[llvm] 8cbbd7e - [llvm-profgen] Ignore broken LBR samples

[Hongtao Yu via llvm-commits]

Author: Hongtao Yu
Date: 2021-09-14T12:11:17-07:00
New Revision: 8cbbd7e0b2aa21ce7e416cfb63d9965518948c35

URL: https://github.com/llvm/llvm-project/commit/8cbbd7e0b2aa21ce7e416cfb63d9965518948c35
DIFF: https://github.com/llvm/llvm-project/commit/8cbbd7e0b2aa21ce7e416cfb63d9965518948c35.diff

LOG: [llvm-profgen] Ignore broken LBR samples

Perf script can sometimes give disordered LBR samples like below.

```
          b022500
          32de0044
          3386e1d1
      7f118e05720c
      7f118df2d81f
 0x2a0b9622/0x2a0b9610/P/-/-/1  0x2a0b79ff/0x2a0b9618/P/-/-/2  0x2a0b7a4a/0x2a0b79e8/P/-/-/1  0x2a0b7a33/0x2a0b7a46/P/-/-/1  0x2a0b7a42/0x2a0b7a23/P/-/-/1  0x2a0b7a21/0x2a0b7a37/P/-/-/2  0x2a0b79e6/0x2a0b7a07/P/-/-/1  0x2a0b79d4/0x2a0b79dc/P/-/-/2  0x2a0b7a03/0x2a0b79aa/P/-/-/1  0x2a0b79a8/0x2a0b7a00/P/-/-/234  0x2a0b9613/0x2a0b7930/P/-/-/1  0x2a0b9622/0x2a0b9610/P/-/-/1  0x2a0b79ff/0x2a0b9618/P/-/-/2  0x2a0b7a4a/0x2aWarning:
Processed 10263226 events and lost 1 chunks!

```
 Note that the last LBR record `0x2a0b7a4a/0x2aWarning:` . Currently llvm-profgen does not detect that and as a result an uninitialized branch target value will be used. The uninitialized value can cause creepy instruction ranges created which which in turn will result in a completely wrong profile. An example is like

```

 .... @ _ZN5folly13loadUnalignedIsEET_PKv]:18446744073709551615:18446744073709551615
 1: 18446744073709551615
 !CFGChecksum: 4294967295
 !Attributes: 0
```

Reviewed By: wenlei, wlei

Differential Revision: https://reviews.llvm.org/D109637

Added: 
    

Modified: 
    llvm/tools/llvm-profgen/PerfReader.cpp
    llvm/tools/llvm-profgen/PerfReader.h
    llvm/tools/llvm-profgen/ProfileGenerator.cpp

Removed: 
    


################################################################################
diff  --git a/llvm/tools/llvm-profgen/PerfReader.cpp b/llvm/tools/llvm-profgen/PerfReader.cpp
index 77baa9bc5a43e..259b23f33d47b 100644
--- a/llvm/tools/llvm-profgen/PerfReader.cpp
+++ b/llvm/tools/llvm-profgen/PerfReader.cpp
@@ -422,8 +422,15 @@ bool PerfReaderBase::extractLBRStack(TraceStream &TraceIt,
     Token.split(Addresses, "/");
     uint64_t Src;
     uint64_t Dst;
-    Addresses[0].substr(2).getAsInteger(16, Src);
-    Addresses[1].substr(2).getAsInteger(16, Dst);
+
+    // Stop at broken LBR records.
+    if (Addresses[0].substr(2).getAsInteger(16, Src) ||
+        Addresses[1].substr(2).getAsInteger(16, Dst)) {
+      WithColor::warning() << "Invalid address in LBR record at line "
+                           << TraceIt.getLineNumber() << ": "
+                           << TraceIt.getCurrentLine() << "\n";
+      break;
+    }
 
     bool SrcIsInternal = Binary->addressIsCode(Src);
     bool DstIsInternal = Binary->addressIsCode(Dst);
@@ -635,11 +642,8 @@ void LBRPerfReader::computeCounterFromLBR(const PerfSample *Sample,
     // If this not the first LBR, update the range count between TO of current
     // LBR and FROM of next LBR.
     uint64_t StartOffset = TargetOffset;
-    if (EndOffeset != 0) {
-      assert(StartOffset <= EndOffeset &&
-             "Bogus range should be filtered ealier!");
+    if (EndOffeset != 0)
       Counter.recordRangeCount(StartOffset, EndOffeset, Repeat);
-    }
     EndOffeset = SourceOffset;
   }
 }

diff  --git a/llvm/tools/llvm-profgen/PerfReader.h b/llvm/tools/llvm-profgen/PerfReader.h
index 0595c7dc5973a..2babcd7246a8b 100644
--- a/llvm/tools/llvm-profgen/PerfReader.h
+++ b/llvm/tools/llvm-profgen/PerfReader.h
@@ -373,6 +373,7 @@ struct SampleCounter {
   BranchSample BranchCounter;
 
   void recordRangeCount(uint64_t Start, uint64_t End, uint64_t Repeat) {
+    assert(Start <= End && "Invalid instruction range");
     RangeCounter[{Start, End}] += Repeat;
   }
   void recordBranchCount(uint64_t Source, uint64_t Target, uint64_t Repeat) {

diff  --git a/llvm/tools/llvm-profgen/ProfileGenerator.cpp b/llvm/tools/llvm-profgen/ProfileGenerator.cpp
index 5b9984cc8986c..5251e3f187c70 100644
--- a/llvm/tools/llvm-profgen/ProfileGenerator.cpp
+++ b/llvm/tools/llvm-profgen/ProfileGenerator.cpp
@@ -183,6 +183,7 @@ void ProfileGenerator::findDisjointRanges(RangeSample &DisjointRanges,
   for (auto Item : Ranges) {
     uint64_t Begin = Item.first.first;
     uint64_t End = Item.first.second;
+    assert(Begin <= End && "Invalid instruction range");
     uint64_t Count = Item.second;
     if (Boundaries.find(Begin) == Boundaries.end())
       Boundaries[Begin] = BoundaryPoint();
@@ -194,7 +195,7 @@ void ProfileGenerator::findDisjointRanges(RangeSample &DisjointRanges,
   }
 
   uint64_t BeginAddress = UINT64_MAX;
-  int Count = 0;
+  uint64_t Count = 0;
   for (auto Item : Boundaries) {
     uint64_t Address = Item.first;
     BoundaryPoint &Point = Item.second;
@@ -208,6 +209,7 @@ void ProfileGenerator::findDisjointRanges(RangeSample &DisjointRanges,
       assert((BeginAddress != UINT64_MAX) &&
              "First boundary point cannot be 'end' point");
       DisjointRanges[{BeginAddress, Address}] = Count;
+      assert(Count >= Point.EndCount && "Mismatched live ranges");
       Count -= Point.EndCount;
       BeginAddress = Address + 1;
     }