[PPCGCodeGeneration] Enable `polly-codegen-perf-monitoring` for PPCGCodegen.

[bollu]

@grosser @philip.pfaffe @Meinersbur @singam-sanjay ping, review please.

[grosser]

Hi Siddharth,

some simple comments, otherwise this LGTM.

lib/CodeGen/PPCGCodeGeneration.cpp

3439 ↗

(On Diff #111888)

Wrong indentation!?

3444 ↗

(On Diff #111888)

I don't think we can keep such a comment. Does not really add information. At the very least explain what you assume and what you do not understand. Also, we should probably think about it once. It is not so complicated in the end.

[annanay25]

This patch seems to result in an unexpected passing test - ScopDetectionDiagnostics/ReportLoopHasNoExit.ll

Could that be changed?

[bollu]

@annanay25 - I believe that was changed by people upstream due to the PostDomTree changes. (specifically, D35851).

polly/trunk/include/polly/CodeGen/CodeGeneration.h

 /// Marks the basic block @p Block unreachable by equipping it with an
 /// UnreachableInst.
 void markBlockUnreachable(BasicBlock &Block, PollyIRBuilder &Builder);
 
 struct CodeGenerationPass : public PassInfoMixin<CodeGenerationPass> {
   PreservedAnalyses run(Scop &S, ScopAnalysisManager &SAM,
                         ScopStandardAnalysisResults &AR, SPMUpdater &U);
 };
+
+extern bool PerfMonitoring;
 } // namespace polly
 
 #endif // POLLY_CODEGENERATION_H

polly/trunk/lib/CodeGen/CodeGeneration.cpp

 
 #define DEBUG_TYPE "polly-codegen"
 
 static cl::opt<bool> Verify("polly-codegen-verify",
                             cl::desc("Verify the function generated by Polly"),
                             cl::Hidden, cl::init(false), cl::ZeroOrMore,
                             cl::cat(PollyCategory));
 
-static cl::opt<bool>
-    PerfMonitoring("polly-codegen-perf-monitoring",
+bool polly::PerfMonitoring;
+static cl::opt<bool, true>
+    XPerfMonitoring("polly-codegen-perf-monitoring",
                     cl::desc("Add run-time performance monitoring"), cl::Hidden,
-                   cl::init(false), cl::ZeroOrMore, cl::cat(PollyCategory));
+                    cl::location(polly::PerfMonitoring), cl::init(false),
+                    cl::ZeroOrMore, cl::cat(PollyCategory));
 
 namespace polly {
 /// Mark a basic block unreachable.
 ///
 /// Marks the basic block @p Block unreachable by equipping it with an
 /// UnreachableInst.
 void markBlockUnreachable(BasicBlock &Block, PollyIRBuilder &Builder) {
   auto *OrigTerminator = Block.getTerminator();

polly/trunk/lib/CodeGen/PPCGCodeGeneration.cpp

 // GPU mapping strategy.
 //
 //===----------------------------------------------------------------------===//
 
 #include "polly/CodeGen/PPCGCodeGeneration.h"
 #include "polly/CodeGen/CodeGeneration.h"
 #include "polly/CodeGen/IslAst.h"
 #include "polly/CodeGen/IslNodeBuilder.h"
+#include "polly/CodeGen/PerfMonitor.h"
 #include "polly/CodeGen/Utils.h"
 #include "polly/DependenceInfo.h"
 #include "polly/LinkAllPasses.h"
 #include "polly/Options.h"
 #include "polly/ScopDetection.h"
 #include "polly/ScopInfo.h"
 #include "polly/Support/SCEVValidator.h"
 #include "llvm/ADT/PostOrderIterator.h"
CudaVersion("polly-acc-cuda-version",
                 cl::desc("The CUDA version to compile for"), cl::Hidden,
                 cl::init("sm_30"), cl::ZeroOrMore, cl::cat(PollyCategory));
 
 static cl::opt<int>
     MinCompute("polly-acc-mincompute",
                cl::desc("Minimal number of compute statements to run on GPU."),
                cl::Hidden, cl::init(10 * 512 * 512));
 
+extern bool polly::PerfMonitoring;
+
 /// Return  a unique name for a Scop, which is the scop region with the
 /// function name.
 std::string getUniqueScopName(const Scop *S) {
   return "Scop Region: " + S->getNameStr() +
          " | Function: " + std::string(S->getFunction().getName());
 }
 
 /// Used to store information PPCG wants for kills. This information is
if (!NodeBuilder.preloadInvariantLoads()) {
       assert(ExitingBB);
 
       DT->changeImmediateDominator(MergeBlock, ExitingBB);
       DT->eraseNode(ExitingBlock);
       isl_ast_expr_free(Condition);
       isl_ast_node_free(Root);
     } else {
 
+      if (polly::PerfMonitoring) {
+        PerfMonitor P(*S, EnteringBB->getParent()->getParent());
+        P.initialize();
+        P.insertRegionStart(SplitBlock->getTerminator());
+
+        // TODO: actually think if this is the correct exiting block to place
+        // the `end` performance marker. Invariant load hoisting changes
+        // the CFG in a way that I do not precisely understand, so I
+        // (Siddharth<[e-mail]>) should come back to this and
+        // think about which exiting block to use.
+        auto *ExitingBlock = StartBlock->getUniqueSuccessor();
+        assert(ExitingBlock);
+        BasicBlock *MergeBlock = ExitingBlock->getUniqueSuccessor();
+        P.insertRegionEnd(MergeBlock->getTerminator());
+      }
+
       NodeBuilder.addParameters(S->getContext().release());
       Value *RTC = NodeBuilder.createRTC(Condition);
       Builder.GetInsertBlock()->getTerminator()->setOperand(0, RTC);
 
       Builder.SetInsertPoint(&*StartBlock->begin());
 
       NodeBuilder.create(Root);
     }