[Debugify] Diagnose mis-sized dbg.values
Report an error in -check-debugify when the size of a dbg.value operand
doesn't match up with the size of the variable it describes.
Eventually this check should be moved into the IR verifier. For the
moment, it's useful to include the check in -check-debugify as a means
of catching regressions and finding existing bugs.
Here are some instances of bugs the new check finds in the -O2 pipeline
(all in InstCombine):
1) A float is used where a double is expected:
ERROR: dbg.value operand has size 32, but its variable has size 64:
call void @llvm.dbg.value(metadata float %expf, metadata !12, metadata
!DIExpression()), !dbg !15
2) An i8 is used where an i32 is expected:
ERROR: dbg.value operand has size 8, but its variable has size 32:
call void @llvm.dbg.value(metadata i8 %t4, metadata !14, metadata
!DIExpression()), !dbg !24
3) A <4 x i32> is used where something twice as large is expected
(perhaps a <4 x i64>, I haven't double-checked):
ERROR: dbg.value operand has size 128, but its variable has size 256:
call void @llvm.dbg.value(metadata <4 x i32> %4, metadata !40, metadata
!DIExpression()), !dbg !95
Differential Revision: https://reviews.llvm.org/D48408
llvm-svn: 335682
2018-06-26 22:46:41 +00:00
|
|
|
; RUN: opt -check-debugify < %s 2>&1 | FileCheck %s
|
|
|
|
|
|
|
|
|
|
define <2 x i64> @test-fun(<2 x i64> %A) !dbg !6 {
|
|
|
|
|
%and = and <2 x i64> %A, <i64 23, i64 42>, !dbg !14
|
|
|
|
|
|
|
|
|
|
; CHECK: ERROR: dbg.value operand has size 128, but its variable has size 256
|
|
|
|
|
call void @llvm.dbg.value(metadata <2 x i64> %and, metadata !12, metadata !DIExpression()), !dbg !15
|
|
|
|
|
|
|
|
|
|
; CHECK: ERROR: dbg.value operand has size 512, but its variable has size 256
|
|
|
|
|
call void @llvm.dbg.value(metadata <2 x i256> zeroinitializer, metadata !12, metadata !DIExpression()), !dbg !15
|
|
|
|
|
|
|
|
|
|
; CHECK: ERROR: dbg.value operand has size 8, but its variable has size 256
|
|
|
|
|
call void @llvm.dbg.value(metadata i8 0, metadata !12, metadata !DIExpression()), !dbg !15
|
|
|
|
|
|
|
|
|
|
; Assume the debugger implicitly uses the lower 256 bits.
|
|
|
|
|
; CHECK-NOT: ERROR: dbg.value operand has size 512, but its variable has size 256
|
|
|
|
|
call void @llvm.dbg.value(metadata i512 0, metadata !12, metadata !DIExpression()), !dbg !15
|
|
|
|
|
|
2018-07-06 17:32:40 +00:00
|
|
|
; Assume the debugger implicitly zero-extends unsigned values.
|
|
|
|
|
; CHECK-NOT: ERROR: dbg.value operand has size 4, but its variable has size 32
|
|
|
|
|
call void @llvm.dbg.value(metadata i8 0, metadata !17, metadata !DIExpression()), !dbg !15
|
|
|
|
|
|
[Debugify] Diagnose mis-sized dbg.values
Report an error in -check-debugify when the size of a dbg.value operand
doesn't match up with the size of the variable it describes.
Eventually this check should be moved into the IR verifier. For the
moment, it's useful to include the check in -check-debugify as a means
of catching regressions and finding existing bugs.
Here are some instances of bugs the new check finds in the -O2 pipeline
(all in InstCombine):
1) A float is used where a double is expected:
ERROR: dbg.value operand has size 32, but its variable has size 64:
call void @llvm.dbg.value(metadata float %expf, metadata !12, metadata
!DIExpression()), !dbg !15
2) An i8 is used where an i32 is expected:
ERROR: dbg.value operand has size 8, but its variable has size 32:
call void @llvm.dbg.value(metadata i8 %t4, metadata !14, metadata
!DIExpression()), !dbg !24
3) A <4 x i32> is used where something twice as large is expected
(perhaps a <4 x i64>, I haven't double-checked):
ERROR: dbg.value operand has size 128, but its variable has size 256:
call void @llvm.dbg.value(metadata <4 x i32> %4, metadata !40, metadata
!DIExpression()), !dbg !95
Differential Revision: https://reviews.llvm.org/D48408
llvm-svn: 335682
2018-06-26 22:46:41 +00:00
|
|
|
ret <2 x i64> %and, !dbg !16
|
|
|
|
|
}
|
|
|
|
|
; CHECK: CheckModuleDebugify: FAIL
|
|
|
|
|
|
|
|
|
|
; Function Attrs: nounwind readnone speculatable
|
|
|
|
|
declare void @llvm.dbg.value(metadata, metadata, metadata) #0
|
|
|
|
|
|
|
|
|
|
attributes #0 = { nounwind readnone speculatable }
|
|
|
|
|
|
|
|
|
|
!llvm.dbg.cu = !{!0}
|
|
|
|
|
!llvm.debugify = !{!3, !4}
|
|
|
|
|
!llvm.module.flags = !{!5}
|
|
|
|
|
|
|
|
|
|
!0 = distinct !DICompileUnit(language: DW_LANG_C, file: !1, producer: "debugify", isOptimized: true, runtimeVersion: 0, emissionKind: FullDebug, enums: !2)
|
|
|
|
|
!1 = !DIFile(filename: "/Users/vsk/src/llvm.org-master/llvm/test/DebugInfo/debugify-bogus-dbg-value.ll", directory: "/")
|
|
|
|
|
!2 = !{}
|
|
|
|
|
!3 = !{i32 4}
|
2018-07-06 17:32:40 +00:00
|
|
|
!4 = !{i32 4}
|
[Debugify] Diagnose mis-sized dbg.values
Report an error in -check-debugify when the size of a dbg.value operand
doesn't match up with the size of the variable it describes.
Eventually this check should be moved into the IR verifier. For the
moment, it's useful to include the check in -check-debugify as a means
of catching regressions and finding existing bugs.
Here are some instances of bugs the new check finds in the -O2 pipeline
(all in InstCombine):
1) A float is used where a double is expected:
ERROR: dbg.value operand has size 32, but its variable has size 64:
call void @llvm.dbg.value(metadata float %expf, metadata !12, metadata
!DIExpression()), !dbg !15
2) An i8 is used where an i32 is expected:
ERROR: dbg.value operand has size 8, but its variable has size 32:
call void @llvm.dbg.value(metadata i8 %t4, metadata !14, metadata
!DIExpression()), !dbg !24
3) A <4 x i32> is used where something twice as large is expected
(perhaps a <4 x i64>, I haven't double-checked):
ERROR: dbg.value operand has size 128, but its variable has size 256:
call void @llvm.dbg.value(metadata <4 x i32> %4, metadata !40, metadata
!DIExpression()), !dbg !95
Differential Revision: https://reviews.llvm.org/D48408
llvm-svn: 335682
2018-06-26 22:46:41 +00:00
|
|
|
!5 = !{i32 2, !"Debug Info Version", i32 3}
|
|
|
|
|
!6 = distinct !DISubprogram(name: "test-fun", linkageName: "test-fun", scope: null, file: !1, line: 1, type: !7, isLocal: false, isDefinition: true, scopeLine: 1, isOptimized: true, unit: !0, retainedNodes: !8)
|
|
|
|
|
!7 = !DISubroutineType(types: !2)
|
|
|
|
|
!8 = !{!9, !11, !12}
|
|
|
|
|
!9 = !DILocalVariable(name: "1", scope: !6, file: !1, line: 1, type: !10)
|
2018-07-06 17:32:40 +00:00
|
|
|
!10 = !DIBasicType(name: "ty64", size: 64, encoding: DW_ATE_signed)
|
[Debugify] Diagnose mis-sized dbg.values
Report an error in -check-debugify when the size of a dbg.value operand
doesn't match up with the size of the variable it describes.
