[PATCH] D61764: [LV] Suppress vectorization in some nontemporal cases
Warren Ristow via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Tue May 28 16:49:57 PDT 2019
wristow marked an inline comment as done.
wristow added inline comments.
Comment at: lib/Target/X86/X86TargetTransformInfo.cpp:3080
+ if (DataType->isFloatTy() || DataType->isDoubleTy())
+ return ST->hasSSE4A();
> RKSimon wrote:
> > SSE4A nt-stores can happen with any alignment, and AFAICT without any perf penalty.
> I didn't realize that. I'll update the patch, and include a test for it.
Looking into this, I'm confused... Are you saying (for example) that with SSE4A, `vmovntps` can do an nt-store with a misaligned address? Looking through the docs, I'm seeing a requirement for the address to be aligned.
Or are you saying (for example) the SSE4A instruction `movntss` (which takes a vector-register operand containing the value to be stored) can take an arbitrary alignment for the memory address? If that's what your point is, then yes I should change the above to allow misaligned `float` and `double` nt-stores. But `movntss` is only storing one `float` element of the vector register (ignoring the other elems), and so it doesn't allow us to vectorize the case. In short, yes I should change that for `float` and `double` nt-stores on SSE4A, but since it doesn't allow us to vectorize, I wonder if I'm misunderstanding your point.
Or are you saying something else? (Like I said, I'm confused.)
CHANGES SINCE LAST ACTION
More information about the llvm-commits