8bb827c0e6
With e8512786fedbfa6ddba70ceddc29d7122173ba5e the for loop that iterates
over MDNode operands was changed to a range-based for loop. This change
surfaces a bug where if the result of MD->operands() is an ArrayRef that
has a size of 0, then iterating over that ArrayRef leads to a
segmentation fault, due to accessing invalid addresses.
This was reverted with 6ce03ff3fef8fb6fa9afe8eb22c6d98bced26d48 but this
test should be added to test that codepath in the future.
26 lines
1.3 KiB
LLVM
26 lines
1.3 KiB
LLVM
; RUN: llvm-as -disable-output %s
|
|
%"class.llvm::StringRef" = type { ptr, i64 }
|
|
define internal void @_ZL30tokenizeWindowsCommandLineImplN4llvm9StringRefERNS_11StringSaverENS_12function_refIFvS0_EEEbNS3_IFvvEEEb() !dbg !12 {
|
|
%7 = alloca %"class.llvm::StringRef", align 8
|
|
%21 = call noundef i64 @_ZNK4llvm9StringRef4sizeEv(ptr noundef nonnull align 8 dereferenceable(16) %7)
|
|
br label %22
|
|
br label %22, !llvm.loop !284 ; This instruction has loop metadata but no operands and should not result in a segmentation fault in the verifier.
|
|
}
|
|
define linkonce_odr noundef i64 @_ZNK4llvm9StringRef4sizeEv() align 2 !dbg !340 {
|
|
%2 = alloca ptr, align 8
|
|
%3 = load ptr, ptr %2, align 8
|
|
%4 = getelementptr inbounds %"class.llvm::StringRef", ptr %3
|
|
%5 = load i64, ptr %4
|
|
ret i64 %5
|
|
}
|
|
!llvm.module.flags = !{!2, !6}
|
|
!llvm.dbg.cu = !{!7}
|
|
!2 = !{i32 2, !"Debug Info Version", i32 3}
|
|
!6 = !{i32 7, !"frame-pointer", i32 1}
|
|
!7 = distinct !DICompileUnit(language: DW_LANG_C_plus_plus_14, file: !8, sdk: "MacOSX14.0.sdk")
|
|
!8 = !DIFile(filename: "file.cpp", directory: "/Users/Dev", checksumkind: CSK_MD5, checksum: "ed7ae158f20f7914bc5fb843291e80da")
|
|
!12 = distinct !DISubprogram(unit: !7, retainedNodes: !36)
|
|
!36 = !{}
|
|
!284 = distinct !{}
|
|
!340 = distinct !DISubprogram(unit: !7, retainedNodes: !36)
|