From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from mail-wm1-f46.google.com (mail-wm1-f46.google.com [209.85.128.46]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 762EF31C58E for ; Mon, 15 Dec 2025 04:45:02 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=209.85.128.46 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1765773904; cv=none; b=e5y21OzwiyTTfUnHIqa/4UZ/rRF8zHTrh09NYa5ZX1M0sgNts56ClD3zS309OO5L55KggRaiHcYxd9A8OllFk+rung3VDPUH4QObDUCXkcTa7BMNsVp0usJc0J64kjer0IH2sEK6Wh62lzAhCo7DA4QYrKw//WyBoH6G+zCRpmU= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1765773904; c=relaxed/simple; bh=7PisnT3hiijUh9oixgDdlnQnza1382MErLdvnXuWzWI=; h=Date:From:To:Cc:Subject:Message-ID:References:MIME-Version: Content-Type:Content-Disposition:In-Reply-To; b=He1Cqz8nWbc7hCmLBve4ci6m3wYyND3vSwUgcIUJnDiApCh0uoal9QBzC8hsSC+wVl0Q1n34ohCBA9M7KkdfNQ+p1D3uBt5dliA11gugyWVGDwCwZzjHZzcqBoeL2iFG2v6Ls8ELCPYnb794ybojCK/M2OOKdfdLdK7Bf+axryA= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=debian.org; spf=pass smtp.mailfrom=gmail.com; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b=hJYtnLXK; arc=none smtp.client-ip=209.85.128.46 Authentication-Results: smtp.subspace.kernel.org; dmarc=none (p=none dis=none) header.from=debian.org Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=gmail.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=gmail.com header.i=@gmail.com header.b="hJYtnLXK" Received: by mail-wm1-f46.google.com with SMTP id 5b1f17b1804b1-4779cc419b2so31361015e9.3 for ; Sun, 14 Dec 2025 20:45:02 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20230601; t=1765773901; x=1766378701; darn=vger.kernel.org; h=in-reply-to:content-transfer-encoding:content-disposition :mime-version:references:message-id:subject:cc:to:from:date:sender :from:to:cc:subject:date:message-id:reply-to; bh=ML0r7ObM9/T5sg62BeFcZqS/ja5rEbtFn4IwdOlHqiw=; b=hJYtnLXK9SUR9/2m3SsBWEfVJj0li8o7nkGWPJVsnI7N1L0QDbiQQiYQKCNNxavXee OKl5NFlXTFeyIOLzx6e49Qls4MvKFwKrvxsqhr1L/I9uiI1SGUOFtV2VYi7E0OXYuRyA 7Qi5TnkNCvi44+YS0fGk14/yG5ej8Lfv3yMzAX2r+lrN426ygBbbDBEDX7BKOV0teTOS BnndKKEg+S1js8H5iX+W7B1rYNxOcuFEQgfB58hgj21y3mBI9E2F0Q9P4ZcpUBh+b+14 diUiOsyilUugf7UvcULpjPajISlENpZqxmcefikM6nQSlOW446m9Ebqa/FD2e7jJ3Yk2 Rk6g== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20230601; t=1765773901; x=1766378701; h=in-reply-to:content-transfer-encoding:content-disposition :mime-version:references:message-id:subject:cc:to:from:date:sender :x-gm-gg:x-gm-message-state:from:to:cc:subject:date:message-id :reply-to; bh=ML0r7ObM9/T5sg62BeFcZqS/ja5rEbtFn4IwdOlHqiw=; b=ADFIp/4Es1ktc3eoF40ut2B4kyetMi/kypzb+3a+AsxPTn2yPGCZ30CPrKtaC/9sHZ 8hh3vejMX+ByD0ywW5mgT9QKhDntDEmy557jeDGR1D8LhgtnEVnOgrNXmW4WVHunfFh+ URmihBvzqwkuYhkTJbXN5i0FitjXyJMK4COj9VuTTOhpApCewWoVOO9lRxGc+qQR1xGn uKnzHwYairi+UT7PDBdHWOxs+FMWN9w43UIyU0CArzNAXSSNmOVTBND8vcAMH1th81SE vqf9R6D97zhCOVRRYioHW/SwAwf4ZbtW7UC5KHUCblg37AZSKi5Ec+bdF5fCnZY8YdkF C23A== X-Forwarded-Encrypted: i=1; AJvYcCXfGYTwdbubeG0rdz16sPP5mMzuOxtRzCDZM7QCN3YBoGp4Nq+Url6h55rWHznEUwTLetAy24n/bWEopIw=@vger.kernel.org X-Gm-Message-State: AOJu0YxF1oRXAMBawj2HnCmnhcepUMecdpF8UT04jNYaEtakFdsGel5E qz7wOJ2cCWuU58mR/YeJv8ix30zoFTXXSmzr8GWPgo6cCgwZyoAPlk9S X-Gm-Gg: