Re: [BUG] NULL pointer dereference in dev_get_drvdata

[Mauro Carvalho Chehab <mchehab@redhat.com>]

Em 29-03-2011 22:28, Steven Rostedt escreveu:
> While running ktest.pl with randconfig tests on latest Linus (before
> 2.6.38-rc9 was released), several tests triggered a bug similar to this:
> 
> Starting udev: udevd-work[880]: '/usr/bin/vmmouse_detect' unexpected exit with status 0x000b
> udevd-work[881]: '/usr/bin/vmmouse_detect' unexpected exit with status 0x000b
> BUG: unable to handle kernel NULL pointer dereference at 0000000000000018
> IP: [<ffffffff818a6e9e>] dev_get_drvdata+0xe/0x39
> PGD 74dc7067 PUD 741ef067 PMD 0 
> Oops: 0000 [#1] SMP 
> last sysfs file: /sys/devices/virtual/net/teql0/address
> CPU 1 
> Modules linked in: ide_pci_generic iTCO_wdt iTCO_vendor_support ide_core
> 
> Pid: 994, comm: v4l_id Not tainted 2.6.38-test-09043-gc585015 #1                  /DG965MQ
> RIP: 0010:[<ffffffff818a6e9e>]  [<ffffffff818a6e9e>] dev_get_drvdata+0xe/0x39
> RSP: 0018:ffff8800743dbb68  EFLAGS: 00010202
> RAX: 0000000000000000 RBX: ffff8800743dbba0 RCX: 0000000000000000
> RDX: ffffffff8408b740 RSI: 0000000000000000 RDI: 0000000000000010
> RBP: ffff8800743dbb68 R08: ffff8800743dbb98 R09: 00000000000023e7
> R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000000
> R13: ffff88007784f888 R14: ffff880076f82140 R15: 0000000000000000
> FS:  00007f4059724700(0000) GS:ffff88007e280000(0000) knlGS:0000000000000000
> CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> CR2: 0000000000000018 CR3: 00000000742bb000 CR4: 00000000000006e0
> DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
> Process v4l_id (pid: 994, threadinfo ffff8800743da000, task ffff880074dac420)
> Stack:
>  ffff8800743dbb88 ffffffff818e4129 ffff880074dac420 ffffffff8408b740
>  ffff8800743dbbc8 ffffffff818e1865 2222222222222222 2222222222222222
>  ffff88007784f000 0000000000000000 ffff88007784f888 ffff880076f82140
> Call Trace:
>  [<ffffffff818e4129>] st_kim_ref+0x2c/0x42
>  [<ffffffff818e1865>] st_register+0x29/0x444
>  [<ffffffff81dc6415>] fmc_prepare+0x75/0x382
>  [<ffffffff81dc9109>] fm_v4l2_fops_open+0x5a/0xce
>  [<ffffffff81d40bab>] v4l2_open+0x180/0x212
>  [<ffffffff811c6fc5>] chrdev_open+0x1c5/0x204
>  [<ffffffff811c6e00>] ? cdev_put+0x45/0x45
>  [<ffffffff811bf61e>] __dentry_open+0x2a1/0x467
>  [<ffffffff811c0bd8>] nameidata_to_filp+0x75/0x83
>  [<ffffffff811d2c5f>] do_last+0x716/0x902
>  [<ffffffff811d02e7>] ? path_init+0x1fd/0x46f
>  [<ffffffff811d3d07>] path_openat+0x102/0x4cd
>  [<ffffffff811d413d>] do_filp_open+0x6b/0xb5
>  [<ffffffff822671ab>] ? _raw_spin_unlock+0x40/0x4c
>  [<ffffffff811e387e>] ? alloc_fd+0x165/0x17e
>  [<ffffffff811c0c7f>] do_sys_open+0x99/0x16a
>  [<ffffffff811c0d77>] sys_open+0x27/0x30
>  [<ffffffff82270a82>] system_call_fastpath+0x16/0x1b
> Code: 90 55 48 89 e5 0f 1f 44 00 00 48 ff 05 ac 70 5d 02 f0 ff 07 48 ff 05 aa 70 5d 02 c9 c3 55 48 89 e5 0f 1f 44 00 00 48 85 ff 74 20
>  8b 47 08 48 ff 05 97 70 5d 02 48 85 c0 74 10 48 8b 80 b8 00 
> RIP  [<ffffffff818a6e9e>] dev_get_drvdata+0xe/0x39
>  RSP <ffff8800743dbb68>
> CR2: 0000000000000018
> 
> As not all configs triggered this, but when a config did trigger it, it
> was consistent. I ran the "config-bisect" of ktest.pl to find the dirty
> culprit, and it came up with:
> 
> ***************************************
> Found bad config: CONFIG_RADIO_WL128X
> ***************************************
> 
> When I removed the config, sure enough, the bug goes away. I added it
> back in and the bug re-appeared.
> 
> The .config can be downloaded here:
> http://rostedt.homelinux.com/private/mxtest-boot-randconfig-fail-20110228232103/config
> 
> and the full dmesg here:
> http://rostedt.homelinux.com/private/mxtest-boot-randconfig-fail-20110228232103/dmesg
> 
> -- Steve

Thanks for reporting it.

There were a change at ST interface, affecting this new driver.

I'm updating the fix today to -next:

commit f3149c7d9d989796306c6e65afbf4e89b3f81436
Author: Manjunatha Halli <manjunatha_halli@ti.com>
Date:   Wed Mar 23 07:44:30 2011 -0300

    [media] radio: wl128x: Update registration process with ST
    
    As underlying ST driver registration API's have changed with
    latest 2.6.38-rc8 kernel this patch will update the FM driver
    accordingly.
    
    Signed-off-by: Manjunatha Halli <manjunatha_halli@ti.com>
    Signed-off-by: Mauro Carvalho Chehab <mchehab@redhat.com>

It is already on my tree:
	ssh://master.kernel.org/pub/scm/linux/kernel/git/mchehab/linux-next.git

It should be available soon at kernel.org mirrors.

Could you please test it?

Thanks!
Mauro