replica
/
hanchenye-llvm-project
2
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
hanchenye-llvm-project/llvm/test/Transforms/LICM/debug-value.ll

66 lines
4.2 KiB
LLVM
Raw Normal View History

Debug Info: enable verifier for testing cases. llvm-svn: 187375
2013-07-30 04:18:19 +08:00
; RUN: opt -licm -basicaa < %s -S | FileCheck %s
While sinking an instruction, do not lose llvm.dbg.value intrinsic. llvm-svn: 127214
2011-03-08 11:06:19 +08:00
define void @dgefa() nounwind ssp {
entry:
br label %for.body
for.body: ; preds = %for.cond.backedge, %entry
br i1 undef, label %if.then, label %for.cond.backedge, !dbg !11
for.cond.backedge: ; preds = %for.body61, %for.body61.us, %for.body
br i1 undef, label %for.end104, label %for.body, !dbg !15
if.then: ; preds = %for.body
br i1 undef, label %if.then27, label %if.end.if.end.split_crit_edge.critedge, !dbg !16
if.then27: ; preds = %if.then
; CHECK: tail call void @llvm.dbg.value
Move the complex address expression out of DIVariable and into an extra argument of the llvm.dbg.declare/llvm.dbg.value intrinsics. Previously, DIVariable was a variable-length field that has an optional reference to a Metadata array consisting of a variable number of complex address expressions. In the case of OpPiece expressions this is wasting a lot of storage in IR, because when an aggregate type is, e.g., SROA'd into all of its n individual members, the IR will contain n copies of the DIVariable, all alike, only differing in the complex address reference at the end. By making the complex address into an extra argument of the dbg.value/dbg.declare intrinsics, all of the pieces can reference the same variable and the complex address expressions can be uniqued across the CU, too. Down the road, this will allow us to move other flags, such as "indirection" out of the DIVariable, too. The new intrinsics look like this: declare void @llvm.dbg.declare(metadata %storage, metadata %var, metadata %expr) declare void @llvm.dbg.value(metadata %storage, i64 %offset, metadata %var, metadata %expr) This patch adds a new LLVM-local tag to DIExpressions, so we can detect and pretty-print DIExpression metadata nodes. What this patch doesn't do: This patch does not touch the "Indirect" field in DIVariable; but moving that into the expression would be a natural next step. http://reviews.llvm.org/D4919 rdar://problem/17994491 Thanks to dblaikie and dexonsmith for reviewing this patch! Note: I accidentally committed a bogus older version of this patch previously. llvm-svn: 218787
2014-10-02 02:55:02 +08:00
tail call void @llvm.dbg.value(metadata !18, i64 0, metadata !19, metadata !{}), !dbg !21
While sinking an instruction, do not lose llvm.dbg.value intrinsic. llvm-svn: 127214
2011-03-08 11:06:19 +08:00
br label %for.body61.us
if.end.if.end.split_crit_edge.critedge: ; preds = %if.then
br label %for.body61
for.body61.us: ; preds = %for.body61.us, %if.then27
br i1 undef, label %for.cond.backedge, label %for.body61.us, !dbg !23
for.body61: ; preds = %for.body61, %if.end.if.end.split_crit_edge.critedge
br i1 undef, label %for.cond.backedge, label %for.body61, !dbg !23
for.end104: ; preds = %for.cond.backedge
ret void, !dbg !24
}
Move the complex address expression out of DIVariable and into an extra argument of the llvm.dbg.declare/llvm.dbg.value intrinsics. Previously, DIVariable was a variable-length field that has an optional reference to a Metadata array consisting of a variable number of complex address expressions. In the case of OpPiece expressions this is wasting a lot of storage in IR, because when an aggregate type is, e.g., SROA'd into all of its n individual members, the IR will contain n copies of the DIVariable, all alike, only differing in the complex address reference at the end. By making the complex address into an extra argument of the dbg.value/dbg.declare