2004-11-18 13:21:44

by Gerd Knorr

[permalink] [raw]
Subject: Re: Fw: Re: Linux 2.6.10-rc2 [dvb-bt8xx unload oops]

> This is vanilla 2.6.10-rc2 on P4. This was a problem with -rc1 but

Yes, looks very simliar ...

> some patches off the list [attached] fixed it. I expected these to be
> in -rc2, I am not able to say which patch is missing.

Uhm, strange. The bttv patches _are_ merged.
Not sure about any for dvb-bt8xx, Michael?

Gerd

--
#define printk(args...) fprintf(stderr, ## args)


2004-11-18 17:28:45

by Michael Hunold

[permalink] [raw]
Subject: Re: Fw: Re: Linux 2.6.10-rc2 [dvb-bt8xx unload oops]

Hi,

On 18.11.2004 14:03, Gerd Knorr wrote:
>>This is vanilla 2.6.10-rc2 on P4. This was a problem with -rc1 but

> Yes, looks very simliar ...

>>some patches off the list [attached] fixed it. I expected these to be
>>in -rc2, I am not able to say which patch is missing.

> Uhm, strange. The bttv patches _are_ merged.
> Not sure about any for dvb-bt8xx, Michael?

Hm, I'm not sure either, because Eyal says that the problem does not
exist with rc2-mm1. But AFAIK all DVB stuff has been merged by Linus so
I'm clueless here... 8-(

> Gerd

CU
Michael.

2004-11-20 01:48:44

by Eyal Lebedinsky

[permalink] [raw]
Subject: Re: Fw: Re: Linux 2.6.10-rc2 [dvb-bt8xx unload oops]

Michael Hunold wrote:
> Hi,
>
> On 18.11.2004 14:03, Gerd Knorr wrote:
>
>>> This is vanilla 2.6.10-rc2 on P4. This was a problem with -rc1 but
>
>
>> Yes, looks very simliar ...
>
>
>>> some patches off the list [attached] fixed it. I expected these to be
>>> in -rc2, I am not able to say which patch is missing.
>
>
>> Uhm, strange. The bttv patches _are_ merged.
>> Not sure about any for dvb-bt8xx, Michael?
>
>
> Hm, I'm not sure either, because Eyal says that the problem does not
> exist with rc2-mm1. But AFAIK all DVB stuff has been merged by Linus so
> I'm clueless here... 8-(

I now managed to get the problem with rc2-mm1/2. They reaction is different
(I do not get a lockup) but the oops shows up. Here is one from rc2-mm2. The
system was still usable at this point.

I now think that the original problem I reported is not the same as this one
which happens with all kernels I tried.

Also, does anyone know how to switch to text console when X locks up (as
it does for me)? sysrq works but does not allow me to switch to another
console. Since the hard lock does not log the oops, and I cannot see the
text from sysrq, I cannot report the details.

