[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[Qemu-devel] [PULL 23/23] qcow2: Make qcow2_alloc_bytes() more explicit
From: |
Kevin Wolf |
Subject: |
[Qemu-devel] [PULL 23/23] qcow2: Make qcow2_alloc_bytes() more explicit |
Date: |
Fri, 11 Sep 2015 21:41:03 +0200 |
From: Max Reitz <address@hidden>
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 <address@hidden>
Signed-off-by: Max Reitz <address@hidden>
Signed-off-by: Kevin Wolf <address@hidden>
---
block/qcow2-refcount.c | 6 ++++++
1 file changed, 6 insertions(+)
diff --git a/block/qcow2-refcount.c b/block/qcow2-refcount.c
index 5f67798..3579c4d 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;
--
1.8.3.1
- [Qemu-devel] [PULL 07/23] block: Allow specifying driver-specific options to reopen, (continued)
- [Qemu-devel] [PULL 07/23] block: Allow specifying driver-specific options to reopen, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 09/23] qemu-io: Add command 'reopen', Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 10/23] qcow2: Improve error message, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 12/23] qcow2: Move qcow2_update_options() call up, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 11/23] qcow2: Factor out qcow2_update_options(), Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 06/23] qcow2: Rename BDRVQcowState to BDRVQcow2State, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 15/23] qcow2: Fix memory leak in qcow2_update_options() error path, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 13/23] qcow2: Move rest of option handling to qcow2_update_options(), Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 14/23] qcow2: Leave s unchanged on qcow2_update_options() failure, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 22/23] vmdk: Fix next_cluster_sector for compressed write, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 23/23] qcow2: Make qcow2_alloc_bytes() more explicit,
Kevin Wolf <=
- [Qemu-devel] [PULL 21/23] iotests: Add test for checking large image files, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 16/23] qcow2: Make qcow2_update_options() suitable for transactions, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 18/23] qemu-iotests: Reopen qcow2 with lazy-refcounts change, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 17/23] qcow2: Support updating driver-specific options in reopen, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 20/23] qcow2: Make size_to_clusters() return uint64_t, Kevin Wolf, 2015/09/11
- [Qemu-devel] [PULL 19/23] qemu-iotests: More qcow2 reopen tests, Kevin Wolf, 2015/09/11
- Re: [Qemu-devel] [PULL 00/23] Block layer patches, Peter Maydell, 2015/09/14