From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.gnu.org (lists.gnu.org [209.51.188.17]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id F2931F483E4 for ; Mon, 23 Mar 2026 18:52:27 +0000 (UTC) Received: from localhost ([::1] helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1w4kNB-0003yd-6d; Mon, 23 Mar 2026 14:51:33 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1w4kN5-0003sN-FQ for qemu-devel@nongnu.org; Mon, 23 Mar 2026 14:51:27 -0400 Received: from us-smtp-delivery-124.mimecast.com ([170.10.129.124]) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1w4kN2-0007Ks-MR for qemu-devel@nongnu.org; Mon, 23 Mar 2026 14:51:27 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1774291883; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: in-reply-to:in-reply-to:references:references; bh=oZ0P3F0LTW02JQ3N8Byicn6pS46ujKJnI4Cb5Xq9t5M=; b=bNp3MBACSHrFry5K8YXyPvlz3SQ918vI2HQO7Y+zVEgiHl9yxKSCWK35meEwD+2ciDIgF3 /hjpWIdKBzv1ImJ3GJ4UGdaxsSryXoR7I05ZxjV2zDbBR04rFylcyOjoZqL+yhiHgt4UUz egxWz+jCQUai7WGelG70Mrqcsahtv0I= Received: from mx-prod-mc-08.mail-002.prod.us-west-2.aws.redhat.com (ec2-35-165-154-97.us-west-2.compute.amazonaws.com [35.165.154.97]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-621-0C0dw-XJPE2K_PbdyFnDWg-1; Mon, 23 Mar 2026 14:51:16 -0400 X-MC-Unique: 0C0dw-XJPE2K_PbdyFnDWg-1 X-Mimecast-MFC-AGG-ID: 0C0dw-XJPE2K_PbdyFnDWg_1774291874 Received: from mx-prod-int-05.mail-002.prod.us-west-2.aws.redhat.com (mx-prod-int-05.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.17]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mx-prod-mc-08.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS id A7598180034F; Mon, 23 Mar 2026 18:51:13 +0000 (UTC) Received: from localhost (unknown [10.44.32.50]) by mx-prod-int-05.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTP id 1AD161955D71; Mon, 23 Mar 2026 18:51:11 +0000 (UTC) Date: Mon, 23 Mar 2026 14:51:09 -0400 From: Stefan Hajnoczi To: JAEHOON KIM Cc: qemu-devel@nongnu.org, qemu-block@nongnu.org, pbonzini@redhat.com, fam@euphon.net, armbru@redhat.com, eblake@redhat.com, berrange@redhat.com, eduardo@habkost.net, dave@treblig.org, sw@weilnetz.de Subject: Re: [PATCH RFC v1 0/3] aio-poll: improve aio-polling efficiency Message-ID: <20260323185109.GB631609@fedora> References: <20260113174824.464720-1-jhkim@linux.ibm.com> <20260219222717.GA1011077@fedora> <79c89de8-68f4-46ff-bcce-dcb817d04ad5@linux.ibm.com> <5a50568c-d7c8-4938-b399-b304c162f6f2@linux.ibm.com> <63173e58-6c6e-4f69-a80d-8ddc4b372198@linux.ibm.com> MIME-Version: 1.0 Content-Type: multipart/signed; micalg=pgp-sha512; protocol="application/pgp-signature"; boundary="Wrk7U+tg6x4DUmcA" Content-Disposition: inline In-Reply-To: <63173e58-6c6e-4f69-a80d-8ddc4b372198@linux.ibm.com> X-Scanned-By: MIMEDefang 3.0 on 10.30.177.17 Received-SPF: pass client-ip=170.10.129.124; envelope-from=stefanha@redhat.com; helo=us-smtp-delivery-124.mimecast.com X-Spam_score_int: -20 X-Spam_score: -2.1 X-Spam_bar: -- X-Spam_report: (-2.1 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.001, DKIM_SIGNED=0.1, DKIM_VALID=-0.1, DKIM_VALID_AU=-0.1, DKIM_VALID_EF=-0.1, RCVD_IN_DNSWL_NONE=-0.0001, RCVD_IN_MSPIKE_H3=0.001, RCVD_IN_MSPIKE_WL=0.001, RCVD_IN_VALIDITY_RPBL_BLOCKED=0.001, RCVD_IN_VALIDITY_SAFE_BLOCKED=0.001, SPF_HELO_PASS=-0.001, SPF_PASS=-0.001 autolearn=ham autolearn_force=no X-Spam_action: no action X-BeenThere: qemu-devel@nongnu.