[X86] Prevent selecting masked aligned load instructions if the load should be non-temporal

Summary: The aligned load predicates don't  suppress themselves if the load is non-temporal the way the unaligned predicates do. For the most part this isn't a problem because the aligned predicates are mostly used for instructions that only load the the non-temporal loads have priority over those. The exception are masked loads.

Reviewers: RKSimon, zvi

Reviewed By: RKSimon

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D35712

llvm-svn: 309079

1 files changed, 6 insertions, 3 deletions

diff --git a/llvm/lib/Target/X86/X86InstrFragmentsSIMD.td b/llvm/lib/Target/X86/X86InstrFragmentsSIMD.td
index 8b5bbf24f6f..e7b2e6bf64c 100644
--- a/llvm/lib/Target/X86/X86InstrFragmentsSIMD.td
+++ b/llvm/lib/Target/X86/X86InstrFragmentsSIMD.td
@@ -698,17 +698,20 @@ def alignedstore512 : PatFrag<(ops node:$val, node:$ptr),
 
 // Like 'load', but always requires 128-bit vector alignment.
 def alignedload : PatFrag<(ops node:$ptr), (load node:$ptr), [{
-  return cast<LoadSDNode>(N)->getAlignment() >= 16;
+  return cast<LoadSDNode>(N)->getAlignment() >= 16 &&
+         (!Subtarget->hasSSE41() || !cast<LoadSDNode>(N)->isNonTemporal());
 }]>;
 
 // Like 'load', but always requires 256-bit vector alignment.
 def alignedload256 : PatFrag<(ops node:$ptr), (load node:$ptr), [{
-  return cast<LoadSDNode>(N)->getAlignment() >= 32;
+  return cast<LoadSDNode>(N)->getAlignment() >= 32 &&
+         (!Subtarget->hasAVX2() || !cast<LoadSDNode>(N)->isNonTemporal());
 }]>;
 
 // Like 'load', but always requires 512-bit vector alignment.
 def alignedload512 : PatFrag<(ops node:$ptr), (load node:$ptr), [{
-  return cast<LoadSDNode>(N)->getAlignment() >= 64;
+  return cast<LoadSDNode>(N)->getAlignment() >= 64 &&
+         (!Subtarget->hasAVX512() || !cast<LoadSDNode>(N)->isNonTemporal());
 }]>;
 
 // 128-bit aligned load pattern fragments