[igt-dev] [PATCH i-g-t 1/2] xe_pm: Kill display for PM tests

[Zbigniew Kempczyński]

On Tue, Mar 28, 2023 at 02:40:43PM +0200, Maarten Lankhorst wrote:
> Display keeps PM alive, do the minimal initialization to disable
> display for the PM tests entirely.

What's the reason of enforcing releasing display before starting PM
tests? Looks like introducing testing gap.

--
Zbigniew

> 
> Signed-off-by: Maarten Lankhorst <maarten.lankhorst at linux.intel.com>
> ---
>  tests/xe/xe_pm.c | 13 +++++++++++++
>  1 file changed, 13 insertions(+)
> 
> diff --git a/tests/xe/xe_pm.c b/tests/xe/xe_pm.c
> index 23b8246ed..861f6b915 100644
> --- a/tests/xe/xe_pm.c
> +++ b/tests/xe/xe_pm.c
> @@ -365,6 +365,8 @@ igt_main
>  	};
>  
>  	igt_fixture {
> +		drmModeResPtr res;
> +
>  		memset(&device, 0, sizeof(device));
>  		device.fd_xe = drm_open_driver(DRIVER_XE);
>  		device.pci_xe = igt_device_get_pci_device(device.fd_xe);
> @@ -376,6 +378,17 @@ igt_main
>  		xe_for_each_hw_engine(device.fd_xe, hwe)
>  			test_exec(device, hwe, 1, 1, NO_SUSPEND, NO_RPM);
>  
> +		/* Kill display, if it exists */
> +		res = drmModeGetResources(device.fd_xe);
> +		if (res) {
> +			igt_display_t display;
> +
> +			drmModeFreeResources(res);
> +			igt_display_require(&display, device.fd_xe);
> +			igt_display_commit2(&display, COMMIT_ATOMIC);
> +			igt_display_fini(&display);
> +		}
> +
>  		get_d3cold_allowed(device.pci_xe, d3cold_allowed);
>  		igt_assert(igt_setup_runtime_pm(device.fd_xe));
>  	}
> -- 
> 2.34.1
>