[igt-dev] [PATCH i-g-t 2/2] tests/kms_hdr: Add negative test to enforce HDR on non-HDR panel

[Kamil Konieczny]

Hi Swati,

I have few nits regrading code style or message,
I will not review KMS parts, for that you already got
review from Uma.

On 2022-12-27 at 14:07:41 +0530, Swati Sharma wrote:
> Negative test is added to ensure HDR is not enabled on non-HDR panel.
> Commit/modeset won't fail. However, in dmesg we can find

Put newline here.

> 	[drm:drm_hdmi_infoframe_set_hdr_metadata [drm_display_helper]] EOTF Not Supported
- ^^^^^^
Remove these spaces or tab.

> 
> Signed-off-by: Swati Sharma <swati2.sharma at intel.com>
> ---
>  tests/kms_hdr.c | 15 +++++++++++++--
>  1 file changed, 13 insertions(+), 2 deletions(-)
> 
> diff --git a/tests/kms_hdr.c b/tests/kms_hdr.c
> index 629a45a7..18ccfce7 100644
> --- a/tests/kms_hdr.c
> +++ b/tests/kms_hdr.c
> @@ -51,6 +51,7 @@ enum {
>  	TEST_SUSPEND = 1 << 2,
>  	TEST_SWAP = 1 << 3,
>  	TEST_INVALID_METADATA_SIZES = 1 << 4,
> +	TEST_INVALID_HDR = 1 << 5,
>  };
>  
>  /* BPC connector state. */
> @@ -414,6 +415,8 @@ static void test_static_toggle(data_t *data, enum pipe pipe,
>  	set_hdr_output_metadata(data, &hdr);
>  	igt_output_set_prop_value(data->output, IGT_CONNECTOR_MAX_BPC, 10);
>  	igt_display_commit_atomic(display, DRM_MODE_ATOMIC_ALLOW_MODESET, NULL);
> +	if (flags & TEST_INVALID_HDR)
> +		igt_assert_eq(system("dmesg|grep -E \"EOTF Not Supported\""), 0);

These looks also strange, why do you not return or goto cleanup here ?
What I mean, you are testing invalid, so in case it was not set
you will follow to HDR test after if() assert();
imho this should look like:

	if (flags & TEST_INVALID_HDR) {
		igt_assert_eq(system("dmesg|grep -E \"EOTF Not Supported\""), 0);
		goto cleanup; /* or just return ??? */
	}

Second concern is that dmesg may contain some old data, so maybe
dmesg | tail -n 1000 | grep -E \"EOTF Not Supported\""
or maybe debugfs or maybe clean dmesg before ?

>  	igt_assert_output_bpc_equal(data->fd, pipe, output->name, 10);
>  
>  	/* Verify that the CRC are equal after DPMS or suspend. */
> @@ -565,7 +568,11 @@ static void test_hdr(data_t *data, uint32_t flags)
>  		if (!has_max_bpc(output) || !has_hdr(output))
>  			continue;
>  
> -		if (!is_panel_hdr(data, output))
> +		/* For negative test, panel should be non-hdr. */
> +		if ((flags & TEST_INVALID_HDR) && is_panel_hdr(data, output))
> +			continue;
> +
> +		if ((flags & ~TEST_INVALID_HDR) && !is_panel_hdr(data, output))
>  			continue;
>  
>  		if (igt_get_output_max_bpc(data->fd, output->name) < 10)
> @@ -587,7 +594,7 @@ static void test_hdr(data_t *data, uint32_t flags)
>  
>  				igt_dynamic_f("pipe-%s-%s",
>  					      kmstest_pipe_name(pipe), output->name) {
> -					if (flags & TEST_NONE || flags & TEST_DPMS || flags & TEST_SUSPEND)
> +					if (flags & TEST_NONE || flags & TEST_DPMS || flags & TEST_SUSPEND || flags & TEST_INVALID_HDR)

Looks strange, why not:
					if (flags & (TEST_NONE | TEST_DPMS | TEST_SUSPEND | TEST_INVALID_HDR))

Regards,
Kamil

>  						test_static_toggle(data, pipe, output, flags);
>  					if (flags & TEST_SWAP)
>  						test_static_swap(data, pipe, output);
> @@ -645,6 +652,10 @@ igt_main
>  	igt_subtest_with_dynamic("invalid-metadata-sizes")
>  		test_hdr(&data, TEST_INVALID_METADATA_SIZES);
>  
> +	igt_describe("Test to ensure HDR is not enabled on non-HDR panel");
> +	igt_subtest_with_dynamic("invalid-hdr")
> +		test_hdr(&data, TEST_INVALID_HDR);
> +
>  	igt_fixture {
>  		igt_display_fini(&data.display);
>  		close(data.fd);
> -- 
> 2.17.1
>