[ previous ] [ next ] [ threads ]
 
 From:  "Sage Telecommunications Pty Ltd" <rod at sages dot com dot au>
 To:  "Marten" <wleiden at xs4all dot nl>, "mono" <m0n0wall at lists dot m0n0 dot ch>
 Subject:  Re: [m0n0wall] Cabletron/Roamabaout/Nokia 910 watchdog timeouts
 Date:  Mon, 17 Nov 2003 21:33:27 +0800
Hi Marten
>
> Hi Rod
> >
> > Investigation suggested that the problem may be to do with conflicting
> > interrupts but a perusal of the logs don't indicate this on either box.
>
> no hardware changed?
> because some pci-pcmcia does not work on all (older) motherboards

exactly the same hardware. I can boot with an old wrp floppy (
http://nocat.net/download/WRP/ )
However that image was modified with a different wireless card driver
specifically for the roamabout wireless card.
So perhaps there is a driver problem. I am guessing that the drivers are
complied into the kernel
>
> try in exe.php

Actually exec.php ( Thanks for reminding me to RTFM ) I had forgotten that
it was possible to run commands under this.

>
> dmesg | grep irq

couldn't find grep but a search of /var/log/system.log doesn't actually show
an interrupt conflict


Nov 17 16:59:50 unisys /kernel: pcic0: <Cirrus Logic PD6729/6730 PCI-PCMCIA
Bridge> port 0x3e0-0x3e3 irq 11 at device 8.0 on pci0
Nov 17 16:59:50 unisys /kernel: pcic0: Polling mode
Nov 17 16:59:50 unisys /kernel: pccard0: <PC Card 16-bit bus (classic)> on
pcic0
Nov 17 16:59:50 unisys /kernel: pccard1: <PC Card 16-bit bus (classic)> on
pcic0
Nov 17 16:59:50 unisys /kernel: pci0: <Cirrus Logic GD5430 SVGA controller>
at 18.0


Nov 17 16:59:50 unisys /kernel: pccard: card inserted, slot 1
Nov 17 16:59:50 unisys /kernel: wi0 at port 0x240-0x27f irq 10 slot 1 on
pccard1
Nov 17 16:59:50 unisys /kernel: wi0: 802.11 address: 00:e0:63:83:15:91
Nov 17 16:59:50 unisys /kernel: wi0: using Lucent Technologies, WaveLAN/IEEE
Nov 17 16:59:50 unisys /kernel: wi0: Lucent Firmware: Station 8.10.01
Nov 17 16:59:53 unisys /kernel: wi0: watchdog timeout


it does show that although an interrupt is assigned to the PCI-PCMIA bridge
it ends up in polling mode, not sure of the relevance though.

rod