> From: Ville Syrjälä <ville.syrj...@linux.intel.com> > On Wed, Dec 15, 2021 at 09:05:03AM +0000, Sarvela, Tomi P wrote: > > > From: Ville Syrjälä <ville.syrj...@linux.intel.com> > > > > > > On Tue, Dec 14, 2021 at 06:25:43PM +0200, Ville Syrjälä wrote: > > > > On Mon, Dec 13, 2021 at 09:54:04PM +0200, Jani Nikula wrote: > > > > > On Mon, 13 Dec 2021, Ville Syrjala <ville.syrj...@linux.intel.com> > wrote: > > > > > > > > > > This one is only used in gvt, anyway. And that actually makes me > wonder > > > > > if this should be breaking the build. Does CI not have gvt enabled? > > > > > > > > Hmm. I thought it was enabled in CI, but maybe not. I've often broken > > > > gvt with register define changes but I've always caught it before > > > > pushing. I think I have gvt enabled in my "make sure all commits build > > > > before I push" test config, so maybe that's where I caught most of > them. > > > > > > > > Tomi, can we enable gvt in ci builds to make sure it at least still > > > > builds? > > > > > > Actually cc Tomi.. > > > > GVT-d is enabled and tested by fi-bdw-gvtdvm. > > We're talking about the other gvt (whatever it was called), ie. > CONFIG_DRM_I915_GVT.
This kconfig entry doesn't exist in default CI kconfig, even as 'is not set' placeholder: https://gitlab.freedesktop.org/gfx-ci/i915-infra/-/blob/master/kconfig/debug If the config entry is exact, I'll probably need to upgrade the default config from 5.13 and add it with requirements. Not today, but maybe soon. Tomi
