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 X-Spam-Level: X-Spam-Status: No, score=-4.8 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS, MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED autolearn=no autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id E1CB5C433E1 for ; Thu, 27 Aug 2020 20:56:24 +0000 (UTC) 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 mail.kernel.org (Postfix) with ESMTPS id A8D0120737 for ; Thu, 27 Aug 2020 20:56:24 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="hUIvrghe" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org A8D0120737 Authentication-Results: mail.kernel.org; dmarc=fail (p=none dis=none) header.from=redhat.com Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Received: from localhost ([::1]:42448 helo=lists1p.gnu.org) by lists.gnu.org with esmtp (Exim 4.90_1) (envelope-from ) id 1kBOwZ-0003QF-TQ for qemu-devel@archiver.kernel.org; Thu, 27 Aug 2020 16:56:23 -0400 Received: from eggs.gnu.org ([2001:470:142:3::10]:54124) by lists.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256) (Exim 4.90_1) (envelope-from ) id 1kBOvu-0002zv-5m for qemu-devel@nongnu.org; Thu, 27 Aug 2020 16:55:42 -0400 Received: from us-smtp-delivery-124.mimecast.com ([63.128.21.124]:29707) by eggs.gnu.org with esmtps (TLS1.2:ECDHE_RSA_AES_256_CBC_SHA1:256) (Exim 4.90_1) (envelope-from ) id 1kBOvr-0003nq-Ij for qemu-devel@nongnu.org; Thu, 27 Aug 2020 16:55:41 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1598561738; 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: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=OIQfc/ICmkIcvXaBVLsSwMjGA1OZ8SgIJ6PkQxNsCFQ=; b=hUIvrgheBh/JcWAVDMc6A4nTSnvmE2rgHH+nPHylls+/NkpbllBzvqEyVwDjEKU9Rfvoum gxcQvWCyEzT5w/5gThGIW6Nj/nKiLOmE30/N5EJ7rnG+x3q8kzo68L/ASx4J9SMU5Sq7qw SkNJNCXMgnPqnAgWLIfkmIi/DsJrKos= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-160-QPLEqi4tMIalrePicSJuWQ-1; Thu, 27 Aug 2020 16:55:36 -0400 X-MC-Unique: QPLEqi4tMIalrePicSJuWQ-1 Received: from smtp.corp.redhat.com (int-mx03.intmail.prod.int.phx2.redhat.com [10.5.11.13]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 370C01029D22; Thu, 27 Aug 2020 20:55:35 +0000 (UTC) Received: from imammedo-mac (unknown [10.40.192.85]) by smtp.corp.redhat.com (Postfix) with ESMTP id 7269774E23; Thu, 27 Aug 2020 20:55:29 +0000 (UTC) Date: Thu, 27 Aug 2020 22:55:26 +0200 From: Igor Mammedov To: Eduardo Habkost Subject: Re: [PATCH v5 0/8] Remove EPYC mode apicid decode and use generic decode Message-ID: <20200827225526.0b1f6d32@imammedo-mac> In-Reply-To: <20200827190752.GK642093@habkost.net> References: <159804762216.39954.15502128500494116468.stgit@naples-babu.amd.com> <20200826143849.59f6970b@redhat.com> <20200826125059.GN168515@redhat.com> <20200826153034.115126cb@redhat.com> <20200826133638.GO168515@redhat.com> <20200826160258.0e9047f4@redhat.com> <20200826150340.GP168515@redhat.com> <20200827190314.717ec788@imammedo-mac> <20200827190752.GK642093@habkost.net> MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.13 Authentication-Results: relay.mimecast.com; auth=pass smtp.auth=CUSA124A263 smtp.mailfrom=imammedo@redhat.com X-Mimecast-Spam-Score: 0.002 X-Mimecast-Originator: redhat.com Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: quoted-printable Received-SPF: pass client-ip=63.128.21.124; envelope-from=imammedo@redhat.com; helo=us-smtp-delivery-124.mimecast.com X-detected-operating-system: by eggs.gnu.org: First seen = 2020/08/27 02:54:02 X-ACL-Warn: Detected OS = Linux 2.2.x-3.x [generic] [fuzzy] X-Spam_score_int: -30 X-Spam_score: -3.1 X-Spam_bar: --- X-Spam_report: (-3.1 / 5.0 requ) BAYES_00=-1.9, DKIMWL_WL_HIGH=-0.959, 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_H5=0.001, RCVD_IN_MSPIKE_WL=0.001, SPF_HELO_NONE=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: 2.1.23 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Cc: "Daniel P. =?UTF-8?Q?Berrang=C3=A9?