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: qemu development List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org --Wrk7U+tg6x4DUmcA Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Mon, Mar 23, 2026 at 09:08:13AM -0500, JAEHOON KIM wrote: > On 3/9/2026 3:46 PM, JAEHOON KIM wrote: > > On 2/26/2026 12:03 AM, JAEHOON KIM wrote: > > > On 2/20/2026 1:00 PM, JAEHOON KIM wrote: > > > > On 2/19/2026 4:27 PM, Stefan Hajnoczi wrote: > > > > > Hi Jaehoon, > > > > > Following the call earlier this week I ran a single fio job to ge= t a > > > > > clearer picture of: > > > > > 1. The QEMU 10.0.0 regression that prompted you to optimize AioCo= ntext > > > > > =C2=A0=C2=A0=C2=A0 polling. > > > > > 2. How the poll-weight parameter affects IOPS. > > > > >=20 > > > > > run=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 rw=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0 bs=C2=A0=C2=A0 numjobs iothreads iops=C2=A0=C2=A0 diff > > > > > v9.2.0=C2=A0=C2=A0 randread=C2=A0 8k=C2=A0=C2=A0 1=C2=A0=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 1=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 174= 944 3.6% > > > > > v10.0.0=C2=A0 randread=C2=A0 8k=C2=A0=C2=A0 1=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0 1=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 174285= 3.2% > > > > > baseline randread=C2=A0 8k=C2=A0=C2=A0 1=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0 1=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 168908 0.0% > > > > > w2=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 randread=C2=A0 8k=C2=A0=C2= =A0 1=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 1=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0 163718 -3.1% > > > > > w3=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 randread=C2=A0 8k=C2=A0=C2= =A0 1=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 1=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0 165805 -1.8% > > > > > w4=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 randread=C2=A0 8k=C2=A0=C2= =A0 1=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 1=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0 167388 -0.9% > > > > >=20 > > > > > This time I only ran randread bs=3D8k iodepth=3D8 numjobs=3D1 wit= h a single > > > > > IOThread. > > > > >=20 > > > > > Observations: > > > > >=20 > > > > > - There might be an IOPS regression between v10.0.0 and the basel= ine > > > > > =C2=A0=C2=A0 (9ad7f544c696) that your patches apply on top of. Th= is is different > > > > > =C2=A0=C2=A0 from the CPU utilization regression that you found i= n v9.2.0 -> > > > > > =C2=A0=C2=A0 v10.0.0. I will bisect it. > > > > >=20 > > > > > - poll-weight=3D3 and 4 improve IOPS to a level that is acceptabl= e. CPU > > > > > =C2=A0=C2=A0 utilization looks like this: > > > > >=20 > > > > > run=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 %usr=C2=A0=C2= =A0=C2=A0=C2=A0 %nice=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 %sys=C2=A0=C2=A0 %iowai= t=C2=A0=C2=A0=C2=A0 %steal > > > > > %irq=C2=A0=C2=A0=C2=A0=C2=A0 %soft=C2=A0=C2=A0=C2=A0 %guest=C2=A0= =C2=A0=C2=A0 %gnice=C2=A0=C2=A0=C2=A0=C2=A0 %idle > > > > > baseline=C2=A0=C2=A0 49.37=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.00=C2= =A0=C2=A0=C2=A0=C2=A0 31.10=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.00=C2=A0=C2=A0= =C2=A0=C2=A0=C2=A0 0.00 > > > > > 11.61=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.04=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0 7.89 > > > > > w2=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 46.24=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0 32.61=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.00 > > > > > 11.84=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.10=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0 9.21 > > > > > w3=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 48.04=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0 32.17=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.00 > > > > > 11.98=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.08=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0 