[PATCH] D16836: [CodeGenPrepare] Remove load-based heuristic

[Junmo Park via llvm-commits]

flyingforyou added a comment.

Hi. Mitch.

  I'm looking into resurrecting the nonlocal dead store elimination in http://reviews.llvm.org/D13363,
  but one of the snags is a performance issue related to select-vs-branch choices. That change enables
  more selects, which unfortunately incur a 10 to 20% perf drop on some benchmarks in the EEMBC automotive
  suite, on various X86 architectures.

This patch will make select to branch lesser. But I uploaded other patch(http://reviews.llvm.org/D17288) for applying select to branch optimization when obvious case.

If you see some regressions regarding this patch(http://reviews.llvm.org/D16836), could you let us know, please?

Junmo.


http://reviews.llvm.org/D16836