Path: csiph.com!1.us.feeder.erje.net!2.us.feeder.erje.net!feeder.erje.net!1.eu.feeder.erje.net!news2.arglkargh.de!news.mixmin.net!aioe.org!bofh.it!news.nic.it!robomod From: Will Deacon Newsgroups: linux.kernel Subject: Re: [PATCH v2] arm64: do not set dma masks that device connection can't handle Date: Tue, 10 Jan 2017 13:00:01 +0100 Message-ID: References: X-Original-To: Nikita Yushchenko MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline User-Agent: Mutt/1.5.23 (2014-03-12) Sender: robomod@news.nic.it List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Approved: robomod@news.nic.it Lines: 44 Organization: linux.* mail to news gateway X-Original-Cc: Arnd Bergmann , linux-arm-kernel@lists.infradead.org, Catalin Marinas , linux-kernel@vger.kernel.org, linux-renesas-soc@vger.kernel.org, Simon Horman , Bjorn Helgaas , artemi.ivanov@cogentembedded.com, robin.murphy@arm.com, fkan@apm.com X-Original-Date: Tue, 10 Jan 2017 11:51:33 +0000 X-Original-Message-ID: <20170110115132.GD21598@arm.com> X-Original-References: <1483947002-16410-1-git-send-email-nikita.yoush@cogentembedded.com> X-Original-Sender: linux-kernel-owner@vger.kernel.org Xref: csiph.com linux.kernel:1555192 On Mon, Jan 09, 2017 at 10:30:02AM +0300, Nikita Yushchenko wrote: > It is possible that device is capable of 64-bit DMA addresses, and > device driver tries to set wide DMA mask, but bridge or bus used to > connect device to the system can't handle wide addresses. > > With swiotlb, memory above 4G still can be used by drivers for streaming > DMA, but *dev->mask and dev->dma_coherent_mask must still keep values > that hardware handles physically. > > This patch enforces that. Based on original version by > Arnd Bergmann , extended with coherent mask hadnling. > > Signed-off-by: Nikita Yushchenko > CC: Arnd Bergmann > --- > Changes since v1: > - fixed issues noted by Sergei Shtylyov > - save mask, not size > - remove doube empty line > > arch/arm64/Kconfig | 3 +++ > arch/arm64/include/asm/device.h | 1 + > arch/arm64/mm/dma-mapping.c | 51 +++++++++++++++++++++++++++++++++++++++++ > 3 files changed, 55 insertions(+) I still don't think this patch is general enough. The problem you're seeing with swiotlb seems to be exactly the same problem reported by Feng Kan over at: http://lkml.kernel.org/r/CAL85gmA_SSCwM80TKdkZqEe+S1beWzDEvdki1kpkmUTDRmSP7g@mail.gmail.com [read on; it was initially thought to be a hardware erratum, but it's actually the inability to restrict the DMA mask of the endpoint that's the problem] The point here is that an IOMMU doesn't solve your issue, and the IOMMU-backed DMA ops need the same treatment. In light of that, it really feels to me like the DMA masks should be restricted in of_dma_configure so that the parent mask is taken into account there, rather than hook into each set of DMA ops to intercept set_dma_mask. We'd still need to do something to stop dma_set_mask widening the mask if it was restricted by of_dma_configure, but I think Robin (cc'd) was playing with that. Will