linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Bug Report: 2.4.23-pre9 / SIS chipset / sundance
@ 2003-11-07 16:48 Stephan von Krawczynski
  2003-11-09 14:29 ` Stephan von Krawczynski
  0 siblings, 1 reply; 2+ messages in thread
From: Stephan von Krawczynski @ 2003-11-07 16:48 UTC (permalink / raw)
  To: linux-kernel

Hello,

I just experienced a problem with a configuration that is known to only work
with ACPI. It consists of:

P-III 1,4 GHz
Mainboard with SIS 630E chipset
Adaptec 29160 Controller
DLINK DFE-580 TX (four port card) (sundance driver)

A slightly different setup is known to work with 2.4.19 and ACPI patch. Problem
with this SIS chipset is interrupt sharing. If you do not use ACPI the box
freezes sometime.
The known-to-work setup is identical besides it contains a tulip-based four
port ethernet card (old DLINK cards).
We stress-tested the working box by NFS-copying data to the local hd over an
ethernet port that shares its interrupt with the aic.

The above failing setup with sundance driver results in network hanging, but no
other negative points. The network hang can be cured by "mii-tool -r <device>"
(restart autonegotiation). After that everything works well, until the next
hang.
During the network hang the shared interrupt (aic + ethernet) still counts up. 
Maybe the sundance driver just misses an interrupt and hangs, and restarting
autonegotiation resets some internals.

I can try whatever patch you like, this is no production setup.

Regards,
Stephan

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: Bug Report: 2.4.23-pre9 / SIS chipset / sundance
  2003-11-07 16:48 Bug Report: 2.4.23-pre9 / SIS chipset / sundance Stephan von Krawczynski
@ 2003-11-09 14:29 ` Stephan von Krawczynski
  0 siblings, 0 replies; 2+ messages in thread
From: Stephan von Krawczynski @ 2003-11-09 14:29 UTC (permalink / raw)
  To: linux-kernel; +Cc: jgarzik

Regarding above topic I can verify that the same hardware setup works if
replacing sundance with tulip ethernet.
So I am pretty confident it has something to do with interrupt sharing (at
least on SIS mb) of sundance driver. Has anybody else experienced troubles there?

Regards,
Stephan

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2003-11-09 14:29 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2003-11-07 16:48 Bug Report: 2.4.23-pre9 / SIS chipset / sundance Stephan von Krawczynski
2003-11-09 14:29 ` Stephan von Krawczynski

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).