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/rematerialize.ll

97 lines
4.5 KiB
LLVM
Raw Normal View History

When the inline spiller rematerializes an instruction, take the debug location from the instruction that immediately follows the rematerialization point. Patch by Andrea DiBiagio. Differential Revision: http://reviews.llvm.org/D23539 llvm-svn: 278822
2016-08-17 01:12:50 +08:00
; REQUIRES: object-emission
; RUN: llc -O2 -filetype=obj -mtriple=x86_64-unknown-linux-gnu < %s \
llvm-dwarfdump: Replace -debug-dump=sect option with individual options. As discussed on llvm-dev in http://lists.llvm.org/pipermail/llvm-dev/2017-September/117301.html this changes the command line interface of llvm-dwarfdump to match the one used by the dwarfdump utility shipping on macOS. In addition to being shorter to type this format also has the advantage of allowing more than one section to be specified at the same time. In a nutshell, with this change $ llvm-dwarfdump --debug-dump=info $ llvm-dwarfdump --debug-dump=apple-objc becomes $ dwarfdump --debug-info --apple-objc Differential Revision: https://reviews.llvm.org/D37714 llvm-svn: 312970
2017-09-12 06:59:45 +08:00
; RUN: | llvm-dwarfdump -debug-line - | FileCheck %s
When the inline spiller rematerializes an instruction, take the debug location from the instruction that immediately follows the rematerialization point. Patch by Andrea DiBiagio. Differential Revision: http://reviews.llvm.org/D23539 llvm-svn: 278822
2016-08-17 01:12:50 +08:00
;
; Generated with clang -O2 -g from
;
; typedef float __m128 __attribute__((__vector_size__(16)));
;
; extern __m128 doSomething(__m128, __m128);
;
;
; __m128 foo(__m128 X) { // line 6
; const __m128 V = {0.5f, 0.5f, 0.5f, 0.5f}; // line 7
; __m128 Sub = X - V; // line 8
; __m128 Add = X + V; // line 9
;
; __m128 Result = doSomething(Add, Sub); // line 11
;
; return V - Result; // line 13
; }
;
;
; We want to see line 13 after line 11 without any other line in between.
; CHECK: 0x{{[0-9a-f]*}} 11
; CHECK-NOT: 0x{{[0-9a-f]*}} 8
; CHECK-NOT: 0x{{[0-9a-f]*}} 9
; CHECK: 0x{{[0-9a-f]*}} 13
; CHECK-NOT: 0x{{[0-9a-f]*}} 8
; CHECK-NOT: 0x{{[0-9a-f]*}} 9
; ModuleID = 'test.c'
source_filename = "test.c"
; target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
; target triple = "x86_64-unknown-linux-gnu"
; Function Attrs: nounwind uwtable
define <4 x float> @foo(<4 x float> %X) local_unnamed_addr #0 !dbg !6 {
entry:
Remove the obsolete offset parameter from @llvm.dbg.value There is no situation where this rarely-used argument cannot be substituted with a DIExpression and removing it allows us to simplify the DWARF backend. Note that this patch does not yet remove any of the newly dead code. rdar://problem/33580047 Differential Revision: https://reviews.llvm.org/D35951 llvm-svn: 309426
2017-07-29 04:21:02 +08:00
tail call void @llvm.dbg.value(metadata <4 x float> %X, metadata !15, metadata !21), !dbg !22
tail call void @llvm.dbg.value(metadata <4 x float> <float 5.000000e-01, float 5.000000e-01, float 5.000000e-01, float 5.000000e-01>, metadata !16, metadata !21), !dbg !23
When the inline spiller rematerializes an instruction, take the debug location from the instruction that immediately follows the rematerialization point. Patch by Andrea DiBiagio. Differential Revision: http://reviews.llvm.org/D23539 llvm-svn: 278822
2016-08-17 01:12:50 +08:00
%sub = fadd <4 x float> %X, <float -5.000000e-01, float -5.000000e-01, float -5.000000e-01, float -5.000000e-01>, !dbg !24
Remove the obsolete offset parameter from @llvm.dbg.value There is no situation where this rarely-used argument cannot be substituted with a DIExpression and removing it allows us to simplify the DWARF backend. Note that this patch does not yet remove any of the newly dead code. rdar://problem/33580047 Differential Revision: https://reviews.llvm.org/D35951 llvm-svn: 309426
2017-07-29 04:21:02 +08:00
tail call void @llvm.dbg.value(metadata <4 x float> %sub, metadata !18, metadata !21), !dbg !25
When the inline spiller rematerializes an instruction, take the debug location from the instruction that immediately follows the rematerialization point. Patch by Andrea DiBiagio. Differential Revision: http://reviews.llvm.org/D23539 llvm-svn: 278822
2016-08-17 01:12:50 +08:00
%add = fadd <4 x float> %X, <float 5.000000e-01, float 5.000000e-01, float 5.000000e-01, float 5.000000e-01>, !dbg !26
Remove the obsolete offset parameter from @llvm.dbg.value There is no situation where this rarely-used argument cannot be substituted with a DIExpression and removing it allows us to simplify the DWARF backend. Note that this patch does not yet remove any of the newly dead code. rdar://problem/33580047 Differential Revision: https://reviews.llvm.org/D35951 llvm-svn: 309426
