On 2020-03-23 08:32, Anshuman Gupta wrote:
> Some times delayed audio codec disabling causes failure
> of test, using igt_wait to check lpsp after disabling all outputs.

Seems like this answers my question on patch 2/5: We are sending audio
to the screen even when we do not have anything to send...

I know there are some screens that take ages to re-enable sound after
not getting any for a while, but waiting 10 or even 30s without sound
before disabling it does not sound like a terrible idea. My TV
automatically disables sound after way less than that anyway.

So, to me, this is a driver bug.

That being said, the patch is fine even if your reason for it is not
acceptable.

Martin
> 
> Signed-off-by: Anshuman Gupta <anshuman.gupta@intel.com>
> ---
>  tests/i915/i915_pm_lpsp.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/tests/i915/i915_pm_lpsp.c b/tests/i915/i915_pm_lpsp.c
> index 0ce694e6..5f15c034 100644
> --- a/tests/i915/i915_pm_lpsp.c
> +++ b/tests/i915/i915_pm_lpsp.c
> @@ -91,7 +91,7 @@ static void screens_disabled_subtest(data_t *data)
>  		igt_display_commit(&data->display);
>  	}
>  
> -	igt_assert(lpsp_is_enabled(data));
> +	igt_assert(igt_wait(lpsp_is_enabled(data), 1000, 100));
>  }
>  
>  static void check_output_lpsp(data_t *data)
>