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=-8.1 required=3.0 tests=DKIM_INVALID,DKIM_SIGNED, HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_PATCH,MAILING_LIST_MULTI,SIGNED_OFF_BY, SPF_HELO_NONE,SPF_PASS,USER_AGENT_SANE_1 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 40899C17441 for ; Mon, 11 Nov 2019 08:12:29 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id DBBF12075C for ; Mon, 11 Nov 2019 08:12:28 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=fail reason="signature verification failed" (2048-bit key) header.d=xenosoft.de header.i=@xenosoft.de header.b="pBEbpYY+" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org DBBF12075C Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=xenosoft.de Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 8ABE66B0005; Mon, 11 Nov 2019 03:12:28 -0500 (EST) Received: by kanga.kvack.org (Postfix, from userid 40) id 85C6B6B0006; Mon, 11 Nov 2019 03:12:28 -0500 (EST) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 771036B0007; Mon, 11 Nov 2019 03:12:28 -0500 (EST) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0240.hostedemail.com [216.40.44.240]) by kanga.kvack.org (Postfix) with ESMTP id 645266B0005 for ; Mon, 11 Nov 2019 03:12:28 -0500 (EST) Received: from smtpin03.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay01.hostedemail.com (Postfix) with SMTP id F3DE1180AD81D for ; Mon, 11 Nov 2019 08:12:27 +0000 (UTC) X-FDA: 76143279534.03.smile21_89c3ee317ea53 X-HE-Tag: smile21_89c3ee317ea53 X-Filterd-Recvd-Size: 10827 Received: from mo4-p01-ob.smtp.rzone.de (mo4-p01-ob.smtp.rzone.de [85.215.255.54]) by imf50.hostedemail.com (Postfix) with ESMTP for ; Mon, 11 Nov 2019 08:12:26 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; t=1573459945; s=strato-dkim-0002; d=xenosoft.de; h=In-Reply-To:Date:Message-ID:References:Cc:To:From:Subject: X-RZG-CLASS-ID:X-RZG-AUTH:From:Subject:Sender; bh=3/MqObd2KBYhkLHMy8MLFEFld3owrwS9kDyklpFMgIE=; b=pBEbpYY+AogIyLkY1Zb8seyfAwgdhwo5B3ufDlYLPdBfQbgngxFINRBBa2UAhozjAG aYtpnbgE625A9bqw8CCpXCYFcjr7tAsHfx1r4JEt7XNyt7HdJ4jb6uQDcQO1Y2O8MRtr Zm4hmyqZ7U3aIgoRHlRDaYHncGdnNRBP9o40oiA4L9Nh8GIAt31mOO4f9ukoMaP/BVQ2 HkN3CU92sDs9vL5hvAoTJRm0Tm6o+JMzOeOw0jU0eUuEpZxv3xMY/mfvRhtgg76JVnkP Mcj/Yoa6UJS1g/ocChLpP4s3m8/OvFA7bliWK7Ffwamkk/Wpy+wbzlL6LHT72z50YMBx rA4g== X-RZG-AUTH: ":L2QefEenb+UdBJSdRCXu93KJ1bmSGnhMdmOod1DhGM4l4Hio94KKxRySfLxnHfJ+Dkjp5DdBJSrwuuqxvPhUdStcud4ztdvxL67fR7ahIrEKaw==" X-RZG-CLASS-ID: mo00 Received: from [IPv6:2a02:8109:89c0:ebfc:7cd8:48a4:7aa4:e860] by smtp.strato.de (RZmta 44.29.0 AUTH) with ESMTPSA id q007c8vAB8CCS3t (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (curve secp521r1 with 521 ECDH bits, eq. 15360 bits RSA)) (Client did not present a certificate); Mon, 11 Nov 2019 09:12:12 +0100 (CET) Subject: Re: Bug 205201 - overflow of DMA mask and bus mask From: Christian Zigotzky To: Christoph Hellwig Cc: linux-arch@vger.kernel.org, linux-kernel@vger.kernel.org, linux-mm@kvack.org, iommu@lists.linux-foundation.org, Benjamin Herrenschmidt , paulus@samba.org, darren@stevens-zone.net, "contact@a-eon.com" , rtd2@xtra.co.nz, mad