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=-1.0 required=3.0 tests=HEADER_FROM_DIFFERENT_DOMAINS, MAILING_LIST_MULTI,SPF_PASS autolearn=unavailable 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 BA6F4C4360F for ; Thu, 7 Mar 2019 14:53:01 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 8946020840 for ; Thu, 7 Mar 2019 14:53:01 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1726357AbfCGOxA (ORCPT ); Thu, 7 Mar 2019 09:53:00 -0500 Received: from foss.arm.com ([217.140.101.70]:46532 "EHLO foss.arm.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726170AbfCGOw7 (ORCPT ); Thu, 7 Mar 2019 09:52:59 -0500 Received: from usa-sjc-imap-foss1.foss.arm.com (unknown [10.72.51.249]) by usa-sjc-mx-foss1.foss.arm.com (Postfix) with ESMTP id E6D45EBD; Thu, 7 Mar 2019 06:52:58 -0800 (PST) Received: from [10.1.196.75] (e110467-lin.cambridge.arm.com [10.1.196.75]) by usa-sjc-imap-foss1.foss.arm.com (Postfix) with ESMTPSA id F03D13F71D; Thu, 7 Mar 2019 06:52:56 -0800 (PST) Subject: Re: [PATCH/RFC] driver core: Postpone DMA tear-down until after devres release To: John Garry , Geert Uytterhoeven Cc: Geert Uytterhoeven , "Rafael J . Wysocki" , Greg Kroah-Hartman , Joerg Roedel , Linux Kernel Mailing List , Linux-Renesas , Linux IOMMU , Christoph Hellwig , Linux ARM , Marek Szyprowski References: <20190207193653.18221-1-geert+renesas@glider.be> <20190208164019.GY32526@8bytes.org> <6e64e564-aa6c-4193-1f57-ebf88d6932d5@arm.com> <1b24a2dc-e9a3-f293-8dd1-360896a554e6@huawei.com> From: Robin Murphy Message-ID: Date: Thu, 7 Mar 2019 14:52:55 +0000 User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:60.0) Gecko/20100101 Thunderbird/60.5.1 MIME-Version: 1.0 In-Reply-To: <1b24a2dc-e9a3-f293-8dd1-360896a554e6@huawei.com> Content-Type: text/plain; charset=windows-1252; format=flowed Content-Language: en-GB Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Hi John, On 07/03/2019 14:45, John Garry wrote: [...] > Hi guys, > > Any idea what happened to this fix? It's been in -next for a while (commit 376991db4b64) - I assume it will land shortly and hit stable thereafter, at which point somebody gets to sort out the manual backport past 4.20. Robin. > I have this on 5.0: > > [    0.000000] Linux version 5.0.0 (john@htsatcamb-server) (gcc version > 4.8.5 (Linaro GCC 4.8-2015.06)) #121 SMP PREEMPT Thu Mar 7 14:28:39 GMT > 2019 > [    0.000000] Kernel command line: BOOT_IMAGE=/john/Image rdinit=/init > crashkernel=256M@32M earlycon console=ttyAMA0,115200 acpi=force > pcie_aspm=off scsi_mod.use_blk_mq=y no_console_suspend pcie-hisi.disable=1 > > ... > > [   26.806856] pci_bus 000c:20: 2-byte config write to 000c:20:00.0 > offset 0x44 may corrupt adjacent RW1C bits > [   26.817521] pcieport 0002:f8:00.0: can't derive routing for PCI INT B > [   26.837167] pci_bus 000c:20: 2-byte config write to 000c:20:00.0 > offset 0x44 may corrupt adjacent RW1C bits > [   26.850091] serial 0002:f9:00.1: PCI INT B: no GSI > [   26.879364] serial 0002:f9:00.1: enabling device (0000 -> 0003) > [   26.913131] 0002:f9:00.1: ttyS3 at I/O 0x1008 (irq = 0, base_baud = > 115200) is a ST16650V2 > [   26.992897] rtc-efi rtc-efi: setting system clock to > 2019-03-07T14:41:48 UTC (1551969708) > [   27.009380] ALSA device list: > [   27.015326]   No soundcards found. > [   27.022549] Freeing unused kernel memory: 1216K > [   27.055567] Run /init as init process > root@(none)$ cd /sys/devices/platform/HISI0162:01 > root@(none)$ echo HISI0162:01 > driver/unbind > [   36.488040] hisi_sas_v2_hw HISI0162:01: dev[9:1] is gone > [   36.561077] hisi_sas_v2_hw HISI0162:01: dev[8:1] is gone > [   36.621061] hisi_sas_v2_hw HISI0162:01: dev[7:1] is gone > [   36.693074] hisi_sas_v2_hw HISI0162:01: dev[6:1] is gone > [   36.753066] hisi_sas_v2_hw HISI0162:01: dev[5:1] is gone > [   36.764276] sd 0:0:3:0: [sdd] Synchronizing SCSI cache > [   36.821106] hisi_sas_v2_hw HISI0162:01: dev[4:1] is gone > [   36.889048] hisi_sas_v2_hw HISI0162:01: dev[3:1] is gone > [   36.993002] hisi_sas_v2_hw HISI0162:01: dev[1:1] is gone > [   37.004276] sd 0:0:0:0: [sda] Synchronizing SCSI cache > [   37.014768] sd 