linux-kernel - Re: mm: mkfs.ext4 invoked oom-killer on i386

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20200617135758.GA548179@chrisdown.name>
Date:   Wed, 17 Jun 2020 14:57:58 +0100
From:   Chris Down <chris@...isdown.name>
To:     Naresh Kamboju <naresh.kamboju@...aro.org>
Cc:     Michal Hocko <mhocko@...nel.org>,
        Yafang Shao <laoar.shao@...il.com>,
        Anders Roxell <anders.roxell@...aro.org>,
        "Linux F2FS DEV, Mailing List" 
        <linux-f2fs-devel@...ts.sourceforge.net>,
        linux-ext4 <linux-ext4@...r.kernel.org>,
        linux-block <linux-block@...r.kernel.org>,
        Andrew Morton <akpm@...ux-foundation.org>,
        open list <linux-kernel@...r.kernel.org>,
        Linux-Next Mailing List <linux-next@...r.kernel.org>,
        linux-mm <linux-mm@...ck.org>, Arnd Bergmann <arnd@...db.de>,
        Andreas Dilger <adilger.kernel@...ger.ca>,
        Jaegeuk Kim <jaegeuk@...nel.org>,
        Theodore Ts'o <tytso@....edu>, Chao Yu <chao@...nel.org>,
        Hugh Dickins <hughd@...gle.com>,
        Andrea Arcangeli <aarcange@...hat.com>,
        Matthew Wilcox <willy@...radead.org>,
        Chao Yu <yuchao0@...wei.com>, lkft-triage@...ts.linaro.org,
        Johannes Weiner <hannes@...xchg.org>,
        Roman Gushchin <guro@...com>, Cgroups <cgroups@...r.kernel.org>
Subject: Re: mm: mkfs.ext4 invoked oom-killer on i386 - pagecache_get_page

Naresh Kamboju writes:
>mkfs -t ext4 /dev/disk/by-id/ata-TOSHIBA_MG04ACA100N_Y8RQK14KF6XF
>mke2fs 1.43.8 (1-Jan-2018)
>Creating filesystem with 244190646 4k blocks and 61054976 inodes
>Filesystem UUID: 7c380766-0ed8-41ba-a0de-3c08e78f1891
>Superblock backups stored on blocks:
>32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208,
>4096000, 7962624, 11239424, 20480000, 23887872, 71663616, 78675968,
>102400000, 214990848
>Allocating group tables:    0/7453 done
>Writing inode tables:    0/7453 done
>Creating journal (262144 blocks): [   51.544525] under min:0 emin:0
>[   51.845304] under min:0 emin:0
>[   51.848738] under min:0 emin:0
>[   51.858147] under min:0 emin:0
>[   51.861333] under min:0 emin:0
>[   51.862034] under min:0 emin:0
>[   51.862442] under min:0 emin:0
>[   51.862763] under min:0 emin:0

Thanks, this helps a lot. Somehow we're entering mem_cgroup_below_min even when 
min/emin is 0 (which should indeed be the case if you haven't set them in the 
hierarchy).

My guess is that page_counter_read(&memcg->memory) is 0, which means 
mem_cgroup_below_min will return 1.

However, I don't know for sure why that should then result in the OOM killer 
coming along. My guess is that since this memcg has 0 pages to scan anyway, we 
enter premature OOM under some conditions. I don't know why we wouldn't have 
hit that with the old version of mem_cgroup_protected that returned 
MEMCG_PROT_* members, though.

Can you please try the patch with the `>=` checks in mem_cgroup_below_min and 
mem_cgroup_below_low changed to `>`? If that fixes it, then that gives a strong 
hint about what's going on here.

Thanks for your help!