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/block-capture.ll

134 lines
10 KiB
LLVM
Raw Normal View History

If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
; RUN: llc -mtriple=x86_64-apple-darwin %s -o %t -filetype=obj
Now that llvm-dwarfdump supports flags to specify which DWARF section to dump, use them in tests that run llvm-dwarfdump. This is in order to make tests as specific as possible. llvm-svn: 173498
2013-01-26 05:44:53 +08:00
; RUN: llvm-dwarfdump -debug-dump=info %t | FileCheck %s
In Dwarf 3 (and Dwarf 2) attributes whose value are offsets into a section use the form DW_FORM_data4 whilst in Dwarf 4 and later they use the form DW_FORM_sec_offset. This patch updates the places where such attributes are generated to use the appropriate form depending on the Dwarf version. The DIE entries affected have the following tags: DW_AT_stmt_list, DW_AT_ranges, DW_AT_location, DW_AT_GNU_pubnames, DW_AT_GNU_pubtypes, DW_AT_GNU_addr_base, DW_AT_GNU_ranges_base It also adds a hidden command line option "--dwarf-version=<uint>" to llc which allows the version of Dwarf to be generated to override what is specified in the metadata; this makes it possible to update existing tests to check the debugging information generated for both Dwarf 4 (the default) and Dwarf 3 using the same metadata. Patch (slightly modified) by Keith Walker! llvm-svn: 195391
2013-11-22 07:46:41 +08:00
; RUN: llc -mtriple=x86_64-apple-darwin %s -o %t -filetype=obj -dwarf-version=3
; RUN: llvm-dwarfdump -debug-dump=info %t | FileCheck %s -check-prefix=DWARF3
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
; Checks that we emit debug info for the block variable declare.
DebugInfo: Generalize a test case to not depend on abbreviation numbering. It's an unnecessary detail for this test and just gets in the way when making unrelated changes to the output in this test. llvm-svn: 209553
2014-05-24 05:07:01 +08:00
; CHECK: DW_TAG_subprogram
; CHECK: DW_TAG_variable
; CHECK: DW_AT_location [DW_FORM_sec_offset]
DebugInfo: Refactor some tests to allow DW_AT_name to not be the first attribute in a local variable. In an effort to fix concrete variables referencing abstract origins where the concrete variable preceeds the first inlined usage, the addition of attributes such as name, file, etc will be delayed until the end of the module (to wait to see if any inlined instances have occurred, thus necessitating an abstract definition that the concrete definition should also reference). These test cases don't actually need to care about this ordering of attributes, so update them to be more resilient to such changes coming in the near future. llvm-svn: 210940
2014-06-14 05:52:33 +08:00
; CHECK: DW_AT_name {{.*}} "block"
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
DebugInfo: Generalize a test case to not depend on abbreviation numbering. It's an unnecessary detail for this test and just gets in the way when making unrelated changes to the output in this test. llvm-svn: 209553
2014-05-24 05:07:01 +08:00
; DWARF3: DW_TAG_subprogram
; DWARF3: DW_TAG_variable
; DWARF3: DW_AT_location [DW_FORM_data4]
DebugInfo: Refactor some tests to allow DW_AT_name to not be the first attribute in a local variable. In an effort to fix concrete variables referencing abstract origins where the concrete variable preceeds the first inlined usage, the addition of attributes such as name, file, etc will be delayed until the end of the module (to wait to see if any inlined instances have occurred, thus necessitating an abstract definition that the concrete definition should also reference). These test cases don't actually need to care about this ordering of attributes, so update them to be more resilient to such changes coming in the near future. llvm-svn: 210940
2014-06-14 05:52:33 +08:00
; DWARF3: DW_AT_name {{.*}} "block"
In Dwarf 3 (and Dwarf 2) attributes whose value are offsets into a section use the form DW_FORM_data4 whilst in Dwarf 4 and later they use the form DW_FORM_sec_offset. This patch updates the places where such attributes are generated to use the appropriate form depending on the Dwarf version. The DIE entries affected have the following tags: DW_AT_stmt_list, DW_AT_ranges, DW_AT_location, DW_AT_GNU_pubnames, DW_AT_GNU_pubtypes, DW_AT_GNU_addr_base, DW_AT_GNU_ranges_base It also adds a hidden command line option "--dwarf-version=<uint>" to llc which allows the version of Dwarf to be generated to override what is specified in the metadata; this makes it possible to update existing tests to check the debugging information generated for both Dwarf 4 (the default) and Dwarf 3 using the same metadata. Patch (slightly modified) by Keith Walker! llvm-svn: 195391