2017-07-29 04:21:02 +08:00
tail call void @llvm.dbg.value(metadata <4 x float> %add, metadata !19, metadata !21), !dbg !27
When the inline spiller rematerializes an instruction, take the debug location from the instruction that immediately follows the rematerialization point. Patch by Andrea DiBiagio. Differential Revision: http://reviews.llvm.org/D23539 llvm-svn: 278822
2016-08-17 01:12:50 +08:00
%call = tail call <4 x float> @doSomething(<4 x float> %add, <4 x float> %sub) #3, !dbg !28
Remove the obsolete offset parameter from @llvm.dbg.value There is no situation where this rarely-used argument cannot be substituted with a DIExpression and removing it allows us to simplify the DWARF backend. Note that this patch does not yet remove any of the newly dead code. rdar://problem/33580047 Differential Revision: https://reviews.llvm.org/D35951 llvm-svn: 309426
2017-07-29 04:21:02 +08:00
tail call void @llvm.dbg.value(metadata <4 x float> %call, metadata !20, metadata !21), !dbg !29
When the inline spiller rematerializes an instruction, take the debug location from the instruction that immediately follows the rematerialization point. Patch by Andrea DiBiagio. Differential Revision: http://reviews.llvm.org/D23539 llvm-svn: 278822
2016-08-17 01:12:50 +08:00
%sub1 = fsub <4 x float> <float 5.000000e-01, float 5.000000e-01, float 5.000000e-01, float 5.000000e-01>, %call, !dbg !30
ret <4 x float> %sub1, !dbg !31
}
declare <4 x float> @doSomething(<4 x float>, <4 x float>) local_unnamed_addr #1
; Function Attrs: nounwind readnone
Remove the obsolete offset parameter from @llvm.dbg.value There is no situation where this rarely-used argument cannot be substituted with a DIExpression and removing it allows us to simplify the DWARF backend. Note that this patch does not yet remove any of the newly dead code. rdar://problem/33580047 Differential Revision: https://reviews.llvm.org/D35951 llvm-svn: 309426
2017-07-29 04:21:02 +08:00
declare void @llvm.dbg.value(metadata, metadata, metadata) #2
When the inline spiller rematerializes an instruction, take the debug location from the instruction that immediately follows the rematerialization point. Patch by Andrea DiBiagio. Differential Revision: http://reviews.llvm.org/D23539 llvm-svn: 278822
2016-08-17 01:12:50 +08:00
attributes #0 = { nounwind uwtable }
attributes #1 = { nounwind uwtable }
attributes #2 = { nounwind readnone }
attributes #3 = { nounwind }
!llvm.dbg.cu = !{!0}
!llvm.module.flags = !{!3, !4}
!llvm.ident = !{!5}
!0 = distinct !DICompileUnit(language: DW_LANG_C99, file: !1, producer: "clang version 4.0.0 (trunk 278291)", isOptimized: true, runtimeVersion: 0, emissionKind: FullDebug, enums: !2)
!1 = !DIFile(filename: "test.c", directory: "/home/test")
!2 = !{}
!3 = !{i32 2, !"Dwarf Version", i32 4}
!4 = !{i32 2, !"Debug Info Version", i32 3}
!5 = !{!"clang version 4.0.0 (trunk 278291)"}
!6 = distinct !DISubprogram(name: "foo", scope: !1, file: !1, line: 6, type: !7, isLocal: false, isDefinition: true, scopeLine: 6, flags: DIFlagPrototyped, isOptimized: true, unit: !0, variables: !14)
!7 = !DISubroutineType(types: !8)
!8 = !{!9, !9}
!9 = !DIDerivedType(tag: DW_TAG_typedef, name: "__m128", file: !1, line: 1, baseType: !10)
!10 = !DICompositeType(tag: DW_TAG_array_type, baseType: !11, size: 128, align: 128, flags: DIFlagVector, elements: !12)
!11 = !DIBasicType(name: "float", size: 32, align: 32, encoding: DW_ATE_float)
!12 = !{!13}
!13 = !DISubrange(count: 4)
!14 = !{!15, !16, !18, !19, !20}
!15 = !DILocalVariable(name: "X", arg: 1, scope: !6, file: !1, line: 6, type: !9)
!16 = !DILocalVariable(name: "V", scope: !6, file: !1, line: 7, type: !17)
!17 = !DIDerivedType(tag: DW_TAG_const_type, baseType: !9)
!18 = !DILocalVariable(name: "Sub", scope: !6, file: !1, line: 8, type: !9)
!19 = !DILocalVariable(name: "Add", scope: !6, file: !1, line: 9, type: !9)
!20 = !DILocalVariable(name: "Result", scope: !6, file: !1, line: 11, type: !9)
!21 = !DIExpression()
!22 = !DILocation(line: 6, column: 19, scope: !6)
!23 = !DILocation(line: 7, column: 16, scope: !6)
!24 = !DILocation(line: 8, column: 18, scope: !6)
!25 = !DILocation(line: 8, column: 10, scope: !6)
!26 = !DILocation(line: 9, column: 18, scope: !6)
!27 = !DILocation(line: 9, column: 10, scope: !6)
!28 = !DILocation(line: 11, column: 19, scope: !6)
!29 = !DILocation(line: 11, column: 10, scope: !6)
!30 = !DILocation(line: 13, column: 12, scope: !6)
!31 = !DILocation(line: 13, column: 3, scope: !6)