linux-kernel - Re: [PATCH 4.19] ext4: fix race condition between ext4_ioctl_setflags and ext4

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: <Ytb+ji56S/de/5Rm@mit.edu>
Date:   Tue, 19 Jul 2022 14:57:18 -0400
From:   "Theodore Ts'o" <tytso@....edu>
To:     Greg KH <gregkh@...uxfoundation.org>
Cc:     Baokun Li <libaokun1@...wei.com>, stable@...r.kernel.org,
        linux-ext4@...r.kernel.org, adilger.kernel@...ger.ca, jack@...e.cz,
        ritesh.list@...il.com, lczerner@...hat.com, enwlinux@...il.com,
        linux-kernel@...r.kernel.org, yi.zhang@...wei.com,
        yebin10@...wei.com, yukuai3@...wei.com,
        Hulk Robot <hulkci@...wei.com>
Subject: Re: [PATCH 4.19] ext4: fix race condition between
 ext4_ioctl_setflags and ext4_fiemap

On Tue, Jul 19, 2022 at 02:25:18PM +0200, Greg KH wrote:
> 
> 95% of the time we take a patch that is not in Linus's tree, it is buggy
> and causes problems in the long run.

So if we really want a 4.19 LTS specific patch, I'd be OK with signing
off on it from an ext4 perspective.... IF AND ONLY IF someone is
willing to tell me that they ran "kvm-xfstests -c ext4/all -g auto" or
the equivalent before and after applying the patch, and is willing to
certify that there are no test regressions.

Helpful links:

  * https://thunk.org/gce-xfstests
  * https://thunk.org/android-xfstests
  * Documentation links from https://github.com/tytso/xfstests-bld
      * https://github.com/tytso/xfstests-bld/blob/master/Documentation/kvm-quickstart.md
      * https://github.com/tytso/xfstests-bld/blob/master/Documentation/gce-xfstests.md

(Note that running "-c ext4/all -g auto" will take some 12+ hours if
the tests are run serially, which is why using gce-xfstests's
lightweight test manager to run the file system test configurations in
parallel is a big win.)

> Or better yet, take the effort here and move off of 4.19 to a newer
> kernel without this problem in it.  What is preventing you from doing
> that today?  4.19 is not going to be around for forever, and will
> probably not even be getting fixes for stuff like RETBLEED, so are you
> _SURE_ you want to keep using it?

Or yeah, maybe it's better/cheaper/time for you to move off of 4.19.  :-)

   	       	    	    	       - Ted

P.S.  If we go down this path, Greg K-H may also insist on getting the
bug fix to the 5.4 LTS kernel, so that a bug isn't just fixed in 4.19
LTS but not 5.4 LTS.  In which case, the same requirement of running
"-c ext4/all -g auto" and showing that there are no test regressions
is going to be a requirement for 5.4 LTS as well.