7.73 > > > > > w4=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 48.56=C2=A0=C2= =A0=C2=A0=C2=A0=C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0 31.23=C2=A0=C2=A0=C2=A0= =C2=A0=C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.00 > > > > > 11.48=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.03=C2=A0=C2=A0=C2=A0=C2=A0= =C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 0.00=C2=A0=C2=A0=C2=A0=C2=A0=C2= =A0 8.69 > > > > >=20 > > > > > poll-weight=3D2 is the winner at CPU utilization. I'm not sure if > > > > > poll-weight=3D3 will produce an acceptable CPU utilization > > > > > improvement for > > > > > you. Do you have data or want to re-run to measure poll-weight=3D= 3? > > > > >=20 > > > > > Stefan > > > >=20 > > > > Thank you very much for sharing the detailed measurement results. > > > > I truly appreciate the effort. > > > >=20 > > > > Regarding w=3D3, I will discuss with our performance team to see > > > > if the CPU consumption levels are acceptable within our internal > > > > test environment. I will get back to you with more definitive data > > > > as soon as possible. > > > >=20 > > > > Thanks again for your thorough analysis. > > > >=20 > > > > Regards, > > > > Jaehoon. > > > >=20 > > > Hello Stefan, > > >=20 > > > Thank you for your patience. I would like to share the observed > > > changes in throughput > > > and CPU consumption in our performance test environment as follows. > > >=20 > > > We observed that when using W=3D3, performance returns to a level > > > comparable to > > > QEMU v9.1, while W=3D2 results in slightly lower CPU consumption. > > >=20 > > > Our initial preference is to use W=3D2 by default. > > > However, we fully understand your concerns regarding the potential > > > performance drop. > > > Should the patch be accepted, we would provide guidance on using the > > > weighted value as > > > a configurable option. > > >=20 > > > For reference, we consider values between -2 and 2 as noise in our > > > analysis. > > > I look forward to your feedback. > > >=20 > > > The table below shows a comparison between: > > > *=C2=A0 - Host:* RHEL10.1-GA+qemu-10.0.0-14.el10_1, *Guest:* RHEL 9.6= GA vs. > > > *=C2=A0 - Host:*=C2=A0RHEL10.1-GA+qemu-10.0.0-14.el10_1 (w=3D2, w=3D3= ), *Guest:* > > > RHEL 9.6 GA > > > =C2=A0 =C2=A0 for FIO FCP and FICON with 1 iothread and 8 iothreads. > > > =C2=A0 =C2=A0 The values shown are the averages for numjobs 1, 4, and= 8. > > >=20 > > > FIO FCP -1 iothread > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0| Seq. Read=C2=A0 | Seq. Read=C2=A0 | Seq. Write | Seq. > > > Write | Rand. Read | Rand. Read | Rand. Write| Rand. Write| > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 =C2= =A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0| =C2=A0w3 > > > (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)= =C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 = =C2=A0| > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > | Throughput avg=C2=A0 =C2=A0 =C2=A0 |=C2=A0 =C2=A0-3.00=C2=A0 =C2=A0= |=C2=A0 =C2=A0-2.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0 0.00=C2=A0 =C2=A0 | > > > =C2=A0-0.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-3.00=C2=A0 =C2=A0 |=C2=A0 =C2= =A0-3.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0 1.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.33= =C2=A0 =C2=A0 | > > > | CPU consumption avg |=C2=A0 =C2=A0-5.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0= -4.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-6.33=C2=A0 =C2=A0 | > > > =C2=A0-5.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-7.33=C2=A0 =C2=A0 |=C2=A0 =C2= =A0-5.33=C2=A0 =C2=A0 |=C2=A0 -10.