Nehalem unaligned memory access is fast.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@100089 91177308-0d34-0410-b5e6-96231b3b80d8

3 files changed, 15 insertions, 2 deletions

diff --git a/lib/Target/X86/X86.td b/lib/Target/X86/X86.td
index b0847a8..6b62795 100644
--- a/lib/Target/X86/X86.td
+++ b/lib/Target/X86/X86.td
@@ -59,6 +59,9 @@ def Feature64Bit   : SubtargetFeature<"64bit", "HasX86_64", "true",
                                       [FeatureCMOV]>;
 def FeatureSlowBTMem : SubtargetFeature<"slow-bt-mem", "IsBTMemSlow", "true",
                                        "Bit testing of memory is slow">;
+def FeatureFastUAMem : SubtargetFeature<"fast-unaligned-mem",
+                                        "IsUAMemFast", "true",
+                                        "Fast unaligned memory access">;
 def FeatureSSE4A   : SubtargetFeature<"sse4a", "HasSSE4A", "true",
                                       "Support SSE 4a instructions">;
 
@@ -98,8 +101,10 @@ def : Proc<"nocona",          [FeatureSSE3,   Feature64Bit, FeatureSlowBTMem]>;
 def : Proc<"core2",           [FeatureSSSE3,  Feature64Bit, FeatureSlowBTMem]>;
 def : Proc<"penryn",          [FeatureSSE41,  Feature64Bit, FeatureSlowBTMem]>;
 def : Proc<"atom",            [FeatureSSE3,   Feature64Bit, FeatureSlowBTMem]>;
-def : Proc<"corei7",          [FeatureSSE42,  Feature64Bit, FeatureSlowBTMem]>;
-def : Proc<"nehalem",         [FeatureSSE42,  Feature64Bit, FeatureSlowBTMem]>;
+def : Proc<"corei7",          [FeatureSSE42,  Feature64Bit, FeatureSlowBTMem,
+                               FeatureFastUAMem]>;
+def : Proc<"nehalem",         [FeatureSSE42,  Feature64Bit, FeatureSlowBTMem,
+                               FeatureFastUAMem]>;
 // Sandy Bridge does not have FMA
 def : Proc<"sandybridge",     [FeatureSSE42,  FeatureAVX,   Feature64Bit]>;
 
diff --git a/lib/Target/X86/X86Subtarget.cpp b/lib/Target/X86/X86Subtarget.cpp
index cd56816..8a0cde4 100644
--- a/lib/Target/X86/X86Subtarget.cpp
+++ b/lib/Target/X86/X86Subtarget.cpp
@@ -266,6 +266,9 @@ void X86Subtarget::AutoDetectSubtargetFeatures() {
     unsigned Model  = 0;
     DetectFamilyModel(EAX, Family, Model);
     IsBTMemSlow = IsAMD || (Family == 6 && Model >= 13);
+    // If it's Nehalem, unaligned memory access is fast.
+    if (Family == 15 && Model == 26)
+      IsUAMemFast = true;
 
     GetCpuIDAndInfo(0x80000001, &EAX, &EBX, &ECX, &EDX);
     HasX86_64 = (EDX >> 29) & 0x1;
@@ -286,6 +289,7 @@ X86Subtarget::X86Subtarget(const std::string &TT, const std::string &FS,
   , HasFMA3(false)
   , HasFMA4(false)
   , IsBTMemSlow(false)
+  , IsUAMemFast(false)
   , HasVectorUAMem(false)
   , DarwinVers(0)
   , stackAlignment(8)
diff --git a/lib/Target/X86/X86Subtarget.h b/lib/Target/X86/X86Subtarget.h
index 56220db..bf30154 100644
--- a/lib/Target/X86/X86Subtarget.h
+++ b/lib/Target/X86/X86Subtarget.h
@@ -78,6 +78,9 @@ protected:
   /// IsBTMemSlow - True if BT (bit test) of memory instructions are slow.
   bool IsBTMemSlow;
 
+  /// IsUAMemFast - True if unaligned memory access is fast.
+  bool IsUAMemFast;
+
   /// HasVectorUAMem - True if SIMD operations can have unaligned memory
   ///                  operands. This may require setting a feature bit in the
   ///                  processor.
@@ -148,6 +151,7 @@ public:
   bool hasFMA3() const { return HasFMA3; }
   bool hasFMA4() const { return HasFMA4; }
   bool isBTMemSlow() const { return IsBTMemSlow; }
+  bool isUnalignedMemAccessFast() const { return IsUAMemFast; }
   bool hasVectorUAMem() const { return HasVectorUAMem; }
 
   bool isTargetDarwin() const { return TargetType == isDarwin; }