| Message ID | 1464943756-14143-15-git-send-email-famz@redhat.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 03.06.2016 10:49, Fam Zheng wrote: > In sync=none the backing image of s->target is s->common.bs, which could > be exclusively locked, the image locking wouldn't work here. > > Later we can update completion code to lock it after the replaced node > has dropped its lock. > > Signed-off-by: Fam Zheng <famz@redhat.com> > --- > blockdev.c | 10 ++++++++-- > 1 file changed, 8 insertions(+), 2 deletions(-) Without having reviewed the patch in the new context, when basing your series on v2 of my "block/mirror: Fix target backing BDS", all iotests pass even without this patch. Without my series, test 041 fails. So I have reason to hope that I was actually able to make this patch superfluous. Max
On Mon, 06/06 17:03, Max Reitz wrote: > On 03.06.2016 10:49, Fam Zheng wrote: > > In sync=none the backing image of s->target is s->common.bs, which could > > be exclusively locked, the image locking wouldn't work here. > > > > Later we can update completion code to lock it after the replaced node > > has dropped its lock. > > > > Signed-off-by: Fam Zheng <famz@redhat.com> > > --- > > blockdev.c | 10 ++++++++-- > > 1 file changed, 8 insertions(+), 2 deletions(-) > > Without having reviewed the patch in the new context, when basing your > series on v2 of my "block/mirror: Fix target backing BDS", all iotests > pass even without this patch. > > Without my series, test 041 fails. > > So I have reason to hope that I was actually able to make this patch > superfluous. Great, then I can drop this patch from v7 later. Fam
diff --git a/blockdev.c b/blockdev.c index fa8b50c..4a9fa7c 100644 --- a/blockdev.c +++ b/blockdev.c @@ -3620,8 +3620,14 @@ void qmp_drive_mirror(const char *device, const char *target, /* Mirroring takes care of copy-on-write using the source's backing * file. */ - target_bs = bdrv_open(target, NULL, options, flags | BDRV_O_NO_BACKING, - errp); + flags |= BDRV_O_NO_BACKING; + if (sync == MIRROR_SYNC_MODE_NONE) { + /* TODO: in mirror complete, after target_bs is switched to and the + * original BDS's lock is dropped, we should enable the lock on + * target_bs. */ + flags |= BDRV_O_NO_LOCK; + } + target_bs = bdrv_open(target, NULL, options, flags, errp); if (!target_bs) { goto out; }
In sync=none the backing image of s->target is s->common.bs, which could be exclusively locked, the image locking wouldn't work here. Later we can update completion code to lock it after the replaced node has dropped its lock. Signed-off-by: Fam Zheng <famz@redhat.com> --- blockdev.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)