intrinsics, all of the pieces can reference the same variable and the complex address expressions can be uniqued across the CU, too. Down the road, this will allow us to move other flags, such as "indirection" out of the DIVariable, too. The new intrinsics look like this: declare void @llvm.dbg.declare(metadata %storage, metadata %var, metadata %expr) declare void @llvm.dbg.value(metadata %storage, i64 %offset, metadata %var, metadata %expr) This patch adds a new LLVM-local tag to DIExpressions, so we can detect and pretty-print DIExpression metadata nodes. What this patch doesn't do: This patch does not touch the "Indirect" field in DIVariable; but moving that into the expression would be a natural next step. http://reviews.llvm.org/D4919 rdar://problem/17994491 Thanks to dblaikie and dexonsmith for reviewing this patch! Note: I accidentally committed a bogus older version of this patch previously. llvm-svn: 218787
2014-10-02 02:55:02 +08:00
declare void @llvm.dbg.value(metadata, i64, metadata, metadata) nounwind readnone
While sinking an instruction, do not lose llvm.dbg.value intrinsic. llvm-svn: 127214
2011-03-08 11:06:19 +08:00
Debug Info: update testing cases to specify the debug info version number. We are going to drop debug info without a version number or with a different version number, to make sure we don't crash when we see bitcode files with different debug info metadata format. Make tests more robust by removing hard-coded metadata numbers in CHECK lines. llvm-svn: 195535
2013-11-23 09:16:29 +08:00
!llvm.module.flags = !{!26}
While sinking an instruction, do not lose llvm.dbg.value intrinsic. llvm-svn: 127214
2011-03-08 11:06:19 +08:00
!llvm.dbg.sp = !{!0, !6, !9, !10}
Revert "Revert "DI: Fold constant arguments into a single MDString"" This reverts commit r218918, effectively reapplying r218914 after fixing an Ocaml bindings test and an Asan crash. The root cause of the latter was a tightened-up check in `DILexicalBlock::Verify()`, so I'll file a PR to investigate who requires the loose check (and why). Original commit message follows. -- This patch addresses the first stage of PR17891 by folding constant arguments together into a single MDString. Integers are stringified and a `\0` character is used as a separator. Part of PR17891. Note: I've attached my testcases upgrade scripts to the PR. If I've just broken your out-of-tree testcases, they might help. llvm-svn: 219010
2014-10-04 04:01:09 +08:00
!0 = metadata !{metadata !"0x2e\00idamax\00idamax\00\00112\000\001\000\006\00256\000\000", metadata !25, metadata !1, metadata !3, i32 0, null, null, null, null} ; [ DW_TAG_subprogram ]
!1 = metadata !{metadata !"0x29", metadata !25} ; [ DW_TAG_file_type ]
!2 = metadata !{metadata !"0x11\0012\00clang version 2.9 (trunk 127169)\001\00\000\00\000", metadata !25, metadata !8, metadata !8, metadata !8, null, null} ; [ DW_TAG_compile_unit ]
!3 = metadata !{metadata !"0x15\00\000\000\000\000\000\000", metadata !25, metadata !1, null, metadata !4, i32 0} ; [ DW_TAG_subroutine_type ]
While sinking an instruction, do not lose llvm.dbg.value intrinsic. llvm-svn: 127214
2011-03-08 11:06:19 +08:00
!4 = metadata !{metadata !5}
Revert "Revert "DI: Fold constant arguments into a single MDString"" This reverts commit r218918, effectively reapplying r218914 after fixing an Ocaml bindings test and an Asan crash. The root cause of the latter was a tightened-up check in `DILexicalBlock::Verify()`, so I'll file a PR to investigate who requires the loose check (and why). Original commit message follows. -- This patch addresses the first stage of PR17891 by folding constant arguments together into a single MDString. Integers are stringified and a `\0` character is used as a separator. Part of PR17891. Note: I've attached my testcases upgrade scripts to the PR. If I've just broken your out-of-tree testcases, they might help. llvm-svn: 219010
