Add cases to the memref.atomicrmw lowering for floating-point min and max, since LLVM supports these. Reviewed By: bondhugula Differential Revision: https://reviews.llvm.org/D158283
Add cases to the memref.atomicrmw lowering for floating-point min and max, since LLVM supports these. Reviewed By: bondhugula Differential Revision: https://reviews.llvm.org/D158283