Nov 19 23:37:44 eyal kernel: f90625f6
Nov 19 23:37:44 eyal kernel: PREEMPT SMP
Nov 19 23:37:44 eyal kernel: Modules linked in: tsdev psmouse mt352 sp887x v4l1_compat dvb_bt8xx dvb_core i810_audio ac97_codec sr_mod sg ide_scsi ide_cd cdrom it87 eeprom i2c_isa i2c_i801 i2c_sensor eth1394 ohci_hcd ohci1394 ieee1394 dc395x scsi_mod bt878 bttv tuner video_buf firmware_class i2c_algo_bit v4l2_common btcx_risc videodev e1000 snd_intel8x0 snd_ac97_codec snd_pcm snd_timer snd snd_page_alloc soundcore i2c_core cfi_cmdset_0002 cfi_util mtdpart jedec_probe cfi_probe gen_probe ichxrom mtdcore chipreg map_funcs ehci_hcd uhci_hcd usbcore shpchp pci_hotplug intel_mch_agp intel_agp agpgart parport_pc parport 8250_pnp 8250 serial_core evdev nls_cp437 msdos fat dm_mod rtc unix
Nov 19 23:37:44 eyal kernel: CPU: 0
Nov 19 23:37:44 eyal kernel: EIP: 0060:[pg0+952432118/1069462528] Not tainted VLI
Nov 19 23:37:44 eyal kernel: EFLAGS: 00210092 (2.6.10-rc2-mm2)
Nov 19 23:37:44 eyal kernel: EIP is at buffer_queue+0x33/0x6f [bttv]
Nov 19 23:37:44 eyal kernel: eax: d6573384 ebx: 00000000 ecx: e7507c80 edx: e7507ce4
Nov 19 23:37:44 eyal kernel: esi: d6572ecc edi: e7507c80 ebp: f907ffa0 esp: d5980c58
Nov 19 23:37:44 eyal kernel: ds: 007b es: 007b ss: 0068
Nov 19 23:37:44 eyal kernel: Process mythbackend (pid: 5981, threadinfo=d5980000 task=eae58020)
Nov 19 23:37:44 eyal kernel: Stack: 00200246 f908035c f9063ca1 d6572e00 e7507c80 f906deb4 00000300 00000240
Nov 19 23:37:44 eyal kernel: 00000004 d6572e00 e7507c80 c01230a8 c011248c 00000000 c010e5b9 0000bae2
Nov 19 23:37:44 eyal kernel: 000dd3d6 00000005 c0103ad8 0000bae2 9347d376 00000361 000dd3d6 00000005
Nov 19 23:37:44 eyal kernel: Call Trace:
Nov 19 23:37:44 eyal kernel: [pg0+952437921/1069462528] bttv_do_ioctl+0x4b4/0x17e4 [bttv]
Nov 19 23:37:44 eyal kernel: [irq_exit+58/60] irq_exit+0x3a/0x3c
Nov 19 23:37:44 eyal kernel: [smp_apic_timer_interrupt+105/222] smp_apic_timer_interrupt+0x69/0xde
Nov 19 23:37:44 eyal kernel: [delay_pmtmr+13/21] delay_pmtmr+0xd/0x15
Nov 19 23:37:44 eyal kernel: [apic_timer_interrupt+28/36] apic_timer_interrupt+0x1c/0x24
Nov 19 23:37:44 eyal kernel: [exit_notify+551/2193] exit_notify+0x227/0x891
Nov 19 23:37:44 eyal kernel: [delay_pmtmr+13/21] delay_pmtmr+0xd/0x15
Nov 19 23:37:44 eyal kernel: [avc_alloc_node+28/358] avc_alloc_node+0x1c/0x166
Nov 19 23:37:44 eyal kernel: [avc_latest_notif_update+62/124] avc_latest_notif_update+0x3e/0x7c
Nov 19 23:37:44 eyal kernel: [avc_node_replace+48/57] avc_node_replace+0x30/0x39
Nov 19 23:37:44 eyal kernel: [avc_insert+250/294] avc_insert+0xfa/0x126
Nov 19 23:37:44 eyal kernel: [avc_has_perm_noaudit+121/338] avc_has_perm_noaudit+0x79/0x152
Nov 19 23:37:44 eyal kernel: [avc_alloc_node+28/358] avc_alloc_node+0x1c/0x166
Nov 19 23:37:44 eyal kernel: [avc_node_replace+48/57] avc_node_replace+0x30/0x39
Nov 19 23:37:44 eyal kernel: [avc_has_perm+110/132] avc_has_perm+0x6e/0x84
Nov 19 23:37:44 eyal kernel: [inode_has_perm+85/131] inode_has_perm+0x55/0x83
Nov 19 23:37:44 eyal kernel: [avc_has_perm+110/132] avc_has_perm+0x6e/0x84
Nov 19 23:37:44 eyal kernel: [selinux_file_ioctl+238/811] selinux_file_ioctl+0xee/0x32b
Nov 19 23:37:44 eyal kernel: [inode_has_perm+85/131] inode_has_perm+0x55/0x83
Nov 19 23:37:44 eyal kernel: [do_no_page+428/689] do_no_page+0x1ac/0x2b1
Nov 19 23:37:44 eyal kernel: [copy_from_user+66/110] copy_from_user+0x42/0x6e
Nov 19 23:37:44 eyal kernel: [pg0+946086937/1069462528] video_usercopy+0x7b/0x135 [videodev]
Nov 19 23:37:44 eyal kernel: [do_mmap_pgoff+1119/1925] do_mmap_pgoff+0x45f/0x785
Nov 19 23:37:44 eyal kernel: [pg0+952442898/1069462528] bttv_ioctl+0x41/0x64 [bttv]
Nov 19 23:37:44 eyal kernel: [pg0+952436717/1069462528] bttv_do_ioctl+0x0/0x17e4 [bttv]
Nov 19 23:37:44 eyal kernel: [sys_ioctl+217/539] sys_ioctl+0xd9/0x21b
Nov 19 23:37:44 eyal kernel: [syscall_call+7/11] syscall_call+0x7/0xb
Nov 19 23:37:44 eyal kernel: Code: 24 04 8b 44 24 0c 8b 80 d0 00 00 00 8b 4c 24 10 8b 30 8d 51 64 c7 41 20 02 00 00 00 8d 86 b8 04 00 00 8b 58 04 89 41 64 89 50 04 <89> 13 89 5a 04 8b 8e d4 04 00 00 85 c9 74 0b 8b 1c 24 8b 74 24
Nov 19 23:37:44 eyal kernel: <6>note: mythbackend[5981] exited with preempt_count 2


