[PATCH] D49262: [DAGCombiner] Call SimplifyDemandedVectorElts from EXTRACT_VECTOR_ELT

[Eli Friedman via Phabricator via llvm-commits]

efriedma added a comment.

Makes sense.



================
Comment at: test/CodeGen/AArch64/aarch64-be-bv.ll:33
-  ; CHECK-NEXT:    movi	   v[[REG2:[0-9]+]].4s, #1, lsl #16
-  ; CHECK-NEXT:    add	   v[[REG1]].8h, v[[REG1]].8h, v[[REG2]].8h
   ; CHECK-NEXT:    umov	   w{{[0-9]+}}, v[[REG1]].h[0]
----------------
This is nice, but it's destroying the intent of the test, which is to check that we generate the correct movi instruction.


================
Comment at: test/CodeGen/ARM/func-argpassing-endian.ll:3
+; RUN: llc -verify-machineinstrs < %s -mtriple=arm-eabi -mattr=v7,neon | FileCheck %s --check-prefixes=CHECK,CHECK-LE
+; RUN: llc -verify-machineinstrs < %s -mtriple=armeb-eabi -mattr=v7,neon | FileCheck %s --check-prefixes=CHECK,CHECK-BE
 
----------------
Regenerating this file LGTM; please commit separately.


Repository:
  rL LLVM

https://reviews.llvm.org/D49262