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-8.67=C2=A0= =C2=A0 | > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > >=20 > > > FIO FCP -8 iothread > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0| Seq. Read=C2=A0 | Seq. Read=C2=A0 | Seq. Write | Seq. > > > Write | Rand. Read | Rand. Read | Rand. Write| Rand. Write| > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 =C2= =A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0| =C2=A0w3 > > > (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)= =C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 = =C2=A0| > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > | Throughput avg=C2=A0 =C2=A0 =C2=A0 |=C2=A0 =C2=A0-5.00=C2=A0 =C2=A0= |=C2=A0 =C2=A0-4.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-3.67=C2=A0 =C2=A0 | > > > =C2=A0-3.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-4.67=C2=A0 =C2=A0 |=C2=A0 =C2= =A0-4.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.67= =C2=A0 =C2=A0 | > > > | CPU consumption avg |=C2=A0 -13.00=C2=A0 =C2=A0 |=C2=A0 -10.67=C2= =A0 =C2=A0 |=C2=A0 -16.00=C2=A0 =C2=A0 | > > > -14.33=C2=A0 =C2=A0 |=C2=A0 -14.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-9.33= =C2=A0 =C2=A0 |=C2=A0 -13.67=C2=A0 =C2=A0 |=C2=A0 -11.00=C2=A0 =C2=A0 | > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > >=20 > > >=20 > > > FIO FICON -1 iothread > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0| Seq. Read=C2=A0 | Seq. Read=C2=A0 | Seq. Write | Seq. > > > Write | Rand. Read | Rand. Read | Rand. Write| Rand. Write| > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 =C2= =A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0| =C2=A0w3 > > > (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)= =C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 = =C2=A0| > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > | Throughput avg=C2=A0 =C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.67=C2=A0 =C2=A0= |=C2=A0 =C2=A0-0.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0-6.33=C2=A0 =C2=A0 | > > > =C2=A0-6.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.67=C2=A0 =C2=A0 |=C2=A0 =C2= =A0 0.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0 1.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0 1.33= =C2=A0 =C2=A0 | > > > | CPU consumption avg |=C2=A0 =C2=A0-7.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0= -7.33=C2=A0 =C2=A0 |=C2=A0 -13.67=C2=A0 =C2=A0 | > > > -13.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-9.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-8= =2E33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-5.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-4.33=C2= =A0 =C2=A0 | > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > >=20 > > > FIO FICON -8 iothread > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0| Seq. Read=C2=A0 | Seq. Read=C2=A0 | Seq. Write | Seq. > > > Write | Rand. Read | Rand. Read | Rand. Write| Rand. Write| > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 =C2= =A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0| =C2=A0w3 > > > (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)= =C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 = =C2=A0| > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > | Throughput avg=C2=A0 =C2=A0 =C2=A0 |=C2=A0 =C2=A0-3.00=C2=A0 =C2=A0= |=C2=A0 =C2=A0-2.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0-7.33=C2=A0 =C2=A0 | > > > =C2=A0-7.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.67=C2=A0 =C2=A0 |=C2=A0 =C2= =A0-1.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0 0.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0 0.67= =C2=A0 =C2=A0 | > > > | CPU consumption avg |=C2=A0 -16.33=C2=A0 =C2=A0 |=C2=A0 -14.33=C2= =A0 =C2=A0 |=C2=A0 -25.33=C2=A0 =C2=A0 | > > > -27.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-8.