[1/1] block: fix libvirt snapshot with existing bitmaps
diff mbox

Message ID 1465821535-6664-1-git-send-email-den@openvz.org
State New
Headers show

Commit Message

Denis V. Lunev June 13, 2016, 12:38 p.m. UTC
From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>

Fix the following bug:

 # virsh start test
 Domain test started

 #  virsh qemu-monitor-command test \
     '{"execute":"block-dirty-bitmap-add",\
      "arguments":{"node":"drive0","name":"ab"}}'
 {"return":{},"id":"libvirt-36"}'}'

 # virsh snapshot-create test
 error: Unable to read from monitor: Connection reset by peer

Actually, assert "assert(pos < hb->size)" in hbitmap_iter_init fires,
because qcow2_save_vmstate just writes to bs (not to bs->file->bs) after
the end of the drive.

Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
Signed-off-by: Denis V. Lunev <den@openvz.org>
CC: Kevin Wolf <kwolf@redhat.com>
CC: Max Reitz <mreitz@redhat.com>
---
 block/dirty-bitmap.c | 6 ++++++
 1 file changed, 6 insertions(+)

Comments

Max Reitz June 14, 2016, 2:12 p.m. UTC | #1
On 13.06.2016 14:38, Denis V. Lunev wrote:
> From: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
> 
> Fix the following bug:
> 
>  # virsh start test
>  Domain test started
> 
>  #  virsh qemu-monitor-command test \
>      '{"execute":"block-dirty-bitmap-add",\
>       "arguments":{"node":"drive0","name":"ab"}}'
>  {"return":{},"id":"libvirt-36"}'}'
> 
>  # virsh snapshot-create test
>  error: Unable to read from monitor: Connection reset by peer
> 
> Actually, assert "assert(pos < hb->size)" in hbitmap_iter_init fires,
> because qcow2_save_vmstate just writes to bs (not to bs->file->bs) after
> the end of the drive.
> 
> Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@virtuozzo.com>
> Signed-off-by: Denis V. Lunev <den@openvz.org>
> CC: Kevin Wolf <kwolf@redhat.com>
> CC: Max Reitz <mreitz@redhat.com>
> ---
>  block/dirty-bitmap.c | 6 ++++++
>  1 file changed, 6 insertions(+)
> 
> diff --git a/block/dirty-bitmap.c b/block/dirty-bitmap.c
> index 4902ca5..7b636ee 100644
> --- a/block/dirty-bitmap.c
> +++ b/block/dirty-bitmap.c
> @@ -364,6 +364,11 @@ void bdrv_set_dirty(BlockDriverState *bs, int64_t cur_sector,
>                      int nr_sectors)
>  {
>      BdrvDirtyBitmap *bitmap;
> +
> +    if (cur_sector >= bdrv_nb_sectors(bs)) {
> +        return;
> +    }

I'd use bitmap->size instead of bdrv_nb_sectors(bs).

In addition, I'd put an assert(cur_sector + nr_sectors <= bitmap->size)
after this conditional block; I understand that we will probably never
write to both the disk and the VM state in a single operation, therefore
asserting this is sufficient.

(If we want to do it right, we'd need to truncate nr_sectors in case
cur_sector < bdrv_nb_sectors(bs) && cur_sector + nr_sectors > bitmap->size)

> +
>      QLIST_FOREACH(bitmap, &bs->dirty_bitmaps, list) {
>          if (!bdrv_dirty_bitmap_enabled(bitmap)) {
>              continue;
> 

Apart from that: I see that this is the only place where it's really
relevant. However, I feel somewhat bad putting it just here and not in
the other functions (bdrv_{re,}set_dirty_bitmap() and bdrv_get_dirty()).

I just want to bring this to your attention. Technically, putting the
check into this function is completely sufficient, so if you think it's
too much work to put it into the other ones, I'm fine with that
(although I think we should at least put an assertion into those other
functions).

Max

Patch
diff mbox

diff --git a/block/dirty-bitmap.c b/block/dirty-bitmap.c
index 4902ca5..7b636ee 100644
--- a/block/dirty-bitmap.c
+++ b/block/dirty-bitmap.c
@@ -364,6 +364,11 @@  void bdrv_set_dirty(BlockDriverState *bs, int64_t cur_sector,
                     int nr_sectors)
 {
     BdrvDirtyBitmap *bitmap;
+
+    if (cur_sector >= bdrv_nb_sectors(bs)) {
+        return;
+    }
+
     QLIST_FOREACH(bitmap, &bs->dirty_bitmaps, list) {
         if (!bdrv_dirty_bitmap_enabled(bitmap)) {
             continue;