On Mon, 31 Jan 2005 David Brownell wrote:
> On Monday 31 January 2005 1:27 pm, Helge Bahmann wrote:
> >
> > well, neither atkbd.c nor i8042.c changed apart from cosmetics between
> > 2.6.7 and 2.6.8; I think this leaves me with either the input system in
> > general, irq issues in general, or usb
>
> Or combinations of some/all of the three ... :)
maybe an interrupt problem, and maybe usb is involved only incidently...
just for the record to summarize what is working or not ("keyboard broken"
precisely means: no interrupts generated for irq1 (i8042 keyboard port)
according to /proc/interrupts)
- 2.6.7: everything is fine
- 2.6.8 before loading uhci-hcd: everything is fine
- 2.6.8 after loading uhci-hcd: keyboard broken
- 2.6.8 with "usb-handoff" after loading uhci-hcd: keyboard broken
- 2.6.8 with "noapic" after loading uhci-hcd: everything is fine
apparently something has changed about irq routing between 2.6.7 and 2.6.8
("PCI-APIC IRQ transform ... -> ..." lines differ)
still there is something very strange going on, maybe there is still
something more than just irq routing
"cat /proc/interrupts | grep uhci" in 2.6.7:
19: 22 36 IO-APIC-level ohci1394, eth3, eth1, uhci_hcd
"cat /proc/interrupts | grep uhci" in 2.6.8 and following:
161: 23 78 IO-APIC-level ohci1394, eth3, eth1, uhci_hcd
as you can see the interrupt is shared - but all other devices on this
interrupt have no influence whatsoever on the ps/2 keyboard, it is only
after loading uhci-hcd that the keyboard becomes unuseable
-------------------------------------------------------
This SF.Net email is sponsored by: IntelliVIEW -- Interactive Reporting
Tool for open source databases. Create drag-&-drop reports. Save time
by over 75%! Publish reports on the web. Export to DOC, XLS, RTF, etc.
Download a FREE copy at http://www.intelliview.com/go/osdn_nl
_______________________________________________
linux-usb-devel@lists.sourceforge.net
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
