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/X86/mem2reg_fp80.ll

58 lines
2.7 KiB
LLVM
Raw Normal View History

Re-apply "[DebugInfo] Check size of variable in ConvertDebugDeclareToDebugValue" This is r334704 (which was reverted in r334732) with a fix for types like x86_fp80. We need to use getTypeAllocSizeInBits and not getTypeStoreSizeInBits to avoid dropping debug info for such types. Original commit msg: > Summary: > Do not convert a DbgDeclare to DbgValue if the store > instruction only refer to a fragment of the variable > described by the DbgDeclare. > > Problem was seen when for example having an alloca for an > array or struct, and there were stores to individual elements. > In the past we inserted a DbgValue intrinsics for each store, > just as if the store wrote the whole variable. > > When handling store instructions we insert a DbgValue that > indicates that the variable is "undefined", as we do not know > which part of the variable that is updated by the store. > > When ConvertDebugDeclareToDebugValue is used with a load/phi > instruction we assert that the referenced value is large enough > to cover the whole variable. Afaict this should be true for all > scenarios where those methods are used on trunk. If the assert > blows in the future I guess we could simply skip to insert a > dbg.value instruction. > > In the future I think we should examine which part of the variable > that is accessed, and add a DbgValue instrinsic with an appropriate > DW_OP_LLVM_fragment expression. > > Reviewers: dblaikie, aprantl, rnk > > Reviewed By: aprantl > > Subscribers: JDevlieghere, llvm-commits > > Tags: #debug-info > > Differential Revision: https://reviews.llvm.org/D48024 llvm-svn: 334830
2018-06-15 21:48:55 +08:00
; RUN: opt < %s -mem2reg -S | FileCheck %s
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnu"
define dso_local x86_fp80 @powixf2() !dbg !1 {
entry:
%r = alloca x86_fp80, align 16
call void @llvm.dbg.declare(metadata x86_fp80* %r, metadata !14, metadata !DIExpression()), !dbg !15
br i1 undef, label %if.then, label %if.end, !dbg !16
if.then: ; preds = %entry
NFC - Various typo fixes in tests llvm-svn: 336268
2018-07-04 21:28:39 +08:00
; CHECK-LABEL: if.then:
Re-apply "[DebugInfo] Check size of variable in ConvertDebugDeclareToDebugValue" This is r334704 (which was reverted in r334732) with a fix for types like x86_fp80. We need to use getTypeAllocSizeInBits and not getTypeStoreSizeInBits to avoid dropping debug info for such types. Original commit msg: > Summary: > Do not convert a DbgDeclare to DbgValue if the store > instruction only refer to a fragment of the variable > described by the DbgDeclare. > > Problem was seen when for example having an alloca for an > array or struct, and there were stores to individual elements. > In the past we inserted a DbgValue intrinsics for each store, > just as if the store wrote the whole variable. > > When handling store instructions we insert a DbgValue that > indicates that the variable is "undefined", as we do not know > which part of the variable that is updated by the store. > > When ConvertDebugDeclareToDebugValue is used with a load/phi > instruction we assert that the referenced value is large enough > to cover the whole variable. Afaict this should be true for all > scenarios where those methods are used on trunk. If the assert > blows in the future I guess we could simply skip to insert a > dbg.value instruction. > > In the future I think we should examine which part of the variable > that is accessed, and add a DbgValue instrinsic with an appropriate > DW_OP_LLVM_fragment expression. > > Reviewers: dblaikie, aprantl, rnk > > Reviewed By: aprantl > > Subscribers: JDevlieghere, llvm-commits > > Tags: #debug-info > > Differential Revision: https://reviews.llvm.org/D48024 llvm-svn: 334830
2018-06-15 21:48:55 +08:00
; CHECK: %mul = fmul x86_fp80
; CHECK: call void @llvm.dbg.value(metadata x86_fp80 %mul, metadata {{.*}}, metadata !DIExpression())
%mul = fmul x86_fp80 undef, undef, !dbg !18
store x86_fp80 %mul, x86_fp80* %r, align 16, !dbg !18
br label %if.end, !dbg !20
if.end: ; preds = %if.then, %entry
NFC - Various typo fixes in tests llvm-svn: 336268
2018-07-04 21:28:39 +08:00
; CHECK-LABEL: if.end:
Re-apply "[DebugInfo] Check size of variable in ConvertDebugDeclareToDebugValue" This is r334704 (which was reverted in r334732) with a fix for types like x86_fp80. We need to use getTypeAllocSizeInBits and not getTypeStoreSizeInBits to avoid dropping debug info for such types. Original commit msg: > Summary: > Do not convert a DbgDeclare to DbgValue if the store > instruction only refer to a fragment of the variable > described by the DbgDeclare. > > Problem was seen when for example having an alloca for an > array or struct, and there were stores to individual elements. > In the past we inserted a DbgValue intrinsics for each store, > just as if the store wrote the whole variable. > > When handling store instructions we insert a DbgValue that > indicates that the variable is "undefined", as we do not know > which part of the variable that is updated by the store. > > When ConvertDebugDeclareToDebugValue is used with a load/phi > instruction we assert that the referenced value is large enough > to cover the whole variable. Afaict this should be true for all > scenarios where those methods are used on trunk. If the assert > blows in the future I guess we could simply skip to insert a > dbg.value instruction. > > In the future I think we should examine which part of the variable > that is accessed, and add a DbgValue instrinsic with an appropriate > DW_OP_LLVM_fragment expression. > > Reviewers: dblaikie, aprantl, rnk > > Reviewed By: aprantl > > Subscribers: JDevlieghere, llvm-commits > > Tags: #debug-info > > Differential Revision: https://reviews.llvm.org/D48024 llvm-svn: 334830
2018-06-15 21:48:55 +08:00
; CHECK: %r.0 = phi x86_fp80
; CHECK: call void @llvm.dbg.value(metadata x86_fp80 %r.0, metadata {{.*}}, metadata !DIExpression())
%out = load x86_fp80, x86_fp80* %r, align 16, !dbg !21
ret x86_fp80 %out, !dbg !22
}
declare void @llvm.dbg.declare(metadata, metadata, metadata) #0
attributes #0 = { nounwind readnone speculatable }
!llvm.dbg.cu = !{}
!llvm.module.flags = !{!0}
!0 = !{i32 2, !"Debug Info Version", i32 3}
!1 = distinct !DISubprogram(name: "__powixf2", scope: !2, file: !2, line: 22, type: !3, isLocal: false, isDefinition: true, scopeLine: 23, flags: DIFlagPrototyped, isOptimized: true, unit: !11, retainedNodes: !13)
!2 = !DIFile(filename: "powixf2.c", directory: "")
!3 = !DISubroutineType(types: !4)
!4 = !{!5, !5, !6}
!5 = !DIBasicType(name: "long double", size: 128, encoding: DW_ATE_float)
!6 = !DIDerivedType(tag: DW_TAG_typedef, name: "si_int", file: !7, line: 28, baseType: !8)
!7 = !DIFile(filename: "int_types.h", directory: "")
!8 = !DIDerivedType(tag: DW_TAG_typedef, name: "int32_t", file: !9, line: 39, baseType: !10)
!9 = !DIFile(filename: "/usr/include/stdint.h", directory: "")
!10 = !DIBasicType(name: "int", size: 32, encoding: DW_ATE_signed)
!11 = distinct !DICompileUnit(language: DW_LANG_C99, file: !2, producer: "clang version 7.0.0 () ()", isOptimized: true, runtimeVersion: 0, emissionKind: FullDebug, enums: !12)
!12 = !{}
!13 = !{!14}
!14 = !DILocalVariable(name: "r", scope: !1, file: !2, line: 25, type: !5)
!15 = !DILocation(line: 25, column: 17, scope: !1)
!16 = !DILocation(line: 28, column: 13, scope: !17)
!17 = distinct !DILexicalBlock(scope: !1, file: !2, line: 27, column: 5)
!18 = !DILocation(line: 29, column: 15, scope: !19)
!19 = distinct !DILexicalBlock(scope: !17, file: !2, line: 28, column: 13)
!20 = !DILocation(line: 29, column: 13, scope: !19)
!21 = !DILocation(line: 35, column: 22, scope: !1)
!22 = !DILocation(line: 35, column: 5, scope: !1)