AY/fxX7vpDgig/Y8ybwH7o7X2KI7aKYDMP+dWf4ZtqzZGYFdk8W7Fwpr8/wr3KGD8Wl YeaPgLTzamV+HIB9Z5zKEkrwu6BirEbLxtZdqlu9666z5zgD/DySL1pNQeg4OFeNpkzWeOlqLGM w3WQZpOttn7SjKbLqloYuEGp3T9rilOG5Tswv4G+n/IYylAHCwxpROX8iEfMifKjhTpEYJHKGz7 fvrBzT20eyN/5mxx1g9ZsEt1grOteFQXdoAOUUltAXQO5dnYBc84kenO0QdA6VRj8Gwp+AzIPvs ksJj9dzS9GclT7zIt+W9efU3k8+hfnnq0XXZIGx2l3KGF8orgJc+JdDS8ZdFyLssOu09uveT8Dl T6dRw1ph8djb2+HlmTrd/krBVle9gvhzbytbMNR+1bC6Mst44aXoZ73VDCjIadDsErHTh6NyEzM G1j8BOphXYgTF8HaUuS65SO3AlL1j7g9iIYgiq3rU1J48q X-Google-Smtp-Source: AGHT+IGW/x/KIsUSoYUE4Tp2R5V0W8WENFntKgFb/J7RE0N6Iq/+aWtRub+6skayu3CX1CkhCxZQTQ== X-Received: by 2002:a05:6000:601:b0:430:fd0f:290c with SMTP id ffacd0b85a97d-430fd0f2c21mr1774879f8f.41.1765773900340; Sun, 14 Dec 2025 20:45:00 -0800 (PST) Received: from eldamar.lan (c-82-192-244-13.customer.ggaweb.ch. [82.192.244.13]) by smtp.gmail.com with ESMTPSA id ffacd0b85a97d-42ff626b591sm14288366f8f.15.2025.12.14.20.44.58 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Sun, 14 Dec 2025 20:44:59 -0800 (PST) Sender: Salvatore Bonaccorso Received: by eldamar.lan (Postfix, from userid 1000) id 02360BE2EE7; Mon, 15 Dec 2025 05:44:57 +0100 (CET) Date: Mon, 15 Dec 2025 05:44:57 +0100 From: Salvatore Bonaccorso To: Joanne Koong Cc: =?iso-8859-1?Q?J=2E_Neusch=E4fer?= , 1120058@bugs.debian.org, Jingbo Xu , Jeff Layton , Miklos Szeredi , linux-fsdevel@vger.kernel.org, linux-kernel@vger.kernel.org, regressions@lists.linux.dev Subject: Re: [regression] 0c58a97f919c ("fuse: remove tmp folio for writebacks and internal rb tree") results in suspend-to-RAM hang on AMD Ryzen 5 5625U on test scenario involving podman containers, x2go and openjdk workload Message-ID: References: <176227232774.2636.13973205036417925311.reportbug@probook> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: 8bit In-Reply-To: Hi Joanne, On Mon, Dec 15, 2025 at 06:37:42AM +0800, Joanne Koong wrote: > On Sun, Dec 14, 2025 at 10:27 PM Salvatore Bonaccorso wrote: > > > > Hi Joanne, > > > > In Debian J. Neuschäfer reported an issue where after 0c58a97f919c > > ("fuse: remove tmp folio for writebacks and internal rb tree") a > > specific, but admittely not very minimal workload, involving podman > > contains, x2goserver and a openjdk application restults in > > suspend-to-ram hang. > > > > The report is at https://bugs.debian.org/1120058 and information on > > bisection and the test setup follows: > > > > On Sun, Nov 30, 2025 at 02:11:13PM +0100, J. Neuschäfer wrote: > > > On Fri, Nov 28, 2025 at 09:10:25PM +0100, Salvatore Bonaccorso wrote: > > > > Control: found -1 6.17.8-1 > > > > > > > > Hi, > > > > > > > > On Fri, Nov 28, 2025 at 11:50:48AM +0100, J. Neuschäfer wrote: > > > > > On Wed, Nov 05, 2025 at 06:09:43AM +0100, Salvatore Bonaccorso wrote: > > > [...] > > > > > I can reproduce the bug fairly reliably on 6.16/17 by running a specific > > > > > podman container plus x2go (not entirely sure which parts of this is > > > > > necessary). > > > > > > > > Okay if you have a very reliable way to reproduce it, would you be > > > > open to make "your hands bit dirty" and do some bisecting on the > > > > issue? > > > > > > Thank you for your detailed instructions! I've already started and completed > > > the git bisect run in the meantime. I had