -949,11 +949,17 @@ int64_t qcow2_alloc_bytes(BlockDriverState *bs, int size)
if (!offset || ROUND_UP(offset, s->cluster_size) != new_cluster) {
offset = new_cluster;
+ free_in_cluster = s->cluster_size;
+ } else {
+ free_in_cluster += s->cluster_size;
}
}
assert(offset);
ret = update_refcount(bs, offset, size, 1, false, QCOW2_DISCARD_NEVER);
+ if (ret < 0) {
+ offset = 0;
+ }
} while (ret == -EAGAIN);
if (ret < 0) {
return ret;
--
2.5.1
