maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/DebugInfo/COFF/inlining-same-name.ll

58 lines
2.6 KiB
LLVM
Raw Normal View History

Modify emitTypeInformation to use MemoryTypeTableBuilder, take 2 This effectively revers commit r270389 and re-lands r270106, but it's almost a rewrite. The behavior change in r270106 was that we could no longer assume that each LF_FUNC_ID record got its own type index. This patch adds a map from DINode* to TypeIndex, so we can stop making that assumption. This change also emits padding bytes between type records similar to the way MSVC does. The size of the type record includes the padding bytes. llvm-svn: 270485
2016-05-24 04:23:46 +08:00
; RUN: llc -mtriple=x86_64-windows-msvc < %s -filetype=obj -o - | llvm-readobj - -codeview | FileCheck %s
; We should only get one func id record, and both inlinees should point to it,
; even though there are two DISubprograms.
; CHECK: FuncId (0x1002) {
; CHECK-NEXT: TypeLeafKind: LF_FUNC_ID (0x1601)
; CHECK-NEXT: ParentScope: 0x0
; CHECK-NEXT: FunctionType: void () (0x1001)
; CHECK-NEXT: Name: same_name
; CHECK-NEXT: }
; CHECK-NOT: Name: same_name
; CHECK: CodeViewDebugInfo [
; CHECK: Section: .debug$S
; CHECK: Subsection [
; CHECK: ProcStart {
; CHECK: DisplayName: main
; CHECK: }
; CHECK: InlineSite {
; CHECK: Inlinee: same_name (0x1002)
; CHECK: }
; CHECK: InlineSiteEnd {
; CHECK: }
; CHECK: InlineSite {
; CHECK: Inlinee: same_name (0x1002)
; CHECK: }
; CHECK: InlineSiteEnd {
; CHECK: }
; CHECK: ProcEnd
; CHECK: ]
target datalayout = "e-m:w-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-pc-windows-msvc"
[CodeView] Don't crash on functions without subprograms A function may have instructions annotated with debug info without having a subprogram. This fixes PR28747. llvm-svn: 276956
2016-07-28 13:03:22 +08:00
define void @main(i32* %i.i) !dbg !16 {
Modify emitTypeInformation to use MemoryTypeTableBuilder, take 2 This effectively revers commit r270389 and re-lands r270106, but it's almost a rewrite. The behavior change in r270106 was that we could no longer assume that each LF_FUNC_ID record got its own type index. This patch adds a map from DINode* to TypeIndex, so we can stop making that assumption. This change also emits padding bytes between type records similar to the way MSVC does. The size of the type record includes the padding bytes. llvm-svn: 270485
2016-05-24 04:23:46 +08:00
store volatile i32 3, i32* %i.i, !dbg !6
store volatile i32 3, i32* %i.i, !dbg !19
ret void
}
!llvm.module.flags = !{!0, !1, !2}
!llvm.dbg.cu = !{!4}
!0 = !{i32 2, !"CodeView", i32 1}
!1 = !{i32 2, !"Debug Info Version", i32 3}
!2 = !{i32 6, !"Linker Options", !{}}
!4 = distinct !DICompileUnit(language: DW_LANG_D, file: !5, producer: "LDC (http://wiki.dlang.org/LDC)", isOptimized: false, runtimeVersion: 1, emissionKind: FullDebug)
!5 = !DIFile(filename: "opover2.d", directory: "C:\5CLDC\5Cninja-ldc\5C..\5Cldc\5Ctests\5Cd2\5Cdmd-testsuite\5Crunnable")
!6 = !DILocation(line: 302, column: 9, scope: !7, inlinedAt: !15)
!7 = distinct !DISubprogram(name: "same_name", linkageName: "same_name", scope: null, file: !5, line: 302, type: !8, isLocal: false, isDefinition: true, scopeLine: 302, flags: DIFlagPrototyped, isOptimized: false, unit: !4, variables: !{})
!8 = !DISubroutineType(types: !{})
!15 = distinct !DILocation(line: 333, column: 5, scope: !16)
!16 = distinct !DISubprogram(name: "main", linkageName: "main", scope: null, file: !5, line: 328, type: !8, isLocal: false, isDefinition: true, scopeLine: 328, flags: DIFlagPrototyped, isOptimized: false, unit: !4, variables: !{})
!19 = !DILocation(line: 308, column: 9, scope: !20, inlinedAt: !25)
!20 = distinct !DISubprogram(name: "same_name", linkageName: "same_name", scope: null, file: !5, line: 308, type: !8, isLocal: false, isDefinition: true, scopeLine: 308, flags: DIFlagPrototyped, isOptimized: false, unit: !4, variables: !{})
!25 = distinct !DILocation(line: 334, column: 5, scope: !16)