=" , mst@redhat.com, Michal Privoznik , qemu-devel@nongnu.org, Babu Moger , pbonzini@redhat.com, rth@twiddle.net Errors-To: qemu-devel-bounces+qemu-devel=archiver.kernel.org@nongnu.org Sender: "Qemu-devel" On Thu, 27 Aug 2020 15:07:52 -0400 Eduardo Habkost wrote: > On Thu, Aug 27, 2020 at 07:03:14PM +0200, Igor Mammedov wrote: > > On Wed, 26 Aug 2020 16:03:40 +0100 > > Daniel P. Berrang=C3=A9 wrote: > >=20 > > > On Wed, Aug 26, 2020 at 04:02:58PM +0200, Igor Mammedov wrote: > > > > On Wed, 26 Aug 2020 14:36:38 +0100 > > > > Daniel P. Berrang=C3=A9 wrote: > > > >=20 > > > > > On Wed, Aug 26, 2020 at 03:30:34PM +0200, Igor Mammedov wrote: > > > > > > On Wed, 26 Aug 2020 13:50:59 +0100 > > > > > > Daniel P. Berrang=C3=A9 wrote: > > > > > > =20 > > > > > > > On Wed, Aug 26, 2020 at 02:38:49PM +0200, Igor Mammedov wrote= : =20 > > > > > > > > On Fri, 21 Aug 2020 17:12:19 -0500 > > > > > > > > Babu Moger wrote: > > > > > > > > =20 > > > > > > > > > To support some of the complex topology, we introduced EP= YC mode apicid decode. > > > > > > > > > But, EPYC mode decode is running into problems. Also it c= an become quite a > > > > > > > > > maintenance problem in the future. So, it was decided to = remove that code and > > > > > > > > > use the generic decode which works for majority of the to= pology. Most of the > > > > > > > > > SPECed configuration would work just fine. With some non-= SPECed user inputs, > > > > > > > > > it will create some sub-optimal configuration. > > > > > > > > > Here is the discussion thread. > > > > > > > > > https://lore.kernel.org/qemu-devel/c0bcc1a6-1d84-a6e7-e46= 8-d5b437c1b254@amd.com/ > > > > > > > > >=20 > > > > > > > > > This series removes all the EPYC mode specific apicid cha= nges and use the generic > > > > > > > > > apicid decode. =20 > > > > > > > >=20 > > > > > > > > the main difference between EPYC and all other CPUs is that > > > > > > > > it requires numa configuration (it's not optional) > > > > > > > > so we need an extra patch on top of this series to enfoce t= hat, i.e: > > > > > > > >=20 > > > > > > > > if (epyc && !numa)=20 > > > > > > > > error("EPYC cpu requires numa to be configured") =20 > > > > > > >=20 > > > > > > > Please no. This will break 90% of current usage of the EPYC C= PU in > > > > > > > real world QEMU deployments. That is way too user hostile to = introduce > > > > > > > as a requirement. > > > > > > >=20 > > > > > > > Why do we need to force this ? People have been successfuly = using > > > > > > > EPYC CPUs without NUMA in QEMU for years now. > > > > > > >=20 > > > > > > > It might not match behaviour of bare metal silicon, but that = hasn't > > > > > > > obviously caused the world to come crashing down. =20 > > > > > > So far it produces warning in linux kernel (RHBZ1728166), > > > > > > (resulting performance might be suboptimal), but I haven't seen > > > > > > anyone reporting crashes yet. > > > > > >=20 > > > > > >=20 > > > > > > What other options do we have? > > > > > > Perhaps we can turn on strict check for new machine types only, > > > > > > so old configs can keep broken topology (CPUID), > > > > > > while new ones would require -numa and produce correct topology= . =20 > > > > >=20 > > > > > No, tieing this to machine types is not viable either. That is st= ill > > > > > going to break essentially every single management application th= at > > > > > exists today using QEMU. > > > > for that we have deprecation process, so users could switch to new = CLI > > > > that would be required. > > >=20 > > > We could, but I don't find the cost/benefit tradeoff is compelling. > > >=20 > > > There are so many places where we diverge from what bare metal would > > > do, that I don't see a good reason to introduce this breakage, even > > > if we notify users via a deprecation message.