From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([140.186.70.92]:44890) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1RIGG3-0003Re-T4 for qemu-devel@nongnu.org; Mon, 24 Oct 2011 04:51:48 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1RIGG2-0001w4-QT for qemu-devel@nongnu.org; Mon, 24 Oct 2011 04:51:47 -0400 Received: from mx1.redhat.com ([209.132.183.28]:10484) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1RIGG2-0001vx-JT for qemu-devel@nongnu.org; Mon, 24 Oct 2011 04:51:46 -0400 Message-ID: <4EA527DC.90902@redhat.com> Date: Mon, 24 Oct 2011 10:54:52 +0200 From: Kevin Wolf MIME-Version: 1.0 References: <1319216912-26964-1-git-send-email-kwolf@redhat.com> <4EA1BD95.8030205@redhat.com> <4EA515B9.8070204@redhat.com> <4EA51963.5060904@redhat.com> <4EA51F16.3030700@redhat.com> <4EA5262D.7090901@redhat.com> In-Reply-To: <4EA5262D.7090901@redhat.com> Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: 7bit Subject: Re: [Qemu-devel] [PATCH 0/2] block: Write out internal caches even with cache=unsafe List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Paolo Bonzini Cc: Alexander Graf , qemu-devel@nongnu.org, avi@redhat.com Am 24.10.2011 10:47, schrieb Paolo Bonzini: > On 10/24/2011 10:17 AM, Kevin Wolf wrote: >>> I think it's not about "why is it there", but rather about "what is it >>> useful for". My interpretation of it is "I do not need the image >>> anymore unless the command exits cleanly": VM installations, qemu-img >>> conversions, BDRV_O_SNAPSHOT (doesn't do it yet, but it could). Even >>> SIGINT and SIGTERM would be excluded from this definition, but they cost >>> nothing so it's nice to include them. >> >> I think another common interpretation is: "I don't run this VM in >> production but for development. I want the VM to go faster and I can >> recreate the image in the unlikely event that power fails during my >> work. But it certainly would be nasty." > > Fair enough. > >> But I think that starting to make exceptions for single block drivers >> isn't a good idea anyway. If we want bdrv_flush() to write out all >> metadata internal to qemu, I think the approach with checking the flag >> in drivers calling things like fsync() is better. The common thing is to >> do the flush. > > I don't know... checking BDRV_O_NO_FLUSH in the drivers rather than in > the generic code sounds like a layering violation. Perhaps what you're > after is a separation of bdrv_co_flush from bdrv_{,co_,aio_}fsync? Then > BDRV_O_NO_FLUSH (better renamed to BDRV_O_NO_FSYNC...) would only > inhibit the latter. Why? All other cache related BDRV_O_* flags are interpreted by the block drivers, so why should BDRV_O_NO_FLUSH be special? Kevin