And later I get

Nov 20 00:26:24 eyal kernel: 0240009a
Nov 20 00:26:24 eyal kernel: PREEMPT SMP
Nov 20 00:26:24 eyal kernel: Modules linked in: tsdev psmouse mt352 sp887x v4l1_compat dvb_bt8xx dvb_core i810_audio ac97_codec sr_mod sg ide_scsi ide_cd cdrom it87 eeprom i2c_isa i2c_i801 i2c_sensor eth1394 ohci_hcd ohci1394 ieee1394 dc395x scsi_mod bt878 bttv tuner video_buf firmware_class i2c_algo_bit v4l2_common btcx_risc videodev e1000 snd_intel8x0 snd_ac97_codec snd_pcm snd_timer snd snd_page_alloc soundcore i2c_core cfi_cmdset_0002 cfi_util mtdpart jedec_probe cfi_probe gen_probe ichxrom mtdcore chipreg map_funcs ehci_hcd uhci_hcd usbcore shpchp pci_hotplug intel_mch_agp intel_agp agpgart parport_pc parport 8250_pnp 8250 serial_core evdev nls_cp437 msdos fat dm_mod rtc unix
Nov 20 00:26:24 eyal kernel: CPU: 0
Nov 20 00:26:24 eyal kernel: EIP: 0060:[phys_startup_32+36700314/-1073741824] Not tainted VLI
Nov 20 00:26:24 eyal kernel: EFLAGS: 00010206 (2.6.10-rc2-mm2)
Nov 20 00:26:24 eyal kernel: EIP is at 0x240009a
Nov 20 00:26:24 eyal kernel: eax: e7507cf4 ebx: e7507ce4 ecx: 0240009a edx: 0000007b
Nov 20 00:26:24 eyal kernel: esi: 00000000 edi: d6f8a380 ebp: e7507ce4 esp: d72e2f8c
Nov 20 00:26:24 eyal kernel: ds: 007b es: 007b ss: 0068
Nov 20 00:26:24 eyal kernel: Process mythbackend (pid: 16038, threadinfo=d72e2000 task=d6d42550)
Nov 20 00:26:24 eyal kernel: Stack: c0159c3d e7507ce4 00000008 d6f8a380 000000e2 d6f8a384 c0159cbf e7507ce4
Nov 20 00:26:24 eyal kernel: d6f8a380 000000e2 000000e3 000003ff d72e2000 c01030db 000000e2 ffffffe0
Nov 20 00:26:24 eyal kernel: b6bb2fcc 000000e3 000003ff bffffcf8 00000006 0000007b 0000007b 00000006
Nov 20 00:26:24 eyal kernel: Call Trace:
Nov 20 00:26:24 eyal kernel: [filp_close+113/134] filp_close+0x71/0x86
Nov 20 00:26:24 eyal kernel: [sys_close+109/143] sys_close+0x6d/0x8f
Nov 20 00:26:24 eyal kernel: [syscall_call+7/11] syscall_call+0x7/0xb
Nov 20 00:26:24 eyal kernel: Code: Bad EIP value.

