View previous topic :: View next topic |
Author |
Message |
seanv775 n00b
Joined: 08 Jun 2009 Posts: 7
|
Posted: Mon Oct 11, 2010 11:41 pm Post subject: atl1c Ethernet Failing [Solved] |
|
|
I recently setup a new personal server with this motherboard. I was able to get the onboard ethernet working with the atl1c driver in the kernel. It worked fine while I was getting the system setup, however, when I started doing some heavy file moving and streaming across the network, it has started failing on me within about an hour. When it fails, the system no longer responds to pings and the system can't ping the router. I'm currently running the hardened profile on the 2.6.35-r2 hardened sources. Here is the output of lspci -vvv, my kernel .config and the message that shows up in dmesg when it fails.
Last edited by seanv775 on Sun Oct 17, 2010 5:48 am; edited 1 time in total |
|
Back to top |
|
|
Jaglover Watchman
Joined: 29 May 2005 Posts: 8291 Location: Saint Amant, Acadiana
|
Posted: Tue Oct 12, 2010 1:51 am Post subject: |
|
|
Sounds like a hardware problem. |
|
Back to top |
|
|
seanv775 n00b
Joined: 08 Jun 2009 Posts: 7
|
Posted: Tue Oct 12, 2010 1:54 am Post subject: |
|
|
Jaglover wrote: | Sounds like a hardware problem. | As in a defect in the motherboard? The thing is that it does work for a while (~1hr), but it will always fail with that same dmesg output. |
|
Back to top |
|
|
chithanh Developer
Joined: 05 Aug 2006 Posts: 2158 Location: Berlin, Germany
|
Posted: Tue Oct 12, 2010 8:15 am Post subject: |
|
|
This could be an interrupt issue, but more likely the hardware is bad. Maybe see if you can reproduce this issue when booting from livecd. |
|
Back to top |
|
|
seanv775 n00b
Joined: 08 Jun 2009 Posts: 7
|
Posted: Fri Oct 15, 2010 1:51 am Post subject: |
|
|
So I booted the system into a live ubuntu environment and proceeded to do the same operations, running a bunch of network traffic and it ran all day. I tried reinstalling the gentoo install with the normal profile and gentoo sources to no avail. If anyone has any ideas as to what would cause this other than hardware problems, I'm all ears. |
|
Back to top |
|
|
Jaglover Watchman
Joined: 29 May 2005 Posts: 8291 Location: Saint Amant, Acadiana
|
Posted: Fri Oct 15, 2010 2:39 am Post subject: |
|
|
Do you overclock. I've seen Gentoo having hardware crashes on overclocked hardware. Practically the only serious crashes I ever had since I switched to Gentoo. |
|
Back to top |
|
|
seanv775 n00b
Joined: 08 Jun 2009 Posts: 7
|
Posted: Fri Oct 15, 2010 3:41 am Post subject: |
|
|
Jaglover wrote: | Do you overclock. | Nope. It's primarily a file server. |
|
Back to top |
|
|
cach0rr0 Bodhisattva
Joined: 13 Nov 2008 Posts: 4123 Location: Houston, Republic of Texas
|
Posted: Fri Oct 15, 2010 5:59 am Post subject: |
|
|
I'm finding much online that seems to point at this being a bug in recent kernels (unrelated to hardened-sources patchset)
example:
https://bugzilla.redhat.com/show_bug.cgi?id=625776
http://www.mail-archive.com/ubuntu-bugs@lists.ubuntu.com/msg2569094.html
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/631545
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/619684
All failing at that same point, same line number, same file, with numerous different network cards.
I would suggest running an older hardened kernel (2.6.32 has been rock solid for me, personally).
If your reason for jumping to 2.6.35 was this http://www.ksplice.com/uptrack/cve-2010-3081
...remember you are running hardened-sources.
I tested that exploit, and also the "cheddar bay" (null pointer deref in tun.c, and others that followed) set of exploits, against both 2.6.29-hardened, and 2.6.32-hardened, and with PaX + grsec properly configured, it would not run.
Code: |
meat@ricker ~ $ ./diagnose
bash: ./diagnose: Permission denied
# tail -n 1 /var/log/grsec.log
Oct 15 00:47:26 ricker kernel: [194350.051437] grsec: From 75.148.243.90: denied untrusted exec of /home/meat/diagnose by /bin/bash[bash:23092] uid/euid:1000/1000 gid/egid:1005/1005, parent /bin/bash[bash:23086] uid/euid:1000/1000 gid/egid:1005/1005
meat@ricker ~ $ uname -a
Linux ricker 2.6.32-hardened-r7 #2 SMP PREEMPT Fri May 28 02:45:56 CDT 2010 x86_64 QEMU Virtual CPU version 0.12.5 AuthenticAMD GNU/Linux
|
Code: |
meat@bauer ~ $ ./diagnose
Diagnostic tool for public CVE-2010-3081 exploit -- Ksplice, Inc.
(see http://www.ksplice.com/uptrack/cve-2010-3081)
$$$ Kernel release: 2.6.29-hardened
!!! Could not find symbol: prepare_creds
A symbol required by the published exploit for CVE-2010-3081 is not
provided by your kernel. The exploit would not work on your system.
|
_________________ Lost configuring your system?
dump lspci -n here | see Pappy's guide | Link Stash |
|
Back to top |
|
|
seanv775 n00b
Joined: 08 Jun 2009 Posts: 7
|
Posted: Sun Oct 17, 2010 5:49 am Post subject: |
|
|
I started rolling the kernel back and got as far as 2.6.31 with no change in the problem. I finally got it to work by rolling the kernel back up to the latest version and downloading the latest linux driver from the Atheros website and using that module instead of the one in the kernel. |
|
Back to top |
|
|
|