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0-7= =2E00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-6.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0-5.00=C2= =A0 =C2=A0 | > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > >=20 > > >=20 > > > The table below shows a comparison between: > > > *=C2=A0- Host:* RHEL 10.0 GA + qemu-9.1.0-15.el10, *Guest:* RHEL 9.6 = GA vs. > > > *=C2=A0- Host:* RHEL 10.1 GA + qemu-10.0.0-14.el10_1 (w=3D2, w=3D3), = *Guest:* > > > RHEL 9.6 GA. > > >=20 > > > FIO FCP -1 iothread > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0| Seq. Read=C2=A0 | Seq. Read=C2=A0 | Seq. Write | Seq. > > > Write | Rand. Read | Rand. Read | Rand. Write| Rand. Write| > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 =C2= =A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0| =C2=A0w3 > > > (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)= =C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 = =C2=A0| > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > | Throughput avg=C2=A0 =C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.67=C2=A0 =C2=A0= |=C2=A0 =C2=A0 0.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.33=C2=A0 =C2=A0 | > > > =C2=A0-1.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-1.00=C2=A0 =C2=A0 |=C2=A0 =C2= =A0-0.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0 3.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0 2.00= =C2=A0 =C2=A0 | > > > | CPU consumption avg |=C2=A0 =C2=A0 0.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0= 2.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0 1.67=C2=A0 =C2=A0 | 3.00=C2=A0 > > > =C2=A0 |=C2=A0 =C2=A0-3.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.33=C2=A0 =C2= =A0 |=C2=A0 =C2=A0-2.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0 0.00=C2=A0 =C2=A0 | > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > >=20 > > > FIO FCP -8 iothread > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0| Seq. Read=C2=A0 | Seq. Read=C2=A0 | Seq. Write | Seq. > > > Write | Rand. Read | Rand. Read | Rand. Write| Rand. Write| > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 =C2= =A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0| =C2=A0w3 > > > (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)= =C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 = =C2=A0| > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > | Throughput avg=C2=A0 =C2=A0 =C2=A0 |=C2=A0 =C2=A0-2.00=C2=A0 =C2=A0= |=C2=A0 =C2=A0-1.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-1.00=C2=A0 =C2=A0 | 0.00= =C2=A0 > > > =C2=A0 |=C2=A0 =C2=A0-0.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0 1.00=C2=A0 =C2= =A0 |=C2=A0 =C2=A0 1.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0 0.67=C2=A0 =C2=A0 | > > > | CPU consumption avg |=C2=A0 =C2=A0-3.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0= -1.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-2.00=C2=A0 =C2=A0 | 0.00=C2=A0 > > > =C2=A0 |=C2=A0 -10.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-5.67=C2=A0 =C2=A0 |= =C2=A0 =C2=A0-6.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0-3.33=C2=A0 =C2=A0 | > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > >=20 > > >=20 > > > FIO FICON -1 iothread > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0| Seq. Read=C2=A0 | Seq. Read=C2=A0 | Seq. Write | Seq. > > > Write | Rand. Read | Rand. Read | Rand. Write| Rand. Write| > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 =C2= =A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0| =C2=A0w3 > > > (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)= =C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 = =C2=A0| > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > | Throughput avg=C2=A0 =C2=A0 =C2=A0 |=C2=A0 =C2=A0-1.67=C2=A0 =C2=A0= |=C2=A0 =C2=A0-1.