[X86] Make CMPXCHG16B feature imply CMPXCHG8B feature.

This fixes znver1 so that it properly enables CMPXHG8B. We can
probably remove explicit CMPXCHG8B from CPUs that also have
CMPXCHG16B, but keeping this simple to allow cherry pick to 9.0.

Fixes PR42935.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@368324 91177308-0d34-0410-b5e6-96231b3b80d8
This commit is contained in:
Craig Topper
2019-08-08 18:11:17 +00:00
parent 1d2260b9e7
commit 3d3e1b8fee
2 changed files with 3 additions and 1 deletions
+2 -1
View File
@@ -95,7 +95,8 @@ def Feature3DNowA : SubtargetFeature<"3dnowa", "X863DNowLevel", "ThreeDNowA",
def Feature64Bit : SubtargetFeature<"64bit", "HasX86_64", "true",
"Support 64-bit instructions">;
def FeatureCMPXCHG16B : SubtargetFeature<"cx16", "HasCmpxchg16b", "true",
"64-bit with cmpxchg16b">;
"64-bit with cmpxchg16b",
[FeatureCMPXCHG8B]>;
def FeatureSlowSHLD : SubtargetFeature<"slow-shld", "IsSHLDSlow", "true",
"SHLD instruction is slow">;
def FeatureSlowPMULLD : SubtargetFeature<"slow-pmulld", "IsPMULLDSlow", "true",
+1
View File
@@ -2,6 +2,7 @@
; RUN: llc < %s -mtriple=i686-unknown- -mcpu=core2 | FileCheck %s --check-prefixes=CHECK,X86
; RUN: llc < %s -mtriple=x86_64-unknown- -mcpu=core2 | FileCheck %s --check-prefixes=CHECK,X64
; RUN: llc < %s -mtriple=i686-unknown- -mcpu=i486 | FileCheck %s --check-prefixes=I486
; RUN: llc < %s -mtriple=i686-unknown- -mcpu=znver1 | FileCheck %s --check-prefixes=CHECK,X86
; Basic 64-bit cmpxchg
define void @t1(i64* nocapture %p) nounwind ssp {