On Mon, 21 May 2012 14:01:52 -0700, Ben Widawsky <ben at bwidawsk.net> wrote: > On Mon, 21 May 2012 09:46:00 +0100 > Chris Wilson <chris at chris-wilson.co.uk> wrote: > > > On Fri, 18 May 2012 09:02:09 -0700, Ben Widawsky <ben at bwidawsk.net> wrote: > > > here are some opportunities to do things a bit better with the naming of > > > things now that our hindsight is better. > > > > The major problem with this is that it breaks the expectations of the > > INTEL_ namespace, masquerading per-gen values as fixed constants for all > > chipsets. > > -Chris > > > > With only a couple of exceptions (BSD user interrupt) there is no need > for per-gen values. I think we shouldn't have much trouble maintaining > this as is. The existing naming scheme wrongly calls things GEN6 when > they were introduced earlier (I'm also not sure GT_ is valid for every > interrupt). So the goal was to have I915_* for old stuff, INTEL_* for > everything, and DEVICENAME_ for special cases. I concur that the current names are rather ad-hoc and not well thought out. I'm just not buying that this replacement is any better. INTEL_BSD is a complete turn off. -Chris -- Chris Wilson, Intel Open Source Technology Centre
