Re: [PATCH i-g-t v3 07/11] tests/intel/kms_fbcon_fbt: update the outdated fbc skip reasons

["Govindapillai, Vinod" <vinod.govindapillai@intel.com>]

On Thu, 2026-04-30 at 09:26 +0530, Reddy Guddati, Santhosh wrote:
> Hi Vinod,
> 
> On 16-04-2026 14:55, Vinod Govindapillai wrote:
> > No fbc reasons have been updated in the driver. Use the same
> > "no fbc reason" being set in the debugfs by the driver.
> > 
> > Signed-off-by: Vinod Govindapillai <vinod.govindapillai@intel.com>
> > ---
> >   tests/intel/kms_fbcon_fbt.c | 22 ++++++++++------------
> >   1 file changed, 10 insertions(+), 12 deletions(-)
> > 
> > diff --git a/tests/intel/kms_fbcon_fbt.c
> > b/tests/intel/kms_fbcon_fbt.c
> > index b36582011..78c6cc4b2 100644
> > --- a/tests/intel/kms_fbcon_fbt.c
> > +++ b/tests/intel/kms_fbcon_fbt.c
> > @@ -297,18 +297,16 @@ static inline void psr_debugfs_enable(int
> > device, int debugfs_fd)
> >   static void fbc_skips_on_fbcon(int debugfs_fd)
> >   {
> >   	const char *reasons[] = {
> > -		"incompatible mode",
> > -		"mode too large for compression",
> > -		"framebuffer not tiled or fenced",
> > -		"pixel format is invalid",
> > -		"rotation unsupported",
> > -		"tiling unsupported",
> > -		"framebuffer stride not supported",
> > -		"per-pixel alpha blending is incompatible with
> > FBC",
> > -		"pixel rate is too big",
> > -		"CFB requirements changed",
> > -		"plane Y offset is misaligned",
> > -		"plane height + offset is non-modulo of 4"
> > +		"pixel format not supported",
> > +		"tiling not supported",
> > +		"rotation not supported",
> > +		"stride not supported",
> > +		"per-pixel alpha not supported",
> > +		"plane size too big",
> > +		"surface size too big",
> > +		"plane start Y offset misaligned",
> > +		"plane end Y offset misaligned",
> > +		"pixel rate too high"
> 
> There are other no fbc reasons being set by the driver for ex:
> VGPU active, plane not visible etc. Shouldn't we consider them as
> well 
> here along with this?
> 
> Thanks,
> Santhosh

Basically my thinking was that these reasons defined in the current IGT
have been uses as skip reasons. Current tests skips based on these and
otherwise treat it as a failure. So in my opinion in case if we have to
treat any other skip reason as a failure, it is better to analyze first
if the reason for the failure is genuine and add it to the list. For
example fbc not enabled because plane not visible is might be an issue
with driver or a specific test sequence etc which we should catch as a
failure and fix.

BR
Vinod