Re: [PATCH i-g-t] tests/kms_content_protection: Skip MST outputs in SST-style tests

["Reddy Guddati, Santhosh" <santhosh.reddy.guddati@intel.com>]

Hi Jason,

On 29-06-2026 07:23, Jason-JH Lin wrote:
> MST (Multi-Stream Transport) outputs require topology-wide HDCP
> authentication which is handled by dedicated dp-mst-* subtests.
> Running SST-style tests on MST outputs can cause issues because:
> 
> 1. MST HDCP authentication involves the entire topology (hub + monitors)
> 2. Per-connector HDCP state management differs from SST
> 3. The dp-mst-lic-type-0/1 subtests are specifically designed to test
>     MST HDCP scenarios
> 
> Skip MST outputs in get_hdcp_outputs() when running non-MST subtests
> to ensure proper test isolation.

IMHO, I dont think we should skip all MST outputs in
the SST-style tests - this removes a lot of coverage. NACK for now

You will see a lot of skips after this any mst setup will not run 
individual test on those monitors. We want to test individual displays 
even though they are connected in MST.

The tests (legacy, type1, lic, uevent, srm, mei-interface) will skip 
even if a single monitor is connected through a MST hub thereby a single 
MST display would then get no HDCP coverage at all.

Please let me know your thoughts.

Thanks,
Santhosh

> 
> Signed-off-by: Jason-JH Lin <jason-jh.lin@mediatek.com>
> ---
>   tests/kms_content_protection.c | 5 +++++
>   1 file changed, 5 insertions(+)
> 
> diff --git a/tests/kms_content_protection.c b/tests/kms_content_protection.c
> index d931ce7682be..9d1728ed82a6 100644
> --- a/tests/kms_content_protection.c
> +++ b/tests/kms_content_protection.c
> @@ -751,6 +751,11 @@ static int get_hdcp_outputs(igt_display_t *display, int content_type,
>   			continue;
>   		}
>   
> +		if (igt_get_dp_mst_connector_id(output) >= 0) {
> +			igt_info("Skipping HDCP test on %s, MST output should use dp-mst-* tests\n",
> +				 output->name);
> +			continue;
> +		}
>   

>   		if (count < MAX_HDCP_OUTPUTS)
>   			hdcp_outputs[count++] = output;