skateman , Rob Herring , linuxppc-dev References: <20181213112511.GA4574@lst.de> <008c981e-bdd2-21a7-f5f7-c57e4850ae9a@xenosoft.de> <20190103073622.GA24323@lst.de> <71A251A5-FA06-4019-B324-7AED32F7B714@xenosoft.de> <1b0c5c21-2761-d3a3-651b-3687bb6ae694@xenosoft.de> <3504ee70-02de-049e-6402-2d530bf55a84@xenosoft.de> <46025f1b-db20-ac23-7dcd-10bc43bbb6ee@xenosoft.de> <20191105162856.GA15402@lst.de> <2f3c81bd-d498-066a-12c0-0a7715cda18f@xenosoft.de> Message-ID: Date: Mon, 11 Nov 2019 09:12:10 +0100 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.9.1 MIME-Version: 1.0 In-Reply-To: Content-Type: text/plain; charset=utf-8; format=flowed Content-Language: de-DE Content-Transfer-Encoding: quoted-printable X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: On 10 November 2019 at 08:27 am, Christian Zigotzky wrote: > On 07 November 2019 at 10:53 am, Christian Zigotzky wrote: >> On 05 November 2019 at 05:28 pm, Christoph Hellwig wrote: >>> On Tue, Nov 05, 2019 at 08:56:27AM +0100, Christian Zigotzky wrote: >>>> Hi All, >>>> >>>> We still have DMA problems with some PCI devices. Since the PowerPC=20 >>>> updates >>>> 4.21-1 [1] we need to decrease the RAM to 3500MB (mem=3D3500M) if we= =20 >>>> want to >>>> work with our PCI devices. The FSL P5020 and P5040 have these proble= ms >>>> currently. >>>> >>>> Error message: >>>> >>>> [=C2=A0=C2=A0 25.654852] bttv 1000:04:05.0: overflow 0x00000000fe077= 000+4096=20 >>>> of DMA >>>> mask ffffffff bus mask df000000 >>>> >>>> All 5.x Linux kernels can't initialize a SCSI PCI card anymore so=20 >>>> booting >>>> of a Linux userland isn't possible. >>>> >>>> PLEASE check the DMA changes in the PowerPC updates 4.21-1 [1]. The=20 >>>> kernel >>>> 4.20 works with all PCI devices without limitation of RAM. >>> Can you send me the .config and a dmesg?=C2=A0 And in the meantime tr= y the >>> patch below? >>> >>> --- >>> >From 4d659b7311bd4141fdd3eeeb80fa2d7602ea01d4 Mon Sep 17 00:00:00 20= 01 >>> From: Nicolas Saenz Julienne >>> Date: Fri, 18 Oct 2019 13:00:43 +0200 >>> Subject: dma-direct: check for overflows on 32 bit DMA addresses >>> >>> As seen on the new Raspberry Pi 4 and sta2x11's DMA implementation=20 >>> it is >>> possible for a device configured with 32 bit DMA addresses and a=20 >>> partial >>> DMA mapping located at the end of the address space to overflow. It >>> happens when a higher physical address, not DMAable, is translated to >>> it's DMA counterpart. >>> >>> For example the Raspberry Pi 4, configurable up to 4 GB of memory, ha= s >>> an interconnect capable of addressing the lower 1 GB of physical memo= ry >>> with a DMA offset of 0xc0000000. It transpires that, any attempt to >>> translate physical addresses higher than the first GB will result in = an >>> overflow which dma_capable() can't detect as it only checks for >>> addresses bigger then the maximum allowed DMA address. >>> >>> Fix this by verifying in dma_capable() if the DMA address range=20 >>> provided >>> is at any point lower than the minimum possible DMA