Boot failure with 2.6.9-rc2-bk latest in usb/hid-core.c

Boot failure with 2.6.9-rc2-bk latest in usb/hid-core.c

[James Bottomley]

I get this out of the machine (an ia64 zx2000 with connected USB
keyboard and mouse):

Discovering hardware: snd-fm801 aic7xxx mptscsih qla1280 ohci-hcd cmd64x
sym53c8xx tulip e1000
Loading snd-fm801:
Linux video capture interface: v1.00
ACPI: PCI interrupt 0000:a0:04.0[A] -> GSI 41 (level, low) -> IRQ 62
Skipping Module aic7xxx. It's already loaded.
Skipping Module mptscsih. It's already loaded.
Skipping Module qla1280. It's already loaded.
Loading ohci-hcd:
usbcore: registered new driver usbfs
usbcore: registered new driver hub
ACPI: PCI interrupt 0000:a0:01.0[A] -> GSI 38 (level, low) -> IRQ 57
ohci_hcd 0000:a0:01.0: NEC Corporation USB
ohci_hcd 0000:a0:01.0: irq 57, pci mem c0000000d0022000
ohci_hcd 0000:a0:01.0: new USB bus registered, assigned bus number 1
usb usb1: Product: NEC Corporation USB
usb usb1: Manufacturer: Linux 2.6.9-rc2 ohci_hcd
usb usb1: SerialNumber: 0000:a0:01.0
hub 1-0:1.0: USB hub found
hub 1-0:1.0: 3 ports detected
ACPI: PCI interrupt 0000:a0:01.1[B] -> GSI 39 (level, low) -> IRQ 58
ohci_hcd 0000:a0:01.1: NEC Corporation USB (#2)
ohci_hcd 0000:a0:01.1: irq 58, pci mem c0000000d0021000
ohci_hcd 0000:a0:01.1: new USB bus registered, assigned bus number 2
usb usb2: Product: NEC Corporation USB (#2)
usb usb2: Manufacturer: Linux 2.6.9-rc2 ohci_hcd
usb usb2: SerialNumber: 0000:a0:01.1
hub 2-0:1.0: USB hub found
hub 2-0:1.0: 2 ports detected
usb 1-2: new low speed USB device using address 2
usb 1-2: Product: Standard USB Keyboard 
usb 1-2: Manufacturer: Silitek
usbcore: registered new driver hiddev
input: USB HID v1.00 Keyboard [Silitek Standard USB Keyboard ] on
usb-0000:a0:01.0-2
usbcore: registered new driver usbhid
drivers/usb/input/hid-core.c: v2.0:USB HID core driver
usb 2-2: new low speed USB device using address 2
usb 2-2: Product: N48
usb 2-2: Manufacturer: Logitech
input: USB HID v1.00 Mouse [Logitech N48] on usb-0000:a0:01.1-2
Loading cmd64x:
Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2
ide: Assuming 33MHz system bus speed for PIO modes; override with
idebus=xx
CMD649: IDE controller at PCI slot 0000:a0:02.0
ACPI: PCI interrupt 0000:a0:02.0[A] -> GSI 43 (level, low) -> IRQ 60
CMD649: chipset revision 2
CMD649: 100% native mode on irq 60
    ide0: BM-DMA at 0xa0d0-0xa0d7, BIOS settings: hda:pio, hdb:pio
    ide1: BM-DMA at 0xa0d8-0xa0df, BIOS settings: hdc:pio, hdd:pio
hdc: HL-DT-ST GCE-8483B, ATAPI CD/DVD-ROM drive
ide1 at 0xa0e0-0xa0e7,0xa0f2 on irq 60
Skipping Module sym53c8xx. It's already loaded.
Loading tulip:
Linux Tulip driver version 1.1.13 (May 11, 2002)
ACPI: PCI interrupt 0000:81:04.0[A] -> GSI 27 (level, low) -> IRQ 55
tulip0:  EEPROM default media type Autosense.
tulip0:  Index #0 - Media MII (#11) described by a 21140 MII PHY (1)
block.
tulip0:  MII transceiver #3 config 3100 status 7809 advertising 01e1.
eth0: Digital DS21140 Tulip rev 32 at 0xc0000000a0281000,
00:E0:29:0F:9F:D9, IRQ 55.
ACPI: PCI interrupt 0000:81:05.0[A] -> GSI 28 (level, low) -> IRQ 56
tulip1:  EEPROM default media type Autosense.
tulip1:  Index #0 - Media MII (#11) described by a 21140 MII PHY (1)
block.
tulip1:  MII transceiver #3 config 3100 status 7809 advertising 01e1.
eth1: Digital DS21140 Tulip rev 32 at 0xc0000000a0280000,
00:E0:29:0F:9F:D8, IRQ 56.
Loading e1000:
e1000: Ignoring new-style parameters in presence of obsolete ones
Intel(R) PRO/1000 Network Driver - version 5.3.19-k2
Copyright (c) 1999-2004 Intel Corporation.
ACPI: PCI interrupt 0000:a0:03.0[A] -> GSI 42 (level, low) -> IRQ 61
e1000: eth2: e1000_probe: Intel(R) PRO/1000 Network Connection
Running 0dns-down to make sure resolv.conf is ok...done.
Cleaning: /etc/network/ifstate.
Starting hotplug subsystem:
   input   
   input    [success]
   isapnp  
   isapnp   [success]
   net     
   net      [success]
   pci     
     ignoring pci display device on 00:00.0
     aic79xx: loaded successfully
     aic7xxx: already loaded
     aic79xx: already loaded
     aic7xxx: already loaded
     mptbase: already loaded
     mptbase: already loaded
     qla1280: already loaded
     tulip: already loaded
     tulip: already loaded
     ohci-hcd: already loaded
     ohci-hcd: already loaded
ACPI: PCI interrupt 0000:a0:01.2[C] -> GSI 40 (level, low) -> IRQ 59
ehci_hcd 0000:a0:01.2: NEC Corporation USB 2.0
ehci_hcd 0000:a0:01.2: irq 59, pci mem c0000000d0020000
ehci_hcd 0000:a0:01.2: new USB bus registered, assigned bus number 3
ehci_hcd 0000:a0:01.2: USB 2.0 enabled, EHCI 0.95, driver 2004-May-10
drivers/usb/input/hid-core.c: input irq status -110 received
drivers/usb/input/hid-core.c: input irq status -110 received
[the last message repeats forever]

