From patchwork Thu Jan 21 19:40:56 2021 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 8bit X-Patchwork-Submitter: Daniel Vetter X-Patchwork-Id: 12037547 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=-18.8 required=3.0 tests=BAYES_00,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_CR_TRAILER, INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED, USER_AGENT_GIT autolearn=ham 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 45658C433E6 for ; Thu, 21 Jan 2021 19:43:40 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 018A623A40 for ; Thu, 21 Jan 2021 19:43:39 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1727340AbhAUTnU (ORCPT ); Thu, 21 Jan 2021 14:43:20 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:35376 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726278AbhAUTmT (ORCPT ); Thu, 21 Jan 2021 14:42:19 -0500 Received: from mail-wm1-x32c.google.com (mail-wm1-x32c.google.com [IPv6:2a00:1450:4864:20::32c]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 99FE4C061788 for ; Thu, 21 Jan 2021 11:41:04 -0800 (PST) Received: by mail-wm1-x32c.google.com with SMTP id m2so2476569wmm.1 for ; Thu, 21 Jan 2021 11:41:04 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=ffwll.ch; s=google; h=from:to:cc:subject:date:message-id:mime-version :content-transfer-encoding; bh=bNL6e0QN+ZR4/EOKF9FDIwkALm5IXytT641Xe0UFkrY=; b=P8ric9b+DEzkRUmhpyp5fhDfW46TI86b2poNLqGZYj3qS6U4sSUnJwXO9DgIU1/7ZD CKZ09IYRErEY3NX1DdvOp/F2B/sSE5opK29JNAgoq/QvIMRF/IILJk/sK4rN/fvGW4Rt XJN9Hiir+rV1rDeOdFpzPu2UqlLf+V1J1aL2M= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:mime-version :content-transfer-encoding; bh=bNL6e0QN+ZR4/EOKF9FDIwkALm5IXytT641Xe0UFkrY=; b=UhsQBxVgUJTBTLGf3cSNzFfU39ymIP8qXLn/lrJKA4x9A/kf45LjN/ya00piqd//Do RHC4TpCgvK0XhtQU3qdAyvvrQXMZMDtiBlGRbCHFLfr89YLesvcXIL/p8zDDwz8pY6hE ysp/npvrSXiKBbg8aEjY9AzudXTZhrCapvKpabBksKdc6pIzVGScHrzpY/I5psfHWi/j y/plucZG/frudS43MrvTd432aDhcB7x7+B4jjXhn04w47zArXy/HrbZfV18G9J8fLDuq Jze4IhBESAlVDc+Og2MwJrcCT9IpzFEVNerGReJEFBusyWgIanOGly2/ab6ZYQI7NC5W Putw== X-Gm-Message-State: AOAM532QHuV8DX5r8CoEvUWfpBvlTb3QWtTl97/+9yxq5r0E+Nj3HkmA NJcrdUYeSaDnoA8X1hytxzAgmg== X-Google-Smtp-Source: ABdhPJz6Ns3zXzoWHuN08CkftiUhSBJXsmHiP4wjoMie+1Gc9+2oPhLWgZ/NzGFdQrNkELplFRSVpw== X-Received: by 2002:a1c:a9ce:: with SMTP id s197mr854645wme.146.1611258063360; Thu, 21 Jan 2021 11:41:03 -0800 (PST) Received: from phenom.ffwll.local ([2a02:168:57f4:0:efd0:b9e5:5ae6:c2fa]) by smtp.gmail.com with ESMTPSA id w13sm9605935wrt.52.2021.01.21.11.41.01 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Thu, 21 Jan 2021 11:41:02 -0800 (PST) From: Daniel Vetter To: DRI Development Cc: Daniel Vetter , Maarten Lankhorst , =?utf-8?q?Thomas_Hel?= =?utf-8?q?lstr=C3=B6m?= , =?utf-8?q?Christian_K?= =?utf-8?q?=C3=B6nig?= , Jerome Glisse , Felix Kuehling , Daniel Vetter , Sumit Semwal , linux-media@vger.kernel.org, linaro-mm-sig@lists.linaro.org Subject: [PATCH] RFC: dma-fence: Document recoverable page fault implications Date: Thu, 21 Jan 2021 20:40:56 +0100 Message-Id: <20210121194056.1734409-1-daniel.vetter@ffwll.ch> X-Mailer: git-send-email 2.30.0 MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: linux-media@vger.kernel.org Recently there was a fairly long thread about recoreable hardware page faults, how they can deadlock, and what to do about that. While the discussion is still fresh I figured good time to try and document the conclusions a bit. References: https://lore.kernel.org/dri-devel/20210107030127.20393-1-Felix.Kuehling@amd.com/ Cc: Maarten Lankhorst Cc: Thomas Hellström Cc: "Christian König" Cc: Jerome Glisse Cc: Felix Kuehling Signed-off-by: Daniel Vetter Cc: Sumit Semwal Cc: linux-media@vger.kernel.org Cc: linaro-mm-sig@lists.linaro.org Reviewed-by: Felix Kuehling --- I'll be away next week, but figured I'll type this up quickly for some comments and to check whether I got this all roughly right. Critique very much wanted on this, so that we can make sure hw which can't preempt (with pagefaults pending) like gfx10 has a clear path to support page faults in upstream. So anything I missed, got wrong or like that would