[BOLT][merge-fdata] Skip truncated lines in raw profile data (#183187)

Raw profile data file may contain lines truncated due to unexpected
app exit. This change is to have merge_fdata check number of fields
in each line of raw profile data file and ignore a line if the number
is not expected.

bolt/test/merge-fdata-bat-no-lbr.test

@@ -8,13 +8,13 @@
 
 # CHECK: boltedcollection
 # CHECK: no_lbr
-# CHECK: 1 main 2
+# CHECK: 1 main 0 2
 
 #--- a.fdata
 boltedcollection
 no_lbr
-1 main 1
+1 main 0 1
 #--- b.fdata
 boltedcollection
 no_lbr
-1 main 1
+1 main 0 1

bolt/test/merge-fdata-mixed-bat-no-lbr.test

@@ -10,7 +10,7 @@
 #--- a.fdata
 boltedcollection
 no_lbr
-1 main 1
+1 main 0 1
 #--- b.fdata
 no_lbr
-1 main 1
+1 main 0 1

bolt/test/merge-fdata-mixed-mode.test

@@ -10,6 +10,6 @@
 
 #--- a.fdata
 no_lbr
-1 main 1
+1 main 0 1
 #--- b.fdata
-1 main 1
+1 main 0 1

bolt/test/merge-fdata-no-lbr-mode.test

@@ -8,11 +8,11 @@
 # RUN: FileCheck %s --input-file %t/merged.fdata
 
 # CHECK: no_lbr
-# CHECK: 1 main 2
+# CHECK: 1 main 0 2
 
 #--- a.fdata
 no_lbr
-1 main 1
+1 main 0 1
 #--- b.fdata
 no_lbr
-1 main 1
+1 main 0 1

bolt/test/merge-fdata-skip-truncated.test

@@ -0,0 +1,20 @@
+## Check that merge-fdata skips truncated lines with a warning
+
+# REQUIRES: system-linux
+
+# RUN: split-file %s %t
+# RUN: merge-fdata %t/a.fdata %t/b.fdata -o %t/merged.fdata 2>&1 | \
+# RUN:     FileCheck -check-prefix=CHECK-WARN %s
+# RUN: FileCheck %s --input-file %t/merged.fdata
+
+# CHECK-WARN: WARNING: {{.*}}b.fdata: ignoring malformed entry
+
+## Only the valid entry should appear in the merged output
+# CHECK: 1 main 0 1 main 2 1 3
+# CHECK-NOT: trunc
+
+#--- a.fdata
+1 main 0 1 main 2 0 1
+#--- b.fdata
+1 main 0 1 main 2 1 2
+1 trunc 0 1

bolt/tools/merge-fdata/merge-fdata.cpp

@@ -318,11 +318,33 @@ void mergeLegacyProfiles(const SmallVectorImpl<std::string> &Filenames) {
 
     do {
       StringRef Line(FdataLine);
+      Line = Line.rtrim();
+      if (Line.empty())
+        continue;
       CounterTy Count;
       unsigned Type = 0;
       if (Line.split(' ').first.getAsInteger(10, Type))
         report_error(Filename, "Malformed / corrupted entry type");
       bool IsBranchEntry = Type < 3;
+
+      // Validate the number of fields in the line. Count only unescaped spaces
+      // as field separators, since function names may contain escaped spaces,
+      // like "foo\ bar".
+      size_t NumFields = 1;
+      for (size_t I = 0; I < Line.size(); ++I) {
+        if (Line[I] == '\\')
+          ++I;
+        else if (Line[I] == ' ')
+          ++NumFields;
+      }
+      size_t ExpectedFields =
+          IsBranchEntry ? (NoLBRCollection.value_or(false) ? 4 : 8) : 7;
+      if (NumFields != ExpectedFields) {
+        errs() << "WARNING: " << Filename << ": ignoring malformed entry with "
+               << NumFields << " fields (expected " << ExpectedFields << ")\n";
+        continue;
+      }
+
       auto [Signature, ExecCount] = Line.rsplit(' ');
       if (ExecCount.getAsInteger(10, Count.Exec))
         report_error(Filename, "Malformed / corrupted execution count");