Removing the BROKEN condition on vc04_services/vchiq?

[Michael Zoran]

On Thu, 2016-10-27 at 09:00 -0700, Eric Anholt wrote:
> Greg KH <gregkh at linuxfoundation.org> writes:
> 
> > On Wed, Oct 26, 2016 at 10:22:37PM -0700, Michael Zoran wrote:
> > > Also, I'm wondering when the default config for bcm2835 should be
> > > added
> > > to arm64 and possibly any 64 bit specific device trees should be
> > > added/changed.
> > > 
> > > Copying the bcm2835_defconfig from arch/arm/configs to
> > > arch/arm64/configs works with almost no modifications.
> > 
> > I'll let the arm maintainers deal with that, it's outside of the
> > staging
> > tree area for me to be messing with.
> 
> (That'll be me)
> 
> We could add VCHI to multi_v7_defconfig and defconfig immediately,
> but
> we should wait until we've got the driver fully working and not
> emitting
> any warnings before adding it to arch/arm64/configs/defconfig.

Thinking about this more, I think it's too soon for that. VCHIQ still
has a bit to go especially for 64 bit.

I was able to get vchiq_test recompiled for arm64 the other day and am
now able to test vchiq better in 64 bit then I was ever able to do. 
I'm seeing some lingering issues even through all the errors are gone
and so are most of the warnings.

Although I've noticed that the userland libraries and test tool both
have a lot of pointer dependent issues with them, so it's a bit hard to
tell at this point what are driver issues and what are test tool
issues.