[Intel-gfx] [PATCH 3/4] drm/i915/perf: only append status when data is available

[Lionel Landwerlin]

On 17/03/2020 00:16, Dixit, Ashutosh wrote:
> On Thu, 12 Mar 2020 16:05:01 -0700, Umesh Nerlige Ramappa wrote:
>> From: Lionel Landwerlin <lionel.g.landwerlin at intel.com>
>>
>> The only bit of the status register we currently report in the
>> i915-perf stream is the "report loss" bit. Only report this when we
>> have some data to report with it. There was a kind of inconsistency
>> here in that we could report report loss without appending the reports
>> associated with the loss.
> Splitting hair a bit, but I am wondering if this is realistic? If reports
> have been lost in the middle of a OA buffer then there /will/ be some data
> from the hardware so head != tail. So is the situation which this patch is
> fixing ever been observed in practice?
>
> Also, if we are doing this, how about moving the entire status handling
> here, including intel_uncore_read() and OABUFFER_OVERFLOW handling (which I
> understand resets the stream so probably doesn't have associated data).
>
> In any case, since these are just random questions, this is:
>
> Reviewed-by: Ashutosh Dixit <ashutosh.dixit at intel.com>

For some reason I thought this writing of lost report was inconsistent.

But it's fairly unrelated to this series.

Maybe drop this patch...


-Lionel