From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Date: Fri, 29 May 2020 05:45:23 -0700 From: Christoph Hellwig Subject: Re: [PATCH 0/8] Convert the intel iommu driver to the dma-iommu api Message-ID: <20200529124523.GA11817@infradead.org> References: <20191221150402.13868-1-murphyt7@tcd.ie> <465815ae-9292-f37a-59b9-03949cb68460@deltatee.com> MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: <465815ae-9292-f37a-59b9-03949cb68460@deltatee.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: To: Logan Gunthorpe Cc: Tom Murphy , iommu@lists.linux-foundation.org, kvm@vger.kernel.org, David Airlie , dri-devel@lists.freedesktop.org, Bjorn Andersson , Matthias Brugger , Julien Grall , Thierry Reding , Will Deacon , Jean-Philippe Brucker , linux-samsung-soc@vger.kernel.org, Marc Zyngier , Krzysztof Kozlowski , Jonathan Hunter , linux-rockchip@lists.infradead.org, Andy Gross , Gerald Schaefer , linux-s390@vger.kernel.org, linux-arm-msm@vger.kernel.org, intel-gfx@lists.freedesktop.org, Alex Williamson , linux-mediatek@lists.infradead.org, Rodrigo Vivi , linux-tegra@vger.kernel.org, Thomas Gleixner , virtualization@lists.linux-foundation.org, linux-arm-kernel@lists.infradead.org, Robin Murphy , Cornelia Huck , linux-kernel@vger.kernel.org, Kukjin Kim , David Woodhouse , Marek Szyprowski , Jani Nikula , Joonas Lahtinen On Thu, May 28, 2020 at 06:00:44PM -0600, Logan Gunthorpe wrote: > > This issue is most likely in the i915 driver and is most likely caused by the driver not respecting the return value of the dma_map_ops::map_sg function. You can see the driver ignoring the return value here: > > https://github.com/torvalds/linux/blob/7e0165b2f1a912a06e381e91f0f4e495f4ac3736/drivers/gpu/drm/i915/gem/i915_gem_dmabuf.c#L51 > > > > Previously this didn’t cause issues because the intel map_sg always returned the same number of elements as the input scatter gather list but with the change to this dma-iommu api this is no longer the case. I wasn’t able to track the bug down to a specific line of code unfortunately. Mark did a big audit into the map_sg API abuse and initially had some i915 patches, but then gave up on them with this comment: "The biggest TODO is DRM/i915 driver and I don't feel brave enough to fix it fully. The driver creatively uses sg_table->orig_nents to store the size of the allocate scatterlist and ignores the number of the entries returned by dma_map_sg function. In this patchset I only fixed the sg_table objects exported by dmabuf related functions. I hope that I didn't break anything there." it would be really nice if the i915 maintainers could help with sorting that API abuse out.