Bug 21362 - "NETDEV WATCHDOG: eth0 (atl1c): transmit queue 0 timed out" if high traffic goes through eth-card
Summary: "NETDEV WATCHDOG: eth0 (atl1c): transmit queue 0 timed out" if high traffic g...
Status: CLOSED CODE_FIX
Alias: None
Product: Drivers
Classification: Unclassified
Component: Network (show other bugs)
Hardware: i386 Linux
: P1 high
Assignee: Jay Cliburn
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-10-28 15:52 UTC by mailrupile
Modified: 2011-02-28 17:03 UTC (History)
4 users (show)

See Also:
Kernel Version: 2.6.35-gentoo-r11
Subsystem:
Regression: Yes
Bisected commit-id:

Attachments
dmesg (2.26 KB, application/octet-stream)
2010-10-28 15:55 UTC, mailrupile 		Details
lspci -vvv as root (27.98 KB, application/octet-stream)
2010-10-28 16:00 UTC, mailrupile 		Details
cpuinfo (1.56 KB, application/octet-stream)
2010-10-28 16:01 UTC, mailrupile 		Details
/proc/iomem (2.18 KB, application/octet-stream)
2010-10-28 16:01 UTC, mailrupile 		Details
/proc/ioports (1.38 KB, application/octet-stream)
2010-10-28 16:02 UTC, mailrupile 		Details
/usr/src/linux/scripts/ver_linux output (787 bytes, application/octet-stream)
2010-10-28 16:03 UTC, mailrupile 		Details
/proc/modules (546 bytes, application/octet-stream)
2010-10-28 16:04 UTC, mailrupile 		Details
Add an attachment (proposed patch, testcase, etc.)

Description mailrupile 2010-10-28 15:52:54 UTC 
My ethernet card (Attansic Technology Corp. Device 1063) freezes on high in\out net traffic. 

If net traffic doesn't high (jabber and web surfing only) all OK. But if I'll run something like rtorrent or just try to download\upload one big file on good speed (>100KB) it will freeze.
Comment 1 mailrupile 2010-10-28 15:55:47 UTC 
Created attachment 35362 [details]
dmesg

dmesg of freeze moment and module restart
Comment 2 mailrupile 2010-10-28 16:00:55 UTC 
Created attachment 35372 [details]
lspci -vvv as root
Comment 3 mailrupile 2010-10-28 16:01:19 UTC 
Created attachment 35382 [details]
cpuinfo
Comment 4 mailrupile 2010-10-28 16:01:42 UTC 
Created attachment 35392 [details]
/proc/iomem
Comment 5 mailrupile 2010-10-28 16:02:04 UTC 
Created attachment 35402 [details]
/proc/ioports
Comment 6 mailrupile 2010-10-28 16:03:46 UTC 
Created attachment 35412 [details]
/usr/src/linux/scripts/ver_linux output
Comment 7 mailrupile 2010-10-28 16:04:29 UTC 
Created attachment 35422 [details]
/proc/modules
Comment 8 Jay Cliburn 2010-10-28 23:37:54 UTC 
On Thu, Oct 28, 2010 at 4:28 PM,  <bugzilla-daemon@bugzilla.kernel.org> wrote:
> https://bugzilla.kernel.org/show_bug.cgi?id=21362
>
>
> Andrew Morton <akpm@linux-foundation.org> changed:
>
>           What    |Removed                     |Added
> ----------------------------------------------------------------------------
>                 CC|                            |akpm@linux-foundation.org,
>                   |                            |csnook@redhat.com
>         AssignedTo|drivers_network@kernel-bugs |jcliburn@gmail.com
>                   |.osdl.org                   |

atl1c is maintained by Atheros directly.  Calling upon Jie Yang to
look at this bug.

Jay
Comment 9 mailrupile 2010-10-31 22:47:14 UTC 
I can add that on windows 7, which I was forced to install due above transcribed problems, everything is OK. With eth-card, at least.
Thus this IS a atl1c driver problem.
Comment 10 mailrupile 2010-11-14 00:08:16 UTC 
Problem solved in kernel 2.6.36-gentoo-r1 by driver version upgrade:

kernel 2.6.35-gentoo-r12
ATL1C_DRV_VERSION "1.0.0.2-NAPI"

kernel 2.6.36-gentoo-r1
ATL1C_DRV_VERSION "1.0.1.0-NAPI"
Comment 11 Giovanni Toraldo 2011-02-28 17:03:56 UTC 
[26782.502214] Hardware name: VPCS13V9E
04:00.0 Ethernet controller: Atheros Communications AR8131 Gigabit Ethernet (rev c0)

Same problem with 2.6.37-1-amd64 on Debian unstable.

The fix isn't applied upstream yet?
Note You need to log in before you can comment on or make changes to this bug.