be good. -Daniel --- Documentation/driver-api/dma-buf.rst | 66 ++++++++++++++++++++++++++++ 1 file changed, 66 insertions(+) diff --git a/Documentation/driver-api/dma-buf.rst b/Documentation/driver-api/dma-buf.rst index a2133d69872c..e924c1e4f7a3 100644 --- a/Documentation/driver-api/dma-buf.rst +++ b/Documentation/driver-api/dma-buf.rst @@ -257,3 +257,69 @@ fences in the kernel. This means: userspace is allowed to use userspace fencing or long running compute workloads. This also means no implicit fencing for shared buffers in these cases. + +Recoverable Hardware Page Faults Implications +~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ + +Modern hardware supports recoverable page faults, which has a lot of +implications for DMA fences. + +First, a pending page fault obviously holds up the work that's running on the +accelerator and a memory allocation is usually required to resolve the fault. +But memory allocations are not allowed to gate completion of DMA fences, which +means any workload using recoverable page faults cannot use DMA fences for +synchronization. Synchronization fences controlled by userspace must be used +instead. + +On GPUs this poses a problem, because current desktop compositor protocols on +Linus rely on DMA fences, which means without an entirely new userspace stack +built on top of userspace fences, they cannot benefit from recoverable page +faults. The exception is when page faults are only used as migration hints and +never to on-demand fill a memory request. For now this means recoverable page +faults on GPUs are limited to pure compute workloads. + +Furthermore GPUs usually have shared resources between the 3D rendering and +compute side, like compute units or command submission engines. If both a 3D +job with a DMA fence and a compute workload using recoverable page faults are +pending they could deadlock: + +- The 3D workload might need to wait for the compute job to finish and release + hardware resources first. + +- The compute workload might be stuck in a page fault, because the memory + allocation is waiting for the DMA fence of the 3D workload to complete. + +There are a few ways to prevent this problem: + +- Compute workloads can always be preempted, even when a page fault is pending + and not yet repaired. Not all hardware supports this. + +- DMA fence workloads and workloads which need page fault handling have + independent hardware resources to guarantee forward progress. This could be + achieved through e.g. through dedicated engines and minimal compute unit + reservations for DMA fence workloads. + +- The reservation approach could be further refined by only reserving the + hardware resources for DMA fence workloads when they are in-flight. This must + cover the time from when the DMA fence is visible to other threads up to + moment when fence is completed through dma_fence_signal(). + +- As a last resort, if the hardware provides no useful reservation mechanics, + all workloads must be flushed from the GPU when switching between jobs + requiring DMA fences or jobs requiring page fault handling: This means all DMA + fences must complete before a compute job with page fault handling can be + inserted into the scheduler queue. And vice versa, before a DMA fence can be + made visible anywhere in the system, all compute workloads must be preempted + to guarantee all pending GPU page faults are flushed. + +Note that workloads that run on independent hardware like copy engines or other +GPUs do not have any impact. This allows us to keep using DMA fences internally +in the kernel even for resolving hardware page faults, e.g. by using copy +engines to clear or copy memory needed to resolve the page fault. + +In some ways this page fault problem is a special case of the `Infinite DMA +Fences` discussions: Infinite fences from compute workloads are allowed to +depend on DMA fences, but not the other way around. And not even the page fault +problem is new, because some other CPU thread in userspace might +hit a page fault which holds up a userspace fence - supporting page faults on +GPUs doesn't anything fundamentally new.