On Thu, Jul 6, 2017 at 5:07 PM, Connor Abbott <cwabbo...@gmail.com> wrote: > FYI, I already have another series which adds ARB_shader_ballot and > ARB_shader_group_vote intrinsics, in addition to adding some more > precise semantics to represent the restrictions on ballotARB() and > similar things [0]. The problem is that marking ballot as > can_eliminate but not can_reorder is overly conservative, since we > won't try to CSE it even in the same basic block. I think Jason had > some concern he was thinking about, so I'll ask him again. > > [0] https://lists.freedesktop.org/archives/mesa-dev/2017-June/158672.html
It seems like your series has been stalled for nearly a month on questions about the semantics of cross-lane attributes. That seems worthwhile, but I'd prefer not to block my series on sorting that out. As far as I can tell, the overlap seems to be almost entirely within your [PATCH 04/17], which itself contains a dependence on cross-lane attributes. Given that my series has minimal overlap with yours, I'd prefer (given review) to commit mine and do the trivial rebase of your 04/17 on top of that, once the cross-lane attributes are settled. _______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/mesa-dev
