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=-0.8 required=3.0 tests=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 AFBFEC32751 for ; Wed, 31 Jul 2019 21:35:17 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 91F8E206B8 for ; Wed, 31 Jul 2019 21:35:17 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729669AbfGaVfQ (ORCPT ); Wed, 31 Jul 2019 17:35:16 -0400 Received: from mail-qk1-f195.google.com ([209.85.222.195]:39780 "EHLO mail-qk1-f195.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1728292AbfGaVfP (ORCPT ); Wed, 31 Jul 2019 17:35:15 -0400 Received: by mail-qk1-f195.google.com with SMTP id w190so50324396qkc.6 for ; Wed, 31 Jul 2019 14:35:15 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:message-id:subject:from:to:cc:date:in-reply-to :references:organization:user-agent:mime-version :content-transfer-encoding; bh=Yt3D9L2XmMJM/hxeEeDRbs7n5AVGloPaWA/yadI+Qrg=; b=l0yVw79e6RDA45Pcs7O8N+uLcWaYEyY2CCK5ORDRjJGrYdfOTv4fVUudOqEEaW48at 6+SkoVRAwxK5RLRnP8VyG9Zbk8gyO+EmtdfVC8YnvZoYR4W+4bxTh+uCHBdGMi+mV4xt yjQCRfgZGJaVc5o3ueJVj5No0sxrKYKz7iUPGou9V4/We4jQ4RkeEGG5SQJuA5A4fKR8 mJJ2AvP3nqzY3EFH9gwR2inUDsv80QprDNaDyVoOEKYeP20h0yrDJO94Mf99OkESi1CI 5A3bbx+jTmzYlrlfTnMqNxJKoflPQVIZc6ndFC0jthFXSa39PpIYtSLcmU+iEl8qT3vW WhrQ== X-Gm-Message-State: APjAAAXiEZ8Uj/1FLERw0+AOdQb3x6lk/HgrRUWwmysaQWcdgIywcHye pj05j+7E2Yn/ri+cM+mg3ffh8Q== X-Google-Smtp-Source: APXvYqwq8iZNfTZjhmXSZpr7SbYmuaQHxbwUhmb4FHrOYyaB4ijjIYu+VCEGw1Epf8zn+a0otAy3Vw== X-Received: by 2002:a05:620a:1411:: with SMTP id d17mr77658658qkj.137.1564608914930; Wed, 31 Jul 2019 14:35:14 -0700 (PDT) Received: from dhcp-10-20-1-11.bss.redhat.com ([144.121.20.162]) by smtp.gmail.com with ESMTPSA id e8sm28934259qkn.95.2019.07.31.14.35.12 (version=TLS1_3 cipher=AEAD-AES256-GCM-SHA384 bits=256/256); Wed, 31 Jul 2019 14:35:13 -0700 (PDT) Message-ID: Subject: Re: [PATCH] Revert "PCI: Enable NVIDIA HDA controllers" From: Lyude Paul To: Karol Herbst , Lukas Wunner Cc: nouveau , Linux PCI , Daniel Drake , Bjorn Helgaas , Aaron Plattner , Peter Wu , Ilia Mirkin , Maik Freudenberg , LKML Date: Wed, 31 Jul 2019 17:35:11 -0400 In-Reply-To: References: <20190731201927.22054-1-lyude@redhat.com> <20190731211842.befvpoyudrm2subf@wunner.de> Organization: Red Hat Content-Type: text/plain; charset="UTF-8" User-Agent: Evolution 3.32.4 (3.32.4-1.fc30) MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Wed, 2019-07-31 at 23:26 +0200, Karol Herbst wrote: > On Wed, Jul 31, 2019 at 11:18 PM Lukas Wunner wrote: > > On Wed, Jul 31, 2019 at 04:19:27PM -0400, Lyude Paul wrote: > > > While this fixes audio for a number of users, this commit has the > > > sideaffect of breaking the BIOS workaround that's required to make the > > > GPU on the nvidia P50 work, by causing the GPU's PCI device function to > > > stop working after it's been set to multifunction mode. > > > > This is missing a reference to the commit introducing the P50 quirk, > > which is e0547c81bfcf ("PCI: Reset Lenovo ThinkPad P50 nvgpu at boot > > if necessary"). > > > > Please describe in more detail how the GPU's PCI function stops working. > > Does it respond with "all ones" when accessing MMIO? > > Do MMIO accesses cause the system to hang? > > > > Could you provide lspci -vvxx output for the GPU and its associated > > HDA controller with and without b516ea586d71? > > > > Does this machine have external display connectors via which audio > > can be streamed? > > > > > > > I'm not really holding my breath on this patch to being accepted: > > > there's a good chance there's a better solution for this (and I'm going > > > to continue investigating for one after sending this patch), this is > > > more just to start a conversation on what the proper way to fix this is. > > > > Posting as an RFC might have been more appropriate then. > > > > no, a revert is actually appropriate. If a commit fixes something, > but breaks something else, it gets either reverted or fixed. If nobody > fixes it, then revert it is. To answer Lukas's question btw: most of the details on how things break are back in the original commit (sorry for forgetting the reference!), there's a _lot_ of explanation there that I'd rather not retype, so just refer back to the commit and bug @ https://bugs.freedesktop.org/show_bug.cgi?id=75985 Additionally, there was some extra discussion providing some more detail in the email thread that I had with Bjorn: https://lkml.org/lkml/2019/2/12/1172 As for how this commit breaks the workaround: it seems that when we enable the HDA controller and put the GPU into multifunction mode, the function-level reset stops working and thus we can't reset the GPU anymore. Currently I can see a couple of solutions (again, please feel free to suggest more!): * Just revert the commit. We should do this if necessary, but of course I'd much rather try finding a fix first * Disable the HDA controller temporarily when a GPU reset is neded in quirk_reset_lenovo_thinkpad_p50_nvgpu(), then call the function level reset, then re-enable the HDA controller. I have no idea if this actually works yet, but I'm about to try this on my system * Get quirk_reset_lenovo_thinkpad_p50_nvgpu() to run before quirk_nvidia_hda(). This would probably be fine, but we would need to rework some stuff in the PCI subsystem (maybe it already has a way to do this? haven't checked yet) so that we could perform an flr probe early enough to perform the quirk > > > > So, I'm kind of confused about why exactly this was implemented as an > > > early boot quirk in the first place. If we're seeing the GPU's PCI > > > device, we already know the GPU is there. Shouldn't we be able to check > > > for the existence of the HDA device once we probe the GPU in nouveau? > > > > I think a motivation to keep this generic was to make it work with > > other drivers besides nouveau, specifically Nvidia's proprietary driver. > > nouveau might not even be enabled. > > > > > > > that still doesn't explain why this was implemented as an early quirk > > > > This isn't an early quirk. Those live in arch/x86/kernel/early-quirks.c. > > This is just a PCI quirk executed on device enumeration and on resume. > > Devices aren't necessarily enumerated only on boot, e.g. think > > Thunderbolt. > > > > Thanks, > > > > Lukas -- Cheers, Lyude Paul