[ 20.717978] ------------[ cut here ]------------
[ 20.718003] WARNING: at fs/proc/base.c:1123 oom_adjust_write+0x286/0x2a0()
[ 20.718011] Hardware name: N150P/N210P/N220P
[ 20.718020] sshd (2910): /proc/2910/oom_adj is deprecated, please use /proc/2910/oom_score_adj instead.
[ 20.718029] Modules linked in: fuse brcmsmac(C) brcmutil(C) [last unloaded: scsi_wait_scan]
[ 20.718052] Pid: 2910, comm: sshd Tainted: G C 3.0.0+ #1
[ 20.718059] Call Trace:
[ 20.718076] [<ffffffff81041f9a>] warn_slowpath_common+0x7a/0xb0
[ 20.718090] [<ffffffff81042071>] warn_slowpath_fmt+0x41/0x50
[ 20.718104] [<ffffffff8104d2d5>] ? ns_capable+0x25/0x50
[ 20.718116] [<ffffffff81292e45>] ? kstrtoint+0x15/0x40
[ 20.718129] [<ffffffff81171bf6>] oom_adjust_write+0x286/0x2a0
[ 20.718145] [<ffffffff8111514e>] vfs_write+0xae/0x180
[ 20.718157] [<ffffffff81115465>] sys_write+0x45/0x90
[ 20.718172] [<ffffffff816f143b>] system_call_fastpath+0x16/0x1b
[ 20.718183] ---[ end trace b16d62b4b2b98e24 ]---
There was a warning there for a year, and the distributions have one
more year to upgrade to the newer software, so it's valid to apply
pressure now.
Could we do it in a way that made it more clear that the problem was
an old version of openssh-server and not the kernel? The information
is there, but it's overwhelmed in stack trace spam. Also the kernel
is tainted now because of sshd. Maybe the message could include a
link to a wiki page that told people which version off sshd uses the
new proc file.
regards,
dan carpenter
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majo...@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
* On 03.08.2011 12:58 AM, Dan Carpenter wrote:
> Debian testing and Fedora 14 (and 15) are still using an old version
> of sshd so in 3.1, I get a popup in gnome that a "kernel crash has
> been detected."
> [...]
> There was a warning there for a year, and the distributions have one
> more year to upgrade to the newer software, so it's valid to apply
> pressure now.
>
> Could we do it in a way that made it more clear that the problem was
> an old version of openssh-server and not the kernel? The information
> is there, but it's overwhelmed in stack trace spam. Also the kernel
> is tainted now because of sshd. Maybe the message could include a
> link to a wiki page that told people which version off sshd uses the
> new proc file.
Please read and add your thoughts to this thread:
https://lkml.org/lkml/2011/8/1/433
(Actually you just created a duplicate discussion. ;))
Best regards,
Mihai
> Could we do it in a way that made it more clear that the problem was
> an old version of openssh-server and not the kernel? The information
> is there, but it's overwhelmed in stack trace spam. Also the kernel
> is tainted now because of sshd. Maybe the message could include a
> link to a wiki page that told people which version off sshd uses the
> new proc file.
>
Would you like to ack http://marc.info/?l=linux-kernel&m=131227886118365?
As a side note, Fedora 15 has new openssh packages with the backport of the
5.7p1 fix to use /proc/<pid>/oom_score_adj.
Kudos to the Fedora dev team for the quick turnaround on my bug report:
https://bugzilla.redhat.com/show_bug.cgi?id=727335
--alessandro
�"There's always a siren singing you to shipwreck"
�� (Radiohead, "There There")