2013-11-22 07:46:41 +08:00
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
%struct.__block_descriptor = type { i64, i64 }
%struct.__block_literal_generic = type { i8*, i32, i32, i8*, %struct.__block_descriptor* }
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.declare(metadata, metadata, metadata) nounwind readnone
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
define hidden void @__foo_block_invoke_0(i8* %.block_descriptor) uwtable ssp {
entry:
%exn.slot = alloca i8*
%ehselector.slot = alloca i32
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
call void @llvm.dbg.value(metadata !{i8* %.block_descriptor}, i64 0, metadata !39, metadata !{metadata !"0x102"}), !dbg !51
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
%block = bitcast i8* %.block_descriptor to <{ i8*, i32, i32, i8*, %struct.__block_descriptor*, void ()* }>*, !dbg !52
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
call void @llvm.dbg.declare(metadata !{<{ i8*, i32, i32, i8*, %struct.__block_descriptor*, void ()* }>* %block}, metadata !53, metadata !65), !dbg !54
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
%block.capture.addr = getelementptr inbounds <{ i8*, i32, i32, i8*, %struct.__block_descriptor*, void ()* }>* %block, i32 0, i32 5, !dbg !55
%0 = load void ()** %block.capture.addr, align 8, !dbg !55
%block.literal = bitcast void ()* %0 to %struct.__block_literal_generic*, !dbg !55
%1 = getelementptr inbounds %struct.__block_literal_generic* %block.literal, i32 0, i32 3, !dbg !55
%2 = bitcast %struct.__block_literal_generic* %block.literal to i8*, !dbg !55
%3 = load i8** %1, !dbg !55
%4 = bitcast i8* %3 to void (i8*)*, !dbg !55
invoke void %4(i8* %2)
to label %invoke.cont unwind label %lpad, !dbg !55
invoke.cont: ; preds = %entry
br label %eh.cont, !dbg !58
eh.cont: ; preds = %catch, %invoke.cont
ret void, !dbg !61
lpad: ; preds = %entry
%5 = landingpad { i8*, i32 } personality i8* bitcast (i32 (...)* @__objc_personality_v0 to i8*)
catch i8* null, !dbg !55
%6 = extractvalue { i8*, i32 } %5, 0, !dbg !55
store i8* %6, i8** %exn.slot, !dbg !55
%7 = extractvalue { i8*, i32 } %5, 1, !dbg !55
store i32 %7, i32* %ehselector.slot, !dbg !55
br label %catch, !dbg !55
catch: ; preds = %lpad
%exn = load i8** %exn.slot, !dbg !62
%exn.adjusted = call i8* @objc_begin_catch(i8* %exn) nounwind, !dbg !62
call void @objc_end_catch(), !dbg !58
br label %eh.cont, !dbg !58
}
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
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
declare i8* @objc_begin_catch(i8*)
declare void @objc_end_catch()
declare i32 @__objc_personality_v0(...)
!llvm.dbg.cu = !{!0}
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. llvm-svn: 195504
2013-11-23 05:49:45 +08:00
!llvm.module.flags = !{!35, !36, !37, !38, !64}
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
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 !"0x11\0016\00clang version 3.1 (trunk 151227)\000\00\002\00\001", metadata !63, metadata !1, metadata !1, metadata !3, metadata !1, metadata !1} ; [ DW_TAG_compile_unit ]
DebugInfo: Remove some unneeded conditionals now that DIBuilder no longer emits zero-length arrays as {i32 0} A bunch of test cases needed to be cleaned up for this, many my fault - when implementid imported modules I updated test cases by simply duplicating the prior metadata field - which wasn't always the empty metadata entry. llvm-svn: 200731
2014-02-04 09:23:52 +08:00
!1 = metadata !{}
Remove the (apparently) unnecessary debug info metadata indirection. The main lists of debug info metadata attached to the compile_unit had an extra layer of metadata nodes they went through for no apparent reason. This patch removes that (& still passes just as much of the GDB 7.5 test suite). If anyone can show evidence as to why these extra metadata nodes are there I'm open to reverting this patch & documenting why they're there. llvm-svn: 174266
