[PowerPC] Implement combineRepeatedFPDivisors

This does not matter on newer cores (where we can use reciprocal estimates in
fast-math mode anyway), but for older cores this allows us to generate better
fast-math code where we have multiple FDIVs with a common divisor.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@222710 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/PowerPC/PPCISelLowering.cpp b/lib/Target/PowerPC/PPCISelLowering.cpp
index 351356028d63e5f456a63295382e0c817d8b98c7..7351d19120a1dd07f5f51356445220657fdea049 100644 (file)
--- a/lib/Target/PowerPC/PPCISelLowering.cpp
+++ b/lib/Target/PowerPC/PPCISelLowering.cpp
@@ -7526,6 +7526,28 @@ SDValue PPCTargetLowering::getRecipEstimate(SDValue Operand,
   return SDValue();
 }
 
+bool PPCTargetLowering::combineRepeatedFPDivisors(unsigned NumUsers) const {
+  // Note: This functionality is used only when unsafe-fp-math is enabled, and
+  // on cores with reciprocal estimates (which are used when unsafe-fp-math is
+  // enabled for division), this functionality is redundant with the default
+  // combiner logic (once the division -> reciprocal/multiply transformation
+  // has taken place). As a result, this matters more for older cores than for
+  // newer ones.
+
+  // Combine multiple FDIVs with the same divisor into multiple FMULs by the
+  // reciprocal if there are two or more FDIVs (for embedded cores with only
+  // one FP pipeline) for three or more FDIVs (for generic OOO cores).
+  switch (Subtarget.getDarwinDirective()) {
+  default:
+    return NumUsers > 2;
+  case PPC::DIR_440:
+  case PPC::DIR_A2:
+  case PPC::DIR_E500mc:
+  case PPC::DIR_E5500:
+    return NumUsers > 1;
+  }
+}
+
 static bool isConsecutiveLSLoc(SDValue Loc, EVT VT, LSBaseSDNode *Base,
                             unsigned Bytes, int Dist,
                             SelectionDAG &DAG) {

diff --git a/lib/Target/PowerPC/PPCISelLowering.h b/lib/Target/PowerPC/PPCISelLowering.h
index bb4d1f1f0197536323fb521f9bd460a93da73e27..4b4d25e3a3ccaffe35ee18c0fd84dfdb0ce41e33 100644 (file)
--- a/lib/Target/PowerPC/PPCISelLowering.h
+++ b/lib/Target/PowerPC/PPCISelLowering.h
@@ -704,6 +704,7 @@ namespace llvm {
                              bool &UseOneConstNR) const override;
     SDValue getRecipEstimate(SDValue Operand, DAGCombinerInfo &DCI,
                              unsigned &RefinementSteps) const override;
+    bool combineRepeatedFPDivisors(unsigned NumUsers) const override;
 
     CCAssignFn *useFastISelCCs(unsigned Flag) const;
   };

diff --git a/test/CodeGen/PowerPC/fdiv-combine.ll b/test/CodeGen/PowerPC/fdiv-combine.ll
new file mode 100644 (file)
index 0000000..d3dc3fe
--- /dev/null
+++ b/test/CodeGen/PowerPC/fdiv-combine.ll
@@ -0,0 +1,39 @@
+; RUN: llc -mcpu=ppc64 < %s | FileCheck %s
+target datalayout = "E-m:e-i64:64-n32:64"
+target triple = "powerpc64-unknown-linux-gnu"
+
+; Following test case checks:
+;   a / D; b / D; c / D;
+;                =>
+;   recip = 1.0 / D; a * recip; b * recip; c * recip;
+
+define void @three_fdiv_double(double %D, double %a, double %b, double %c) #0 {
+; CHECK-LABEL: three_fdiv_double:
+; CHECK: fdiv
+; CHECK-NEXT-NOT: fdiv
+; CHECK: fmul
+; CHECK: fmul
+; CHECK: fmul
+  %div = fdiv double %a, %D
+  %div1 = fdiv double %b, %D
+  %div2 = fdiv double %c, %D
+  tail call void @foo_3d(double %div, double %div1, double %div2)
+  ret void
+}
+
+define void @two_fdiv_double(double %D, double %a, double %b) #0 {
+; CHECK-LABEL: two_fdiv_double:
+; CHECK: fdiv
+; CHECK: fdiv
+; CHECK-NEXT-NOT: fmul
+  %div = fdiv double %a, %D
+  %div1 = fdiv double %b, %D
+  tail call void @foo_2d(double %div, double %div1)
+  ret void
+}
+
+declare void @foo_3d(double, double, double)
+declare void @foo_3_2xd(<2 x double>, <2 x double>, <2 x double>)
+declare void @foo_2d(double, double)
+
+attributes #0 = { "unsafe-fp-math"="true" }