From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:45244) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WDfoD-0003X1-BC for qemu-devel@nongnu.org; Wed, 12 Feb 2014 14:49:30 -0500 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1WDfo8-0006TE-5t for qemu-devel@nongnu.org; Wed, 12 Feb 2014 14:49:25 -0500 Received: from paradis.irqsave.net ([62.212.105.220]:49813) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1WDfo7-0006Sx-Rb for qemu-devel@nongnu.org; Wed, 12 Feb 2014 14:49:20 -0500 Date: Wed, 12 Feb 2014 20:49:18 +0100 From: =?iso-8859-1?Q?Beno=EEt?= Canet Message-ID: <20140212194918.GC4225@irqsave.net> References: <1392068957-14184-1-git-send-email-benoit.canet@irqsave.net> <1392068957-14184-2-git-send-email-benoit.canet@irqsave.net> <20140211081217.GB12675@T430.nay.redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline In-Reply-To: <20140211081217.GB12675@T430.nay.redhat.com> Content-Transfer-Encoding: quoted-printable Subject: Re: [Qemu-devel] [FIX V2] block: Fix device snapshots broken by the block filter snapshots patchset. List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Fam Zheng Cc: =?iso-8859-1?Q?Beno=EEt?= Canet , kwolf@redhat.com, armbru@redhat.com, jcody@redhat.com, qemu-devel@nongnu.org, mreitz@redhat.com, stefanha@redhat.com The Tuesday 11 Feb 2014 =E0 16:12:17 (+0800), Fam Zheng wrote : > On Mon, 02/10 22:49, Beno=EEt Canet wrote: > > Take into account the fact that a block filter like quorum will be in= bs->file > > while a regular block driver device is really on the top level. > >=20 > > Signed-off-by: Benoit Canet > > --- > > block.c | 14 ++++++++------ > > 1 file changed, 8 insertions(+), 6 deletions(-) > >=20 > > diff --git a/block.c b/block.c > > index 07ac50a..d04f535 100644 > > --- a/block.c > > +++ b/block.c > > @@ -5400,14 +5400,16 @@ bool bdrv_is_first_non_filter(BlockDriverStat= e *candidate) > > =20 > > /* walk down the bs forest recursively */ > > QTAILQ_FOREACH(bs, &bdrv_states, device_list) { > > - bool perm; > > - > > - if (!bs->file) { > > - continue; > > + bool perm =3D false; > > + > > + if (bs->file && > > + bs->file->drv && > > + bs->file->drv->authorizations[BS_IS_A_FILTER]) { > > + perm =3D bdrv_recurse_is_first_non_filter(bs->file, cand= idate); > > + } else if (bs =3D=3D candidate) { > > + perm =3D true; > > } > > =20 > > - perm =3D bdrv_recurse_is_first_non_filter(bs->file, candidat= e); > > - > > /* candidate is the first non filter */ > > if (perm) { > > return true; >=20 > With this change, if the top level driver has ->file, its implementatio= n of > .bdrv_recurse_is_first_non_filter() is skipped and bs->file is the star= t point. >=20 > So we have an implication that single child block drivers (that has ->f= ile) > doesn't need to, and shouldn't implement this operation, as commentted = above > bdrv_generic_is_first_non_filter. >=20 > Tested that this patch fixes the external snapshot problem, but didn't = test > the "quorum as bs->file case". >=20 > Thanks, >=20 > Reviewed-by: Fam Zheng After extensive testing of all type of quorum instanciations and snapshot= s I discovered that we are not done yet with this issue. When instantiating quorum from the command line the quorum driver is in bs->file->drv. When using QMP's blockdev_add at once or by using references the quorum d= river is in bs->drv. As a result this patch work in half the cases: regular file and command l= ine quorum but not when quorum is instantiated via QMP. I don't understand why the block layer have this irregular behavior regar= ding quorum instantiation. What would be the best way to fix this ? Best regards Beno=EEt