Re: [PATCH i-g-t v2 01/23] tests/intel/kms_psr: Don't pass uninitialized 'pipe' to intel_fbc_supported_on_chipset()

["Ville Syrjälä" <ville.syrjala@linux.intel.com>]

On Mon, Feb 23, 2026 at 01:22:08PM +0200, Jani Nikula wrote:
> On Sat, 21 Feb 2026, Ville Syrjala <ville.syrjala@linux.intel.com> wrote:
> > From: Ville Syrjälä <ville.syrjala@linux.intel.com>
> >
> > Instead passing stack garbage in 'pipe' to
> > intel_fbc_supported_on_chipset() iterate over all the CRTCs
> > and properly check if any of them supports FBC.
> >
> > Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
> > ---
> >  tests/intel/kms_psr.c | 11 ++++++++---
> >  1 file changed, 8 insertions(+), 3 deletions(-)
> >
> > diff --git a/tests/intel/kms_psr.c b/tests/intel/kms_psr.c
> > index c411981c0104..c730a6f23944 100644
> > --- a/tests/intel/kms_psr.c
> > +++ b/tests/intel/kms_psr.c
> > @@ -774,7 +774,6 @@ int igt_main()
> >  {
> >  	int z, y;
> >  	enum operations op;
> > -	enum pipe pipe;
> >  	const char *append_subtest_name[3] = {
> >  		"psr-",
> >  		"psr2-",
> > @@ -787,10 +786,12 @@ int igt_main()
> >  	int modes[] = {PSR_MODE_1, PSR_MODE_2, PR_MODE};
> >  	int fbc_status[] = {FBC_DISABLED, FBC_ENABLED};
> >  	igt_output_t *output;
> > -	bool fbc_chipset_support;
> > +	bool fbc_chipset_support = true;
> >  	int disp_ver;
> >  
> >  	igt_fixture() {
> > +		igt_crtc_t *crtc;
> > +
> >  		data.drm_fd = drm_open_driver_master(DRIVER_INTEL | DRIVER_XE);
> >  		data.debugfs_fd = igt_debugfs_dir(data.drm_fd);
> >  		kmstest_set_vt_graphics_mode();
> > @@ -799,7 +800,11 @@ int igt_main()
> >  		igt_display_require(&data.display, data.drm_fd);
> >  		igt_require_f(output_supports_psr(&data), "Sink does not support PSR/PSR2/PR\n");
> >  		disp_ver = intel_display_ver(data.devid);
> > -		fbc_chipset_support = intel_fbc_supported_on_chipset(data.drm_fd, pipe);
> > +
> > +		for_each_crtc(&data.display, crtc) {
> > +			if (!intel_fbc_supported_on_chipset(data.drm_fd, crtc->pipe))
> > +				fbc_chipset_support = false;
> 
> The commit message says "if any of them", but this requires all of them
> to support FBC, which is unlikely.

Hmm, yeah that is definitely a problem for hsw+. I guess I'll just
reverse the logic here.

intel_fbc_supported_on_chipset() should probably itself iterate all
the CRTCs, but with the current low level API that'd be a bit
inconvenient to implement. We should probably look into changing 
the intel_fbc*() API to take igt_display_t/igt_crtc_t/etc. as well...

-- 
Ville Syrjälä
Intel