From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:54994) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1coQJc-0003xo-R2 for qemu-devel@nongnu.org; Thu, 16 Mar 2017 03:59:21 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1coQJb-0000ej-Ue for qemu-devel@nongnu.org; Thu, 16 Mar 2017 03:59:20 -0400 From: Juan Quintela In-Reply-To: <20170316013301.GB21865@lemon.lan> (Fam Zheng's message of "Thu, 16 Mar 2017 09:33:01 +0800") References: <1489549053-31728-1-git-send-email-jemmy858585@gmail.com> <20170315111021.GD3088@lemon.lan> <20170315173108.GB2701@work-vm> <20170316013301.GB21865@lemon.lan> Reply-To: quintela@redhat.com Date: Thu, 16 Mar 2017 08:59:11 +0100 Message-ID: <87inn9y7n4.fsf@secure.mitica> MIME-Version: 1.0 Content-Type: text/plain Subject: Re: [Qemu-devel] [PATCH v2] migration/block: Avoid invoking blk_drain too frequently List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Fam Zheng Cc: "Dr. David Alan Gilbert" , Lidong Chen , qemu-devel@nongnu.org, stefanha@redhat.com, qemu-block@nongnu.org Fam Zheng wrote: > On Wed, 03/15 17:31, Dr. David Alan Gilbert wrote: >> * Fam Zheng (famz@redhat.com) wrote: >> > On Wed, 03/15 11:37, Lidong Chen wrote: >> > > Increase bmds->cur_dirty after submit io, so reduce the frequency >> > > involve into blk_drain, and improve the performance obviously >> > > when block migration. >> > > >> > > The performance test result of this patch: >> > > >> > > During the block dirty save phase, this patch improve guest os IOPS >> > > from 4.0K to 9.5K. and improve the migration speed from >> > > 505856 rsec/s to 855756 rsec/s. >> > > >> > > Signed-off-by: Lidong Chen >> > > --- >> > > migration/block.c | 3 +++ >> > > 1 file changed, 3 insertions(+) >> > > >> > > diff --git a/migration/block.c b/migration/block.c >> > > index 6741228..7734ff7 100644 >> > > --- a/migration/block.c >> > > +++ b/migration/block.c >> > > @@ -576,6 +576,9 @@ static int mig_save_device_dirty(QEMUFile *f, BlkMigDevState *bmds, >> > > } >> > > >> > > bdrv_reset_dirty_bitmap(bmds->dirty_bitmap, sector, nr_sectors); >> > > + sector += nr_sectors; >> > > + bmds->cur_dirty = sector; >> > > + >> > > break; >> > > } >> > > sector += BDRV_SECTORS_PER_DIRTY_CHUNK; >> > > -- >> > > 1.8.3.1 >> > > >> > >> > Nice catch above all, thank you! >> > >> > Reviewed-by: Fam Zheng >> >> Are you taking that via a block pull? > > I can do that, but I'm not sure whether it should go to 2.9. This is a > performance improvement, which usually doesn't qualify as bug fixes. But this > also looks like a mistake in original code. > > Fam I am taking it through migration and push it. I agree with your description.