mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Bernd Schubert <bernd-schubert@gmx.de>
To: John M Flinchbaugh <john@hjsoft.com>, reiserfs-list@namesys.com
Cc: Sam Vilain <sam@vilain.net>, linux-kernel@vger.kernel.org
Subject: Re: 2.6.15 Bug? New security model?
Date: Wed, 8 Feb 2006 13:14:59 +0100	[thread overview]
Message-ID: <200602081314.59639.bernd-schubert@gmx.de> (raw)
In-Reply-To: <20060208053732.GA13560@butterfly.hjsoft.com>

CC'ed also the reiser-list


On Wednesday 08 February 2006 06:37, John M Flinchbaugh wrote:
> On Wed, Feb 08, 2006 at 02:31:46PM +1300, Sam Vilain wrote:
> > Bernd Schubert wrote:
> > >With 2.6.15:
> > >bathl:~# touch /var/run/test
> > >touch: cannot touch `/var/run/test': Permission denied
> > >With 2.6.13:
> > >bathl:~# touch /var/run/test
> > >(No error message)
> >
> > Some ideas; ACLs, SELinux, Attributes, Capabilities.
>
> lsattr -d /var/run && lsattr /var/run

Indeed, with 2.6.13

bathl:~# lsattr -d /var/run
lsattr: Inappropriate ioctl for device While reading flags on /var/run

with 2.6.15.3

bathl:~# cat lsatr.out.2.6.15
--S-ia-AcBZXEj-t- /var/run


After the problem came up, I already suspected something like this and 
therefore already had the kernel recompiled without xattr support, so I  
don't know why lsattr shows something for 2.6.15 and nothing for 2.6.13.

here the reiser part from the kernel config of 2.6.13 

CONFIG_REISERFS_FS=y
# CONFIG_REISERFS_CHECK is not set
CONFIG_REISERFS_PROC_INFO=y
CONFIG_REISERFS_FS_XATTR=y
CONFIG_REISERFS_FS_POSIX_ACL=y
CONFIG_REISERFS_FS_SECURITY=y
# CONFIG_JFS_FS is not set
CONFIG_FS_POSIX_ACL=y

here 2.6.15

CONFIG_REISERFS_FS=y
# CONFIG_REISERFS_CHECK is not set
CONFIG_REISERFS_PROC_INFO=y
# CONFIG_REISERFS_FS_XATTR is not set
# CONFIG_JFS_FS is not set
CONFIG_FS_POSIX_ACL=y
# CONFIG_XFS_FS is not set
# CONFIG_MINIX_FS is not set
# CONFIG_ROMFS_FS is not set


Anyway, I never set any attributes using chattr, so these seem to be random 
attributes, nice.



>
> I saw very similar things going from 2.6.15.1 to 2.6.15.2.  2.6.15.2's
> changelog advertises a fix to reenable extended attributes on reiserfs.
> On one machine this is fine, and lsattr shows no attributes enabled
> (----------), but on another machine, I ended up with all sorts of crazy
> attributes set seemingly randomly -- compression, experimental flags,
> immutable, append-only, all over the map.
>
> I tried clearing them (chattr -R = /var ...etc), but I still found a
> file here and there which refused to be removed, even though lsattr
> showed no flags for it.  After a restart or 2, I saw some attributes
> revert back and I started having trouble removing files from /var/run
> and other places again.
>
> I ended up reverting back to 2.6.15.1 until I have a chance to
> investigate further and try to come up with something reportable.  In
> 2.6.15.1, attributes didn't work at all, giving an ioctl error, though
> the same kernel options were used.  I suspect this is the fix to which
> the Changelog is referring.
>
> I must wonder if I'm suffering from some sort of fs corruption which
> only manifests itself in the attribute settings, and which a reisefsck
> doesn't recognize or correct.  I could be tempted to recreate the
> filesystems from scratch to see if they still have issues.

-- 
Bernd Schubert
PCI / Theoretische Chemie
Universität Heidelberg
INF 229
69120 Heidelberg


  reply	other threads:[~2006-02-08 12:15 UTC|newest]

Thread overview: 18+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-02-08  1:12 Bernd Schubert
2006-02-08  1:31 ` Sam Vilain
2006-02-08  5:37   ` John M Flinchbaugh
2006-02-08 12:14     ` Bernd Schubert [this message]
2006-02-08 20:50       ` Chris Wright
2006-02-08 21:46         ` Bernd Schubert
2006-02-08 22:11           ` Chris Wright
2006-02-11 21:55             ` Sergey Vlasov
2006-02-11 23:50               ` Bernd Schubert
2006-02-12 18:44                 ` Sergey Vlasov
2006-02-12 17:57               ` Jeff Mahoney
2006-02-12 19:21                 ` Sergey Vlasov
2006-02-12 23:03                   ` Bernd Schubert
2006-02-13 17:52                     ` Sergey Vlasov
2006-02-12 23:10                 ` Bernd Schubert
2006-02-13  5:47               ` Hans Reiser
2006-02-13 15:24                 ` Jeff Mahoney
2006-02-13 16:39                   ` Hans Reiser

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=200602081314.59639.bernd-schubert@gmx.de \
    --to=bernd-schubert@gmx.de \
    --cc=john@hjsoft.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=reiserfs-list@namesys.com \
    --cc=sam@vilain.net \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

Powered by JetHome