=20 > > I find (3) and (4) good enough reasons to use deprecation. > >=20 > > > If QEMU wants to require NUMA for EPYC, then QEMU could internally > > > create a single NUMA node if none was specified for new machine > > > types, such that there is no visible change or breakage to any > > > mgmt apps. =20 > >=20 > > (1) for configs that started without -numa &&|| without -smp dies>1, > > QEMU can do just that (enable auto_enable_numa). >=20 > Why exactly do we need auto_enable_numa with dies=3D1? >=20 > If I understand correctly, Babu said earlier in this thread[1] > that we don't need auto_enable_numa. >=20 > [1] https://lore.kernel.org/qemu-devel/11489e5f-2285-ddb4-9c35-c9f522d603= a0@amd.com/ in case of 1 die, -numa is not must have as it's one numa node only. Though having auto_enable_numa, will allow to reuse the CPU.node-id propert= y to compose CPUID_Fn8000001E_ECX. i.e only code one path vs numa|non-numa va= riant. =20 > > (2) As for configs that are out of spec, I do not care much (junk in - = junk out) > > (though not having to spend time on bug reports and debug issues, just = to say > > it's not supported in the end, makes deprecation sound like a reasonabl= e > > choice) > >=20 > > (3) However if config matches bare metal i.e. CPU has more than 1 die a= nd within > > dies limits (spec wise), QEMU has to produce valid CPUs. > > In this case QEMU can't make up multiple numa nodes and mappings of RAM= /CPUs > > on user's behalf. That's where we have to error out and ask for explici= t > > numa configuration. > >=20 > > For such configs, current code (since 5.0), will produce in the best ca= se > > performance issues due to mismatching data in APICID, CPUID and ACPI t= ables, > > in the worst case issues might be related to invalid APIC ID if running= on EPYC host > > and HW takes in account subfields of APIC ID (according to Babu real CP= U uses > > die_id(aka node_id) internally). > > I'd rather error out on nonsense configs earlier than debug such issues > > and than error out anyways later (upsetting more users). > >=20 >=20 > The requirements are not clear to me. Is this just about making > CPU die_id match the NUMA node ID, or are there additional > constraints? die_id is per socket numa node index, so it's not numa node id in a sense we use it in qemu (that's where all the confusion started that led to current code) I understood that each die in EPYC chip is a numa node, which encodes NUMA node ID (system wide) in CPUID_Fn8000001E_ECX, that's why I wrote earlier that EPYC makes -numa non optional. In case of only one die we can either use auto_enable_numa to ensure that we have consistent code or special case it and just hardcode=20 CPUID_Fn8000001E_ECX value which is hackish but will let us avoid enabling numa (explicitly or implictly). in case of multiple dies, CPUID_Fn8000001E_ECX (encodes number of nodes + systemwide numa node id looking at CPUID of real EPYC machine) shall match -numa mapping (otherwise it's a bug where CPUID and ACPI mismatch). Here we can go to ways: 1) ask user to provide sane config with -numa (I'd prefer that) and use that info to fill in CPUID_Fn8000001E_ECX 2) pretend that it's non numa machine, skip ACPI SRAT table but make up CPUID_Fn8000001E (i.e. another special case) (requires another code path and addition to -numa one) >=20 >=20 > > (4) > > If I were non hobby user, I'd hate if QEMU allowed me to start invalid = config, > > that I'd have to spend time on debugging issues (including performance = ones), > > instead of clearly telling me what's wrong and how config should be cor= rected. > > I'd probably jump to another hypervisor that does the job right, > > instead of digging into QEMU codebase and CPU specs to figure out how > > to hack and configure it. > >=20 >=20