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.33=C2=A0 =C2=A0 | 0.00= =C2=A0 > > > =C2=A0 |=C2=A0 =C2=A0-1.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-1.00=C2=A0 =C2= =A0 |=C2=A0 =C2=A0-1.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0-2.33=C2=A0 =C2=A0 | > > > | CPU consumption avg |=C2=A0 =C2=A0-0.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0= 1.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0 1.00=C2=A0 =C2=A0 | 2.00=C2=A0 > > > =C2=A0 |=C2=A0 =C2=A0-2.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-1.33=C2=A0 =C2= =A0 |=C2=A0 =C2=A0-1.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.33=C2=A0 =C2=A0 | > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > >=20 > > > FIO FICON -8 iothread > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0| Seq. Read=C2=A0 | Seq. Read=C2=A0 | Seq. Write | Seq. > > > Write | Rand. Read | Rand. Read | Rand. Write| Rand. Write| > > > |=C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 =C2= =A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0| =C2=A0w3 > > > (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)= =C2=A0 =C2=A0|=C2=A0 =C2=A0w2 (%)=C2=A0 =C2=A0|=C2=A0 =C2=A0w3 (%)=C2=A0 = =C2=A0| > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > > | Throughput avg=C2=A0 =C2=A0 =C2=A0 |=C2=A0 =C2=A0-1.00=C2=A0 =C2=A0= |=C2=A0 =C2=A0-0.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0-0.33=C2=A0 =C2=A0 | 0.33= =C2=A0 > > > =C2=A0 |=C2=A0 =C2=A0 0.67=C2=A0 =C2=A0 |=C2=A0 =C2=A0 0.33=C2=A0 =C2= =A0 |=C2=A0 =C2=A0 0.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0 0.33=C2=A0 =C2=A0 | > > > | CPU consumption avg |=C2=A0 =C2=A0-1.33=C2=A0 =C2=A0 |=C2=A0 =C2=A0= 1.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0 4.33=C2=A0 =C2=A0 | 2.33=C2=A0 > > > =C2=A0 |=C2=A0 =C2=A0-2.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0 0.00=C2=A0 =C2= =A0 |=C2=A0 =C2=A0-3.00=C2=A0 =C2=A0 |=C2=A0 =C2=A0-1.33=C2=A0 =C2=A0 | > > > +---------------------+------------+------------+------------+-------= -----+------------+------------+------------+------------+ > > >=20 > > >=20 > > >=20 > > > Regards, > > > Jaehoon. > > >=20 > > >=20 > > Hello Stefan, > >=20 > > I=E2=80=99m following up to see if you=E2=80=99ve had a chance to revie= w the performance > > results I shared. > > As I mentioned, we are completely fine with using w=3D3 as the default > > value. > > It effectively restores performance to QEMU v9.1 levels, and we will > > provide guidance > > for users who need further CPU savings to use w=3D2. > >=20 > > I=E2=80=99ve updated the patch to set the default value to 3 and includ= e the > > feedback so far. > > Please let me know if you have any comments or if I should post the new > > patch. > >=20 > > Regards, > > Jaehoon > >=20 > >=20 > Hello, >=20 > I'd like to follow up on this thread. I have posted a v2 incorporating the > feedback so far. > I'd really appreciate your thoughts when you have a chance. >=20 > Thanks again for your time. Hi Jaehoon, Sorry that I haven't replied yet. I will review your emails and the v2 patches tomorrow. Stefan --Wrk7U+tg6x4DUmcA Content-Type: application/pgp-signature; name=signature.asc -----BEGIN PGP SIGNATURE----- iQEzBAEBCgAdFiEEhpWov9P5fNqsNXdanKSrs4Grc8gFAmnBi50ACgkQnKSrs4Gr c8gnmAgAg3miVGla9LThE7mwTfweZa9D7sMmsRTdnOQWvYQjoiAlSA5XRwa/vYXM FM3Z8wZ0VfOC9ScjYwztaNW5v53UMCljB5omueDkI7Pa2pJM0Da3uy+9f7OXJi9Y DfG6v/c+rCl/xbZzfUxjfQNUUpVSPoLg/3iitgC/RSBguZFPmQD33Oe0oWkQmnpy x9/CtzynxhRIjjH+6d2fOuNzuzKIDm9AO6Lvaa8DE6cQ9uYQEJWr8wYBkaOt87XM HciY05wkysrqjAnv2u/jIPwffcEca+eSVuD9FiHZQ3i6DE7OT9/F/cu8uyXKttzx 9OXNXOw/IwBLOFpN28EQ0rTQG/ky5g== =qNIp -----END PGP SIGNATURE----- --Wrk7U+tg6x4DUmcA--