2014-10-04 04:01:09 +08:00
!5 = metadata !{metadata !"0x24\00int\000\0032\0032\000\000\005", null, metadata !2} ; [ DW_TAG_base_type ]
!6 = metadata !{metadata !"0x2e\00dscal\00dscal\00\00206\000\001\000\006\00256\000\000", metadata !25, metadata !1, metadata !7, i32 0, null, null, null, null} ; [ DW_TAG_subprogram ]
!7 = metadata !{metadata !"0x15\00\000\000\000\000\000\000", metadata !25, metadata !1, null, metadata !8, null, null, null} ; [ DW_TAG_subroutine_type ] [line 0, size 0, align 0, offset 0] [from ]
While sinking an instruction, do not lose llvm.dbg.value intrinsic. llvm-svn: 127214
2011-03-08 11:06:19 +08:00
!8 = metadata !{null}
Revert "Revert "DI: Fold constant arguments into a single MDString"" This reverts commit r218918, effectively reapplying r218914 after fixing an Ocaml bindings test and an Asan crash. The root cause of the latter was a tightened-up check in `DILexicalBlock::Verify()`, so I'll file a PR to investigate who requires the loose check (and why). Original commit message follows. -- This patch addresses the first stage of PR17891 by folding constant arguments together into a single MDString. Integers are stringified and a `\0` character is used as a separator. Part of PR17891. Note: I've attached my testcases upgrade scripts to the PR. If I've just broken your out-of-tree testcases, they might help. llvm-svn: 219010
2014-10-04 04:01:09 +08:00
!9 = metadata !{metadata !"0x2e\00daxpy\00daxpy\00\00230\000\001\000\006\00256\000\000", metadata !25, metadata !1, metadata !7, i32 0, null, null, null, null} ; [ DW_TAG_subprogram ]
!10 = metadata !{metadata !"0x2e\00dgefa\00dgefa\00\00267\000\001\000\006\00256\000\000", metadata !25, metadata !1, metadata !7, null, null, null, null, null} ; [ DW_TAG_subprogram ] [line 267] [def] [scope 0] [dgefa]
While sinking an instruction, do not lose llvm.dbg.value intrinsic. llvm-svn: 127214
2011-03-08 11:06:19 +08:00
!11 = metadata !{i32 281, i32 9, metadata !12, null}
Revert "Revert "DI: Fold constant arguments into a single MDString"" This reverts commit r218918, effectively reapplying r218914 after fixing an Ocaml bindings test and an Asan crash. The root cause of the latter was a tightened-up check in `DILexicalBlock::Verify()`, so I'll file a PR to investigate who requires the loose check (and why). Original commit message follows. -- This patch addresses the first stage of PR17891 by folding constant arguments together into a single MDString. Integers are stringified and a `\0` character is used as a separator. Part of PR17891. Note: I've attached my testcases upgrade scripts to the PR. If I've just broken your out-of-tree testcases, they might help. llvm-svn: 219010
2014-10-04 04:01:09 +08:00
!12 = metadata !{metadata !"0xb\00272\005\0032", metadata !25, metadata !13} ; [ DW_TAG_lexical_block ]
!13 = metadata !{metadata !"0xb\00271\005\0031", metadata !25, metadata !14} ; [ DW_TAG_lexical_block ]
!14 = metadata !{metadata !"0xb\00267\001\0030", metadata !25, metadata !10} ; [ DW_TAG_lexical_block ]
While sinking an instruction, do not lose llvm.dbg.value intrinsic. llvm-svn: 127214
2011-03-08 11:06:19 +08:00
!15 = metadata !{i32 271, i32 5, metadata !14, null}
!16 = metadata !{i32 284, i32 10, metadata !17, null}
Revert "Revert "DI: Fold constant arguments into a single MDString"" This reverts commit r218918, effectively reapplying r218914 after fixing an Ocaml bindings test and an Asan crash. The root cause of the latter was a tightened-up check in `DILexicalBlock::Verify()`, so I'll file a PR to investigate who requires the loose check (and why). Original commit message follows. -- This patch addresses the first stage of PR17891 by folding constant arguments together into a single MDString. Integers are stringified and a `\0` character is used as a separator. Part of PR17891. Note: I've attached my testcases upgrade scripts to the PR. If I've just broken your out-of-tree testcases, they might help. llvm-svn: 219010
