4 files changed, 23 insertions, 9 deletions
diff --git a/lib/Target/ARM/ARM.td b/lib/Target/ARM/ARM.td
index 46915eecf6..68380847a0 100644
--- a/lib/Target/ARM/ARM.td
+++ b/lib/Target/ARM/ARM.td
@@ -143,14 +143,12 @@ include "ARMSchedule.td"
 // ARM processor families.
 def ProcA5      : SubtargetFeature<"a5", "ARMProcFamily", "CortexA5",
                                    "Cortex-A5 ARM processors",
-                                   [FeatureSlowFPBrcc, FeatureNEONForFP,
-                                    FeatureHasSlowFPVMLx, FeatureVMLxForwarding,
-                                    FeatureT2XtPk]>;
+                                   [FeatureSlowFPBrcc, FeatureHasSlowFPVMLx,
+                                    FeatureVMLxForwarding, FeatureT2XtPk]>;
 def ProcA8      : SubtargetFeature<"a8", "ARMProcFamily", "CortexA8",
                                    "Cortex-A8 ARM processors",
-                                   [FeatureSlowFPBrcc, FeatureNEONForFP,
-                                    FeatureHasSlowFPVMLx, FeatureVMLxForwarding,
-                                    FeatureT2XtPk]>;
+                                   [FeatureSlowFPBrcc, FeatureHasSlowFPVMLx,
+                                    FeatureVMLxForwarding, FeatureT2XtPk]>;
 def ProcA9      : SubtargetFeature<"a9", "ARMProcFamily", "CortexA9",
                                    "Cortex-A9 ARM processors",
                                    [FeatureVMLxForwarding,
diff --git a/lib/Target/ARM/ARMSubtarget.cpp b/lib/Target/ARM/ARMSubtarget.cpp
index e11314d4fc..739300e4ef 100644
--- a/lib/Target/ARM/ARMSubtarget.cpp
+++ b/lib/Target/ARM/ARMSubtarget.cpp
@@ -19,6 +19,7 @@
 #include "llvm/IR/Function.h"
 #include "llvm/Support/CommandLine.h"
 #include "llvm/Target/TargetInstrInfo.h"
+#include "llvm/Target/TargetOptions.h"
 
 #define GET_SUBTARGETINFO_TARGET_DESC
 #define GET_SUBTARGETINFO_CTOR
@@ -42,12 +43,13 @@ StrictAlign("arm-strict-align", cl::Hidden,
             cl::desc("Disallow all unaligned memory accesses"));
 
 ARMSubtarget::ARMSubtarget(const std::string &TT, const std::string &CPU,
-                           const std::string &FS)
+                           const std::string &FS, const TargetOptions &Options)
   : ARMGenSubtargetInfo(TT, CPU, FS)
   , ARMProcFamily(Others)
   , stackAlignment(4)
   , CPUString(CPU)
   , TargetTriple(TT)
+  , Options(Options)
   , TargetABI(ARM_ABI_APCS) {
   initializeEnvironment();
   resetSubtargetFeatures(CPU, FS);
@@ -92,6 +94,7 @@ void ARMSubtarget::initializeEnvironment() {
   AllowsUnalignedMem = false;
   Thumb2DSP = false;
   UseNaClTrap = false;
+  UnsafeFPMath = false;
 }
 
 void ARMSubtarget::resetSubtargetFeatures(const MachineFunction *MF) {
@@ -162,6 +165,12 @@ void ARMSubtarget::resetSubtargetFeatures(StringRef CPU, StringRef FS) {
   // configuration.
   if (!StrictAlign && hasV6Ops() && isTargetDarwin())
     AllowsUnalignedMem = true;
+
+  // NEON f32 ops are non-IEEE 754 compliant. Darwin is ok with it by default.
+  uint64_t Bits = getFeatureBits();
+  if ((Bits & ARM::ProcA5 || Bits & ARM::ProcA8) && // Where this matters
+      (Options.UnsafeFPMath || isTargetDarwin()))
+    UseNEONForSinglePrecisionFP = true;
 }
 
 /// GVIsIndirectSymbol - true if the GV will be accessed via an indirect symbol.
diff --git a/lib/Target/ARM/ARMSubtarget.h b/lib/Target/ARM/ARMSubtarget.h
index 8ce22e1de2..5b5ee6aeb8 100644
--- a/lib/Target/ARM/ARMSubtarget.h
+++ b/lib/Target/ARM/ARMSubtarget.h
@@ -26,6 +26,7 @@
 namespace llvm {
 class GlobalValue;
 class StringRef;
+class TargetOptions;
 
 class ARMSubtarget : public ARMGenSubtargetInfo {
 protected:
@@ -159,6 +160,9 @@ protected:
   /// NaCl TRAP instruction is generated instead of the regular TRAP.
   bool UseNaClTrap;
 
+  /// Target machine allowed unsafe FP math (such as use of NEON fp)
+  bool UnsafeFPMath;
+
   /// stackAlignment - The minimum alignment known to hold of the stack frame on
   /// entry to the function and which must be maintained by every function.
   unsigned stackAlignment;
@@ -175,6 +179,9 @@ protected:
   /// Selected instruction itineraries (one entry per itinerary class.)
   InstrItineraryData InstrItins;
 
+  /// Options passed via command line that could influence the target
+  const TargetOptions &Options;
+
  public:
   enum {
     isELF, isDarwin
@@ -189,7 +196,7 @@ protected:
   /// of the specified triple.
   ///
   ARMSubtarget(const std::string &TT, const std::string &CPU,
-               const std::string &FS);
+               const std::string &FS, const TargetOptions &Options);
 
   /// getMaxInlineSizeThreshold - Returns the maximum memset / memcpy size
   /// that still makes it profitable to inline the call.
diff --git a/lib/Target/ARM/ARMTargetMachine.cpp b/lib/Target/ARM/ARMTargetMachine.cpp
index 3003760df7..b0f9e56db7 100644
--- a/lib/Target/ARM/ARMTargetMachine.cpp
+++ b/lib/Target/ARM/ARMTargetMachine.cpp
@@ -48,7 +48,7 @@ ARMBaseTargetMachine::ARMBaseTargetMachine(const Target &T, StringRef TT,
                                            Reloc::Model RM, CodeModel::Model CM,
                                            CodeGenOpt::Level OL)
   : LLVMTargetMachine(T, TT, CPU, FS, Options, RM, CM, OL),
-    Subtarget(TT, CPU, FS),
+    Subtarget(TT, CPU, FS, Options),
     JITInfo(),
     InstrItins(Subtarget.getInstrItineraryData()) {
   // Default to soft float ABI