It boots just fine with 2.6.9-rc2 and the only difference to the usb
input subsystem appears to be your latest merge.

James

Re: Boot failure with 2.6.9-rc2-bk latest in usb/hid-core.c

[Vojtech Pavlik]

On Sun, Sep 19, 2004 at 12:08:04PM -0400, James Bottomley wrote:

> I get this out of the machine (an ia64 zx2000 with connected USB
> keyboard and mouse):

> usb usb1: Product: NEC Corporation USB
> usb usb1: Manufacturer: Linux 2.6.9-rc2 ohci_hcd
> usb usb2: Product: NEC Corporation USB (#2)
> usb usb2: Manufacturer: Linux 2.6.9-rc2 ohci_hcd
> usb 1-2: Product: Standard USB Keyboard 
> usb 1-2: Manufacturer: Silitek
> usbcore: registered new driver hiddev
> input: USB HID v1.00 Keyboard [Silitek Standard USB Keyboard ] on usb-0000:a0:01.0-2
> usb 2-2: Product: N48
> usb 2-2: Manufacturer: Logitech
> input: USB HID v1.00 Mouse [Logitech N48] on usb-0000:a0:01.1-2

> ACPI: PCI interrupt 0000:a0:01.2[C] -> GSI 40 (level, low) -> IRQ 59
> ehci_hcd 0000:a0:01.2: NEC Corporation USB 2.0
> ehci_hcd 0000:a0:01.2: USB 2.0 enabled, EHCI 0.95, driver 2004-May-10
> drivers/usb/input/hid-core.c: input irq status -110 received
> drivers/usb/input/hid-core.c: input irq status -110 received
> [the last message repeats forever]

> It boots just fine with 2.6.9-rc2 and the only difference to the usb
> input subsystem appears to be your latest merge.

There were changes in the function that prints the above message,
however they were indentation only. I really doubt it could be the HID
changes I did.

It looks like there is either a problem with ACPI IRQ routing that when
enabling the EHCI controller IRQ does something bad to the OHCI
controllers, or the EHCI driver itself does something bad to the OHCI
controllers. (Afte all, the controllers share their ports.)

Try disabling EHCI in your config to confirm my theory.

-- 
Vojtech Pavlik
SuSE Labs, SuSE CR

Re: Boot failure with 2.6.9-rc2-bk latest in usb/hid-core.c

[James Bottomley]

On Tue, 2004-09-21 at 04:10, Vojtech Pavlik wrote:
> There were changes in the function that prints the above message,
> however they were indentation only. I really doubt it could be the HID
> changes I did.

Well, I verified that it works before this merge

torvalds@ppc970.osdl.org|ChangeSet|20040916140404|15905

and doesn't with it (this was Linus pulling in your input tree).

> It looks like there is either a problem with ACPI IRQ routing that when
> enabling the EHCI controller IRQ does something bad to the OHCI
> controllers, or the EHCI driver itself does something bad to the OHCI
> controllers. (Afte all, the controllers share their ports.)
> 
> Try disabling EHCI in your config to confirm my theory.

Yes, disabling EHCI in the config allows boot.

This is what lspci says about the USB controller:

0000:a0:01.0 USB Controller: NEC Corporation USB (rev 41) (prog-if 10
[OHCI])
        Subsystem: NEC Corporation USB
        Flags: bus master, medium devsel, latency 128, IRQ 57
        Memory at 00000000d0022000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: [40] Power Management version 2

0000:a0:01.1 USB Controller: NEC Corporation USB (rev 41) (prog-if 10
[OHCI])
        Subsystem: NEC Corporation USB
        Flags: bus master, medium devsel, latency 128, IRQ 58
        Memory at 00000000d0021000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: [40] Power Management version 2

0000:a0:01.2 USB Controller: NEC Corporation USB 2.0 (rev 02) (prog-if
20 [EHCI])
        Subsystem: NEC Corporation USB 2.0
        Flags: bus master, medium devsel, latency 128, IRQ 59
        Memory at 00000000d0020000 (32-bit, non-prefetchable) [size=256]
        Capabilities: [40] Power Management version 2

So it looks like ACPI is routing the IRQ's corectly.  Just to confirm,
here's the ACPI routing probe from boot:

ACPI: PCI interrupt 0000:a0:01.0[A] -> GSI 38 (level, low) -> IRQ 57
GSI 39 (level, low) -> CPU 0 (0x0000) vector 58
ACPI: PCI interrupt 0000:a0:01.1[B] -> GSI 39 (level, low) -> IRQ 58
GSI 40 (level, low) -> CPU 0 (0x0000) vector 59
ACPI: PCI interrupt 0000:a0:01.2[C] -> GSI 40 (level, low) -> IRQ 59

Anything else you need?

James