New Iteration of subregister liveness patches

[Renato Golin]

On 5 December 2014 at 19:19, Tom Stellard <tom at stellard.net> wrote:
> I tested these patches a few months ago and there were some regressions
> on R600.  I'm not sure when I'll get a chance to test them again, but I
> am in favor of merging these patches as soon as possible, since sub-reg
> liveness is hidden behind a subtarget flag and it's easier to fully test
> once it is in ToT.

If this is not the default behaviour, and it can be turned on by just
adding a flag, then I don't see why not, after addressing Quentin's
comments. I agree it'll be much easier to test this once it lands.

cheers,
--renato