[PATCH] D65961: AMDGPU/SILoadStoreOptimizer: Optimize scanning for mergeable instructions

Mon Dec 2 05:34:06 PST 2019

tstellar added a comment.

In D65961#1763143 <https://reviews.llvm.org/D65961#1763143>, @foad wrote:

> Hi @tstellar, I'm looking into a case where this patch slowed down a shader by 10%. Before I go too far, was this patch supposed to change the behaviour at all, or was it supposed to be purely a compile time improvement?

The intention was to not change the behavior at all.

> In the case I'm looking at it seems to do the same amount of load merging as before, but the merged loads are inserted at different places in the basic block.

Do you have a MIR or .ll dump of the shader I could look at ?  Also, does https://reviews.llvm.org/D65966 help?

Repository:
  rL LLVM

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D65961/new/

https://reviews.llvm.org/D65961