[igt-dev] [[PATCH i-g-t]] tests/kms_setmode: disable DRRS if eDP support variable refresh rate

[Mark Yacoub]

On Wed, Oct 18, 2023 at 4:05 AM Lee Shawn C <shawn.c.lee at intel.com> wrote:
>
> Below is a failed case while testing kms_setmode.
>
> CRTC[80] [Pipe A] Mode: 1920x1200 at 60Hz Connectors: eDP-1[236]
> Expected frametime: 16659us; measured 20824.4us +- 4.799us accuracy 0.01% [0.10 scanlines]
>
> The expected frame time is 16659us (60Hz refresh rate).
> But kms_setmode report measured time was 20824.4us (48Hz) then
> it failed. We found legacy DRRS mode was active at that time
> then refresh rate fall into lower RR automatically to cause this problem.
> Disable DRRS after crtc setmode can avoid driver switch to lower RR mode

Is it okay for the driver to do so? Would it be that the driver should
not do that if it wasn't asked?
Thanks,
-Mark

> and get pass for this test as well.
>
> Issue: https://gitlab.freedesktop.org/drm/igt-gpu-tools/-/issues/148
> Cc: Tvrtko Ursulin <tvrtko.ursulin at linux.intel.com>
> Cc: Vidya Srinivas <vidya.srinivas at intel.com>
> Signed-off-by: Lee Shawn C <shawn.c.lee at intel.com>
> ---
>  tests/kms_setmode.c | 4 ++++
>  1 file changed, 4 insertions(+)
>
> diff --git a/tests/kms_setmode.c b/tests/kms_setmode.c
> index 5fc4421c0080..b6fe3f9c159a 100644
> --- a/tests/kms_setmode.c
> +++ b/tests/kms_setmode.c
> @@ -31,6 +31,7 @@
>  #include <sys/time.h>
>  #include <math.h>
>  #include "xe/xe_query.h"
> +#include "i915/intel_drrs.h"
>  /**
>   * TEST: kms setmode
>   * Category: Display
> @@ -651,6 +652,9 @@ retry:
>                                              crtc->fb_info.fb_id, 0, 0, ids,
>                                              crtc->connector_count, &crtc->mode);
>
> +               if (is_intel_device(drm_fd))
> +                       intel_drrs_disable(drm_fd, crtc->pipe_id);
> +
>                 free(ids);
>
>                 if (ret < 0) {
> --
> 2.34.1
>