kernel cannot initialize OHCI chip

nukkel · 06-09-2005, 03:07 PM

Hi all,

when the kernel initializes my OHCI usb host controller, it fails with the following message:

Code:

...
ACPI: Local APIC address 0xfee00000
ACPI: LAPIC (acpi_id[0x00] lapic_id[0x01] enabled)
Processor #1 6:8 APIC version 17
ACPI: LAPIC (acpi_id[0x01] lapic_id[0x00] enabled)
Processor #0 6:8 APIC version 17
ACPI: LAPIC_NMI (acpi_id[0x00] high edge lint[0x1])
ACPI: LAPIC_NMI (acpi_id[0x01] high edge lint[0x1])
Using ACPI for processor (LAPIC) configuration information
Intel MultiProcessor Specification v1.4
    Virtual Wire compatibility mode.
OEM ID: HP       Product ID: MARSHAL      APIC at: 0xFEE00000
I/O APIC #2 Version 17 at 0xFEC00000.
I/O APIC #3 Version 17 at 0xFEC01000.
Enabling APIC mode:  Flat.  Using 2 I/O APICs
...
PCI->APIC IRQ transform: 0000:00:02.0[A] -> IRQ 18
PCI->APIC IRQ transform: 0000:00:03.0[A] -> IRQ 20
PCI->APIC IRQ transform: 0000:00:07.0[A] -> IRQ 26
PCI->APIC IRQ transform: 0000:00:08.0[A] -> IRQ 27
PCI->APIC IRQ transform: 0000:00:0f.2[A] -> IRQ 33
PCI->APIC IRQ transform: 0000:01:00.0[A] -> IRQ 28
...
ohci_hcd: 2004 Nov 08 USB 1.1 'Open' Host Controller (OHCI) Driver (PCI)
ohci_hcd 0000:00:0f.2: ServerWorks OSB4/CSB5 OHCI USB Controller
ohci_hcd 0000:00:0f.2: request interrupt 33 failed
ohci_hcd 0000:00:0f.2: init 0000:00:0f.2 fail, -38
ohci_hcd: probe of 0000:00:0f.2 failed with error -38

Anyone have an idea _why_ the interrupt reservation may fail?
It would be nice to get usb up 'n' running without putting in an extra board

Thanks a lot in advance!

Additional info:
parentboard chipset: ServerWorks CNB20HE
OHCI controller is on-board
lspci: 0000:00:0f.2 USB Controller: ServerWorks OSB4/CSB5 OHCI USB Controller (rev 04)
kernel: 2.6.11 SMP
using IO-APIC, ACPI, no APM

finegan · 06-09-2005, 03:31 PM

I keep seeing warnings about this in my 2.6 kernel boots, haven't bumped into an indcedent where it happens myself.

http://www.ussg.iu.edu/hypermail/lin...09.2/1896.html

Is a pretty good example of where it's happening to someone else too, apparently passing the option: "pci=routeirq" fixes it, but is evidently enabling the older, slower PCI IRQ routing behaviour.

Code:

PCI: Using ACPI for IRQ routing
** PCI interrupts are no longer routed automatically.  If this
** causes a device to stop working, it is probably because the
** driver failed to call pci_enable_device().  As a temporary
** workaround, the "pci=routeirq" argument restores the old
** behavior.  If this argument makes the device work again,
** please email the output of "lspci" to bjorn.helgaas@hp.com
** so I can fix the driver.

Cheers,

Finegan

nukkel · 06-10-2005, 01:24 AM

Thanks for the pointer!, no luck here though, still giving the same message.
I actually haven't tried if it works on a 2.4 kernel yet; maybe I should check that with some old installer cd.

Do your devices work properly (despite the warnings you get)?

Skimmed through ohci-hcd.c, no sign of pci_enable_device()... If I find some spare time I'll try to add that in the right place (study some similar drivers), see if I can get it to work...

Greets, nukkel