Using branch probability to guide critical edge splitting.
Summary:
The original heuristic to break critical edge during machine sink is relatively conservertive: when there is only one instruction sinkable to the critical edge, it is likely that the machine sink pass will not break the critical edge. This leads to many speculative instructions executed at runtime. However, with profile info, we could model the splitting benefits: if the critical edge has 50% taken rate, it would always be beneficial to split the critical edge to avoid the speculated runtime instructions. This patch uses profile to guide critical edge splitting in machine sink pass.
The performance impact on speccpu2006 on Intel sandybridge machines:
spec/2006/fp/C++/444.namd 25.3 +0.26%
spec/2006/fp/C++/447.dealII 45.96 -0.10%
spec/2006/fp/C++/450.soplex 41.97 +1.49%
spec/2006/fp/C++/453.povray 36.83 -0.96%
spec/2006/fp/C/433.milc 23.81 +0.32%
spec/2006/fp/C/470.lbm 41.17 +0.34%
spec/2006/fp/C/482.sphinx3 48.13 +0.69%
spec/2006/int/C++/471.omnetpp 22.45 +3.25%
spec/2006/int/C++/473.astar 21.35 -2.06%
spec/2006/int/C++/483.xalancbmk 36.02 -2.39%
spec/2006/int/C/400.perlbench 33.7 -0.17%
spec/2006/int/C/401.bzip2 22.9 +0.52%
spec/2006/int/C/403.gcc 32.42 -0.54%
spec/2006/int/C/429.mcf 39.59 +0.19%
spec/2006/int/C/445.gobmk 26.98 -0.00%
spec/2006/int/C/456.hmmer 24.52 -0.18%
spec/2006/int/C/458.sjeng 28.26 +0.02%
spec/2006/int/C/462.libquantum 55.44 +3.74%
spec/2006/int/C/464.h264ref 46.67 -0.39%
geometric mean +0.20%
Manually checked 473 and 471 to verify the diff is in the noise range.
Reviewers: rengolin, davidxl
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D24818
llvm-svn: 284757
diff --git a/llvm/lib/CodeGen/MachineSink.cpp b/llvm/lib/CodeGen/MachineSink.cpp
index d2650db..27d6c20 100644
--- a/llvm/lib/CodeGen/MachineSink.cpp
+++ b/llvm/lib/CodeGen/MachineSink.cpp
@@ -24,6 +24,7 @@
#include "llvm/Analysis/AliasAnalysis.h"
#include "llvm/CodeGen/MachineBasicBlock.h"
#include "llvm/CodeGen/MachineBlockFrequencyInfo.h"
+#include "llvm/CodeGen/MachineBranchProbabilityInfo.h"
#include "llvm/CodeGen/MachineDominators.h"
#include "llvm/CodeGen/MachineFunction.h"
#include "llvm/CodeGen/MachineFunctionPass.h"
@@ -60,6 +61,15 @@
cl::desc("Use block frequency info to find successors to sink"),
cl::init(true), cl::Hidden);
+static cl::opt<unsigned> SplitEdgeProbabilityThreshold(
+ "machine-sink-split-probability-threshold",
+ cl::desc(
+ "Percentage threshold for splitting single-instruction critical edge. "
+ "If the branch threshold is higher than this threshold, we allow "
+ "speculative execution of up to 1 instruction to avoid branching to "
+ "splitted critical edge"),
+ cl::init(40), cl::Hidden);
+
STATISTIC(NumSunk, "Number of machine instructions sunk");
STATISTIC(NumSplit, "Number of critical edges split");
STATISTIC(NumCoalesces, "Number of copies coalesced");
@@ -74,6 +84,7 @@
MachinePostDominatorTree *PDT; // Machine post dominator tree
MachineLoopInfo *LI;
const MachineBlockFrequencyInfo *MBFI;
+ const MachineBranchProbabilityInfo *MBPI;
AliasAnalysis *AA;
// Remember which edges have been considered for breaking.
@@ -105,6 +116,7 @@
AU.addRequired<MachineDominatorTree>();
AU.addRequired<MachinePostDominatorTree>();
AU.addRequired<MachineLoopInfo>();
+ AU.addRequired<MachineBranchProbabilityInfo>();
AU.addPreserved<MachineDominatorTree>();
AU.addPreserved<MachinePostDominatorTree>();
AU.addPreserved<MachineLoopInfo>();
@@ -163,6 +175,7 @@
char &llvm::MachineSinkingID = MachineSinking::ID;
INITIALIZE_PASS_BEGIN(MachineSinking, "machine-sink",
"Machine code sinking", false, false)
+INITIALIZE_PASS_DEPENDENCY(MachineBranchProbabilityInfo)
INITIALIZE_PASS_DEPENDENCY(MachineDominatorTree)
INITIALIZE_PASS_DEPENDENCY(MachineLoopInfo)
INITIALIZE_PASS_DEPENDENCY(AAResultsWrapperPass)
@@ -283,6 +296,7 @@
PDT = &getAnalysis<MachinePostDominatorTree>();
LI = &getAnalysis<MachineLoopInfo>();
MBFI = UseBlockFreqInfo ? &getAnalysis<MachineBlockFrequencyInfo>() : nullptr;
+ MBPI = &getAnalysis<MachineBranchProbabilityInfo>();
AA = &getAnalysis<AAResultsWrapperPass>().getAAResults();
bool EverMadeChange = false;
@@ -383,6 +397,10 @@
if (!MI.isCopy() && !TII->isAsCheapAsAMove(MI))
return true;
+ if (From->isSuccessor(To) && MBPI->getEdgeProbability(From, To) <=
+ BranchProbability(SplitEdgeProbabilityThreshold, 100))
+ return true;
+
// MI is cheap, we probably don't want to break the critical edge for it.
// However, if this would allow some definitions of its source operands
// to be sunk then it's probably worth it.