Hello Roger,
Hello Ingo,
Hello Rokas,
Hello World,
I experienced a strange problem when buying an Athlon from a friend;
I just re-configured and re-compiled the kernel when the NIC stopped
working. It was possible to ifconfig it, but it would not talk to the
world, just as if there would be no interrupts signalling data was there.
Googling I found
http://www.scyld.com/pipermail/netdrivers/2003-May/000022.html
where Rokas gives a lot of debug output about the problem, although
he switched from 2.2.x to 2.4.x, not from one config to another.
Finally, I think I found the problem:
on Rokas' 2.2.x the card is configured on a standard interrupt
(see /proc/interrupts), while on 2.4 it's on some special (?)
interrupt 16 (must be something "new" (not IBM-AT-386) with APIC stuff).
Here's output from my machine; it's a VIA-based one.
lspci -n
00:00.0 Class 0600: 1106:3099
00:01.0 Class 0604: 1106:b099
00:0e.0 Class 0100: 1000:0001 (rev 23)
00:10.0 Class 0c03: 1106:3038 (rev 80)
00:10.1 Class 0c03: 1106:3038 (rev 80)
00:10.2 Class 0c03: 1106:3038 (rev 80)
00:10.3 Class 0c03: 1106:3104 (rev 82)
00:11.0 Class 0601: 1106:3177
00:11.1 Class 0101: 1106:0571 (rev 06)
00:11.5 Class 0401: 1106:3059 (rev 50)
00:12.0 Class 0200: 1106:3065 (rev 74)
01:00.0 Class 0300: 10de:0110 (rev b2)
cat /proc/interrupts
This (IO-APIC configured) does not work:
CPU0
0: 47654 IO-APIC-edge timer
1: 1577 IO-APIC-edge keyboard
2: 0 XT-PIC cascade
8: 4 IO-APIC-edge rtc
12: 2908 IO-APIC-edge PS/2 Mouse
14: 22993 IO-APIC-edge ide0
15: 7 IO-APIC-edge ide1
16: 0 IO-APIC-level eth0
17: 14 IO-APIC-level sym53c8xx
19: 0 IO-APIC-level ehci_hcd
21: 522 IO-APIC-level usb-uhci, usb-uhci, usb-uhci
22: 13752 IO-APIC-level via82cxxx
NMI: 0
LOC: 47606
ERR: 0
MIS: 0
This, however, DOES work (IO APIC NOT configured)
CPU0
0: 73916 XT-PIC timer
1: 135 XT-PIC keyboard
2: 0 XT-PIC cascade
5: 991 XT-PIC ehci_hcd, usb-uhci, eth0
8: 4 XT-PIC rtc
11: 5141 XT-PIC sym53c8xx, via82cxxx, usb-uhci, usb-uhci
12: 4280 XT-PIC PS/2 Mouse
14: 9576 XT-PIC ide0
15: 5 XT-PIC ide1
NMI: 0
ERR: 0
I'd be glad to assist you in tracking this darn bug down; I'll even
go for a 2.6.x soon.
Greetings+thanks,
Enver
On Thu, 29 Jan 2004 10:50:02 +0100, Enver Haase wrote:
> Here's output from my machine; it's a VIA-based one.
> [...]
> This (IO-APIC configured) does not work:
> [...]
> This, however, DOES work (IO APIC NOT configured)
I've seen plenty of reports from people who complain about IO APIC
breakage on VIA. It might be fixed in 2.6, but I wouldn't know since I
don't have any IO APIC hardware. Or maybe the reporters found it easier
to turn off that option than to get somebody to fix it. If 2.6 fails
as well I suggest you try bugzilla.kernel.org so at least the report
won't go away.
Roger