[PATCH] D52813: [WebAssembly] Saturating arithmetic intrinsics

[Thomas Lively via Phabricator via llvm-commits]

tlively added inline comments.


================
Comment at: test/CodeGen/WebAssembly/simd-arith.ll:6
+; RUN: llc < %s -asm-verbose=false -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=-simd128 | FileCheck %s --check-prefixes CHECK,NO-SIMD128,NO-SIMD128-SLOW
+; RUN: llc < %s -asm-verbose=false -disable-wasm-fallthrough-return-opt -wasm-disable-explicit-locals -wasm-keep-registers -mattr=-simd128 -fast-isel | FileCheck %s --check-prefixes CHECK,NO-SIMD128,NO-SIMD128-FAST
 
----------------
aheejin wrote:
> aheejin wrote:
> > Do we need both the slow and fast versions for `NO-SIMD128` too? Aren't they the same anyway?
> Oh what I meant is... do we need to test both with and without fastisel for `NO-SIMD128`? With `NO-SIMD128` we are just testing the pre-existing scalar expansion behavior and checking the SIMD instructions do not kick in.
Sounds like fast isel is responsible for complying with architecture features itself, so we still need this test.


Repository:
  rL LLVM

https://reviews.llvm.org/D52813