Re: Linuv 2.6.15-rc1

From: Zilvinas Valinskas
Date: Wed Nov 16 2005 - 13:16:48 EST


Hello Pekka,

I've tried to reproduce a freeze. For about an hour tortures gave no
results. Usual messages:

Nov 16 20:02:00 evo800N kernel: ipw2200: Firmware error detected. Restarting.
Nov 16 20:02:00 evo800N kernel: ipw2200: Sysfs 'error' log already exists.
Nov 16 20:02:42 evo800N kernel: ipw2200: Firmware error detected. Restarting.
Nov 16 20:02:42 evo800N kernel: ipw2200: Sysfs 'error' log already exists.
Nov 16 20:03:01 evo800N kernel: ipw2200: Firmware error detected. Restarting.
Nov 16 20:03:01 evo800N kernel: ipw2200: Sysfs 'error' log already exists.
Nov 16 20:04:10 evo800N kernel: ipw2200: Firmware error detected. Restarting.
Nov 16 20:04:10 evo800N kernel: ipw2200: Sysfs 'error' log already exists.

Other than that nothing unusual :) Even under X running, I cannot freeze
the laptop. Anything else to try ?

On Wed, Nov 16, 2005 at 06:33:32PM +0200, Pekka Enberg wrote:
> Hi Zilvinas,
>
> On Wed, 2005-11-16 at 15:15 +0200, Zilvinas Valinskas wrote:
> > Also this time I saw f/w loading problem and as a result, oops:
> > http://www.gemtek.lt/~zilvinas/dumps/trace.4.
> >
> > F/W load has failed as a result followed oops in yenta driver +
> > slabcorruption ... Rebooted several times, no f/w loading problems and
> > now I am running 2.6.15-rc1 in runlevel 2 (X started & services ...).
>
> The oops is not in yenta but most likely in the release() function of
> fs/sysfs/bin.c and related to ipw firmware loading. The function is
> touching already freed memory. As I am mostly clueless of sysfs and
> hotplug, perhaps someone can give us a hint?
>

I think you are right - I remmember seeing EIP: release+ (numbers ...)
earlier in the Ooops output.

http://www.gemtek.lt/~zilvinas/dumps/oops.1

CPU: 0
EIP: 0060:[<c0181cec>] Not tainted VLI
EFLAGS: 00010202 (2.6.15-rc1)
EIP is at release+0x2a/0x4f <---

It happened right after f/w upload had failed as it seems.

pw2200: ipw-2.4-bss.fw load failed: Reason -2
ipw2200: Unable to load firmware: -2
ipw2200: failed to register network device
ACPI: PCI interrupt for device 0000:02:04.0 disabled
ipw2200: probe of 0000:02:04.0 failed with error -5


Zilvinas Valinskas

> The oops from Zilvinas logs is the following:
>
> Unable to handle kernel paging request at virtual address 6b6b6c6b
> printing eip:
> c0181cec
> *pde = 00000000
> Oops: 0002 [#1]
> DEBUG_PAGEALLOC
> Modules linked in: ehci_hcd yenta_socket rsrc_nonstatic pcmcia_core ohci_hcd snd_intel8x0 snd_ac97_codec snd_ac97_bus snd_pcm_oss snd_mixer_oss irtty_sir snd_pcm snd_timer sir_dev ipw2200 irda snd soundcore ieee80211 ieee80211_crypt crc_ccitt snd_page_alloc floppy firmware_class 8250_pnp intel_agp agpgart pcspkr 8250 serial_core ide_cd cdrom
> CPU: 0
> EIP: 0060:[<c0181cec>] Not tainted VLI
> EFLAGS: 00010202 (2.6.15-rc1)
> EIP is at release+0x2a/0x4f
> eax: 6b6b6b6b ebx: ddb23f00 ecx: 1dcde000 edx: 00000001
> esi: dd900ddc edi: deff2000 ebp: deb0bf74 esp: deb5ff74
> ds: 007b es: 007b ss: 0068
> Process cat (pid: 1926, threadinfo=deb5f000 task=dd845b10)
> Stack: 00000008 de36bfd4 ddb54eb4 c0149a4d 00000000 de36bf70 c14d9e38 ddb54eb4
> deb0bf74 00000000 df8d2e48 deb5f000 c01484ef 00000001 00000001 b7f275c0
> ffffffff c0102bd3 00000001 00000000 b7f26ff4 b7f275c0 ffffffff bfe649e8
> Call Trace:
> [<c0149a4d>] __fput+0xa0/0x14f
> [<c01484ef>] filp_close+0x3e/0x62
> [<c0102bd3>] sysenter_past_esp+0x54/0x75
> Code: e8 57 56 53 8b 4a 08 8b 41 14 8b 40 50 8b 58 14 8b 41 50 8b 70 14 8b 7a 74 85 db 74 07 89 d8 e8 26 4b 0a 00 8b 46 04 85 c0 74 0b <ff> 88 00 01 00 00 83 38 02 74 0d 89 f8 e8 7e 4e fb ff 31 c0 5b
>
> Pekka
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@xxxxxxxxxxxxxxx
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/