[PATCH i-g-t v1 1/1] tests/kms_cursor_legacy: don't inhibit pipe CRC calculation

Sharma, Swati2 swati2.sharma at intel.com
Mon Jul 21 18:12:18 UTC 2025


Hi Michal,

On 15-07-2025 06:06 pm, Michał Grzelak wrote:
> Running either flip-vs-cursor-busy-crc-(legacy|atomic) subtest from
> kms_cursor_legacy results in entering PSR1 after the test. Even with
> PSR2 enabled panel, it is impossible to switch to PSR2 via
> i915_edp_psr_debug. Kernel does so because it prevents inhibiting pipe
> CRC calculation which is allowed in PSR1 but prohibited in PSR2.
> Therefore subtests from kms_psr testing PSR2 feature (e.g. psr2-no-drrs,
> psr2-primary-blt etc.) are failing after running any of those
> kms_cursor_legacy subtests.
>
> Restore enabling Panel Replay Selective Update by cleaning pipe CRC
> calculation after running the subtest.
>
> Signed-off-by: Michał Grzelak <michal.grzelak at intel.com>
> ---
>   tests/kms_cursor_legacy.c | 2 +-
>   1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/tests/kms_cursor_legacy.c b/tests/kms_cursor_legacy.c
> index 9f6e644e1..2ef9690ea 100644
> --- a/tests/kms_cursor_legacy.c
> +++ b/tests/kms_cursor_legacy.c
> @@ -1751,7 +1751,7 @@ static void flip_vs_cursor_busy_crc(igt_display_t *display, bool atomic)
>   	/* Clean-up */
>   	igt_plane_set_fb(plane_primary, NULL);
>   	igt_plane_set_fb(cursor, NULL);
> -	igt_output_set_pipe(output, PIPE_NONE);
> +	igt_output_set_pipe(output, pipe);
During clean-up we set PIPE_NONE. This is standard clean-up process.
>   	igt_display_commit2(display, display->is_atomic ? COMMIT_ATOMIC : COMMIT_LEGACY);
>   
>   	igt_pipe_crc_stop(pipe_crc);
Can you re-order? Put crc_stop() first, followed by cleanup? Check if it 
helps.



More information about the igt-dev mailing list