Rename -allow-excess-fp-precision flag to -fuse-fp-ops, and switch from a

boolean flag to an enum: { Fast, Standard, Strict } (default = Standard).

This option controls the creation by optimizations of fused FP ops that store
intermediate results in higher precision than IEEE allows (E.g. FMAs). The
behavior of this option is intended to match the behaviour specified by a
soon-to-be-introduced frontend flag: '-ffuse-fp-ops'.

Fast mode - allows formation of fused FP ops whenever they're profitable.

Standard mode - allow fusion only for 'blessed' FP ops. At present the only
blessed op is the fmuladd intrinsic. In the future more blessed ops may be
added.

Strict mode - allow fusion only if/when it can be proven that the excess
precision won't effect the result.

Note: This option only controls formation of fused ops by the optimizers.  Fused
operations that are explicitly requested (e.g. FMA via the llvm.fma.* intrinsic)
will always be honored, regardless of the value of this option.

Internally TargetOptions::AllowExcessFPPrecision has been replaced by
TargetOptions::AllowFPOpFusion.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@158956 91177308-0d34-0410-b5e6-96231b3b80d8

1 files changed, 14 insertions, 6 deletions

diff --git a/tools/llc/llc.cpp b/tools/llc/llc.cpp
index b303cec3b5..e08bad93b0 100644
--- a/tools/llc/llc.cpp
+++ b/tools/llc/llc.cpp
@@ -156,11 +156,6 @@ DisableFPElimNonLeaf("disable-non-leaf-fp-elim",
   cl::init(false));
 
 static cl::opt<bool>
-EnableExcessPrecision("enable-excess-fp-precision",
-  cl::desc("Enable optimizations that may increase FP precision"),
-  cl::init(false));
-
-static cl::opt<bool>
 EnableUnsafeFPMath("enable-unsafe-fp-math",
   cl::desc("Enable optimizations that may decrease FP precision"),
   cl::init(false));
@@ -199,6 +194,19 @@ FloatABIForCalls("float-abi",
                "Hard float ABI (uses FP registers)"),
     clEnumValEnd));
 
+static cl::opt<llvm::FPOpFusion::FPOpFusionMode>
+FuseFPOps("fuse-fp-ops",
+  cl::desc("Enable aggresive formation of fused FP ops"),
+  cl::init(FPOpFusion::Standard),
+  cl::values(
+    clEnumValN(FPOpFusion::Fast, "fast",
+               "Fuse FP ops whenever profitable"),
+    clEnumValN(FPOpFusion::Standard, "standard",
+               "Only fuse 'blessed' FP ops."),
+    clEnumValN(FPOpFusion::Strict, "strict",
+               "Only fuse FP ops when the result won't be effected."),
+    clEnumValEnd));
+
 static cl::opt<bool>
 DontPlaceZerosInBSS("nozero-initialized-in-bss",
   cl::desc("Don't place zero-initialized symbols into bss section"),
@@ -404,7 +412,7 @@ int main(int argc, char **argv) {
   Options.LessPreciseFPMADOption = EnableFPMAD;
   Options.NoFramePointerElim = DisableFPElim;
   Options.NoFramePointerElimNonLeaf = DisableFPElimNonLeaf;
-  Options.AllowExcessFPPrecision = EnableExcessPrecision;
+  Options.AllowFPOpFusion = FuseFPOps;
   Options.UnsafeFPMath = EnableUnsafeFPMath;
   Options.NoInfsFPMath = EnableNoInfsFPMath;
   Options.NoNaNsFPMath = EnableNoNaNsFPMath;