Skip to content

[llvm-debuginfo-analyzer] Incorrect directory path (.debug_line) #143849

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

CarlosAlbertoEnciso
Copy link
Member

If any 'name' entry in the 'file_names' table from the .debug_line
section contains directory information, the llvm-debuginfo-analyzer
tool will display duplicated directory information when using the
'--attribute=directories' option.

.debug_line contents:
Line table prologue:
...
include_directories[ 0] = "path"
file_names[ 0]:
name: "path/test.cpp"
dir_index: 0
...

the '--attribute=directories' will show:

{Directory} 'path/path' <-- Duplicated 'path'

If any 'name' entry in the 'file_names' table from the .debug_line
section contains directory information, the llvm-debuginfo-analyzer
tool will display duplicated directory information when using the
'--attribute=directories' option.

.debug_line contents:
Line table prologue:
  ...
include_directories[  0] = "path"
file_names[  0]:
           name: "path/test.cpp"
      dir_index: 0
  ...

the '--attribute=directories' will show:

  {Directory} 'path/path'  <-- Duplicated 'path'
@llvmbot
Copy link
Member

llvmbot commented Jun 12, 2025

@llvm/pr-subscribers-debuginfo

Author: Carlos Alberto Enciso (CarlosAlbertoEnciso)

Changes

If any 'name' entry in the 'file_names' table from the .debug_line
section contains directory information, the llvm-debuginfo-analyzer
tool will display duplicated directory information when using the
'--attribute=directories' option.

.debug_line contents:
Line table prologue:
...
include_directories[ 0] = "path"
file_names[ 0]:
name: "path/test.cpp"
dir_index: 0
...

the '--attribute=directories' will show:

{Directory} 'path/path' <-- Duplicated 'path'


Full diff: https://github.com/llvm/llvm-project/pull/143849.diff

2 Files Affected:

  • (modified) llvm/lib/DebugInfo/LogicalView/Readers/LVDWARFReader.cpp (+6-4)
  • (added) llvm/test/tools/llvm-debuginfo-analyzer/DWARF/dwarf-duplicated-directory-path.ll (+67)
diff --git a/llvm/lib/DebugInfo/LogicalView/Readers/LVDWARFReader.cpp b/llvm/lib/DebugInfo/LogicalView/Readers/LVDWARFReader.cpp
index 7ff96ae90a7fd..f0b9da8c3fe80 100644
--- a/llvm/lib/DebugInfo/LogicalView/Readers/LVDWARFReader.cpp
+++ b/llvm/lib/DebugInfo/LogicalView/Readers/LVDWARFReader.cpp
@@ -522,14 +522,16 @@ void LVDWARFReader::createLineAndFileRecords(
     for (const DWARFDebugLine::FileNameEntry &Entry :
          Lines->Prologue.FileNames) {
       std::string Directory;
-      if (Lines->getDirectoryForEntry(Entry, Directory))
-        Directory = transformPath(Directory);
+      Lines->getDirectoryForEntry(Entry, Directory);
       if (Directory.empty())
         Directory = std::string(CompileUnit->getCompilationDirectory());
-      std::string File = transformPath(dwarf::toStringRef(Entry.Name));
+      // Take just the filename component, as it may be contain the full
+      // path that would be added to the already existing path in Directory.
+      StringRef File =
+          llvm::sys::path::filename(dwarf::toStringRef(Entry.Name));
       std::string String;
       raw_string_ostream(String) << Directory << "/" << File;
-      CompileUnit->addFilename(String);
+      CompileUnit->addFilename(transformPath(String));
     }
 
