On Wed, 31 Oct 2018, Rodrigo Vivi <[email protected]> wrote: > On Wed, Oct 31, 2018 at 11:00:54AM +0200, Jani Nikula wrote: >> On Wed, 31 Oct 2018, Tvrtko Ursulin <[email protected]> wrote: >> > I saw some mention somewhere on IS_GEN_RANGE, which looked clearer than >> > IS_GEN(dev_priv, s, e). Presumably that did not go anywhere since now >> > the proposal is the above? I have to say I am not sure it reads >> > completely intuitive when seen near in code: >> > >> > IS_GEN(dev_priv, 9) >> > IS_GEN(dev_priv, 8, 9) >> > >> > Looks like a variable arg list and the difference in semantics does not >> > come through. As such I am leaning towards thinking it is too much churn >> > for unclear benefit. Or in other words I thought IS_GEN_RANGE was a >> > better direction. >> >> Okay, thanks for the feedback. I'm not locked into any resolution yet, >> apart from not churning anything until we have a better picture where >> we're going. > > I believe we have 2 orthogonal discussions here where they shouldn't block > each other. > > 1. The addition of DISPLAY_GEN checks to group platforms and prefer display > gen checks over platform codenames. By doing this all platform enabling work > for > next platforms gets easier and less bureaucratic. > > 2. consolidated IS_GEN macro vs GEN_RANGE vs leave the way it currently is.
IMO if we add some display gen macro, it better be aligned with whatever will be done with IS_GEN and friends from the start. BR, Jani. > >> >> BR, >> Jani. >> >> -- >> Jani Nikula, Intel Open Source Graphics Center -- Jani Nikula, Intel Open Source Graphics Center _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
