[Intel-gfx] [PATCH 1/2] drm/i915: Slice/Subslice/EU info via GETPARAM

[Daniel Vetter]

On Wed, Aug 06, 2014 at 09:43:44AM -0500, Jeff McGee wrote:
> On Tue, Aug 05, 2014 at 04:50:16PM +0200, Daniel Vetter wrote:
> > On Tue, Aug 5, 2014 at 4:03 PM, Jeff McGee <jeff.mcgee at intel.com> wrote:
> > >> Also, usual broken record request: I need open-source userspace using
> > >> this (mesa, ddx, libva).
> > >> -Daniel
> > >>
> > >
> > > This is kind of chicken-and-egg problem that I haven't been through. I
> > > assume that we build new interfaces up the stack (kernel->libdrm->usermode).
> > > Any tips or docs on how to proceed?
> > 
> > You need to have the full thing ready with kernel, igt, libdrm, and
> > umd driver patches (and maybe even testcases for the new umd feature,
> > e.g. for some of the arb robustness stuff we've done). Then you post
> > them all to relevant mailing lists and go through the review process
> > for all parts. Once that's done and _all_ parts are reviewed, merging
> > starts with kernel+igt, then libdrm + libdrm release, then umd.
> > -Daniel
> 
> Thanks Daniel. Is it acceptable to get the kernel part merged with only
> an igt test (technically an open-source userspace component)? I think
> probably not, but have to ask.

Not really. The entire idea behind this is that only once we have the full
stack picture available we can properly assess whether the interface
fullfills the goal. That requires an actual implementation of the feature
for the entire stack.

And Dave Airlie (as the drm upstream maintainer) requires that that entire
pile is open-source.

So we really don't have wiggle room here. And personally I don't want to
open the opportunity for discussions for each specific feature on whether
a bare-bones demonstration test in igt is good enough or not, it really
should be something shippable to users.
-Daniel
-- 
Daniel Vetter
Software Engineer, Intel Corporation
+41 (0) 79 365 57 48 - http://blog.ffwll.ch