Hi,

A couple of kernels ago, I ran into an issue with the sky2 driver.
Basically, I began getting repeated errors on my console of the
following:

Oct 21 16:33:39 holly kernel: sky2 0000:04:00.0: error interrupt
status=0x80000000
Oct 21 16:33:39 holly kernel: sky2 0000:04:00.0: error interrupt
status=0xc0000000

Which eventually led to:
Oct 21 16:33:49 holly kernel: printk: 92 messages suppressed.

Here is a section of the log that may be useful:
Oct 21 16:33:37 holly kernel: sky2 eth1: enabling interface
Oct 21 16:33:38 holly dhclient: DHCPDISCOVER on eth1 to
255.255.255.255 port 67 interval 8
Oct 21 16:33:39 holly kernel: sky2 0000:04:00.0: error interrupt
status=0x80000010
Oct 21 16:33:39 holly kernel: sky2 eth1: Link is up at 1000 Mbps, full
duplex, flow control both
Oct 21 16:33:39 holly kernel: sky2 0000:04:00.0: error interrupt
status=0x80000000
Oct 21 16:33:39 holly kernel: sky2 0000:04:00.0: error interrupt
status=0xc0000000
Oct 21 16:33:40 holly kernel: sky2 0000:04:00.0: error interrupt
status=0x80000000
Oct 21 16:33:40 holly kernel: sky2 0000:04:00.0: error interrupt
status=0xc0000000
Oct 21 16:33:41 holly kernel: sky2 0000:04:00.0: error interrupt
status=0x80000000
Oct 21 16:33:41 holly kernel: sky2 0000:04:00.0: error interrupt
status=0xc0000000
Oct 21 16:33:42 holly kernel: sky2 0000:04:00.0: error interrupt
status=0x80000000

These messages are appearing on all the VTs, and because they keep
popping up, I can't reasonably use the consoles to test if the network
driver is working.

The kernel I've got going at the moment is 2.6.23-rc9 with git
revision g804b3f9a, and it's fine.  The error surfaced _sometime_
after this.  Attached is my .config.

I'll try rebooting and pinging the machine to see if the network
driver is working, and this is just (unnecessary?) spew from the
kernel.

-- 
Regards,
Tom Spink
University of Edinburgh
[ message continues ]

Hi again,

So it seems the driver and the network card is functioning, but I'm
getting the 'error interrupt' message plastered over every console.

I can ping and ssh into the computer, but the message is still
scrolling everything off my screens.

-- 
Regards,
Tom Spink
University of Edinburgh
-

[ message continues ]

On Wed, 24 Oct 2007 19:06:37 +0100

Your system is not supporting MMCONFIG access to allow driver to turn off
PCI express error value.  It looks like the hw error irq mask doesn't work
as expected either.

What is your kernel config?


-- 
Stephen Hemminger <shemminger@linux-foundation.org>

-

[ message continues ]

I bisected the kernel, with the following results:

git-bisect start
# bad: [c9927c2bf4f45bb85e8b502ab3fb79ad6483c244] Linux 2.6.24-rc1
git-bisect bad c9927c2bf4f45bb85e8b502ab3fb79ad6483c244
# good: [3146b39c185f8a436d430132457e84fa1d8f8208] Linux 2.6.23-rc9
git-bisect good 3146b39c185f8a436d430132457e84fa1d8f8208
# bad: [a9957449b08ab561a33e1e038df06843b8d8dd9f] sched: uninline scheduler
git-bisect bad a9957449b08ab561a33e1e038df06843b8d8dd9f
# bad: [038a5008b2f395c85e6e71d6ddf3c684e7c405b0] Merge branch
'master' of git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6
git-bisect bad 038a5008b2f395c85e6e71d6ddf3c684e7c405b0
# bad: [87bdc48d304191313203df9b98d783e1ab5a55ab] [IPSEC]: Get rid of
ipv6_{auth,esp,comp}_hdr
git-bisect bad 87bdc48d304191313203df9b98d783e1ab5a55ab
# bad: [1aafee265723818d903766212015b6116885bc6f] cxgb3 - Add T3C rev
git-bisect bad 1aafee265723818d903766212015b6116885bc6f
# good: [d6f2da5b33911a31eb61e1790ef8e555e9605837] [MAC80211]: Remove
bitfields from struct ieee80211_if_sta
git-bisect good d6f2da5b33911a31eb61e1790ef8e555e9605837
# good: [e2aa334bee4ae9fe27f26b0732c0443d2f11a8af] libertas: remove
adapter->scanprobes
git-bisect good e2aa334bee4ae9fe27f26b0732c0443d2f11a8af
# good: [64f222cce2605420470d7a1f678783bdc2754af6] zd1211rw:
consistent handling of ZD1211 specific rates
git-bisect good 64f222cce2605420470d7a1f678783bdc2754af6
# good: [c9a6ce500d78932c43361eae28c3de81b3660c77] cxgb3 - tighten
checks on TID values
git-bisect good c9a6ce500d78932c43361eae28c3de81b3660c77
# bad: [26fcfa95aef980cab4ff1ea55979c30e772dd0dd] pasemi_mac: RX
performance tweaks
git-bisect bad 26fcfa95aef980cab4ff1ea55979c30e772dd0dd
# bad: [555382cbfc6d2187b53888190755e56f52308cd6] sky2: advanced error reporting
git-bisect bad 555382cbfc6d2187b53888190755e56f52308cd6
# good: [3f61e4278c8489c1d026a45b0903cd4230d1ec07] cxgb3 - Update
internal memory management
git-bisect good 3f61e4278c8489c1d026a45b0903cd4230d1ec07
# good: ...
[ message continues ]

Whoops. And of course, the first bad revision is:

555382cbfc6d2187b53888190755e56f52308cd6: sky2: advanced error reporting

-- 
Regards,
Tom Spink
University of Edinburgh
-

[ message continues ]