Re: 2.6.30-rc1: OOPS on usb disk insert

[Justin Madru <jdm64@gawab.com>]

Andrew Morton wrote:
> (cc linux-scsi and linux-usb)
>
> On Tue, 07 Apr 2009 22:05:24 -0700 Justin Madru <jdm64@gawab.com> wrote:
>
>   
>> Hello,
>>
>> Testing out .30-rc1 and usb devises don't work. I insert my flash disk 
>> and get this oops.
>> More information (dmesg, config, etc.) at 
>> http://jdserver.homelinux.org/bugreports/003/
>>     
>
> It looks the async code blew up.
>
>   
>> usb 5-3: new high speed USB device using ehci_hcd and address 2
>> usb 5-3: configuration #1 chosen from 1 choice
>> Initializing USB Mass Storage driver...
>> scsi2 : SCSI emulation for USB Mass Storage devices
>> usb-storage: device found at 2
>> usb-storage: waiting for device to settle before scanning
>> usbcore: registered new interface driver usb-storage
>> USB Mass Storage support registered.
>> usb-storage: device scan complete
>> scsi 2:0:0:0: Direct-Access     SanDisk  Cruzer Micro     8.01 PQ: 0 
>> ANSI: 0 CCS
>> BUG: unable to handle kernel NULL pointer dereference at (null)
>> IP: [<(null)>] (null)
>>     
>
> The kernel tried to execute code at 0x00000000
>
>   
>> *pde = 00000000
>> Oops: 0000 [#2] PREEMPT SMP
>> last sysfs file: 
>> /sys/devices/pci0000:00/0000:00:1e.0/0000:03:00.0/ssb0:0/net/eth0/statistics/collisions
>> Modules linked in: usb_storage i915 cfbcopyarea cfbimgblt cfbfillrect 
>> binfmt_misc acpi_cpufreq container sbs cpufreq_powersave pci_slot 
>> cpufreq_stats cpufreq_userspace sbshc cpufreq_conservative ext3 jbd 
>> mbcache firewire_sbp2 ecb cryptomgr aead pcompress crypto_blkcipher 
>> crypto_hash crypto_algapi dell_laptop snd_hda_codec_idt evdev iwl3945 
>> iwlcore rfkill sdhci_pci dcdbas video backlight rtc_cmos snd_hda_intel 
>> fb output sdhci rtc_core rtc_lib mac80211 snd_hda_codec snd_hwdep 
>> snd_pcm_oss snd_mixer_oss mmc_core psmouse snd_pcm battery ac button 
>> processor intel_agp snd_page_alloc reiserfs crc32 sr_mod cdrom sg 
>> firewire_ohci firewire_core crc_itu_t ata_piix ehci_hcd uhci_hcd usbcore 
>> thermal fan
>>
>> Pid: 5579, comm: scsi_scan_2 Tainted: G      D    (2.6.30-rc1-git #1) 
>> MM061                          
>> EIP: 0060:[<00000000>] EFLAGS: 00010046 CPU: 0
>> EIP is at 0x0
>> EAX: f7117fb0 EBX: f7117fb0 ECX: 00000000 EDX: 00000003
>> ESI: b8023874 EDI: 00000000 EBP: e4432dd4 ESP: e4432dac
>>  DS: 007b ES: 007b FS: 00d8 GS: 0000 SS: 0068
>> Process scsi_scan_2 (pid: 5579, ti=e4432000 task=ee7b3bb0 task.ti=e4432000)
>> Stack:
>>  c012a8af 00000000 00000001 00000003 c04f22d8 00000282 c04f22dc 00000005
>>  00000000 ee42ea20 e4432e00 c015e8b9 00000000 c0240d8d c04f22f8 f6527b00
>>  c02ff2d0 00000080 ffffffff f6527b00 e445c60e e4432e08 c015e93d e4432e44
>> Call Trace:
>>  [<c012a8af>] ? __wake_up+0x4f/0x80
>>  [<c015e8b9>] ? __async_schedule+0x139/0x1a0
>>  [<c0240d8d>] ? ida_pre_get+0x11d/0x150
>>  [<c02ff2d0>] ? sd_probe_async+0x0/0x270
>>  [<c015e93d>] ? async_schedule+0xd/0x10
>>  [<c02ff748>] ? sd_probe+0x208/0x2a0
>>  [<c02e6df6>] ? really_probe+0x156/0x220
>>  [<c02e6fbc>] ? __device_attach+0x4c/0x60
>>  [<c02e5b0b>] ? bus_for_each_drv+0x5b/0x80
>>  [<c02e6c7b>] ? device_attach+0xeb/0x110
>>  [<c02e6f70>] ? __device_attach+0x0/0x60
>>  [<c02e5a7f>] ? bus_attach_device+0x3f/0x70
>>  [<c02e442e>] ? device_add+0x65e/0x830
>>  [<c02ea8e0>] ? transport_configure+0x0/0x30
>>  [<c02fb3b2>] ? scsi_sysfs_add_sdev+0x62/0x2a0
>>  [<c0241592>] ? kobject_get+0x12/0x20
>>  [<c02f779d>] ? scsi_finish_async_scan+0x11d/0x160
>>  [<c02f9920>] ? do_scan_async+0x0/0x90
>>  [<c02f9978>] ? do_scan_async+0x58/0x90
>>  [<c02f9920>] ? do_scan_async+0x0/0x90
>>  [<c015558c>] ? kthread+0x3c/0x70
>>  [<c0155550>] ? kthread+0x0/0x70
>>  [<c0103f5f>] ? kernel_thread_helper+0x7/0x18
>> Code:  Bad EIP value.
>> EIP: [<00000000>] 0x0 SS:ESP 0068:e4432dac
>> CR2: 0000000000000000
>>     
>
> The trace appears to be claiming that kernel/async.c:async_new.func is
> NULL.
>
> But it cannot be - it's initialised at compilation time.  Unless
> something scribbled on it of course.
>
> Cute.  This might require a bisection search unless someone can see
> something which I missed?
>
>   
Well, since it's consistently repeatable, a bisection shouldn't be too 
painful.
I'll try to do that tomorrow.

Justin Madru