[PATCH] Fix PR19657 : SLP vectorization doesn't combine scalar load to vector loads
echristo at gmail.com
Fri May 16 18:21:21 PDT 2014
Any reason we don't want to check opt level when running to see how
aggressive we should be and how much we should spend compile time? Is
there some other approach you'd like to get optimizations like this?
On Fri, May 16, 2014 at 8:01 AM, Nadav Rotem <nrotem at apple.com> wrote:
> I am sorry but this is the wrong approach. You copy a ton of code that we will need to maintain if we want to add support to new instructions.
> You also increase the compile time by scanning the tree twice. In the SLP vectorizer there are a bunch of features we did not implement because we did not want to increase compile time. Remember that we run this pass on all optimization levels.
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
More information about the llvm-commits