to restart a few times due to > > > mistakes, but I was able to identify the following upstream commit as the > > > commit that introduced the issue: > > > > > > https://git.kernel.org/linus/0c58a97f919c24fe4245015f4375a39ff05665b6 > > > > > > fuse: remove tmp folio for writebacks and internal rb tree > > > > > > The relevant commit history is as follows: > > > > > > * 2619a6d413f4c3 Merge tag 'fuse-update-6.16' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse <-- bad > > > |\ > > > | * dabb9039102879 fuse: increase readdir buffer size > > > | * 467e245d47e666 readdir: supply dir_context.count as readdir buffer size hint > > > | * c31f91c6af96a5 fuse: don't allow signals to interrupt getdents copying > > > | * f3cb8bd908c72e fuse: support large folios for writeback > > > | * 906354c87f4917 fuse: support large folios for readahead > > > | * ff7c3ee4842d87 fuse: support large folios for queued writes > > > | * c91440c89fbd9d fuse: support large folios for stores > > > | * cacc0645bcad3e fuse: support large folios for symlinks > > > | * 351a24eb48209b fuse: support large folios for folio reads > > > | * d60a6015e1a284 fuse: support large folios for writethrough writes > > > | * 63c69ad3d18a80 fuse: refactor fuse_fill_write_pages() > > > | * 3568a956932621 fuse: support large folios for retrieves > > > | * 394244b24fdd09 fuse: support copying large folios > > > | * f09222980d7751 fs: fuse: add dev id to /dev/fuse fdinfo > > > | * 18ee43c398af0b docs: filesystems: add fuse-passthrough.rst > > > | * 767c4b82715ad3 MAINTAINERS: update filter of FUSE documentation > > > | * 69efbff69f89c9 fuse: fix race between concurrent setattrs from multiple nodes > > > | * 0c58a97f919c24 fuse: remove tmp folio for writebacks and internal rb tree <-- first bad commit > > > | * 0c4f8ed498cea1 mm: skip folio reclaim in legacy memcg contexts for deadlockable mappings > > > | * 4fea593e625cd5 fuse: optimize over-io-uring request expiration check > > > | * 03a3617f92c2a7 fuse: use boolean bit-fields in struct fuse_copy_state > > > | * a5c4983bb90759 fuse: Convert 'write' to a bit-field in struct fuse_copy_state > > > | * 2396356a945bb0 fuse: add more control over cache invalidation behaviour > > > | * faa794dd2e17e7 fuse: Move prefaulting out of hot write path > > > | * 0486b1832dc386 fuse: change 'unsigned' to 'unsigned int' > > > * 0fb34422b5c223 Merge tag 'vfs-6.16-rc1.netfs' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs <-- good > > > > > > The first and last commits shown are merge commits done by Linus Torvalds. The > > > fuse-update branch was based on v6.15-rc1, under which I can't run my test due > > > to an unrelated bug, so I ended up merging in 0fb34422b5c223 to test the > > > commits within the fuse-update branch. e.g.: > > > > > > git reset --hard 394244b24fdd09 && git merge 0fb34422b5c223 && make clean && make > > > > > > > > > I have also verified that the issue still happens on v6.18-rc7 but I wasn't > > > able to revert 0c58a97f919 on top of this release, because a trivial revert > > > is not possible. > > > > > > My test case consists of a few parts: > > > > > > - A podman container based on the "debian:13" image (which points to > > > docker.io/library/debian via /etc/containers/registries.conf.d/shortnames.conf), > > > where I installed x2goserver and a openjdk-21-based application; It runs the > > > OpenSSH server and port 22 is exposed as localhost:2001 > > > - x2goclient to start a desktop session in the container > > > > > > Source code: https://codeberg.org/neuschaefer/re-workspace > > > > > > I suspect, but haven't verified, that the X server in the container somehow > > > uses the FUSE-emulated filesystem in the container to create a file that is > > > used with mmap (perhaps to create shared pages as frame buffers). > > > > > > > > > Raw bisect notes: > > > > > > good: > > > - v6.12.48+deb13-amd64 > > > - v6.12.59 > > > - v6.12 > > > - v6.14 > > > - v6.15-1304-g14418ddcc2c205 > > > - v6.15-10380-gec71f661a572 > > > - v6.15-10888-gb509c16e1d7cba > > > - v6.15-rc7-357-g8e86e73626527e > > > - v6.15-10933-g4c3b7df7844340 > > > - v6.15-10954-gd00a83477e7a8f > > > - v6.15-rc7-366-g438e22801b1958 (CONFIG_X86_5LEVEL=y) > > > - v6.15-rc4-126-g07212d16adc7a0 > > > - v6.15-10958-gdf7b9b4f6bfeb1 <-- first parent, 5LEVEL doesn't exist > > > - v6.15-rc4-00127-g4d62121ce9b5 > > > - v6.15-rc7-375-g61374cc145f4a5 <-- second parent, `X86_5LEVEL=y` > > > - v6.15-rc7-375-g61374cc145f4a5 <-- second parent, `X86_5LEVEL=n` > > > - v6.15-11061-g7f9039c524a351: "first bad", actually good. merge of df7b9b4f6bfeb1 61374cc145f4a5 > > > - v6.15-11093-g0fb34422b5c223 > > > - v6.15-rc1-7-g0c4f8ed498cea1 + merge = v6.15-11101-gaec20ffad33068 > > > > > > testing: > > > - v6.18-rc7 + revert: doesn't apply > > > > > > weird (ssh doesn't work): > > > - v6.15-rc1-1-g0486b1832dc386 > > > - v6.15-rc1-10-g767c4b82715ad3 > > > - v6.15-rc1-13-g394244b24fdd09: folio stuff > > > - v6.15-rc1-22-gf3cb8bd908c72e > > > - v6.15-rc1-23-gc31f91c6af96a5 > > > - next-20251128 > > > > > > bad: > > > - v6.15-rc1-8-g0c58a97f919c24 + merge = v6.15-11102-gdfc4869c8ef1f0 first bad commit > > > - v6.15-rc1-9-g69efbff69f89c9 + merge = v6.15-11103-ga7b103c57680ce > > > - v6.15-rc1-11-g18ee43c398af0b + merge = v6.15-11105-g4ad0d4fa61974c > > > - v6.15-rc1-13-g394244b24fdd09 + merge = v6.15-11107-g37da056b3b873b > > > - v6.15-11119-g2619a6d413f4c3: merge of 0fb34422b5c223 (last good) dabb9039102879 (fuse branch) > > > - v6.15-11165-gfd1f8473503e5b: confirmed bad > > > - v6.15-11401-g69352bd52b2667 > > > - v6.15-12422-g2c7e4a2663a1ab > > > - regulator-fix-v6.16-rc2-372-g5c00eca95a9a20 > > > - v6.16.12 > > > - v6.16.12 again > > > - v6.16.12+deb14+1-amd64 > > > - v6.18-rc7 > > > > Would that ring some bells to you which make this tackable? > > Hi Salvatore, > > This looks like the same issue reported in this thread [1]. The lockup > occurs when there's a faulty fuse server on the system that doesn't > complete a write request. Prior to commit 0c58a97f919c24 ("fuse: > remove tmp folio for writebacks and internal rb tree"), syncs on fuse > filesystems were effectively no-ops. This patch upstream [2] reverts > the behavior back to that. I'll send v2 of that patch and work on > getting it merged as soon as possible. Thank you and apologies I did miss the already reported issue. J. Neuschäfer might it be possible you test the patch and report back if that fixes your issue? Regards, Salvatore