linux-ext4 - Re: LTP: statx06: FAIL: Birth time

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]

Date:   Fri, 17 Jan 2020 16:33:52 -0500
From:   "Theodore Y. Ts'o" <tytso@....edu>
To:     Jan Stancek <jstancek@...hat.com>
Cc:     Naresh Kamboju <naresh.kamboju@...aro.org>, adilger@...ger.ca,
        LTP List <ltp@...ts.linux.it>,
        Daniel Lezcano <daniel.lezcano@...aro.org>,
        Arnd Bergmann <arnd@...db.de>,
        open list <linux-kernel@...r.kernel.org>,
        lkft-triage@...ts.linaro.org, chrubis <chrubis@...e.cz>,
        linux-ext4@...r.kernel.org
Subject: Re: LTP: statx06: FAIL: Birth time < before time

On Fri, Jan 17, 2020 at 12:24:47PM -0500, Jan Stancek wrote:
> > LTP syscalls statx06 test case getting failed from linux next 20200115
> > tag onwards on all x86_64, i386, arm and arm64 devices
> > 
> > Test output:
> > statx06.c:152: FAIL: Birth time < before time
> 
> [CC Theo & linux-ext4]
> 
> It's returning '0' in stx_btime for STATX_ALL or STATX_BTIME.
> 
> Looking at changes, I suspect:
>   commit 927353987d503b24e1813245563cde0c6167af6e
>   Author: Theodore Ts'o <tytso@....edu>
>   Date:   Thu Nov 28 22:26:51 2019 -0500
>     ext4: avoid fetching btime in ext4_getattr() unless requested
> 
> and that perhaps it should be instead...
> 
> diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c
> index c8355f022e6e..6d76eb6d2e7f 100644
> --- a/fs/ext4/inode.c
> +++ b/fs/ext4/inode.c
> @@ -5398,7 +5398,7 @@ int ext4_getattr(const struct path *path, struct kstat *stat,
>         struct ext4_inode_info *ei = EXT4_I(inode);
>         unsigned int flags;
> 
> -       if ((query_flags & STATX_BTIME) &&
> +       if ((request_mask & STATX_BTIME) &&
>             EXT4_FITS_IN_INODE(raw_inode, ei, i_crtime)) {
>                 stat->result_mask |= STATX_BTIME;
>                 stat->btime.tv_sec = ei->i_crtime.tv_sec;

Yep, nice catch!  Unfortunately we don't have a test like this in
xfstests, or I would have caught this sooner.

I've fixed this in the dev branch, so it will hopefully be fixed in
the future linux-next tgs.

							- Ted