Re: [PATCH] rate limit drm:radeon_cp_idle/reset errors

[Roberto Oppedisano <roberto.oppedisano@infracom.it>]

Dave Airlie ha scritto, Il 09/09/2008 10.15:
>>>>> Sep  6 10:24:31 poppero1 kernel: [  186.138774] [drm:radeon_cp_start] *ERROR* radeon_cp_start called without lock held, held  0 owner f726bc80 f68f6840
>>>>> Sep  6 10:24:31 poppero1 kernel: [  186.138968] [drm:radeon_cp_idle] *ERROR* radeon_cp_idle called without lock held, held  0 owner f726bc80 f68f6840
>>>>> Sep  6 10:24:31 poppero1 kernel: [  186.139214] [drm:radeon_cp_reset] *ERROR* radeon_cp_reset called without lock held, held  0 owner f726bc80 f68f6840
>>>>> Sep  6 10:24:31 poppero1 kernel: [  186.139408] [drm:radeon_cp_start] *ERROR* radeon_cp_start called without lock held, held  0 owner f726bc80 f68f6840
>>>>> Sep  6 10:24:31 poppero1 kernel: [  186.139601] [drm:radeon_cp_idle] *ERROR* radeon_cp_idle called without lock held, held  0 owner f726bc80 f68f6840
>>>>> Sep  6 10:24:31 poppero1 kernel: [  186.139866] [drm:radeon_cp_reset] *ERROR* radeon_cp_reset called without lock held, held  0 owner f726bc80 f68f6840
>>>>> Sep  6 10:24:31 poppero1 kernel: [  186.140072] [drm:radeon_cp_start] *ERROR* radeon_cp_start called without lock held, held  0 owner f726bc80 f68f6840
>>>>> Sep  6 10:24:31 poppero1 kernel: [  186.140467] [drm:radeon_cp_idle] *ERROR* radeon_cp_idle called without lock held, held  0 owner f726bc80 f68f6840
>>>>> Sep  6 10:24:36 poppero1 kernel: [  191.139019] __ratelimit: 253431 callbacks suppressed
>>>>> Sep  6 10:24:36 poppero1 kernel: [  191.139030] [drm:radeon_cp_reset] *ERROR* radeon_cp_reset called without lock held, held  0 owner f726bc80 f68f6840
>>>>> Sep  6 10:24:36 poppero1 kernel: [  191.139314] [drm:radeon_cp_start] *ERROR* radeon_cp_start called without lock held, held  0 owner f726bc80 f68f6840
>>>>>
>>>>> Also if it doesn't solve a bug I think it may still be worth applying it. 
>>>>> Patch is against current git.
>>>>>
>>>>> Signed-off-by: Roberto Oppedisano <roberto.oppedisano@infracom.it>
>>>>>
>>>>> diff --git a/include/drm/drmP.h b/include/drm/drmP.h
>>>>> index 1c1b13e..1107361 100644
>>>>> --- a/include/drm/drmP.h
>>>>> +++ b/include/drm/drmP.h
>>>>> @@ -162,7 +162,8 @@ struct drm_device;
>>>>>   * \param arg arguments
>>>>>   */
>>>>>  #define DRM_ERROR(fmt, arg...) \
>>>>> -	printk(KERN_ERR "[" DRM_NAME ":%s] *ERROR* " fmt , __func__ , ##arg)
>>>>> +	if (printk_ratelimit()) \
>>>>> +		printk(KERN_ERR "[" DRM_NAME ":%s] *ERROR* " fmt , __func__ , ##arg)
>>>>>  
>>>>>  /**
>>>>>   * Memory error output.
>>>>>
>>>> Which kernel version(s)?
>>> This is against vanilla current git. 
>>>
>>> Linux poppero1 2.6.27-rc5-00000-g7686ad5-dirty #1 PREEMPT Sun Sep 7 08:47:08 CEST 2008 i686 GNU/Linux
>>>
>> Thanks.  I should have asked earlier: was 2.6.26 OK?  Any other kernels
>> tested?
> 
> This looks like a userspace problem, some app is accessing the drm file 
> descriptror without holding the drm lock, this is very illegal.
> 
> Its probably a race between the X server and the AIGLX driver.
> 
> Dave.

I also think it's a userspace problem. I found it while playing with
KDE4.1 kwin/Compiz, less than a month ago, always running on a current
vanilla git kernel: I noticed the syslog storm only recently (my bad,
because judging from the disk light I think it has always been there).

If you think it's useful I can do some test with older kernels (will
take some time btw).
R