Eventually this check should be moved into the IR verifier. For the
moment, it's useful to include the check in -check-debugify as a means
of catching regressions and finding existing bugs.
Here are some instances of bugs the new check finds in the -O2 pipeline
(all in InstCombine):
1) A float is used where a double is expected:
ERROR: dbg.value operand has size 32, but its variable has size 64:
call void @llvm.dbg.value(metadata float %expf, metadata !12, metadata
!DIExpression()), !dbg !15
2) An i8 is used where an i32 is expected:
ERROR: dbg.value operand has size 8, but its variable has size 32:
call void @llvm.dbg.value(metadata i8 %t4, metadata !14, metadata
!DIExpression()), !dbg !24
3) A <4 x i32> is used where something twice as large is expected
(perhaps a <4 x i64>, I haven't double-checked):
ERROR: dbg.value operand has size 128, but its variable has size 256:
call void @llvm.dbg.value(metadata <4 x i32> %4, metadata !40, metadata
!DIExpression()), !dbg !95
Differential Revision: https://reviews.llvm.org/D48408
llvm-svn: 335682
2018-06-26 22:46:41 +00:00
|
|
|
!11 = !DILocalVariable(name: "2", scope: !6, file: !1, line: 2, type: !10)
|
|
|
|
|
!12 = !DILocalVariable(name: "3", scope: !6, file: !1, line: 3, type: !13)
|
|
|
|
|
|
|
|
|
|
; Set the size here to an incorrect value to check that the size diagnostic
|
|
|
|
|
; triggers.
|
2018-07-06 17:32:40 +00:00
|
|
|
!13 = !DIBasicType(name: "ty128", size: 256, encoding: DW_ATE_signed)
|
[Debugify] Diagnose mis-sized dbg.values
Report an error in -check-debugify when the size of a dbg.value operand
doesn't match up with the size of the variable it describes.
Eventually this check should be moved into the IR verifier. For the
moment, it's useful to include the check in -check-debugify as a means
of catching regressions and finding existing bugs.
Here are some instances of bugs the new check finds in the -O2 pipeline
(all in InstCombine):
1) A float is used where a double is expected:
ERROR: dbg.value operand has size 32, but its variable has size 64:
call void @llvm.dbg.value(metadata float %expf, metadata !12, metadata
!DIExpression()), !dbg !15
2) An i8 is used where an i32 is expected:
ERROR: dbg.value operand has size 8, but its variable has size 32:
call void @llvm.dbg.value(metadata i8 %t4, metadata !14, metadata
!DIExpression()), !dbg !24
3) A <4 x i32> is used where something twice as large is expected
(perhaps a <4 x i64>, I haven't double-checked):
ERROR: dbg.value operand has size 128, but its variable has size 256:
call void @llvm.dbg.value(metadata <4 x i32> %4, metadata !40, metadata
!DIExpression()), !dbg !95
Differential Revision: https://reviews.llvm.org/D48408
llvm-svn: 335682
2018-06-26 22:46:41 +00:00
|
|
|
|
|
|
|
|
!14 = !DILocation(line: 2, column: 1, scope: !6)
|
|
|
|
|
!15 = !DILocation(line: 3, column: 1, scope: !6)
|
|
|
|
|
!16 = !DILocation(line: 4, column: 1, scope: !6)
|
2018-07-06 17:32:40 +00:00
|
|
|
|
|
|
|
|
!17 = !DILocalVariable(name: "4", scope: !6, file: !1, line: 3, type: !18)
|
|
|
|
|
!18 = !DIBasicType(name: "ty32_unsigned", size: 32, encoding: DW_ATE_unsigned)
|