   // In DWARF5 the file indexes start at 0;
diff --git a/llvm/test/tools/llvm-debuginfo-analyzer/DWARF/dwarf-duplicated-directory-path.ll b/llvm/test/tools/llvm-debuginfo-analyzer/DWARF/dwarf-duplicated-directory-path.ll
new file mode 100644
index 0000000000000..87b383b16c721
--- /dev/null
+++ b/llvm/test/tools/llvm-debuginfo-analyzer/DWARF/dwarf-duplicated-directory-path.ll
@@ -0,0 +1,67 @@
+; REQUIRES: x86-registered-target
+
+; When the 'filename' field in DIFile entry contains the full pathname, such as:
+;
+; !1 = !DIFile(filename: "path/test.cpp", directory: "path")
+;
+; llvm-debuginfo-analyzer displays incorrect information for the directories
+; extracted from the '.debug_line' section, when using '--attribute=directories'
+;
+; llvm-debuginfo-analyzer --attribute=directories --print=scopes test.o
+;
+; {CompileUnit} 'test.cpp'
+;   {Directory} 'path/path'  <------- Duplicated 'path'
+
+; The test case was produced by the following steps:
+; // test.cpp
+; void foo() {
+; }
+
+; 1) clang++ --target=x86_64-pc-linux-gnu -Xclang -disable-O0-optnone
+;            -Xclang -disable-llvm-passes -fno-discard-value-names
+;            -emit-llvm -S -g -O0 test.cpp -o test.ll
+;
+; 2) Manually adding directory information to the DIFile 'filename' field:
+;
+; !1 = !DIFile(filename: "test.cpp", directory: "path")
+; -->
+; !1 = !DIFile(filename: "path/test.cpp", directory: "path")
+
+; RUN: llc --filetype=obj %p/dwarf-duplicated-directory-path.ll \
+; RUN:                    -o %t.dwarf-duplicated-directory-path.o
+
+; RUN: llvm-debuginfo-analyzer --attribute=directories,files               \
+; RUN:                         --print=scopes                              \
+; RUN:                         %t.dwarf-duplicated-directory-path.o 2>&1 | \
+; RUN: FileCheck --strict-whitespace -check-prefix=ONE %s
+
+; ONE: Logical View:
+; ONE-NEXT:           {File} '{{.*}}dwarf-duplicated-directory-path.ll{{.*}}'
+; ONE-EMPTY:
+; ONE-NEXT:             {CompileUnit} 'test.cpp'
+; ONE-NEXT:               {Directory} 'path'
+; ONE-NEXT:               {File} 'test.cpp'
+; ONE-NEXT:     1         {Function} extern not_inlined 'foo' -> 'void'
+
+source_filename = "test.cpp"
+target triple = "x86_64-pc-linux-gnu"
+
+define dso_local void @_Z3foov() !dbg !10 {
+entry:
+  ret void, !dbg !13
+}
+
+!llvm.dbg.cu = !{!0}
+!llvm.module.flags = !{!2, !3, !4}
+!llvm.ident = !{!9}
+
+!0 = distinct !DICompileUnit(language: DW_LANG_C_plus_plus_14, file: !1, producer: "clang", isOptimized: false, runtimeVersion: 0, emissionKind: FullDebug, splitDebugInlining: false, nameTableKind: None)
+!1 = !DIFile(filename: "path/test.cpp", directory: "path")
+!2 = !{i32 7, !"Dwarf Version", i32 5}
+!3 = !{i32 2, !"Debug Info Version", i32 3}
+!4 = !{i32 1, !"wchar_size", i32 4}
+!9 = !{!"clang"}
+!10 = distinct !DISubprogram(name: "foo", scope: !1, file: !1, line: 1, type: !11, scopeLine: 1, flags: DIFlagPrototyped, spFlags: DISPFlagDefinition, unit: !0)
+!11 = !DISubroutineType(types: !12)
+!12 = !{null}
+!13 = !DILocation(line: 2, column: 1, scope: !10)

@llvmbot
Copy link
Member

llvmbot commented Jun 12, 2025

@llvm/pr-subscribers-llvm-binary-utilities

Author: Carlos Alberto Enciso (CarlosAlbertoEnciso)

Changes

If any 'name' entry in the 'file_names' table from the .debug_line
section contains directory information, the llvm-debuginfo-analyzer
tool will display duplicated directory information when using the
'--attribute=directories' option.

.debug_line contents:
Line table prologue:
...
include_directories[ 0] = "path"
file_names[ 0]:
name: "path/test.cpp"
dir_index: 0
...

the '--attribute=directories' will show:

{Directory} 'path/path' <-- Duplicated 'path'


Full diff: https://github.com/llvm/llvm-project/pull/143849.diff

2 Files Affected:

  • (modified) llvm/lib/DebugInfo/LogicalView/Readers/LVDWARFReader.cpp (+6-4)
  • (added) llvm/test/tools/llvm-debuginfo-analyzer/DWARF/dwarf-duplicated-directory-path.ll (+67)
diff --git a/llvm/lib/DebugInfo/LogicalView/Readers/LVDWARFReader.cpp b/llvm/lib/DebugInfo/LogicalView/Readers/LVDWARFReader.cpp
index 7ff96ae90a7fd..f0b9da8c3fe80 100644
--- a/llvm/lib/DebugInfo/LogicalView/Readers/LVDWARFReader.cpp
+++ b/llvm/lib/DebugInfo/LogicalView/Readers/LVDWARFReader.cpp
@@ -522,14 +522,16 @@ void LVDWARFReader::createLineAndFileRecords(
     for (const DWARFDebugLine::FileNameEntry &Entry :
          Lines->Prologue.FileNames) {
       std::string Directory;
-      if (Lines->getDirectoryForEntry(Entry, Directory))
-        Directory = transformPath(Directory);
+      Lines->getDirectoryForEntry(Entry, Directory);
       if (Directory.empty())
         Directory = std::string(CompileUnit->getCompilationDirectory());
-      std::string File = transformPath(dwarf::toStringRef(Entry.Name));
+      // Take just the filename component, as it may be contain the full
+      // path that would be added to the already existing path in Directory.
+      StringRef File =
+          llvm::sys::path::filename(dwarf::toStringRef(Entry.Name));
       std::string String;
       raw_string_ostream(String) << Directory << "/" << File;
-      CompileUnit->addFilename(String);
+      CompileUnit->addFilename(transformPath(String));
     }
 
