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 Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 0BC32C433EF for ; Fri, 27 May 2022 13:04:55 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1351866AbiE0NEx (ORCPT ); Fri, 27 May 2022 09:04:53 -0400 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:51552 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S239275AbiE0NEv (ORCPT ); Fri, 27 May 2022 09:04:51 -0400 Received: from mail-lf1-x132.google.com (mail-lf1-x132.google.com [IPv6:2a00:1450:4864:20::132]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 298A55DA25 for ; Fri, 27 May 2022 06:04:50 -0700 (PDT) Received: by mail-lf1-x132.google.com with SMTP id u23so6840634lfc.1 for ; Fri, 27 May 2022 06:04:50 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20210112; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=YjJnblK4U3NCIbGYCswwX+gF+Gis3UKa2COyN188bL4=; b=N6O/fB2x/BSZYMQNgEXYJwA4qx0VYKXVwTj6kA4G2UWe+WYJh4Aae+2Kx0HsKZTtiu VFPgovLvGNeKNefUh4fKrBByTIhbZaNwfmGSWranS8+/cPeMSLG5v8Um8hnBzYIo0T76 iRJZ178kIluZkkwqBfT4adiMxgvOFsMeBR6f+U7pxRDMuHlA2n8jD78C2hSPLwbYyi/M mRRjc9T4GhPnSVmQYm8BU8ah6cY1rxwEHVNhJF1iyzPbow0VMlbU+TrkncFD0GsAJC5x Pv4+Dh289o783zY9BtJWr2CcKFQGyvwE+OEHkm7QfZijfM91cJgJS4XhoRgFT79lZp66 HJXg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=YjJnblK4U3NCIbGYCswwX+gF+Gis3UKa2COyN188bL4=; b=tiqCjLuhb8DswfjBST+P+ODu7HylE3MlnsnpfPGJIAoct9aD9fLPiHDb9DPkTRl+nk ndPk6bvcfFUqg3uRrjPmuqg6Stb3hCv/ND8CIT8mRLI9wZWrd3JsewAcw1d+BpQeP1vG B+iLR8EQIQQmQCNdq1uTY+rVW3VFhgdkIrgTzE/A4iVuhMgCL0uz+kkhEiCcTDdSAOWv AgvXGX5ydOmnnvj/+gCcDl1P8n51dLjNJM4/fMcT/ExWPMIdzYkucBZjfHKg+68ctVyz XVDM730NghiFdeCQhXD31Efjnlwa/eHDTpmbNQox++tVTP5IzcnNZNIgu4AU5kp/l/Oh k8aw== X-Gm-Message-State: AOAM532p8dB4QxZAxrM366KHBNt1spKO0oAM/jvf/ORj8awipe7VQk2n DciOahwu90fNxJuyz62LHMoN82phvNZKqWBbr92vCg== X-Google-Smtp-Source: ABdhPJwU+3pJiFMDOEcQfB65J9GtQ+fKq2dh6yOiyUMLeFk46Mqa08BRNL4oZP4rMJFiPy9KJhSdbcvHO7TcLX9OWfg= X-Received: by 2002:a05:6512:3c94:b0:477:ba25:de54 with SMTP id h20-20020a0565123c9400b00477ba25de54mr30098024lfv.137.1653656688036; Fri, 27 May 2022 06:04:48 -0700 (PDT) MIME-Version: 1.0 References: <00000000000038779505d5d8b372@google.com> <5e702296-9ce0-f1e6-dae8-cc719bc040b9@ghiti.fr> In-Reply-To: From: Dmitry Vyukov Date: Fri, 27 May 2022 15:04:36 +0200 Message-ID: Subject: Re: [syzbot] riscv/fixes boot error: can't ssh into the instance To: Alexandre Ghiti Cc: Atish Patra , Alexandre Ghiti , Aleksandr Nogikh , linux-riscv , kasan-dev , Palmer Dabbelt , syzbot , LKML , syzkaller-bugs@googlegroups.com Content-Type: text/plain; charset="UTF-8" Precedence: bulk List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, 27 May 2022 at 09:33, Alexandre Ghiti wrote: > > Hi Atish, > > > > On 5/12/22 13:48, Dmitry Vyukov wrote: > > > > > On Fri, 18 Feb 2022 at 14:45, Alexandre Ghiti > > > > > wrote: > > > > >> Hi Aleksandr, > > > > >> > > > > >> On Thu, Feb 17, 2022 at 6:08 PM Aleksandr Nogikh wrote: > > > > >>> Hi Alex, > > > > >>> > > > > >>> On Thu, Feb 17, 2022 at 5:53 PM Alexandre Ghiti > > > > >>> wrote: > > > > >>>> Aleksandr, > > > > >>>> > > > > >>>> On Wed, Feb 16, 2022 at 5:58 PM Alexandre Ghiti > > > > >>>> wrote: > > > > >>>>> First, thank you for working on this. > > > > >>>>> > > > > >>>>> On Wed, Feb 16, 2022 at 5:17 PM Aleksandr Nogikh wrote: > > > > >>>>>> If I use just defconfig + DEBUG_VIRTUAL, without any KASAN, it begins > > > > >>>>>> to boot, but overwhelms me with tons of `virt_to_phys used for > > > > >>>>>> non-linear address:` errors. > > > > >>>>>> > > > > >>>>>> Like that > > > > >>>>>> > > > > >>>>>> [ 2.701271] virt_to_phys used for non-linear address: > > > > >>>>>> 00000000b59e31b6 (0xffffffff806c2000) > > > > >>>>>> [ 2.701727] WARNING: CPU: 0 PID: 1 at arch/riscv/mm/physaddr.c:16 > > > > >>>>>> __virt_to_phys+0x7e/0x86 > > > > >>>>>> [ 2.702207] Modules linked in: > > > > >>>>>> [ 2.702393] CPU: 0 PID: 1 Comm: swapper/0 Tainted: G W > > > > >>>>>> 5.17.0-rc1 #1 > > > > >>>>>> [ 2.702806] Hardware name: riscv-virtio,qemu (DT) > > > > >>>>>> [ 2.703051] epc : __virt_to_phys+0x7e/0x86 > > > > >>>>>> [ 2.703298] ra : __virt_to_phys+0x7e/0x86 > > > > >>>>>> [ 2.703547] epc : ffffffff80008448 ra : ffffffff80008448 sp : > > > > >>>>>> ffff8f800021bde0 > > > > >>>>>> [ 2.703977] gp : ffffffff80ed9b30 tp : ffffaf8001230000 t0 : > > > > >>>>>> ffffffff80eea56f > > > > >>>>>> [ 2.704704] t1 : ffffffff80eea560 t2 : 0000000000000000 s0 : > > > > >>>>>> ffff8f800021be00 > > > > >>>>>> [ 2.705153] s1 : ffffffff806c2000 a0 : 000000000000004f a1 : > > > > >>>>>> ffffffff80e723d8 > > > > >>>>>> [ 2.705555] a2 : 0000000000000010 a3 : fffffffffffffffe a4 : > > > > >>>>>> 0000000000000000 > > > > >>>>>> [ 2.706027] a5 : 0000000000000000 a6 : 0000000000000005 a7 : > > > > >>>>>> ffffffffffffffff > > > > >>>>>> [ 2.706474] s2 : ffffffff80b80b08 s3 : 00000000000000c2 s4 : > > > > >>>>>> ffffffff806c2000 > > > > >>>>>> [ 2.706891] s5 : ffffffff80edba10 s6 : ffffffff80edb960 s7 : > > > > >>>>>> 0000000000000001 > > > > >>>>>> [ 2.707290] s8 : 00000000000000ff s9 : ffffffff80b80b40 s10: > > > > >>>>>> 00000000000000cc > > > > >>>>>> [ 2.707689] s11: ffffaf807e1fcf00 t3 : 0000000000000076 t4 : > > > > >>>>>> ffffffffffffffff > > > > >>>>>> [ 2.708092] t5 : 00000000000001f2 t6 : ffff8f800021bb48 > > > > >>>>>> [ 2.708433] status: 0000000000000120 badaddr: 0000000000000000 > > > > >>>>>> cause: 0000000000000003 > > > > >>>>>> [ 2.708919] [] free_reserved_area+0x72/0x19a > > > > >>>>>> [ 2.709296] [] free_initmem+0x6c/0x7c > > > > >>>>>> [ 2.709648] [] kernel_init+0x3a/0x10a > > > > >>>>>> [ 2.709993] [] ret_from_exception+0x0/0xc > > > > >>>>>> [ 2.710310] ---[ end trace 0000000000000000 ]--- > > > > >>>>>> > > > > >>>>> I was able to reproduce this: the first one regarding init_zero_pfn is > > > > >>>>> legit but not wrong, I have to check when it was introduced and how to > > > > >>>>> fix this. > > > > >>>>> Regarding the huge batch that follows, at first sight, I would say > > > > >>>>> this is linked to my sv48 patchset but that does not seem important as > > > > >>>>> the address is a kernel mapping address so the use of virt_to_phys is > > > > >>>>> right. > > > > >>>>> > > > > >>>>>> On Wed, Feb 16, 2022 at 5:09 PM Aleksandr Nogikh wrote: > > > > >>>>>>> On Wed, Feb 16, 2022 at 12:56 PM Dmitry Vyukov wrote: > > > > >>>>>>>> On Wed, 16 Feb 2022 at 12:47, Aleksandr Nogikh wrote: > > > > >>>>>>>>> On Wed, Feb 16, 2022 at 11:37 AM Aleksandr Nogikh wrote: > > > > >>>>>>>>>> Hi Alex, > > > > >>>>>>>>>> > > > > >>>>>>>>>> On Wed, Feb 16, 2022 at 5:14 AM Alexandre Ghiti wrote: > > > > >>>>>>>>>>> Hi Dmitry, > > > > >>>>>>>>>>> > > > > >>>>>>>>>>> On 2/15/22 18:12, Dmitry Vyukov wrote: > > > > >>>>>>>>>>>> On Wed, 2 Feb 2022 at 14:18, Alexandre Ghiti > > > > >>>>>>>>>>>> wrote: > > > > >>>>>>>>>>>>> Hi Aleksandr, > > > > >>>>>>>>>>>>> > > > > >>>>>>>>>>>>> On Wed, Feb 2, 2022 at 12:08 PM Aleksandr Nogikh wrote: > > > > >>>>>>>>>>>>>> Hello, > > > > >>>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> syzbot has already not been able to fuzz its RISC-V instance for 97 > > > > >>>>>>>>>>>>> That's a longtime, I'll take a look more regularly. > > > > >>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> days now because the compiled kernel cannot boot. I bisected the issue > > > > >>>>>>>>>>>>>> to the following commit: > > > > >>>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> commit 54c5639d8f507ebefa814f574cb6f763033a72a5 > > > > >>>>>>>>>>>>>> Author: Alexandre Ghiti > > > > >>>>>>>>>>>>>> Date: Fri Oct 29 06:59:27 2021 +0200 > > > > >>>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> riscv: Fix asan-stack clang build > > > > >>>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> Apparently, the problem appears on GCC-built RISC-V kernels with KASAN > > > > >>>>>>>>>>>>>> enabled. In the previous message syzbot mentions > > > > >>>>>>>>>>>>>> "riscv64-linux-gnu-gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU > > > > >>>>>>>>>>>>>> Binutils for Debian) 2.35.2", but the issue also reproduces finely on > > > > >>>>>>>>>>>>>> a newer GCC compiler: "riscv64-linux-gnu-gcc (Debian 11.2.0-10) > > > > >>>>>>>>>>>>>> 11.2.0, GNU ld (GNU Binutils for Debian) 2.37". > > > > >>>>>>>>>>>>>> For convenience, I also duplicate the .config file from the bot's > > > > >>>>>>>>>>>>>> message: https://syzkaller.appspot.com/x/.config?x=522544a2e0ef2a7d > > > > >>>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> Can someone with KASAN and RISC-V expertise please take a look? > > > > >>>>>>>>>>>>> I'll take a look at that today. > > > > >>>>>>>>>>>>> > > > > >>>>>>>>>>>>> Thanks for reporting the issue, > > > > >>>>>>>>>>> I took a quick look, not enough to fix it but I know the issue comes > > > > >>>>>>>>>>> from the inline instrumentation, I have no problem with the outline > > > > >>>>>>>>>>> instrumentation. I need to find some cycles to work on this, my goal is > > > > >>>>>>>>>>> to fix this for 5.17. > > > > >>>>>>>>>> Thanks for the update! > > > > >>>>>>>>>> > > > > >>>>>>>>>> Can you please share the .config with which you tested the outline > > > > >>>>>>>>>> instrumentation? > > > > >>>>>>>>>> I updated the syzbot config to use KASAN_OUTLINE instead of KASAN_INLINE, > > > > >>>>>>>>>> but it still does not boot :( > > > > >>>>>>>>>> > > > > >>>>>>>>>> Here's what I used: > > > > >>>>>>>>>> https://gist.github.com/a-nogikh/279c85c2d24f47efcc3e865c08844138 > > > > >>>>>>>>> Update: it doesn't boot with that big config, but boots if I generate > > > > >>>>>>>>> a simple one with KASAN_OUTLINE: > > > > >>>>>>>>> > > > > >>>>>>>>> make defconfig ARCH=riscv CROSS_COMPILE=riscv64-linux-gnu- > > > > >>>>>>>>> ./scripts/config -e KASAN -e KASAN_OUTLINE > > > > >>>>>>>>> make olddefconfig ARCH=riscv CROSS_COMPILE=riscv64-linux-gnu- > > > > >>>>>>>>> > > > > >>>>>>>>> And it indeed doesn't work if I use KASAN_INLINE. > > > > >>>>>>>> It may be an issue with code size. Full syzbot config + KASAN + KCOV > > > > >>>>>>>> produce hugely massive .text. It may be hitting some limitation in the > > > > >>>>>>>> bootloader/kernel bootstrap code. > > > > >>>>> I took a quick glance and it traps on a KASAN address that is not > > > > >>>>> mapped, either because it is too soon or because the mapping failed > > > > >>>>> somehow. > > > > >>>>> > > > > >>>>> I'll definitely dive into that tomorrow, sorry for being slow here and > > > > >>>>> thanks again for all your work, that helps a lot. > > > > >>>>> > > > > >>>>> Thanks, > > > > >>>>> > > > > >>>>> Alex > > > > >>>>> > > > > >>>>>>> I bisected the difference between the config we use on syzbot and the > > > > >>>>>>> simple one that was generated like I described above. > > > > >>>>>>> Turns out that it's the DEBUG_VIRTUAL config that makes the difference. > > > > >>>>>>> > > > > >>>>>>> make defconfig ARCH=riscv CROSS_COMPILE=riscv64-linux-gnu- > > > > >>>>>>> ./scripts/config -e KASAN -e KASAN_OUTLINE -e DEBUG_VIRTUAL > > > > >>>>>>> make olddefconfig ARCH=riscv CROSS_COMPILE=riscv64-linux-gnu- > > > > >>>>>>> > > > > >>>>>>> And the resulting kernel does not boot. > > > > >>>>>>> My env: the `riscv/fixes` branch, commit > > > > >>>>>>> 6df2a016c0c8a3d0933ef33dd192ea6606b115e3, qemu 6.2.0. > > > > >>>> I fixed a few things today: KASAN + SPARSE_VMEMMAP, DEBUG_VIRTUAL and > > > > >>>> maybe KASAN + KCOV. > > > > >>>> > > > > >>>> With those small fixes, I was able to boot your large dotconfig with > > > > >>>> KASAN_OUTLINE, the inline version still fails, this is my next target > > > > >>>> :) > > > > >>>> I'll push that tomorrow! > > > > >>> Awesome, thank you very much! > > > > >>> Looking forward to finally seeing the instance run :) > > > > >> I sent a patchset which should fix your config with *outline* instrumentation. > > > > > Was this fix merged? The riscv instance still does not boot: > > > > > https://syzkaller.appspot.com/bug?id=5f2ff52ad42cba9f222202219baebd4e63e35127 > > > > > > > > > > > > Yes it has been in Linus tree since 5.18-rc1. I'll take a look at that > > > > this week. > > > > > > > > > > Are you seeing this error or a different one ? I used the > > > syzkaller_defconfig from the patch below on v5.18. > > > > > > https://lore.kernel.org/all/20220419174952.699-1-palmer@rivosinc.com/ > > > > > > [ 15.076116][ T1] Mandatory Access Control activated. > > > [ 15.158241][ T1] AppArmor: AppArmor Filesystem Enabled > > > [ 16.150870][ T1] NET: Registered PF_INET protocol family > > > [ 16.166167][ T1] IP idents hash table entries: 32768 (order: 6, > > > 262144 bytes, linear) > > > [ 16.188727][ T1] Unable to handle kernel paging request at > > > virtual address ffebfffeffff2000 > > > [ 16.192727][ T1] Oops [#1] > > > [ 16.193479][ T1] Modules linked in: > > > [ 16.194687][ T1] CPU: 3 PID: 1 Comm: swapper/0 Not tainted > > > 5.18.0-00001-g37ac279268bf-dirty #9 > > > [ 16.196486][ T1] Hardware name: riscv-virtio,qemu (DT) > > > [ 16.197836][ T1] epc : kasan_check_range+0x9e/0x14e > > > [ 16.199104][ T1] ra : memset+0x1e/0x4c > > > [ 16.200091][ T1] epc : ffffffff804787e0 ra : ffffffff80478f30 sp > > > : ff600000073ffb70 > > > [ 16.201420][ T1] gp : ffffffff85879e80 tp : ff600000073f0000 t0 > > > : 7300000000000000 > > > [ 16.202762][ T1] t1 : ffebfffeffff21ff t2 : 73746e6564692050 s0 > > > : ff600000073ffba0 > > > [ 16.204047][ T1] s1 : 0000000000001000 a0 : ffebfffeffff2200 a1 > > > : 0000000000001000 > > > [ 16.205312][ T1] a2 : 0000000000000001 a3 : ffffffff803a4f32 a4 > > > : ff5ffffffff90000 > > > [ 16.206592][ T1] a5 : ffebfffeffff2000 a6 : 0000004000000000 a7 > > > : ff5ffffffff90fff > > > [ 16.207865][ T1] s2 : ff5ffffffff90000 s3 : 0000000000000000 s4 > > > : ffffffff8467ea90 > > > [ 16.209134][ T1] s5 : 0000000000000000 s6 : ff5ffffffff90000 s7 > > > : 0000000000000000 > > > [ 16.210394][ T1] s8 : 0000000000001000 s9 : ffffffff8587ca40 > > > s10: 0000000000000004 > > > [ 16.211952][ T1] s11: ffffffff858a03a0 t3 : 0000000000000000 t4 > > > : 0000000000000040 > > > [ 16.213469][ T1] t5 : ffebfffeffff2200 t6 : ff600000073ff738 > > > [ 16.214853][ T1] status: 0000000200000120 badaddr: > > > ffebfffeffff2000 cause: 000000000000000d > > > [ 16.216910][ T1] Call Trace: > > > [ 16.217816][ T1] [] pcpu_alloc+0x844/0x1254 > > > [ 16.219110][ T1] [] __alloc_percpu+0x28/0x34 > > > [ 16.220244][ T1] [] ip_rt_init+0x17e/0x382 > > > [ 16.221606][ T1] [] ip_init+0x18/0x30 > > > [ 16.222719][ T1] [] inet_init+0x2a6/0x550 > > > [ 16.223863][ T1] [] do_one_initcall+0x130/0x7dc > > > [ 16.225002][ T1] [] kernel_init_freeable+0x510/0x5b4 > > > [ 16.226273][ T1] [] kernel_init+0x28/0x21c > > > [ 16.227337][ T1] [] ret_from_exception+0x0/0x10 > > > [ 16.229910][ T1] ---[ end trace 0000000000000000 ]--- > > > [ 16.231880][ T1] Kernel panic - not syncing: Fatal exception > > > > > > > > > > Enabling CONFIG_KASAN_VMALLOC=y solves the issue and I am able to boot > > to the userspace. > > I have tried enabling/disabling CONFIG_VMAP_STACK as well. Both works fine. > > > > Looking at the ARM64 Kconfig, KASAN_VMALLOC is enabled if KASAN is enabled. > > This diff seems to work for me. > > diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig > > index 00fd9c548f26..cbf0fe227c77 100644 > > --- a/arch/riscv/Kconfig > > +++ b/arch/riscv/Kconfig > > @@ -122,6 +122,7 @@ config RISCV > > select TRACE_IRQFLAGS_SUPPORT > > select UACCESS_MEMCPY if !MMU > > select ZONE_DMA32 if 64BIT > > + select KASAN_VMALLOC if KASAN > > > > I am not a kasan expert so I am not sure if this is the correct fix or > > just hides the real issue. pcpu_alloc seems to use vmalloc though. > > When this type of thing happens, generally this is because of an error > in the kasan page table, I'll take a look this time, sorry I did not > do this before. > > Thanks for finding this, > > Alex Unfortunately riscv still does not work for syzbot, other bugs piled up. Something has changes wrt user-space pointers: https://groups.google.com/g/syzkaller-bugs/c/lU0GQTZoNQQ/m/iuRBTgU3AAAJ https://syzkaller.appspot.com/bug?extid=2c5da6a0a16a0c4f34aa 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 Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.133]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id 96EB5C433F5 for ; Fri, 27 May 2022 13:05:06 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; q=dns/txt; c=relaxed/relaxed; d=lists.infradead.org; s=bombadil.20210309; h=Sender: Content-Transfer-Encoding:Content-Type:List-Subscribe:List-Help:List-Post: List-Archive:List-Unsubscribe:List-Id:Cc:To:Subject:Message-ID:Date:From: In-Reply-To:References:MIME-Version:Reply-To:Content-ID:Content-Description: Resent-Date:Resent-From:Resent-Sender:Resent-To:Resent-Cc:Resent-Message-ID: List-Owner; bh=d0NxvJ4C1Ajf2mWxWAb6URdBa13MdI6cWdvmktXezzU=; b=1Ln3iVJC2T7VK+ s0V99zqNG9StWsbmXTD05DOVGIm8GhI9Jan73bvYWT7tFwAVjkKtejADJQt2T7WCDUR9eCAbkR+cg z8zVovoiXr7FIrVBvPwN/sM1P/+qMmhmMx8mchL18V/Rr55SeHEhp7GvcMqq3OdMVRWdBYU8eUjaX coPkRBUJ6x5yQpXsZ1rcE1Q2UW8zGLmiwzmhpTvPiSHPVzlXRxzAErIbkvMAZxv5AIekW3BS6mr3b AUR2FvKvrgpTvBLgBhLexttS6LIdj9ZaWfnsxqJDv7/BpixYx6vn87AZ2e99ASDpi/ohamB4OgKYl nhzdBhBwwuoGV8whKIxw==; Received: from localhost ([::1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.94.2 #2 (Red Hat Linux)) id 1nuZeB-0005uO-TE; Fri, 27 May 2022 13:04:55 +0000 Received: from mail-lf1-x132.google.com ([2a00:1450:4864:20::132]) by bombadil.infradead.org with esmtps (Exim 4.94.2 #2 (Red Hat Linux)) id 1nuZe8-0005th-5O for linux-riscv@lists.infradead.org; Fri, 27 May 2022 13:04:54 +0000 Received: by mail-lf1-x132.google.com with SMTP id p4so6825715lfg.4 for ; Fri, 27 May 2022 06:04:49 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=google.com; s=20210112; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=YjJnblK4U3NCIbGYCswwX+gF+Gis3UKa2COyN188bL4=; b=N6O/fB2x/BSZYMQNgEXYJwA4qx0VYKXVwTj6kA4G2UWe+WYJh4Aae+2Kx0HsKZTtiu VFPgovLvGNeKNefUh4fKrBByTIhbZaNwfmGSWranS8+/cPeMSLG5v8Um8hnBzYIo0T76 iRJZ178kIluZkkwqBfT4adiMxgvOFsMeBR6f+U7pxRDMuHlA2n8jD78C2hSPLwbYyi/M mRRjc9T4GhPnSVmQYm8BU8ah6cY1rxwEHVNhJF1iyzPbow0VMlbU+TrkncFD0GsAJC5x Pv4+Dh289o783zY9BtJWr2CcKFQGyvwE+OEHkm7QfZijfM91cJgJS4XhoRgFT79lZp66 HJXg== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=YjJnblK4U3NCIbGYCswwX+gF+Gis3UKa2COyN188bL4=; b=Dyf6c0OdzL7oW+jTREi+OTQIxEXBV1EB4hfAk0OzViTvejo43ipGQLwbBRd3noodt7 cePkzjJBt/frDUb187CD5gOdZvlfz09ZlskH1SNu3L1VJL+oJQFr0eexAzdfE1HSlhOF +0wAskPcZgz3S8Vc9IHaNwzSEZpUC0CpVTzFTBl39AyxIZMc9UKo7eamd1LpGtTr4Spj fQ0exXwJpbOceHlA2vhUYTXClcrwyNxPDykjw6FaQejobaz4rrTKvu2CIs90907R75kK 3VNR5aOyLAcGQ8YxDGVPZoIJg+PzGR/GkfoUnzjRHB6Vci+7RM8Sd8mMMnbIUnu3vDau XvDQ== X-Gm-Message-State: AOAM533p+KcMxiKoW/PwVifV3Kwi2FqCgpNjvRWQ3bWJyQ7aMxjpLSTB 3/0ZJ8aGkwkMgJNMYUm7bsArKgSEjyWxX6Ua16gx6g== X-Google-Smtp-Source: ABdhPJwU+3pJiFMDOEcQfB65J9GtQ+fKq2dh6yOiyUMLeFk46Mqa08BRNL4oZP4rMJFiPy9KJhSdbcvHO7TcLX9OWfg= X-Received: by 2002:a05:6512:3c94:b0:477:ba25:de54 with SMTP id h20-20020a0565123c9400b00477ba25de54mr30098024lfv.137.1653656688036; Fri, 27 May 2022 06:04:48 -0700 (PDT) MIME-Version: 1.0 References: <00000000000038779505d5d8b372@google.com> <5e702296-9ce0-f1e6-dae8-cc719bc040b9@ghiti.fr> In-Reply-To: From: Dmitry Vyukov Date: Fri, 27 May 2022 15:04:36 +0200 Message-ID: Subject: Re: [syzbot] riscv/fixes boot error: can't ssh into the instance To: Alexandre Ghiti Cc: Atish Patra , Alexandre Ghiti , Aleksandr Nogikh , linux-riscv , kasan-dev , Palmer Dabbelt , syzbot , LKML , syzkaller-bugs@googlegroups.com X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20220527_060452_273005_04771497 X-CRM114-Status: GOOD ( 50.00 ) X-BeenThere: linux-riscv@lists.infradead.org X-Mailman-Version: 2.1.34 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: "linux-riscv" Errors-To: linux-riscv-bounces+linux-riscv=archiver.kernel.org@lists.infradead.org On Fri, 27 May 2022 at 09:33, Alexandre Ghiti wrote: > > Hi Atish, > > > > On 5/12/22 13:48, Dmitry Vyukov wrote: > > > > > On Fri, 18 Feb 2022 at 14:45, Alexandre Ghiti > > > > > wrote: > > > > >> Hi Aleksandr, > > > > >> > > > > >> On Thu, Feb 17, 2022 at 6:08 PM Aleksandr Nogikh wrote: > > > > >>> Hi Alex, > > > > >>> > > > > >>> On Thu, Feb 17, 2022 at 5:53 PM Alexandre Ghiti > > > > >>> wrote: > > > > >>>> Aleksandr, > > > > >>>> > > > > >>>> On Wed, Feb 16, 2022 at 5:58 PM Alexandre Ghiti > > > > >>>> wrote: > > > > >>>>> First, thank you for working on this. > > > > >>>>> > > > > >>>>> On Wed, Feb 16, 2022 at 5:17 PM Aleksandr Nogikh wrote: > > > > >>>>>> If I use just defconfig + DEBUG_VIRTUAL, without any KASAN, it begins > > > > >>>>>> to boot, but overwhelms me with tons of `virt_to_phys used for > > > > >>>>>> non-linear address:` errors. > > > > >>>>>> > > > > >>>>>> Like that > > > > >>>>>> > > > > >>>>>> [ 2.701271] virt_to_phys used for non-linear address: > > > > >>>>>> 00000000b59e31b6 (0xffffffff806c2000) > > > > >>>>>> [ 2.701727] WARNING: CPU: 0 PID: 1 at arch/riscv/mm/physaddr.c:16 > > > > >>>>>> __virt_to_phys+0x7e/0x86 > > > > >>>>>> [ 2.702207] Modules linked in: > > > > >>>>>> [ 2.702393] CPU: 0 PID: 1 Comm: swapper/0 Tainted: G W > > > > >>>>>> 5.17.0-rc1 #1 > > > > >>>>>> [ 2.702806] Hardware name: riscv-virtio,qemu (DT) > > > > >>>>>> [ 2.703051] epc : __virt_to_phys+0x7e/0x86 > > > > >>>>>> [ 2.703298] ra : __virt_to_phys+0x7e/0x86 > > > > >>>>>> [ 2.703547] epc : ffffffff80008448 ra : ffffffff80008448 sp : > > > > >>>>>> ffff8f800021bde0 > > > > >>>>>> [ 2.703977] gp : ffffffff80ed9b30 tp : ffffaf8001230000 t0 : > > > > >>>>>> ffffffff80eea56f > > > > >>>>>> [ 2.704704] t1 : ffffffff80eea560 t2 : 0000000000000000 s0 : > > > > >>>>>> ffff8f800021be00 > > > > >>>>>> [ 2.705153] s1 : ffffffff806c2000 a0 : 000000000000004f a1 : > > > > >>>>>> ffffffff80e723d8 > > > > >>>>>> [ 2.705555] a2 : 0000000000000010 a3 : fffffffffffffffe a4 : > > > > >>>>>> 0000000000000000 > > > > >>>>>> [ 2.706027] a5 : 0000000000000000 a6 : 0000000000000005 a7 : > > > > >>>>>> ffffffffffffffff > > > > >>>>>> [ 2.706474] s2 : ffffffff80b80b08 s3 : 00000000000000c2 s4 : > > > > >>>>>> ffffffff806c2000 > > > > >>>>>> [ 2.706891] s5 : ffffffff80edba10 s6 : ffffffff80edb960 s7 : > > > > >>>>>> 0000000000000001 > > > > >>>>>> [ 2.707290] s8 : 00000000000000ff s9 : ffffffff80b80b40 s10: > > > > >>>>>> 00000000000000cc > > > > >>>>>> [ 2.707689] s11: ffffaf807e1fcf00 t3 : 0000000000000076 t4 : > > > > >>>>>> ffffffffffffffff > > > > >>>>>> [ 2.708092] t5 : 00000000000001f2 t6 : ffff8f800021bb48 > > > > >>>>>> [ 2.708433] status: 0000000000000120 badaddr: 0000000000000000 > > > > >>>>>> cause: 0000000000000003 > > > > >>>>>> [ 2.708919] [] free_reserved_area+0x72/0x19a > > > > >>>>>> [ 2.709296] [] free_initmem+0x6c/0x7c > > > > >>>>>> [ 2.709648] [] kernel_init+0x3a/0x10a > > > > >>>>>> [ 2.709993] [] ret_from_exception+0x0/0xc > > > > >>>>>> [ 2.710310] ---[ end trace 0000000000000000 ]--- > > > > >>>>>> > > > > >>>>> I was able to reproduce this: the first one regarding init_zero_pfn is > > > > >>>>> legit but not wrong, I have to check when it was introduced and how to > > > > >>>>> fix this. > > > > >>>>> Regarding the huge batch that follows, at first sight, I would say > > > > >>>>> this is linked to my sv48 patchset but that does not seem important as > > > > >>>>> the address is a kernel mapping address so the use of virt_to_phys is > > > > >>>>> right. > > > > >>>>> > > > > >>>>>> On Wed, Feb 16, 2022 at 5:09 PM Aleksandr Nogikh wrote: > > > > >>>>>>> On Wed, Feb 16, 2022 at 12:56 PM Dmitry Vyukov wrote: > > > > >>>>>>>> On Wed, 16 Feb 2022 at 12:47, Aleksandr Nogikh wrote: > > > > >>>>>>>>> On Wed, Feb 16, 2022 at 11:37 AM Aleksandr Nogikh wrote: > > > > >>>>>>>>>> Hi Alex, > > > > >>>>>>>>>> > > > > >>>>>>>>>> On Wed, Feb 16, 2022 at 5:14 AM Alexandre Ghiti wrote: > > > > >>>>>>>>>>> Hi Dmitry, > > > > >>>>>>>>>>> > > > > >>>>>>>>>>> On 2/15/22 18:12, Dmitry Vyukov wrote: > > > > >>>>>>>>>>>> On Wed, 2 Feb 2022 at 14:18, Alexandre Ghiti > > > > >>>>>>>>>>>> wrote: > > > > >>>>>>>>>>>>> Hi Aleksandr, > > > > >>>>>>>>>>>>> > > > > >>>>>>>>>>>>> On Wed, Feb 2, 2022 at 12:08 PM Aleksandr Nogikh wrote: > > > > >>>>>>>>>>>>>> Hello, > > > > >>>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> syzbot has already not been able to fuzz its RISC-V instance for 97 > > > > >>>>>>>>>>>>> That's a longtime, I'll take a look more regularly. > > > > >>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> days now because the compiled kernel cannot boot. I bisected the issue > > > > >>>>>>>>>>>>>> to the following commit: > > > > >>>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> commit 54c5639d8f507ebefa814f574cb6f763033a72a5 > > > > >>>>>>>>>>>>>> Author: Alexandre Ghiti > > > > >>>>>>>>>>>>>> Date: Fri Oct 29 06:59:27 2021 +0200 > > > > >>>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> riscv: Fix asan-stack clang build > > > > >>>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> Apparently, the problem appears on GCC-built RISC-V kernels with KASAN > > > > >>>>>>>>>>>>>> enabled. In the previous message syzbot mentions > > > > >>>>>>>>>>>>>> "riscv64-linux-gnu-gcc (Debian 10.2.1-6) 10.2.1 20210110, GNU ld (GNU > > > > >>>>>>>>>>>>>> Binutils for Debian) 2.35.2", but the issue also reproduces finely on > > > > >>>>>>>>>>>>>> a newer GCC compiler: "riscv64-linux-gnu-gcc (Debian 11.2.0-10) > > > > >>>>>>>>>>>>>> 11.2.0, GNU ld (GNU Binutils for Debian) 2.37". > > > > >>>>>>>>>>>>>> For convenience, I also duplicate the .config file from the bot's > > > > >>>>>>>>>>>>>> message: https://syzkaller.appspot.com/x/.config?x=522544a2e0ef2a7d > > > > >>>>>>>>>>>>>> > > > > >>>>>>>>>>>>>> Can someone with KASAN and RISC-V expertise please take a look? > > > > >>>>>>>>>>>>> I'll take a look at that today. > > > > >>>>>>>>>>>>> > > > > >>>>>>>>>>>>> Thanks for reporting the issue, > > > > >>>>>>>>>>> I took a quick look, not enough to fix it but I know the issue comes > > > > >>>>>>>>>>> from the inline instrumentation, I have no problem with the outline > > > > >>>>>>>>>>> instrumentation. I need to find some cycles to work on this, my goal is > > > > >>>>>>>>>>> to fix this for 5.17. > > > > >>>>>>>>>> Thanks for the update! > > > > >>>>>>>>>> > > > > >>>>>>>>>> Can you please share the .config with which you tested the outline > > > > >>>>>>>>>> instrumentation? > > > > >>>>>>>>>> I updated the syzbot config to use KASAN_OUTLINE instead of KASAN_INLINE, > > > > >>>>>>>>>> but it still does not boot :( > > > > >>>>>>>>>> > > > > >>>>>>>>>> Here's what I used: > > > > >>>>>>>>>> https://gist.github.com/a-nogikh/279c85c2d24f47efcc3e865c08844138 > > > > >>>>>>>>> Update: it doesn't boot with that big config, but boots if I generate > > > > >>>>>>>>> a simple one with KASAN_OUTLINE: > > > > >>>>>>>>> > > > > >>>>>>>>> make defconfig ARCH=riscv CROSS_COMPILE=riscv64-linux-gnu- > > > > >>>>>>>>> ./scripts/config -e KASAN -e KASAN_OUTLINE > > > > >>>>>>>>> make olddefconfig ARCH=riscv CROSS_COMPILE=riscv64-linux-gnu- > > > > >>>>>>>>> > > > > >>>>>>>>> And it indeed doesn't work if I use KASAN_INLINE. > > > > >>>>>>>> It may be an issue with code size. Full syzbot config + KASAN + KCOV > > > > >>>>>>>> produce hugely massive .text. It may be hitting some limitation in the > > > > >>>>>>>> bootloader/kernel bootstrap code. > > > > >>>>> I took a quick glance and it traps on a KASAN address that is not > > > > >>>>> mapped, either because it is too soon or because the mapping failed > > > > >>>>> somehow. > > > > >>>>> > > > > >>>>> I'll definitely dive into that tomorrow, sorry for being slow here and > > > > >>>>> thanks again for all your work, that helps a lot. > > > > >>>>> > > > > >>>>> Thanks, > > > > >>>>> > > > > >>>>> Alex > > > > >>>>> > > > > >>>>>>> I bisected the difference between the config we use on syzbot and the > > > > >>>>>>> simple one that was generated like I described above. > > > > >>>>>>> Turns out that it's the DEBUG_VIRTUAL config that makes the difference. > > > > >>>>>>> > > > > >>>>>>> make defconfig ARCH=riscv CROSS_COMPILE=riscv64-linux-gnu- > > > > >>>>>>> ./scripts/config -e KASAN -e KASAN_OUTLINE -e DEBUG_VIRTUAL > > > > >>>>>>> make olddefconfig ARCH=riscv CROSS_COMPILE=riscv64-linux-gnu- > > > > >>>>>>> > > > > >>>>>>> And the resulting kernel does not boot. > > > > >>>>>>> My env: the `riscv/fixes` branch, commit > > > > >>>>>>> 6df2a016c0c8a3d0933ef33dd192ea6606b115e3, qemu 6.2.0. > > > > >>>> I fixed a few things today: KASAN + SPARSE_VMEMMAP, DEBUG_VIRTUAL and > > > > >>>> maybe KASAN + KCOV. > > > > >>>> > > > > >>>> With those small fixes, I was able to boot your large dotconfig with > > > > >>>> KASAN_OUTLINE, the inline version still fails, this is my next target > > > > >>>> :) > > > > >>>> I'll push that tomorrow! > > > > >>> Awesome, thank you very much! > > > > >>> Looking forward to finally seeing the instance run :) > > > > >> I sent a patchset which should fix your config with *outline* instrumentation. > > > > > Was this fix merged? The riscv instance still does not boot: > > > > > https://syzkaller.appspot.com/bug?id=5f2ff52ad42cba9f222202219baebd4e63e35127 > > > > > > > > > > > > Yes it has been in Linus tree since 5.18-rc1. I'll take a look at that > > > > this week. > > > > > > > > > > Are you seeing this error or a different one ? I used the > > > syzkaller_defconfig from the patch below on v5.18. > > > > > > https://lore.kernel.org/all/20220419174952.699-1-palmer@rivosinc.com/ > > > > > > [ 15.076116][ T1] Mandatory Access Control activated. > > > [ 15.158241][ T1] AppArmor: AppArmor Filesystem Enabled > > > [ 16.150870][ T1] NET: Registered PF_INET protocol family > > > [ 16.166167][ T1] IP idents hash table entries: 32768 (order: 6, > > > 262144 bytes, linear) > > > [ 16.188727][ T1] Unable to handle kernel paging request at > > > virtual address ffebfffeffff2000 > > > [ 16.192727][ T1] Oops [#1] > > > [ 16.193479][ T1] Modules linked in: > > > [ 16.194687][ T1] CPU: 3 PID: 1 Comm: swapper/0 Not tainted > > > 5.18.0-00001-g37ac279268bf-dirty #9 > > > [ 16.196486][ T1] Hardware name: riscv-virtio,qemu (DT) > > > [ 16.197836][ T1] epc : kasan_check_range+0x9e/0x14e > > > [ 16.199104][ T1] ra : memset+0x1e/0x4c > > > [ 16.200091][ T1] epc : ffffffff804787e0 ra : ffffffff80478f30 sp > > > : ff600000073ffb70 > > > [ 16.201420][ T1] gp : ffffffff85879e80 tp : ff600000073f0000 t0 > > > : 7300000000000000 > > > [ 16.202762][ T1] t1 : ffebfffeffff21ff t2 : 73746e6564692050 s0 > > > : ff600000073ffba0 > > > [ 16.204047][ T1] s1 : 0000000000001000 a0 : ffebfffeffff2200 a1 > > > : 0000000000001000 > > > [ 16.205312][ T1] a2 : 0000000000000001 a3 : ffffffff803a4f32 a4 > > > : ff5ffffffff90000 > > > [ 16.206592][ T1] a5 : ffebfffeffff2000 a6 : 0000004000000000 a7 > > > : ff5ffffffff90fff > > > [ 16.207865][ T1] s2 : ff5ffffffff90000 s3 : 0000000000000000 s4 > > > : ffffffff8467ea90 > > > [ 16.209134][ T1] s5 : 0000000000000000 s6 : ff5ffffffff90000 s7 > > > : 0000000000000000 > > > [ 16.210394][ T1] s8 : 0000000000001000 s9 : ffffffff8587ca40 > > > s10: 0000000000000004 > > > [ 16.211952][ T1] s11: ffffffff858a03a0 t3 : 0000000000000000 t4 > > > : 0000000000000040 > > > [ 16.213469][ T1] t5 : ffebfffeffff2200 t6 : ff600000073ff738 > > > [ 16.214853][ T1] status: 0000000200000120 badaddr: > > > ffebfffeffff2000 cause: 000000000000000d > > > [ 16.216910][ T1] Call Trace: > > > [ 16.217816][ T1] [] pcpu_alloc+0x844/0x1254 > > > [ 16.219110][ T1] [] __alloc_percpu+0x28/0x34 > > > [ 16.220244][ T1] [] ip_rt_init+0x17e/0x382 > > > [ 16.221606][ T1] [] ip_init+0x18/0x30 > > > [ 16.222719][ T1] [] inet_init+0x2a6/0x550 > > > [ 16.223863][ T1] [] do_one_initcall+0x130/0x7dc > > > [ 16.225002][ T1] [] kernel_init_freeable+0x510/0x5b4 > > > [ 16.226273][ T1] [] kernel_init+0x28/0x21c > > > [ 16.227337][ T1] [] ret_from_exception+0x0/0x10 > > > [ 16.229910][ T1] ---[ end trace 0000000000000000 ]--- > > > [ 16.231880][ T1] Kernel panic - not syncing: Fatal exception > > > > > > > > > > Enabling CONFIG_KASAN_VMALLOC=y solves the issue and I am able to boot > > to the userspace. > > I have tried enabling/disabling CONFIG_VMAP_STACK as well. Both works fine. > > > > Looking at the ARM64 Kconfig, KASAN_VMALLOC is enabled if KASAN is enabled. > > This diff seems to work for me. > > diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig > > index 00fd9c548f26..cbf0fe227c77 100644 > > --- a/arch/riscv/Kconfig > > +++ b/arch/riscv/Kconfig > > @@ -122,6 +122,7 @@ config RISCV > > select TRACE_IRQFLAGS_SUPPORT > > select UACCESS_MEMCPY if !MMU > > select ZONE_DMA32 if 64BIT > > + select KASAN_VMALLOC if KASAN > > > > I am not a kasan expert so I am not sure if this is the correct fix or > > just hides the real issue. pcpu_alloc seems to use vmalloc though. > > When this type of thing happens, generally this is because of an error > in the kasan page table, I'll take a look this time, sorry I did not > do this before. > > Thanks for finding this, > > Alex Unfortunately riscv still does not work for syzbot, other bugs piled up. Something has changes wrt user-space pointers: https://groups.google.com/g/syzkaller-bugs/c/lU0GQTZoNQQ/m/iuRBTgU3AAAJ https://syzkaller.appspot.com/bug?extid=2c5da6a0a16a0c4f34aa _______________________________________________ linux-riscv mailing list linux-riscv@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-riscv