Hi,
it seems that whether the netdevice nor the 8139too parts of the kernel cause
my problems but there must have been some bad changes to the interrupt
routing or the IO-APIC parts.
I don't have -mm7 handy currently (but at the moment it compiles again) but I
attached the contents of /proc/interrupts of 2.4.10 (the system I currently
run) and 2.5.70-mm8 again.
As you can see there, the -mm8 code does not assign more than 16 interrupt
sinks but the 2.4.10 (and if I remember correctly -mm7, too) has 22. Perhaps
even more problematic is that -mm8 seems not to use level-triggered
interrupts...!
I could try to revert some IRQ or APIC changes if someone could tell me which
Changeset may have caused that...
Best reagrds
Thomas Schlichter
CPU0
0: 184641 IO-APIC-edge timer
1: 2037 IO-APIC-edge keyboard
2: 0 XT-PIC cascade
4: 2 IO-APIC-edge serial
8: 2 IO-APIC-edge rtc
12: 78366 IO-APIC-edge PS/2 Mouse
14: 188581 IO-APIC-edge ide0
15: 6 IO-APIC-edge ide1
16: 180617 IO-APIC-level nvidia
17: 14155 IO-APIC-level eth0
18: 82091 IO-APIC-level EMU10K1
19: 457792 IO-APIC-level fcpci
21: 0 IO-APIC-level usb-uhci, usb-uhci, usb-uhci
NMI: 0
LOC: 184592
ERR: 0
MIS: 0
CPU0
0: 52599 IO-APIC-edge timer
1: 1075 IO-APIC-edge i8042
2: 0 XT-PIC cascade
4: 11 IO-APIC-edge serial
5: 0 IO-APIC-edge uhci-hcd
8: 2 IO-APIC-edge rtc
9: 0 IO-APIC-edge acpi
10: 0 IO-APIC-edge ehci-hcd
11: 0 IO-APIC-edge eth0, EMU10K1, bttv0, uhci-hcd, uhci-hcd
12: 79 IO-APIC-edge i8042
14: 6217 IO-APIC-edge ide0
15: 10 IO-APIC-edge ide1
NMI: 0
LOC: 52557
ERR: 0
MIS: 0
This archive was generated by hypermail 2b29 : Sun Jun 15 2003 - 22:00:27 EST