Displaying 2 results from an estimated 2 matches for "gurardwidening".
2018 May 16
2
RFC: separating guards and implicit control flow
...earlier guard to be "widened" (i.e. made to fail
more often) if doing so allows the elimination of a later guard or
otherwise is useful to prune possible code paths. At this point, the
basic notion of a guard is fairly well proven with both critical
transformations (LoopPredication, GurardWidening) available upstream,
and multiple downstream speculative optimizations built on top.
As we've explored the design, we've stumbled on a implementation
challenge. Today, guards are modeled as call, not invokes. As it turns
out, LLVM is not terribly good at dealing with potential implici...
2018 May 17
0
RFC: separating guards and implicit control flow
...o be "widened" (i.e. made to fail
> more often) if doing so allows the elimination of a later guard or
> otherwise is useful to prune possible code paths. At this point, the
> basic notion of a guard is fairly well proven with both critical
> transformations (LoopPredication, GurardWidening) available upstream,
> and multiple downstream speculative optimizations built on top.
>
> As we've explored the design, we've stumbled on a implementation
> challenge. Today, guards are modeled as call, not invokes. As it
> turns out, LLVM is not terribly good at dealing w...