0:0:0:0: [sda] Stopping disk > [   37.709094] hisi_sas_v2_hw HISI0162:01: dev[2:5] is gone > [   37.721231] hisi_sas_v2_hw HISI0162:01: dev[0:2] is gone > [   37.803774] BUG: Bad page state in process sh  pfn:11356 > [   37.814444] page:ffff7e000044d580 count:1 mapcount:0 > mapping:0000000000000000 index:0x0 > [   37.830525] flags: 0xfffc00000001000(reserved) > [   37.839443] raw: 0fffc00000001000 ffff7e000044d588 ffff7e000044d588 > 0000000000000000 > [   37.854998] raw: 0000000000000000 0000000000000000 00000001ffffffff > 0000000000000000 > [   37.870552] page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set > [   37.883485] bad because of flags: 0x1000(reserved) > [   37.893098] Modules linked in: > [   37.899221] CPU: 5 PID: 2691 Comm: sh Not tainted 5.0.0 #121 > [   37.910578] Hardware name: Huawei Taishan 2280 /D05, BIOS Hisilicon > D05 IT21 Nemo 2.0 RC0 04/18/2018 > [   37.928924] Call trace: > [   37.933825]  dump_backtrace+0x0/0x150 > [   37.941166]  show_stack+0x14/0x1c > [   37.947808]  dump_stack+0x8c/0xb0 > [   37.954451]  bad_page+0xe4/0x144 > [   37.960918]  free_pages_check_bad+0x7c/0x84 > [   37.969307]  __free_pages_ok+0x284/0x290 > [   37.977173]  __free_pages+0x30/0x44 > [   37.984163]  __dma_direct_free_pages+0x68/0x6c > [   37.993076]  dma_direct_free+0x24/0x38 > [   38.000591]  dma_free_attrs+0x84/0xc0 > [   38.007930]  dmam_release+0x20/0x28 > [   38.014924]  release_nodes+0x128/0x1f8 > [   38.022439]  devres_release_all+0x34/0x4c > [   38.030478]  device_release_driver_internal+0x190/0x208 > [   38.040963]  device_release_driver+0x14/0x1c > [   38.049526]  unbind_store+0xbc/0xf4 > [   38.056517]  drv_attr_store+0x20/0x30 > [   38.063859]  sysfs_kf_write+0x44/0x4c > [   38.071200]  kernfs_fop_write+0xd0/0x1c4 > [   38.079065]  __vfs_write+0x2c/0x158 > [   38.086055]  vfs_write+0xa8/0x19c > [   38.092696]  ksys_write+0x44/0xa0 > [   38.099338]  __arm64_sys_write+0x1c/0x24 > [   38.107203]  el0_svc_common+0xb0/0x100 > [   38.114718]  el0_svc_handler+0x70/0x88 > [   38.122232]  el0_svc+0x8/0x7c0 > [   38.128356] Disabling lock debugging due to kernel taint > [   38.139019] BUG: Bad page state in process sh  pfn:11355 > [   38.149682] page:ffff7e000044d540 count:0 mapcount:0 > mapping:0000000000000000 index:0x0 > [   38.165760] flags: 0xfffc00000001000(reserved) > [   38.174676] raw: 0fffc00000001000 ffff7e000044d548 ffff7e000044d548 > 0000000000000000 > [   38.190230] raw: 0000000000000000 0000000000000000 00000000ffffffff > 0000000000000000 > [   38.205783] page dumped because: PAGE_FLAGS_CHECK_AT_FREE flag(s) set > [   38.218716] bad because of flags: 0x1000(reserved) > [   38.228329] Modules linked in: > [   38.234451] CPU: 5 PID: 2691 Comm: sh Tainted: G    B 5.0.0 #121 > [   38.248604] Hardware name: Huawei Taishan 2280 /D05, BIOS Hisilicon > D05 IT21 Nemo 2.0 RC0 04/18/2018 > [   38.266949] Call trace: > [   38.271844]  dump_backtrace+0x0/0x150 > [   38.279185]  show_stack+0x14/0x1c > [   38.285826]  dump_stack+0x8c/0xb0 > [   38.292468]  bad_page+0xe4/0x144 > [   38.298936]  free_pages_check_bad+0x7c/0x84 > > ... > > Thanks, > John > >> Thanks, >> Robin. >> >>>> There aren't many drivers using dmam_alloc_*(), let alone which would >>>> also find themselves behind an IOMMU on an Arm system, but it turns out >>>> I actually have another one which can reproduce the BUG() with 5.0-rc. >>> >>> SATA core uses dmam_alloc_*(). >>> >>>> I've tried a 4.12 kernel with a bit of instrumentation[1] and sure >>>> enough the devres-managed buffer is freed with the wrong ops[2] even >>>> then. How it manages not to blow up more catastrophically I have no >>>> idea... I guess at best it just leaks the buffers and IOMMU mappings, >>>> and at worst quietly frees random other pages instead. >>> >>> May depend on the actual ops, and whether CMA is used or not. >>> >>> Gr{oetje,eeting}s, >>> >>>                          Geert >>> >> >> _______________________________________________ >> linux-arm-kernel mailing list >> linux-arm-kernel@lists.infradead.org >> http://lists.infradead.org/mailman/listinfo/linux-arm-kernel >> >> . >> > >