[igt-dev] [PATCH i-g-t 1/2] tests/kms_plane_lowres: Skip Yf tiling tests from gen12 onwards

[Kahola, Mika]

On Thu, 2020-01-30 at 12:30 +0200, Jani Nikula wrote:
> On Mon, 27 Jan 2020, Mika Kahola <mika.kahola at intel.com> wrote:
> > Yf tiling is not supported by gen12. The patch proposes that we
> > intentionally skip this subtest.
> > 
> > Signed-off-by: Mika Kahola <mika.kahola at intel.com>
> > ---
> >  tests/kms_plane_lowres.c | 6 ++++++
> >  1 file changed, 6 insertions(+)
> > 
> > diff --git a/tests/kms_plane_lowres.c b/tests/kms_plane_lowres.c
> > index 4c3f5636..f4a9ff8d 100644
> > --- a/tests/kms_plane_lowres.c
> > +++ b/tests/kms_plane_lowres.c
> > @@ -37,6 +37,7 @@ IGT_TEST_DESCRIPTION("Test atomic mode setting
> > with a plane by switching between
> >  typedef struct {
> >  	int drm_fd;
> >  	igt_display_t display;
> > +	uint32_t devid;
> >  	struct igt_fb fb_primary;
> >  	struct igt_fb fb_plane[2];
> >  	struct {
> > @@ -236,6 +237,9 @@ test_planes_on_pipe(data_t *data, enum pipe
> > pipe, uint64_t modifier)
> >  	igt_skip_on(!igt_display_has_format_mod(&data->display,
> >  						DRM_FORMAT_XRGB8888,
> > modifier));
> >  
> > +	igt_skip_on(intel_gen(data->devid) >= 12 &&
> 
> In general I don't think we should skip based on platform/gen. Check
> for
> the *feature*, not the *platform*. Don't accumulate platform specific
> knowledge to igt.
Ok. I'll drop this platform specific check. 

Cheers,
Mika

> 
> BR,
> Jani.
> 
> > +		    modifier == LOCAL_I915_FORMAT_MOD_Yf_TILED);
> > +
> >  	for_each_valid_output_on_pipe(&data->display, pipe, output)
> >  		tested += test_planes_on_pipe_with_output(data, pipe,
> > output,
> >  							  modifier);
> > @@ -250,6 +254,8 @@ igt_main
> >  
> >  	igt_fixture {
> >  		data.drm_fd = drm_open_driver_master(DRIVER_ANY);
> > +		data.devid = is_i915_device(data.drm_fd) ?
> > +			intel_get_drm_devid(data.drm_fd) : 0;
> >  
> >  		kmstest_set_vt_graphics_mode();
> 
>