address on the bu= s. >>> >>> Signed-off-by: Nicolas Saenz Julienne >>> --- >>> =C2=A0 include/linux/dma-direct.h | 8 ++++++++ >>> =C2=A0 1 file changed, 8 insertions(+) >>> >>> diff --git a/include/linux/dma-direct.h b/include/linux/dma-direct.h >>> index adf993a3bd58..6ad9e9ea7564 100644 >>> --- a/include/linux/dma-direct.h >>> +++ b/include/linux/dma-direct.h >>> @@ -3,6 +3,7 @@ >>> =C2=A0 #define _LINUX_DMA_DIRECT_H 1 >>> =C2=A0 =C2=A0 #include >>> +#include /* for min_low_pfn */ >>> =C2=A0 #include >>> =C2=A0 =C2=A0 #ifdef CONFIG_ARCH_HAS_PHYS_TO_DMA >>> @@ -27,6 +28,13 @@ static inline bool dma_capable(struct device=20 >>> *dev, dma_addr_t addr, size_t size) >>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 if (!dev->dma_mask) >>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return false; >>> =C2=A0 +#ifndef CONFIG_ARCH_DMA_ADDR_T_64BIT >>> +=C2=A0=C2=A0=C2=A0 /* Check if DMA address overflowed */ >>> +=C2=A0=C2=A0=C2=A0 if (min(addr, addr + size - 1) < >>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 __phys_to_dma(dev, (phys_= addr_t)(min_low_pfn << PAGE_SHIFT))) >>> +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return false; >>> +#endif >>> + >>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return addr + size - 1 <=3D >>> =C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 min_not_zero(*= dev->dma_mask, dev->bus_dma_mask); >>> =C2=A0 } >> Hello Christoph, >> >> Thanks a lot for your patch! Unfortunately this patch doesn't solve=20 >> the issue. >> >> Error messages: >> >> [=C2=A0=C2=A0=C2=A0 6.041163] bttv: driver version 0.9.19 loaded >> [=C2=A0=C2=A0=C2=A0 6.041167] bttv: using 8 buffers with 2080k (520 pa= ges) each for=20 >> capture >> [=C2=A0=C2=A0=C2=A0 6.041559] bttv: Bt8xx card found (0) >> [=C2=A0=C2=A0=C2=A0 6.041609] bttv: 0: Bt878 (rev 17) at 1000:04:05.0,= irq: 19,=20 >> latency: 128, mmio: 0xc20001000 >> [=C2=A0=C2=A0=C2=A0 6.041622] bttv: 0: using: Typhoon TView RDS + FM S= tereo / KNC1=20 >> TV Station RDS [card=3D53,insmod option] >> [=C2=A0=C2=A0=C2=A0 6.042216] bttv: 0: tuner type=3D5 >> [=C2=A0=C2=A0=C2=A0 6.111994] bttv: 0: audio absent, no audio device f= ound! >> [=C2=A0=C2=A0=C2=A0 6.176425] bttv: 0: Setting PLL: 28636363 =3D> 3546= 8950 (needs up=20 >> to 100ms) >> [=C2=A0=C2=A0=C2=A0 6.200005] bttv: PLL set ok >> [=C2=A0=C2=A0=C2=A0 6.209351] bttv: 0: registered device video0 >> [=C2=A0=C2=A0=C2=A0 6.211576] bttv: 0: registered device vbi0 >> [=C2=A0=C2=A0=C2=A0 6.214897] bttv: 0: registered device radio0 >> [=C2=A0 114.218806] bttv 1000:04:05.0: overflow 0x00000000ff507000+409= 6 of=20 >> DMA mask ffffffff bus mask df000000 >> [=C2=A0 114.218848] Modules linked in: rfcomm bnep tuner_simple=20 >> tuner_types tea5767 tuner tda7432 tvaudio msp3400 bttv tea575x=20 >> tveeprom videobuf_dma_sg videobuf_core rc_core videodev mc btusb=20 >> btrtl btbcm btintel bluetooth uio_pdrv_genirq uio ecdh_generic ecc >> [=C2=A0 114.219012] [c0000001ecddf720] [80000000008ff6e8]=20 >> .buffer_prepare+0x150/0x268 [bttv] >> [=C2=A0 114.219029] [c0000001ecddf860] [80000000008fff6c]=20 >> .bttv_qbuf+0x50/0x64 [bttv] >> >> ----- >> >> Trace: >> >> [=C2=A0 462.783184] Call Trace: >> [=C2=A0 462.783187] [c0000001c6c67420] [c0000000000b3358]=20 >> .report_addr+0xb8/0xc0 (unreliable) >> [=C2=A0 462.783192] [c0000001c6c67490] [c0000000000b351c]=20 >> .dma_direct_map_page+0xf0/0x128 >> [=C2=A0 462.783195] [c0000001c6c67530] [c0000000000b35b0]=20 >> .dma_direct_map_sg+0x5c/0xac >> [=C2=A0 462.783205] [c0000001c6c675e0] [8000000000862e88]=20 >> .__videobuf_iolock+0x660/0x6d8 [videobuf_dma_sg] >> [=C2=A0 462.783220] [c0000001c6c676b0] [8000000000854274]=20 >> .videobuf_iolock+0x98/0xb4 [videobuf_core] >> [=C2=A0 462.783271] [c0000001c6c67720] [80000000008686e8]=20 >> .buffer_prepare+0x150/0x268 [bttv] >> [=C2=A0 462.783276] [c0000001c6c677c0] [8000000000854afc]=20 >> .videobuf_qbuf+0x2b8/0x428 [videobuf_core] >> [=C2=A0 462.783288] [c0000001c6c67860] [8000000000868f6c]=20 >> .bttv_qbuf+0x50/0x64 [bttv] >> [=C2=A0 462.783383] [c0000001c6c678e0] [80000000007bf208]=20 >> .v4l_qbuf+0x54/0x60 [videodev] >> [=C2=A0 462.783402] [c0000001c6c67970] [80000000007c1eac]=20 >> .__video_do_ioctl+0x30c/0x3f8 [videodev] >> [=C2=A0 462.783421] [c0000001c6c67a80] [80000000007c3c08]=20 >> .video_usercopy+0x18c/0x3dc [videodev] >> [=C2=A0 462.783440] [c0000001c6c67c00] [80000000007bb14c]=20 >> .v4l2_ioctl+0x60/0x78 [videodev] >> [=C2=A0 462.783460] [c0000001c6c67c90] [80000000007d3c48]=20 >> .v4l2_compat_ioctl32+0x9b4/0x1850 [videodev] >> [=C2=A0 462.783468] [c0000001c6c67d70] [c0000000001ad9cc]=20 >> .__se_compat_sys_ioctl+0x284/0x127c >> [=C2=A0 462.783473] [c0000001c6c67e20] [c00000000000067c]=20 >> system_call+0x60/0x6c >> [=C2=A0 462.783475] Instruction dump: >> [=C2=A0 462.783477] 40fe0044 60000000 892255d0 2f890000 40fe0020 3c82f= fc5=20 >> 39200001 60000000 >> [=C2=A0 462.783483] 38842029 992255d0 485ad0d9 60000000 <0fe00000>=20 >> 38210070 e8010010 7c0803a6 >> [=C2=A0 462.783490] ---[ end trace b677d4a00458e277 ]--- >> >> ----- >> >> dmesg fsl p5040: https://bugzilla.kernel.org/attachment.cgi?id=3D28581= 3 >> >> Kernel 5.4-rc6 config for the Cyrus+ board and for the QEMU ppce500=20 >> board (CPU: P5040 and P5020):=20 >> https://bugzilla.kernel.org/attachment.cgi?id=3D285815 >> >> Bug report: https://bugzilla.kernel.org/show_bug.cgi?id=3D205201 >> >> Thanks for your help, >> >> Christian > > Christoph, > > Do you have another patch for testing or shall I bisect? > > Thanks, > Christian Hi Christoph, I have seen that I have activated the kernel config option=20 CONFIG_ARCH_DMA_ADDR_T_64BIT. That means your code in your patch won't=20 work if this kernel option is enabled. +#ifndef CONFIG_ARCH_DMA_ADDR_T_64BIT +=C2=A0=C2=A0=C2=A0 /* Check if DMA address overflowed */ +=C2=A0=C2=A0=C2=A0 if (min(addr, addr + size - 1) < +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 __phys_to_dma(dev, (phys_addr= _t)(min_low_pfn << PAGE_SHIFT))) +=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0=C2=A0 return false; +#endif I will delete the lines with ifndef and endif and will try it again. Cheers, Christian