[BOLT][NFCI] Skip validation in parseLBRSample (#143288)

Parsed branches and fall-throughs are validated in `doBranch` and
`doTrace` respectively. Simplify parseLBRSample by omitting the
validation. This also speeds up perf data processing as checks are only
done once for aggregated branches/fall-throughs and not individual LBR
entries.

Since invalid/external addresses are no longer sanitized during parsing,
sanitize them in `doBranch`.

Test Plan: updated X86/pre-aggregated-perf.test

Author: aaupov

bolt/lib/Profile/DataAggregator.cpp

@@ -735,8 +735,10 @@ bool DataAggregator::doBranch(uint64_t From, uint64_t To, uint64_t Count,
   // corresponds to a return (if \p IsFrom) or a call continuation (otherwise).
   auto handleAddress = [&](uint64_t &Addr, bool IsFrom) {
     BinaryFunction *Func = getBinaryFunctionContainingAddress(Addr);
-    if (!Func)
+    if (!Func) {
+      Addr = 0;
       return std::pair{Func, false};
+    }
 
     Addr -= Func->getAddress();
 
@@ -1432,54 +1434,16 @@ void DataAggregator::parseLBRSample(const PerfBranchSample &Sample,
       const uint64_t TraceTo = NextLBR->From;
       const BinaryFunction *TraceBF =
           getBinaryFunctionContainingAddress(TraceFrom);
-      if (opts::HeatmapMode == opts::HeatmapModeKind::HM_Exclusive) {
-        FTInfo &Info = FallthroughLBRs[Trace(TraceFrom, TraceTo)];
+      FTInfo &Info = FallthroughLBRs[Trace(TraceFrom, TraceTo)];
+      if (TraceBF && TraceBF->containsAddress(LBR.From))
         ++Info.InternCount;
-      } else if (TraceBF && TraceBF->containsAddress(TraceTo)) {
-        FTInfo &Info = FallthroughLBRs[Trace(TraceFrom, TraceTo)];
-        if (TraceBF->containsAddress(LBR.From))
-          ++Info.InternCount;
-        else
-          ++Info.ExternCount;
-      } else {
-        const BinaryFunction *ToFunc =
-            getBinaryFunctionContainingAddress(TraceTo);
-        if (TraceBF && ToFunc) {
-          LLVM_DEBUG({
-            dbgs() << "Invalid trace starting in " << TraceBF->getPrintName()
-                   << formatv(" @ {0:x}", TraceFrom - TraceBF->getAddress())
-                   << formatv(" and ending @ {0:x}\n", TraceTo);
-          });
-          ++NumInvalidTraces;
-        } else {
-          LLVM_DEBUG({
-            dbgs() << "Out of range trace starting in "
-                   << (TraceBF ? TraceBF->getPrintName() : "None")
-                   << formatv(" @ {0:x}",
-                              TraceFrom - (TraceBF ? TraceBF->getAddress() : 0))
-                   << " and ending in "
-                   << (ToFunc ? ToFunc->getPrintName() : "None")
-                   << formatv(" @ {0:x}\n",
-                              TraceTo - (ToFunc ? ToFunc->getAddress() : 0));
-          });
-          ++NumLongRangeTraces;
-        }
-      }
+      else
+        ++Info.ExternCount;
       ++NumTraces;
     }
     NextLBR = &LBR;
 
-    // Record branches outside binary functions for heatmap.
-    if (opts::HeatmapMode == opts::HeatmapModeKind::HM_Exclusive) {
-      TakenBranchInfo &Info = BranchLBRs[Trace(LBR.From, LBR.To)];
-      ++Info.TakenCount;
-      continue;
-    }
-    uint64_t From = getBinaryFunctionContainingAddress(LBR.From) ? LBR.From : 0;
-    uint64_t To = getBinaryFunctionContainingAddress(LBR.To) ? LBR.To : 0;
-    if (!From && !To)
-      continue;
-    TakenBranchInfo &Info = BranchLBRs[Trace(From, To)];
+    TakenBranchInfo &Info = BranchLBRs[Trace(LBR.From, LBR.To)];
     ++Info.TakenCount;
     Info.MispredCount += LBR.Mispred;
   }

bolt/test/X86/pre-aggregated-perf.test

@@ -69,9 +69,9 @@ CHECK-BASIC-NL: no_lbr cycles
 
 PERF2BOLT: 1 frame_dummy/1 1e 1 frame_dummy/1 0 0 1
 PERF2BOLT-NEXT: 1 main 451 1 SolveCubic 0 0 2
-PERF2BOLT-NEXT: 1 main 490 0 [unknown] 4005f0 0 1
-PERF2BOLT-NEXT: 1 main 537 0 [unknown] 400610 0 1
-PERF2BOLT-NEXT: 0 [unknown] 7f36d18d60c0 1 main 53c 0 2
+PERF2BOLT-NEXT: 1 main 490 0 [unknown] 0 0 1
+PERF2BOLT-NEXT: 1 main 537 0 [unknown] 0 0 1
+PERF2BOLT-NEXT: 0 [unknown] 0 1 main 53c 0 2
 PERF2BOLT-NEXT: 1 usqrt a 1 usqrt 10 0 22
 PERF2BOLT-NEXT: 1 usqrt 30 1 usqrt 32 0 22
 PERF2BOLT-NEXT: 1 usqrt 30 1 usqrt 39 4 33