2014-10-04 04:01:09 +08:00
!17 = metadata !{metadata !"0xb\00282\009\0033", metadata !25, metadata !12} ; [ DW_TAG_lexical_block ]
While sinking an instruction, do not lose llvm.dbg.value intrinsic. llvm-svn: 127214
2011-03-08 11:06:19 +08:00
!18 = metadata !{double undef}
Revert "Revert "DI: Fold constant arguments into a single MDString"" This reverts commit r218918, effectively reapplying r218914 after fixing an Ocaml bindings test and an Asan crash. The root cause of the latter was a tightened-up check in `DILexicalBlock::Verify()`, so I'll file a PR to investigate who requires the loose check (and why). Original commit message follows. -- This patch addresses the first stage of PR17891 by folding constant arguments together into a single MDString. Integers are stringified and a `\0` character is used as a separator. Part of PR17891. Note: I've attached my testcases upgrade scripts to the PR. If I've just broken your out-of-tree testcases, they might help. llvm-svn: 219010
2014-10-04 04:01:09 +08:00
!19 = metadata !{metadata !"0x100\00temp\00268\000", metadata !14, metadata !1, metadata !20} ; [ DW_TAG_auto_variable ]
!20 = metadata !{metadata !"0x24\00double\000\0064\0064\000\000\004", null, metadata !2} ; [ DW_TAG_base_type ]
While sinking an instruction, do not lose llvm.dbg.value intrinsic. llvm-svn: 127214
2011-03-08 11:06:19 +08:00
!21 = metadata !{i32 286, i32 14, metadata !22, null}
Revert "Revert "DI: Fold constant arguments into a single MDString"" This reverts commit r218918, effectively reapplying r218914 after fixing an Ocaml bindings test and an Asan crash. The root cause of the latter was a tightened-up check in `DILexicalBlock::Verify()`, so I'll file a PR to investigate who requires the loose check (and why). Original commit message follows. -- This patch addresses the first stage of PR17891 by folding constant arguments together into a single MDString. Integers are stringified and a `\0` character is used as a separator. Part of PR17891. Note: I've attached my testcases upgrade scripts to the PR. If I've just broken your out-of-tree testcases, they might help. llvm-svn: 219010
2014-10-04 04:01:09 +08:00
!22 = metadata !{metadata !"0xb\00285\0013\0034", metadata !25, metadata !17} ; [ DW_TAG_lexical_block ]
While sinking an instruction, do not lose llvm.dbg.value intrinsic. llvm-svn: 127214
2011-03-08 11:06:19 +08:00
!23 = metadata !{i32 296, i32 13, metadata !17, null}
!24 = metadata !{i32 313, i32 1, metadata !14, null}
Debug Info Verifier: verify SPs in llvm.dbg.sp. Also always add DIType, DISubprogram and DIGlobalVariable to the list in DebugInfoFinder without checking them, so we can verify them later on. llvm-svn: 187285
2013-07-27 09:26:08 +08:00
!25 = metadata !{metadata !"/Volumes/Lalgate/work/llvm/projects/llvm-test/SingleSource/Benchmarks/CoyoteBench/lpbench.c", metadata !"/private/tmp"}
Revert "Revert "DI: Fold constant arguments into a single MDString"" This reverts commit r218918, effectively reapplying r218914 after fixing an Ocaml bindings test and an Asan crash. The root cause of the latter was a tightened-up check in `DILexicalBlock::Verify()`, so I'll file a PR to investigate who requires the loose check (and why). Original commit message follows. -- This patch addresses the first stage of PR17891 by folding constant arguments together into a single MDString. Integers are stringified and a `\0` character is used as a separator. Part of PR17891. Note: I've attached my testcases upgrade scripts to the PR. If I've just broken your out-of-tree testcases, they might help. llvm-svn: 219010
2014-10-04 04:01:09 +08:00
!26 = metadata !{i32 1, metadata !"Debug Info Version", i32 2}