2013-02-02 13:56:24 +08:00
!3 = metadata !{metadata !5, metadata !28, metadata !31, metadata !34}
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 !"0x2e\00foo\00foo\00\005\000\001\000\006\00256\000\005", metadata !6, metadata !6, metadata !7, null, null, null, null, null} ; [ DW_TAG_subprogram ]
!6 = metadata !{metadata !"0x29", metadata !63} ; [ DW_TAG_file_type ]
!7 = metadata !{metadata !"0x15\00\000\000\000\000\000\000", i32 0, null, null, metadata !8, null, null, null} ; [ DW_TAG_subroutine_type ] [line 0, size 0, align 0, offset 0] [from ]
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
!8 = metadata !{null, metadata !9}
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 !"0x16\00dispatch_block_t\001\000\000\000\000", metadata !63, null, metadata !10} ; [ DW_TAG_typedef ]
!10 = metadata !{metadata !"0xf\00\000\0064\000\000\000", null, null, metadata !11} ; [ DW_TAG_pointer_type ]
!11 = metadata !{metadata !"0x13\00__block_literal_generic\005\00256\000\000\008\000", metadata !63, metadata !6, null, metadata !12, null, null, null} ; [ DW_TAG_structure_type ] [__block_literal_generic] [line 5, size 256, align 0, offset 0] [def] [from ]
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
!12 = metadata !{metadata !13, metadata !15, metadata !17, metadata !18, metadata !19}
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
!13 = metadata !{metadata !"0xd\00__isa\000\0064\0064\000\000", metadata !63, metadata !6, metadata !14} ; [ DW_TAG_member ]
!14 = metadata !{metadata !"0xf\00\000\0064\0064\000\000", null, null, null} ; [ DW_TAG_pointer_type ]
!15 = metadata !{metadata !"0xd\00__flags\000\0032\0032\0064\000", metadata !63, metadata !6, metadata !16} ; [ DW_TAG_member ]
!16 = metadata !{metadata !"0x24\00int\000\0032\0032\000\000\005", null, null} ; [ DW_TAG_base_type ]
!17 = metadata !{metadata !"0xd\00__reserved\000\0032\0032\0096\000", metadata !63, metadata !6, metadata !16} ; [ DW_TAG_member ]
!18 = metadata !{metadata !"0xd\00__FuncPtr\000\0064\0064\00128\000", metadata !63, metadata !6, metadata !14} ; [ DW_TAG_member ]
!19 = metadata !{metadata !"0xd\00__descriptor\005\0064\0064\00192\000", metadata !63, metadata !6, metadata !20} ; [ DW_TAG_member ]
!20 = metadata !{metadata !"0xf\00\000\0064\000\000\000", null, null, metadata !21} ; [ DW_TAG_pointer_type ]
!21 = metadata !{metadata !"0x13\00__block_descriptor\005\00128\000\000\008\000", metadata !63, metadata !6, null, metadata !22, null, null, null} ; [ DW_TAG_structure_type ] [__block_descriptor] [line 5, size 128, align 0, offset 0] [def] [from ]
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
!22 = metadata !{metadata !23, metadata !25}
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
!23 = metadata !{metadata !"0xd\00reserved\000\0064\0064\000\000", metadata !63, metadata !6, metadata !24} ; [ DW_TAG_member ]
!24 = metadata !{metadata !"0x24\00long unsigned int\000\0064\0064\000\000\007", null, null} ; [ DW_TAG_base_type ]
!25 = metadata !{metadata !"0xd\00Size\000\0064\0064\0064\000", metadata !63, metadata !6, metadata !24} ; [ DW_TAG_member ]
!28 = metadata !{metadata !"0x2e\00__foo_block_invoke_0\00__foo_block_invoke_0\00\007\001\001\000\006\00256\000\007", metadata !6, metadata !6, metadata !29, null, void (i8*)* @__foo_block_invoke_0, null, null, null} ; [ DW_TAG_subprogram ]
!29 = metadata !{metadata !"0x15\00\000\000\000\000\000\000", i32 0, null, null, metadata !30, null, null, null} ; [ DW_TAG_subroutine_type ] [line 0, size 0, align 0, offset 0] [from ]
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
!30 = metadata !{null, metadata !14}
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
!31 = metadata !{metadata !"0x2e\00__copy_helper_block_\00__copy_helper_block_\00\0010\001\001\000\006\00256\000\0010", metadata !6, metadata !6, metadata !32, null, null, null, null, null} ; [ DW_TAG_subprogram ]
!32 = metadata !{metadata !"0x15\00\000\000\000\000\000\000", i32 0, null, null, metadata !33, null, null, null} ; [ DW_TAG_subroutine_type ] [line 0, size 0, align 0, offset 0] [from ]
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
!33 = metadata !{null, metadata !14, metadata !14}
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
