On Thu, 2005-03-24 at 14:03 -0500, Lee Revell wrote:
> On Thu, 2005-03-24 at 13:49 -0500, Lee Revell wrote:
> > Is the USB irq shared with anything?
>
> Never mind, I see that it is shared, and with a pretty "heavy" device
> (the network card).
There's two similar (except for hardware) user experiences being
reported here... in my case I was able to eliminate "shared IRQ" issues.
By issuing the "acpi_irq_balance" kernel parameter. Currently my irq
assignment is:
CPU0
0: 228818 XT-PIC timer 0/28818
1: 617 XT-PIC i8042 0/617
2: 0 XT-PIC cascade 0/0
7: 63 XT-PIC Intel 82801DB-ICH4, eth0 0/63
8: 2 XT-PIC rtc 0/2
9: 11329 XT-PIC acpi, radeon@email-addr-hidden:0000:01:00.0 0/11329
10: 727 XT-PIC uhci_hcd:usb2, ehci_hcd:usb4 0/727
11: 3 XT-PIC uhci_hcd:usb1, uhci_hcd:usb3, ohci1394
0/3
12: 13706 XT-PIC i8042 1/13706
14: 1701 XT-PIC ide0 0/1701
15: 5553 XT-PIC ide1 0/5553
It doesn't seem to matter what USB bus it's on (or for that matter if
irq's are "reshuffled" or not) so I'm suspicious of some intrinsic
kernel or driver issue (barring some weird hardware hangup). I've got
the week off so I'll do some further testing (I also have the builtin
ac97 chip and another USB device (emi 2|6).
> Is this the case for everyone who is having the problem?
>
> Does disconnecting the network cable help?
As reported above, eth0 is on another IRQ. Also all my kernels are
patched with Ingo Molnar's RP patches (2.6.8.1 through 2.6.12.rcX). I'm
unthreading whatever IRQ the device is on (10 or 11 and sometimes both).
thanks,
-ry
-- Ryan Gallagher <ruinaudio@email-addr-hidden>Received on Fri Mar 25 00:15:10 2005
This archive was generated by hypermail 2.1.8 : Fri Mar 25 2005 - 00:15:11 EET