Re: [PATCH 3/5] drm/i915: Increase busyspin limit before a context-switch

[Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>]

On 02/08/2018 15:40, Tvrtko Ursulin wrote:
> 
> On 28/07/2018 17:46, Chris Wilson wrote:
>> Looking at the distribution of i915_wait_request for a set of GL
> 
> What was the set?
> 
>> benchmarks, we see:
>>
>> broadwell# python bcc/tools/funclatency.py -u i915_wait_request
>>     usecs               : count     distribution
>>         0 -> 1          : 29184    
>> |****************************************|
>>         2 -> 3          : 5767     
>> |*******                                 |
>>         4 -> 7          : 3000     
>> |****                                    |
>>         8 -> 15         : 491      
>> |                                        |
>>        16 -> 31         : 140      
>> |                                        |
>>        32 -> 63         : 203      
>> |                                        |
>>        64 -> 127        : 543      
>> |                                        |
>>       128 -> 255        : 881      
>> |*                                       |
>>       256 -> 511        : 1209     
>> |*                                       |
>>       512 -> 1023       : 1739     
>> |**                                      |
>>      1024 -> 2047       : 22855    
>> |*******************************         |
>>      2048 -> 4095       : 1725     
>> |**                                      |
>>      4096 -> 8191       : 5813     
>> |*******                                 |
>>      8192 -> 16383      : 5348     
>> |*******                                 |
>>     16384 -> 32767      : 1000     
>> |*                                       |
>>     32768 -> 65535      : 4400     
>> |******                                  |
>>     65536 -> 131071     : 296      
>> |                                        |
>>    131072 -> 262143     : 225      
>> |                                        |
>>    262144 -> 524287     : 4        
>> |                                        |
>>    524288 -> 1048575    : 1        
>> |                                        |
>>   1048576 -> 2097151    : 1        
>> |                                        |
>>   2097152 -> 4194303    : 1        
>> |                                        |
>>
>> broxton# python bcc/tools/funclatency.py -u i915_wait_request
>>     usecs               : count     distribution
>>         0 -> 1          : 5523     
>> |*************************************   |
>>         2 -> 3          : 1340     
>> |*********                               |
>>         4 -> 7          : 2100     
>> |**************                          |
>>         8 -> 15         : 755      
>> |*****                                   |
>>        16 -> 31         : 211      
>> |*                                       |
>>        32 -> 63         : 53       
>> |                                        |
>>        64 -> 127        : 71       
>> |                                        |
>>       128 -> 255        : 113      
>> |                                        |
>>       256 -> 511        : 262      
>> |*                                       |
>>       512 -> 1023       : 358      
>> |**                                      |
>>      1024 -> 2047       : 1105     
>> |*******                                 |
>>      2048 -> 4095       : 848      
>> |*****                                   |
>>      4096 -> 8191       : 1295     
>> |********                                |
>>      8192 -> 16383      : 5894     
>> |****************************************|
>>     16384 -> 32767      : 4270     
>> |****************************            |
>>     32768 -> 65535      : 5622     
>> |**************************************  |
>>     65536 -> 131071     : 306      
>> |**                                      |
>>    131072 -> 262143     : 50       
>> |                                        |
>>    262144 -> 524287     : 76       
>> |                                        |
>>    524288 -> 1048575    : 34       
>> |                                        |
>>   1048576 -> 2097151    : 0        
>> |                                        |
>>   2097152 -> 4194303    : 1        
>> |                                        |
>>
>> Picking 20us for the context-switch busyspin has the dual advantage of
>> catching most frequent short waits while avoiding the cost of a context
>> switch. 20us is a typical latency of 2 context-switches, i.e. the cost
>> of taking the sleep, without the secondary effects of cache flushing.
>>
>> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
>> Cc: Sagar Kamble <sagar.a.kamble@intel.com>
>> Cc: Eero Tamminen <eero.t.tamminen@intel.com>
>> Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
>> Cc: Ben Widawsky <ben@bwidawsk.net>
>> Cc: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
>> Cc: Michał Winiarski <michal.winiarski@intel.com>
>> ---
>>   drivers/gpu/drm/i915/Kconfig.profile | 2 +-
>>   1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/gpu/drm/i915/Kconfig.profile 
>> b/drivers/gpu/drm/i915/Kconfig.profile
>> index 63cb744d920d..de394dea4a14 100644
>> --- a/drivers/gpu/drm/i915/Kconfig.profile
>> +++ b/drivers/gpu/drm/i915/Kconfig.profile
>> @@ -14,7 +14,7 @@ config DRM_I915_SPIN_REQUEST_IRQ
>>   config DRM_I915_SPIN_REQUEST_CS
>>       int
>> -    default 2 # microseconds
>> +    default 20 # microseconds
>>       help
>>         After sleeping for a request (GPU operation) to complete, we will
>>         be woken up on the completion of every request prior to the one
>>
> 
> I'd be more tempted to pick 10us given the histograms. It would avoid 
> wasting cycles on Broadwell and keep the majority of the benefit on 
> Broxton.

Actually the first spin is 5us so are you sure bumping of the second 
spin should be the first step? In other words, wouldn't bumping the 
first one to 10us eliminate most the the low bars from the histogram?

Regards,

Tvrtko

> However.. it also raises the question if we perhaps want to have this 
> initialized per-platform at runtime.. ? That would open up the way of 
> auto-tuning it, if the goal is to eliminate the low part of the histogram.
> 
> Also, please add to the commit what kind of perf/watt or something 
> effect on the benchmarks we get with it.
> 
> Regards,
> 
> Tvrtko
> _______________________________________________
> Intel-gfx mailing list
> Intel-gfx@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/intel-gfx
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx