[PATCH] D20494: [CUDA] Add section to docs about controlling fp optimizations.

Justin Lebar via llvm-commits llvm-commits at lists.llvm.org
Wed May 25 16:07:12 PDT 2016


jlebar added inline comments.

================
Comment at: docs/CompileCudaWithLLVM.rst:166-168
@@ +165,5 @@
+    and add instructions.
+  * ``on``: fuse multiplies and adds within a single statement, but never
+    across statements (C11 semantics).  Prevent ptxas from fusing other
+    multiplies and adds.
+  * ``fast``: fuse multiplies and adds wherever profitable, even across
----------------
rnk wrote:
> Do we actually know how to honor the C11 semantics?
My understanding from talking to Chandler was that this was the intent of "on".  But there is a good chance I misunderstood.


http://reviews.llvm.org/D20494





More information about the llvm-commits mailing list