linux-kernel - Re: [PATCH] mm: memory-failure: convert to pr

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] [day] [month] [year] [list]

Message-ID: <20220727022406.GA3350995@hori.linux.bs1.fc.nec.co.jp>
Date:   Wed, 27 Jul 2022 02:24:12 +0000
From:   HORIGUCHI NAOYA(堀口　直也) 
        <naoya.horiguchi@....com>
To:     Kefeng Wang <wangkefeng.wang@...wei.com>
CC:     Miaohe Lin <linmiaohe@...wei.com>,
        Andrew Morton <akpm@...ux-foundation.org>,
        "linux-mm@...ck.org" <linux-mm@...ck.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] mm: memory-failure: convert to pr_fmt()

On Wed, Jul 27, 2022 at 10:09:50AM +0800, Kefeng Wang wrote:
> 
> On 2022/7/27 9:12, HORIGUCHI NAOYA(堀口 直也) wrote:
> > On Tue, Jul 26, 2022 at 04:10:46PM +0800, Kefeng Wang wrote:
> > > Use pr_fmt to prefix pr_<level> output.
> > > 
> > > Signed-off-by: Kefeng Wang <wangkefeng.wang@...wei.com>
> > There're other classes of prefixes like "Unpoison:" and "soft offline:" in
> > mm/memory-failure.c.  So simply adding subsystem-wide pr_fmt() changes to
> > print out two prefixes like below:
> > 
> >    [ 1479.892348] Memory failure: soft offline: 0x1b8200: page migration failed 1, type 0x57ffffc00a000c(uptodate|dirty|mappedtodisk|swapbacked|node=1|zone=2|lastcpupid=0x1fffff)
> > 
> >    [ 1486.871130] Memory failure: Unpoison: Software-unpoisoned page 0x14f00
> > 
> > Do you intend this change?
> 
> 
> Firstly I think they are all in memory failure, with this prefix is no bad
> impact,
> 
> soft_offline_page() and unpoison_memory()  are used by error injection,

Yes, this is right, and I think that testing systems might save console
output to detect issues afterward, so the longer prefix might consume more
storage of such systems.  So I'd like to keep the output shorter if possible.

> 
> but soft_offline_page() could be used in memory hotremove,
> 
> so let's add
> 
> '#undef pr_fmt' before unpoison_pr_info definition, options? thanks.

Yes, I agree with this approach, thank you.

- Naoya Horiguchi