[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

Wrong indentation!?

3444

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).

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

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();
static bool CodeGen(Scop &S, IslAstInfo &AI, LoopInfo &LI, DominatorTree &DT,
   IslNodeBuilder NodeBuilder(Builder, Annotator, DL, LI, SE, DT, S, StartBlock);
 
   // All arrays must have their base pointers known before
   // ScopAnnotator::buildAliasScopes.
   NodeBuilder.allocateNewArrays(StartExitBlocks);
   Annotator.buildAliasScopes(S);
 
   if (PerfMonitoring) {
       PerfMonitor P(S, EnteringBB->getParent()->getParent());
       P.initialize();
       P.insertRegionStart(SplitBlock->getTerminator());
 
       BasicBlock *MergeBlock = ExitBlock->getUniqueSuccessor();
       P.insertRegionEnd(MergeBlock->getTerminator());
   }
 
   // First generate code for the hoisted invariant loads and transitively the
   // parameters they reference. Afterwards, for the remaining parameters that
   // might reference the hoisted loads. Finally, build the runtime check
   // that might reference both hoisted loads as well as parameters.
   // If the hoisting fails we have to bail and execute the original code.
   Builder.SetInsertPoint(SplitBlock->getTerminator());

lib/CodeGen/PPCGCodeGeneration.cpp

 #include "llvm/IR/Verifier.h"
 #include "llvm/IRReader/IRReader.h"
 #include "llvm/Linker/Linker.h"
 #include "llvm/Support/TargetRegistry.h"
 #include "llvm/Support/TargetSelect.h"
 #include "llvm/Target/TargetMachine.h"
 #include "llvm/Transforms/IPO/PassManagerBuilder.h"
 #include "llvm/Transforms/Utils/BasicBlockUtils.h"
+#include "polly/CodeGen/PerfMonitor.h"
 
 #include "isl/union_map.h"
 
 extern "C" {
 #include "ppcg/cuda.h"
 #include "ppcg/gpu.h"
 #include "ppcg/gpu_print.h"
 #include "ppcg/ppcg.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());
+
+            // Not sure if this is the correct "exiting" BB.
+            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);
     }