Nov 20 00:26:31 eyal kernel: <1>Unable to handle kernel paging request at virtual address 0240009a
Nov 20 00:26:31 eyal kernel: 0240009a
Nov 20 00:26:31 eyal kernel: PREEMPT SMP
Nov 20 00:26:31 eyal kernel: Modules linked in: tsdev psmouse mt352 sp887x v4l1_compat dvb_bt8xx dvb_core i810_audio ac97_codec sr_mod sg ide_scsi ide_cd cdrom it87 eeprom i2c_isa i2c_i801 i2c_sensor eth1394 ohci_hcd ohci1394 ieee1394 dc395x scsi_mod bt878 bttv tuner video_buf firmware_class i2c_algo_bit v4l2_common btcx_risc videodev e1000 snd_intel8x0 snd_ac97_codec snd_pcm snd_timer snd snd_page_alloc soundcore i2c_core cfi_cmdset_0002 cfi_util mtdpart jedec_probe cfi_probe gen_probe ichxrom mtdcore chipreg map_funcs ehci_hcd uhci_hcd usbcore shpchp pci_hotplug intel_mch_agp intel_agp agpgart parport_pc parport 8250_pnp 8250 serial_core evdev nls_cp437 msdos fat dm_mod rtc unix
Nov 20 00:26:31 eyal kernel: CPU: 0
Nov 20 00:26:31 eyal kernel: EIP: 0060:[phys_startup_32+36700314/-1073741824] Not tainted VLI
Nov 20 00:26:31 eyal kernel: EFLAGS: 00010206 (2.6.10-rc2-mm2)
Nov 20 00:26:31 eyal kernel: EIP is at 0x240009a
Nov 20 00:26:31 eyal kernel: eax: e7507cf4 ebx: e7507ce4 ecx: 0240009a edx: 0000007b
Nov 20 00:26:31 eyal kernel: esi: 00000000 edi: d6f8a580 ebp: e7507ce4 esp: d72e2f8c
Nov 20 00:26:31 eyal kernel: ds: 007b es: 007b ss: 0068
Nov 20 00:26:31 eyal kernel: Process mythbackend (pid: 16050, threadinfo=d72e2000 task=d6d42550)
Nov 20 00:26:31 eyal kernel: Stack: c0159c3d e7507ce4 00000008 d6f8a580 000000e2 d6f8a584 c0159cbf e7507ce4
Nov 20 00:26:31 eyal kernel: d6f8a580 000000e2 000000e3 000003ff d72e2000 c01030db 000000e2 ffffffe0
Nov 20 00:26:31 eyal kernel: b6bb2fcc 000000e3 000003ff bffffcf8 00000006 0000007b 0000007b 00000006
Nov 20 00:26:31 eyal kernel: Call Trace:
Nov 20 00:26:31 eyal kernel: [filp_close+113/134] filp_close+0x71/0x86
Nov 20 00:26:31 eyal kernel: [sys_close+109/143] sys_close+0x6d/0x8f
Nov 20 00:26:31 eyal kernel: [syscall_call+7/11] syscall_call+0x7/0xb
Nov 20 00:26:31 eyal kernel: Code: Bad EIP value.


--
Eyal Lebedinsky ([email protected]) <http://samba.org/eyal/>

2004-11-22 10:03:08

by Gerd Knorr

[permalink] [raw]
Subject: Re: Fw: Re: Linux 2.6.10-rc2 [dvb-bt8xx unload oops]

> Also, does anyone know how to switch to text console when X locks up (as
> it does for me)? sysrq works but does not allow me to switch to another
> console. Since the hard lock does not log the oops, and I cannot see the
> text from sysrq, I cannot report the details.

Setting up a serial console works best for that ...

> Nov 19 23:37:44 eyal kernel: EIP is at buffer_queue+0x33/0x6f [bttv]

That most likely a bug in bttv, I've pinned it down and queued a patch
to fix that end of last week. I've also seen mails from Andrew
forwarding stuff to Linus on a quick scan of my inbox, so the latest
-bk snapshots might already have that fix.

Gerd

--
#define printk(args...) fprintf(stderr, ## args)

2004-11-22 12:54:25

by Eyal Lebedinsky

[permalink] [raw]
Subject: Re: Fw: Re: Linux 2.6.10-rc2 [dvb-bt8xx unload oops]

Gerd Knorr wrote:
>>This is vanilla 2.6.10-rc2 on P4. This was a problem with -rc1 but
>
> Yes, looks very simliar ...
>
>>some patches off the list [attached] fixed it. I expected these to be
>>in -rc2, I am not able to say which patch is missing.
>
> Uhm, strange. The bttv patches _are_ merged.
> Not sure about any for dvb-bt8xx, Michael?

I now have the details, captured off a serial console. This is vanilla
2.6.10-rc2-bk6.

Attaches is a sysrq session that has the tasks list dump.

Unable to handle kernel NULL pointer dereference at virtual address 00000000
** PCI interrupts are no longer routed automatically. If this
printing eip:
f919e5f6
*pde = 00000000device to stop
Oops: 0002 [#1]robably because
PREEMPT SMP
Modules linked in: mt352 sp887x dvb_bt8xx dvb_core bt878 bttv tuner video_buf fievice(). As a temporary
_i801 i2c_sensor eth1394 ohci_hcd ohci1394 ieee1394 dc395x scsi_mod snd_bt87x e1ce pool size: 64 pages
Total HugeTLB memory al
000 snd_intel8x0 snd_ac97_codec snd_pcm snd_timer snd snd_page_alloc soundcore i
Dquot-cache hash table entries: 1024 (order 0, 4096 bytes)
2c_core cfi_cmdset_0002 cfi_util mtdpart jedec_probe cfi_probe gen_probe ichxromd Gooch ([email protected])
mtdcore chipreg map_funcs ehci_hcd uhci_hcd usbcore shpchp pci_hotplug intel_mclter hooks
Initializing Cryptographic API
EFLAGS: 00210092 (2.6.10-rc2-bk6)
serio: i8042 KBD port at 0x60,0x64
EIP is at buffer_queue+0x33/0x6f [bttv]
Serial: 8250/16550 driver $Revis
eax: f5b98d54 ebx: 00000000 ecx: d528a080 edx: d528a0e4

ttyS0
esi: f5b988c8 edi: d528a080 ebp: f91bbda0 esp: f7b7cc58
ttyS1 at I/O 0x2f8 (irq = 3) is a 16550A
ds: 007b es: 007b ss: 0068
ttyS0 at I/O 0x3f8 (irq = 4)
[<c013874d>] irq_exit+0x3a/0x3c
[<c0103a94>] apic_timer_interrupt+0x1c/0x24
[<c010e409>] delay_pmtmr+0xd/0x15
[<c01c2362>] __delay+0x12/0x16
[<f8a50446>] i2c_outb+0x1ff/0x2bb [i2c_algo_bit]
[<c01c2362>] __delay+0x12/0x16
[<c01ab2df>] avc_has_perm_noaudit+0x9a/0x1ad
[<c01ab468>] avc_has_perm+0x76/0x8e
[<c01ac7d0>] inode_has_perm+0x65/0x93
[<c01ab468>] avc_has_perm+0x76/0x8e
[<c01aef69>] selinux_file_ioctl+0xf9/0x352
[<c0149119>] do_no_page+0x195/0x284
[<c01c2948>] copy_from_user+0x42/0x6e
[<f8a47419>] video_usercopy+0x7b/0x135 [videodev]
[<c014ae63>] do_mmap_pgoff+0x442/0x77a
[<f91a1012>] bttv_ioctl+0x41/0x64 [bttv]
[<f919f7ed>] bttv_do_ioctl+0x0/0x17e4 [bttv]
[<c01696d2>] sys_ioctl+0xd9/0x21b
[<c0103097>] syscall_call+0x7/0xb
Code: 24 04 8b 44 24 0c 8b 80 cc 00 00 00 8b 4c 24 10 8b 30 8d 51 64 c7 41 20 02 00 00 00 8d 86 8c 04 00 00 8b 58 04 89 41 64 89 50 04 <89> 13 89 5a 04 8b 8e a8 04 00 00 85 c9 74 0b 8b 1c 24 8b 74 24
<6>note: mythbackend[14038] exited with preempt_count 2

--
Eyal Lebedinsky ([email protected]) <http://samba.org/eyal/>


Attachments:
minilog.zip (10.49 kB)