Re: [PATCH i-g-t 3/3] tests/perf_pmu: Verify engine busyness accuracy

From: Tvrtko Ursulin <tvrtko.ursulin@linux.intel.com>
To: Chris Wilson <chris@chris-wilson.co.uk>,
	Tvrtko Ursulin <tursulin@ursulin.net>,
	Intel-gfx@lists.freedesktop.org
Subject: Re: [PATCH i-g-t 3/3] tests/perf_pmu: Verify engine busyness accuracy
Date: Wed, 20 Dec 2017 09:48:04 +0000	[thread overview]
Message-ID: <bd5eae76-222f-dcc7-3281-8c3becc4ce7b@linux.intel.com> (raw)
In-Reply-To: <151372350597.3125.1855432770553851194@mail.alporthouse.com>

On 19/12/2017 22:45, Chris Wilson wrote:
> Quoting Tvrtko Ursulin (2017-12-19 15:45:43)
>> +static void debug_error(const char *str, double val, double ref)
>> +{
>> +       igt_debug("%s=%.2f%% (%.2f/%.2f)\n", str, __error(val, ref), val, ref);
>> +}
>> +
>> +static void log_error(const char *str, double val, double ref)
>> +{
>> +       debug_error(str, val, ref);
>> +       igt_info("%s=%.2f%%\n", str, __error(val, ref));
>> +}
> 
> Looking at the CI output, these could really do with some explanation.

Yes a very disappointing think to look at in the morning. Much worse 
than the previous version which used calibrated batches. While at the 
same time new version is much better on my SKL..

Regards,

Tvrtko
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/intel-gfx