[linux-audio-user] preempt patch and USB audio

[Hector Centeno]

Hi,

I've being experiencing system freezes and IRQs dying with the kernels 
that include Ingo's preempt patch using FC3+CCRMA and Demudi. I'm using 
a USB audio interface (M-Audio Duo). I would like to post some 
information hoping somebody could point me to a fix. This are my USB 
controllers:

02:0b.0 USB Controller: NEC Corporation USB (rev 43) (prog-if 10 [OHCI])
        Subsystem: Unknown device 3083:0035
        Flags: bus master, medium devsel, latency 64, IRQ 23
        Memory at feafd000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: [40] Power Management version 2

02:0b.1 USB Controller: NEC Corporation USB (rev 43) (prog-if 10 [OHCI])
        Subsystem: Unknown device 3083:0035
        Flags: bus master, medium devsel, latency 64, IRQ 20
        Memory at feafe000 (32-bit, non-prefetchable) [size=4K]
        Capabilities: [40] Power Management version 2

02:0b.2 USB Controller: NEC Corporation USB 2.0 (rev 04) (prog-if 20 [EHCI])
        Subsystem: Unknown device 3083:00e0
        Flags: bus master, medium devsel, latency 64, IRQ 21
        Memory at feaff400 (32-bit, non-prefetchable) [size=256]
        Capabilities: [40] Power Management version 2

and this is dmesg after the tragedy (here in Demudi as in FC3 the system 
hard locks and I can't get a dmesg):

usbcore: registered new driver snd-usb-audio
ohci_hcd 0000:02:0b.0: bad entry 34e630c0
ohci_hcd 0000:02:0b.0: OHCI Unrecoverable Error, disabled
ohci_hcd 0000:02:0b.0: HC died; cleaning up
usb 5-1: USB disconnect, address 2
irq 193: nobody cared!
 [<c013c52a>] __report_bad_irq+0x2a/0xa0 (8)
 [<c013c630>] note_interrupt+0x70/0xb0 (20)
 [<c013c3c4>] do_hardirq+0xe4/0xf0 (28)
 [<c013c3d0>] do_irqd+0x0/0xb0 (32)
 [<c013c449>] do_irqd+0x79/0xb0 (4)
 [<c013055a>] kthread+0xaa/0xb0 (28)
 [<c01304b0>] kthread+0x0/0xb0 (28)
 [<c010133d>] kernel_thread_helper+0x5/0x18 (16)
handlers:
[<f8a4cfc0>] (usb_hcd_irq+0x0/0x70 [usbcore])
[<f8a4cfc0>] (usb_hcd_irq+0x0/0x70 [usbcore])
Disabling IRQ #193
ohci_hcd 0000:02:0b.0: leak ed f490e080 (#03) state 0 (has tds)
ohci_hcd 0000:02:0b.0: leak ed f490e040 (#85) state 0 (has tds)
BUG: Unable to handle kernel NULL pointer dereference at virtual address 
00000000
 printing eip:
c01a35f6
*pde = 00000000
Oops: 0000 [#2]
PREEMPT
Modules linked in: snd_usb_audio snd_usb_lib radeon drm thermal fan 
button processor ac battery realtime commoncap ipv6 af_packet ohci_hcd 
sk98lin ohci1394 snd_intel8x0 snd_ac97_codec snd_pcm_oss snd_mixer_oss 
snd_pcm snd_page_alloc i2c_i801 i2c_core ehci_hcd usbhid uhci_hcd 
usbcore shpchp pci_hotplug intel_agp agpgart parport_pc parport floppy 
pcspkr tsdev mousedev evdev sr_mod sbp2 scsi_mod ieee1394 psmouse ide_cd 
cdrom snd_seq_dummy snd_seq_oss snd_seq_midi snd_rawmidi 
snd_seq_midi_event snd_seq snd_timer snd_seq_device snd soundcore genrtc 
ext3 jbd mbcache ide_disk ide_generic via82cxxx trm290 triflex slc90e66 
sis5513 siimage serverworks sc1200 rz1000 piix pdc202xx_old opti621 
ns87415 hpt366 hpt34x generic cy82c693 cs5530 cs5520 cmd64x atiixp 
amd74xx alim15x3 aec62xx pdc202xx_new ide_core unix
CPU:    0
EIP:    0060:[<c01a35f6>]    Not tainted VLI
EFLAGS: 00010246   (2.6.12-3-multimedia-686)
EIP is at get_kobj_path_length+0x26/0x40
eax: 00000000   ebx: 00000000   ecx: ffffffff   edx: f48b5098
esi: 00000001   edi: 00000000   ebp: ffffffff   esp: c18e5d6c
ds: 007b   es: 007b   ss: 0068   preempt: 00000001
Process events/0 (pid: 4, threadinfo=c18e4000 task=dfc045a0)
Stack: f48b5058 f48b5074 f48c2798 f48b5098 c01a368f f48b5098 c014612b 
dffef080
       dffef0c4 f48b5058 f48b5074 f48c2798 000003a8 c0203dd4 f48b5098 
000000d0
       c0145f8e dffefac4 f497bd40 00000016 ffffffff fffffffd f48b502c 
00000000
Call Trace:
 [<c01a368f>] kobject_get_path+0x1f/0x80 (20)
 [<c014612b>] __kmalloc+0x7b/0x140 (8)
 [<c0203dd4>] class_hotplug+0x134/0x200 (28)
 [<c0145f8e>] kmem_cache_alloc+0x4e/0x120 (12)
 [<c01a438e>] kobject_hotplug+0x1ee/0x300 (60)
 [<c0134f79>] rt_up+0x59/0x60 (60)
 [<c020427f>] class_device_del+0x8f/0xe0 (20)
 [<c02042e3>] class_device_unregister+0x13/0x30 (32)
 [<f88eb5ee>] snd_unregister_device+0x9e/0xf0 [snd] (12)
 [<f8ac683a>] snd_pcm_dev_unregister+0x5a/0xe0 [snd_pcm] (32)
 [<f88f0477>] snd_device_free+0xa7/0xc0 [snd] (32)
 [<f88f06c0>] snd_device_free_all+0x60/0x70 [snd] (16)
 [<f88ebe4c>] snd_card_free+0x12c/0x260 [snd] (28)
 [<c0130ab0>] autoremove_wake_function+0x0/0x60 (20)
 [<c0130ab0>] autoremove_wake_function+0x0/0x60 (32)
 [<f88ebfcf>] snd_card_free_thread+0x4f/0xc0 [snd] (36)
 [<c012bdbd>] worker_thread+0x1cd/0x270 (24)
 [<f88ebf80>] snd_card_free_thread+0x0/0xc0 [snd] (20)
 [<c0117390>] default_wake_function+0x0/0x30 (32)
 [<c0117390>] default_wake_function+0x0/0x30 (32)
 [<c012bbf0>] worker_thread+0x0/0x270 (36)
 [<c013055a>] kthread+0xaa/0xb0 (4)
 [<c01304b0>] kthread+0x0/0xb0 (28)
 [<c010133d>] kernel_thread_helper+0x5/0x18 (16)
Code: 90 8d 74 26 00 55 bd ff ff ff ff 57 56 be 01 00 00 00 53 8b 54 24 
14 31 db 8d b6 00 00 00 00 8d bf 00 00 00 00 8b 3a 89 e9 89 d8 <f2> ae 
f7 d1 49 8b 52 24 8d 74 31 01 85 d2 75 ea 5b 89 f0 5e 5f

I've tried changing the IRQ latency as well as the priority using chrt 
pidof and setpci but nothing improves. I tried also starting the kernel 
with no acpi using pci=noacpi and acpi=off (one at a time). The problem 
usually is triggered by starting and/or closing an audio application 
while running jack. Probably the easiest solution would be to stop 
trying to use a USB audio interface... but anyway, I would like to try a 
little more.

thank you in advance!


Hector