   // In DWARF5 the file indexes start at 0;
diff --git a/llvm/test/tools/llvm-debuginfo-analyzer/DWARF/dwarf-duplicated-directory-path.ll b/llvm/test/tools/llvm-debuginfo-analyzer/DWARF/dwarf-duplicated-directory-path.ll
new file mode 100644
index 0000000000000..87b383b16c721
--- /dev/null
+++ b/llvm/test/tools/llvm-debuginfo-analyzer/DWARF/dwarf-duplicated-directory-path.ll
@@ -0,0 +1,67 @@
+; REQUIRES: x86-registered-target
+
+; When the 'filename' field in DIFile entry contains the full pathname, such as:
+;
+; !1 = !DIFile(filename: "path/test.cpp", directory: "path")
+;
+; llvm-debuginfo-analyzer displays incorrect information for the directories
+; extracted from the '.debug_line' section, when using '--attribute=directories'
+;
+; llvm-debuginfo-analyzer --attribute=directories --print=scopes test.o
+;
+; {CompileUnit} 'test.cpp'
+;   {Directory} 'path/path'  <------- Duplicated 'path'
+
+; The test case was produced by the following steps:
+; // test.cpp
+; void foo() {
+; }
+
+; 1) clang++ --target=x86_64-pc-linux-gnu -Xclang -disable-O0-optnone
+;            -Xclang -disable-llvm-passes -fno-discard-value-names
+;            -emit-llvm -S -g -O0 test.cpp -o test.ll
+;
+; 2) Manually adding directory information to the DIFile 'filename' field:
+;
+; !1 = !DIFile(filename: "test.cpp", directory: "path")
+; -->
+; !1 = !DIFile(filename: "path/test.cpp", directory: "path")
+
+; RUN: llc --filetype=obj %p/dwarf-duplicated-directory-path.ll \
+; RUN:                    -o %t.dwarf-duplicated-directory-path.o
+
+; RUN: llvm-debuginfo-analyzer --attribute=directories,files               \
+; RUN:                         --print=scopes                              \
+; RUN:                         %t.dwarf-duplicated-directory-path.o 2>&1 | \
+; RUN: FileCheck --strict-whitespace -check-prefix=ONE %s
+
+; ONE: Logical View:
+; ONE-NEXT:           {File} '{{.*}}dwarf-duplicated-directory-path.ll{{.*}}'
+; ONE-EMPTY:
+; ONE-NEXT:             {CompileUnit} 'test.cpp'
+; ONE-NEXT:               {Directory} 'path'
+; ONE-NEXT:               {File} 'test.cpp'
+; ONE-NEXT:     1         {Function} extern not_inlined 'foo' -> 'void'
+
+source_filename = "test.cpp"
+target triple = "x86_64-pc-linux-gnu"
+
+define dso_local void @_Z3foov() !dbg !10 {
+entry:
+  ret void, !dbg !13
+}
+
+!llvm.dbg.cu = !{!0}
+!llvm.module.flags = !{!2, !3, !4}
+!llvm.ident = !{!9}
+
+!0 = distinct !DICompileUnit(language: DW_LANG_C_plus_plus_14, file: !1, producer: "clang", isOptimized: false, runtimeVersion: 0, emissionKind: FullDebug, splitDebugInlining: false, nameTableKind: None)
+!1 = !DIFile(filename: "path/test.cpp", directory: "path")
+!2 = !{i32 7, !"Dwarf Version", i32 5}
+!3 = !{i32 2, !"Debug Info Version", i32 3}
+!4 = !{i32 1, !"wchar_size", i32 4}
+!9 = !{!"clang"}
+!10 = distinct !DISubprogram(name: "foo", scope: !1, file: !1, line: 1, type: !11, scopeLine: 1, flags: DIFlagPrototyped, spFlags: DISPFlagDefinition, unit: !0)
+!11 = !DISubroutineType(types: !12)
+!12 = !{null}
+!13 = !DILocation(line: 2, column: 1, scope: !10)

@CarlosAlbertoEnciso
Copy link
Member Author

This problem was discovered during the IR reader development (#135440), while comparing canonical views generated from IR and object files.

Copy link
Contributor

@jalopezg-git jalopezg-git left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thank you, @CarlosAlbertoEnciso! Looks clean; just one comment (see above).

Comment on lines +528 to +531
// Take just the filename component, as it may be contain the full
// path that would be added to the already existing path in Directory.
StringRef File =
llvm::sys::path::filename(dwarf::toStringRef(Entry.Name));
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Could it happen that the filename part of a !DIFile does not have the directory part as a prefix? E.g.

!1 = !DIFile(filename: "foo/test.cpp", directory: "bar")

Should this be interpreted as bar/foo/test.cpp instead?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants