DivergenceAnalysis: added debug output

[tpr]

Summary

This commit does two things:

1. modified the existing DivergenceAnalysis::dump() so it dumps the whole function with added DIVERGENT: annotations;

2. added code to do that dump if the appropriate -debug-only option is on.

Change-Id: Id97b605aab1fc6f5a11a20c58a99bbe8c565bf83

[arsenm]

lib/Analysis/DivergenceAnalysis.cpp

304–307 ↗

(On Diff #149699)

Doesn't -print-after=divergence already get you this?

[tpr]

lib/Analysis/DivergenceAnalysis.cpp

304–307 ↗

(On Diff #149699)

No:

1. -print-after=divergence (or -print-after-all) does not print anything after this pass, perhaps because it is an analysis that says it has not modified the code.

2. Even if it did, it would surely be using the standard pass printer that doesn't know anything about annotating with DIVERGENT:.

[nhaehnle]

I guess it would be nice to have something like opt's -analyze output in an llc flow as well, but I don't think we do, so I think this is fine.

llvm/trunk/lib/Analysis/DivergenceAnalysis.cpp

 #include "llvm/IR/InstIterator.h"
 #include "llvm/IR/Instructions.h"
 #include "llvm/IR/Value.h"
 #include "llvm/Support/Debug.h"
 #include "llvm/Support/raw_ostream.h"
 #include <vector>
 using namespace llvm;
 
+#define DEBUG_TYPE "divergence"
+
 namespace {
 
 class DivergencePropagator {
 public:
   DivergencePropagator(Function &F, TargetTransformInfo &TTI, DominatorTree &DT,
                        PostDominatorTree &PDT, DenseSet<const Value *> &DV)
       : F(F), TTI(TTI), DT(DT), PDT(PDT), DV(DV) {}
   void populateWithSourcesOfDivergence();
bool DivergenceAnalysis::runOnFunction(Function &F) {
 
   DivergentValues.clear();
   auto &PDT = getAnalysis<PostDominatorTreeWrapperPass>().getPostDomTree();
   DivergencePropagator DP(F, TTI,
                           getAnalysis<DominatorTreeWrapperPass>().getDomTree(),
                           PDT, DivergentValues);
   DP.populateWithSourcesOfDivergence();
   DP.propagate();
+  LLVM_DEBUG(
+    dbgs() << "\nAfter divergence analysis on " << F.getName() << ":\n";
+    print(dbgs(), F.getParent())
+  );
   return false;
 }
 
 void DivergenceAnalysis::print(raw_ostream &OS, const Module *) const {
   if (DivergentValues.empty())
     return;
   const Value *FirstDivergentValue = *DivergentValues.begin();
   const Function *F;
   if (const Argument *Arg = dyn_cast<Argument>(FirstDivergentValue)) {
     F = Arg->getParent();
   } else if (const Instruction *I =
                  dyn_cast<Instruction>(FirstDivergentValue)) {
     F = I->getParent()->getParent();
   } else {
     llvm_unreachable("Only arguments and instructions can be divergent");
   }
 
   // Dumps all divergent values in F, arguments and then instructions.
   for (auto &Arg : F->args()) {
-    if (DivergentValues.count(&Arg))
-      OS << "DIVERGENT:  " << Arg << "\n";
+    OS << (DivergentValues.count(&Arg) ? "DIVERGENT: " : "           ");
+    OS << Arg << "\n";
   }
   // Iterate instructions using instructions() to ensure a deterministic order.
-  for (auto &I : instructions(F)) {
-    if (DivergentValues.count(&I))
-      OS << "DIVERGENT:" << I << "\n";
+  for (auto BI = F->begin(), BE = F->end(); BI != BE; ++BI) {
+    auto &BB = *BI;
+    OS << "\n           " << BB.getName() << ":\n";
+    for (auto &I : BB.instructionsWithoutDebug()) {
+      OS << (DivergentValues.count(&I) ? "DIVERGENT:     " : "               ");
+      OS << I << "\n";
+    }
   }
+  OS << "\n";
 }

llvm/trunk/test/Analysis/DivergenceAnalysis/AMDGPU/phi-undef.ll

 ; RUN: opt -mtriple=amdgcn-- -analyze -divergence %s | FileCheck %s
 
 ; CHECK-LABEL: 'test1':
 ; CHECK-NEXT: DIVERGENT: i32 %bound
+; CHECK: {{^  *}}%counter =
 ; CHECK-NEXT: DIVERGENT: %break = icmp sge i32 %counter, %bound
 ; CHECK-NEXT: DIVERGENT: br i1 %break, label %footer, label %body
-; CHECK-NEXT: DIVERGENT: br i1 %break, label %end, label %header
+; CHECK: {{^  *}}%counter.next =
+; CHECK: {{^  *}}%counter.footer =
+; CHECK: DIVERGENT: br i1 %break, label %end, label %header
 ; Note: %counter is not divergent!
 define amdgpu_ps void @test1(i32 %bound) {
 entry:
   br label %header
 
 header:
   %counter = phi i32 [ 0, %entry ], [ %counter.footer, %footer ]
   %break = icmp sge i32 %counter, %bound