[llvm] [llvm-profgen] Improve sample profile density (PR #92144)

[via llvm-commits]

================
@@ -768,9 +745,79 @@ void ProfileGenerator::populateBoundarySamplesForAllFunctions(
   }
 }
 
+void ProfileGeneratorBase::calculateDensity(
+    const FunctionSamples &FSamples,
+    std::vector<std::pair<double, uint64_t>> &DensityList,
+    uint64_t &TotalProfileSamples) {
+  uint64_t TotalBodySamples = 0;
+  uint64_t FuncBodySize = 0;
+  for (const auto &I : FSamples.getBodySamples()) {
+    TotalBodySamples += I.second.getSamples();
+    FuncBodySize++;
+  }
+
+  // The whole function could be inlined and optimized out, use the callsite
+  // head samples instead to estimate the body count.
+  if (FuncBodySize == 0) {
+    for (const auto &CallsiteSamples : FSamples.getCallsiteSamples()) {
+      FuncBodySize++;
+      for (const auto &Callee : CallsiteSamples.second) {
+        calculateDensity(Callee.second, DensityList, TotalProfileSamples);
----------------
WenleiHe wrote:

Ok, the recursion here covers inlinee samples. Good to know it's not dropped. 
 
Though if we decided to compute density after pre-inline, it might make more sense to include inlinees directly? But yes, data will tell us which one is better. If they are similar, then we can aim for implementation simplicity.

https://github.com/llvm/llvm-project/pull/92144