Implement a feature (-vector-unaligned-mem) to allow targets to
ignore alignment requirements for SIMD memory operands. This
is useful on architectures like the AMD 10h that do not trap on
unaligned references if a status bit is twiddled at startup time.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@93151 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/Target/X86/X86.td b/lib/Target/X86/X86.td
index a6e1ca3..7919559 100644
--- a/lib/Target/X86/X86.td
+++ b/lib/Target/X86/X86.td
@@ -23,6 +23,7 @@
def FeatureCMOV : SubtargetFeature<"cmov","HasCMov", "true",
"Enable conditional move instructions">;
+
def FeatureMMX : SubtargetFeature<"mmx","X86SSELevel", "MMX",
"Enable MMX instructions">;
def FeatureSSE1 : SubtargetFeature<"sse", "X86SSELevel", "SSE1",
@@ -66,6 +67,9 @@
"Enable three-operand fused multiple-add">;
def FeatureFMA4 : SubtargetFeature<"fma4", "HasFMA4", "true",
"Enable four-operand fused multiple-add">;
+def FeatureVectorUAMem : SubtargetFeature<"vector-unaligned-mem",
+ "HasVectorUAMem", "true",
+ "Allow unaligned memory operands on vector/SIMD instructions">;
//===----------------------------------------------------------------------===//
// X86 processors supported.