Improve instruction scheduling for the PPC POWER7

Aside from a few minor latency corrections, the major change here is a new
hazard recognizer which focuses on better dispatch-group formation on the
POWER7. As with the PPC970's hazard recognizer, the most important thing it
does is avoid load-after-store hazards within the same dispatch group. It uses
the POWER7's special dispatch-group-terminating nop instruction (instead of
inserting multiple regular nop instructions). This new hazard recognizer makes
use of the scheduling dependency graph itself, built using AA information, to
robustly detect the possibility of load-after-store hazards.

significant test-suite performance changes (the error bars are 99.5% confidence
intervals based on 5 test-suite runs both with and without the change --
speedups are negative):

speedups:

MultiSource/Benchmarks/FreeBench/pcompress2/pcompress2
	-0.55171% +/- 0.333168%

MultiSource/Benchmarks/TSVC/CrossingThresholds-dbl/CrossingThresholds-dbl
	-17.5576% +/- 14.598%

MultiSource/Benchmarks/TSVC/Reductions-dbl/Reductions-dbl
	-29.5708% +/- 7.09058%

MultiSource/Benchmarks/TSVC/Reductions-flt/Reductions-flt
	-34.9471% +/- 11.4391%

SingleSource/Benchmarks/BenchmarkGame/puzzle
	-25.1347% +/- 11.0104%

SingleSource/Benchmarks/Misc/flops-8
	-17.7297% +/- 9.79061%

SingleSource/Benchmarks/Shootout-C++/ary3
	-35.5018% +/- 23.9458%

SingleSource/Regression/C/uint64_to_float
	-56.3165% +/- 25.4234%

SingleSource/UnitTests/Vectorizer/gcc-loops
	-18.5309% +/- 6.8496%

regressions:

MultiSource/Benchmarks/ASCI_Purple/SMG2000/smg2000
	18.351% +/- 12.156%

SingleSource/Benchmarks/Shootout-C++/methcall
	27.3086% +/- 14.4733%

llvm-svn: 197099

1 files changed, 26 insertions, 0 deletions

diff --git a/llvm/lib/Target/PowerPC/PPCHazardRecognizers.h b/llvm/lib/Target/PowerPC/PPCHazardRecognizers.h
index 9ee042b63e1..6b7fe41e574 100644
--- a/llvm/lib/Target/PowerPC/PPCHazardRecognizers.h
+++ b/llvm/lib/Target/PowerPC/PPCHazardRecognizers.h
@@ -21,6 +21,32 @@
 
 namespace llvm {
 
+/// PPCDispatchGroupSBHazardRecognizer - This class implements a scoreboard-based
+/// hazard recognizer for PPC ooo processors with dispatch-group hazards.
+class PPCDispatchGroupSBHazardRecognizer : public ScoreboardHazardRecognizer {
+  const ScheduleDAG *DAG;
+  SmallVector<SUnit *, 7> CurGroup;
+  unsigned CurSlots, CurBranches;
+
+  bool isLoadAfterStore(SUnit *SU);
+  bool isBCTRAfterSet(SUnit *SU);
+  bool mustComeFirst(const MCInstrDesc *MCID, unsigned &NSlots);
+public:
+  PPCDispatchGroupSBHazardRecognizer(const InstrItineraryData *ItinData,
+                         const ScheduleDAG *DAG_) :
+    ScoreboardHazardRecognizer(ItinData, DAG_), DAG(DAG_),
+    CurSlots(0), CurBranches(0) {}
+
+  virtual HazardType getHazardType(SUnit *SU, int Stalls);
+  virtual bool ShouldPreferAnother(SUnit* SU);
+  virtual unsigned PreEmitNoops(SUnit *SU);
+  virtual void EmitInstruction(SUnit *SU);
+  virtual void AdvanceCycle();
+  virtual void RecedeCycle();
+  virtual void Reset();
+  virtual void EmitNoop();
+};
+
 /// PPCHazardRecognizer970 - This class defines a finite state automata that
 /// models the dispatch logic on the PowerPC 970 (aka G5) processor.  This
 /// promotes good dispatch group formation and implements noop insertion to