!34 = metadata !{metadata !"0x2e\00__destroy_helper_block_\00__destroy_helper_block_\00\0010\001\001\000\006\00256\000\0010", metadata !6, metadata !6, metadata !29, null, null, null, null, null} ; [ DW_TAG_subprogram ]
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
!35 = metadata !{i32 1, metadata !"Objective-C Version", i32 2}
!36 = metadata !{i32 1, metadata !"Objective-C Image Info Version", i32 0}
!37 = metadata !{i32 1, metadata !"Objective-C Image Info Section", metadata !"__DATA, __objc_imageinfo, regular, no_dead_strip"}
!38 = metadata !{i32 4, metadata !"Objective-C Garbage Collection", i32 0}
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
!39 = metadata !{metadata !"0x101\00.block_descriptor\0016777223\0064", metadata !28, metadata !6, metadata !40} ; [ DW_TAG_arg_variable ]
!40 = metadata !{metadata !"0xf\00\000\0064\000\000\000", null, null, metadata !41} ; [ DW_TAG_pointer_type ]
!41 = metadata !{metadata !"0x13\00__block_literal_1\007\00320\0064\000\000\000", metadata !63, metadata !6, null, metadata !42, null, null, null} ; [ DW_TAG_structure_type ] [__block_literal_1] [line 7, size 320, align 64, offset 0] [def] [from ]
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
!42 = metadata !{metadata !43, metadata !44, metadata !45, metadata !46, metadata !47, metadata !50}
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
!43 = metadata !{metadata !"0xd\00__isa\007\0064\0064\000\000", metadata !63, metadata !6, metadata !14} ; [ DW_TAG_member ]
!44 = metadata !{metadata !"0xd\00__flags\007\0032\0032\0064\000", metadata !63, metadata !6, metadata !16} ; [ DW_TAG_member ]
!45 = metadata !{metadata !"0xd\00__reserved\007\0032\0032\0096\000", metadata !63, metadata !6, metadata !16} ; [ DW_TAG_member ]
!46 = metadata !{metadata !"0xd\00__FuncPtr\007\0064\0064\00128\000", metadata !63, metadata !6, metadata !14} ; [ DW_TAG_member ]
!47 = metadata !{metadata !"0xd\00__descriptor\007\0064\0064\00192\000", metadata !63, metadata !6, metadata !48} ; [ DW_TAG_member ]
!48 = metadata !{metadata !"0xf\00\000\0064\0064\000\000", null, null, metadata !49} ; [ DW_TAG_pointer_type ]
!49 = metadata !{metadata !"0x13\00__block_descriptor_withcopydispose\007\000\000\000\004\000", metadata !63, null, null, null, null, null, null} ; [ DW_TAG_structure_type ] [__block_descriptor_withcopydispose] [line 7, size 0, align 0, offset 0] [decl] [from ]
!50 = metadata !{metadata !"0xd\00block\007\0064\0064\00256\000", metadata !63, metadata !6, metadata !9} ; [ DW_TAG_member ]
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
!51 = metadata !{i32 7, i32 18, metadata !28, null}
!52 = metadata !{i32 7, i32 19, metadata !28, 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
!53 = metadata !{metadata !"0x100\00block\005\000", metadata !28, metadata !6, metadata !9} ; [ DW_TAG_auto_variable ]
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
!54 = metadata !{i32 5, i32 27, metadata !28, null}
!55 = metadata !{i32 8, i32 22, metadata !56, 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
!56 = metadata !{metadata !"0xb\007\0026\002", metadata !6, metadata !57} ; [ DW_TAG_lexical_block ]
!57 = metadata !{metadata !"0xb\007\0019\001", metadata !6, metadata !28} ; [ DW_TAG_lexical_block ]
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
!58 = metadata !{i32 10, i32 20, metadata !59, 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
!59 = metadata !{metadata !"0xb\009\0035\004", metadata !6, metadata !60} ; [ DW_TAG_lexical_block ]
!60 = metadata !{metadata !"0xb\009\0035\003", metadata !6, metadata !57} ; [ DW_TAG_lexical_block ]
If the Address of a variable is an argument then treat the entire variable declaration as an argument because we want that address anyhow for our debug information. This seems to fix rdar://9965111, at least we have more debug information than before and from reading the assembly it appears to be the correct location. llvm-svn: 151335
2012-02-24 09:59:08 +08:00
!61 = metadata !{i32 10, i32 21, metadata !28, null}
!62 = metadata !{i32 9, i32 20, metadata !56, null}
Split out filename & directory from DIFile to start generalizing over DIScopes This is the first step to making all DIScopes have a common metadata prefix (so that things (using directives, for example) that can appear in any scope can be added to that common prefix). DIFile is itself a DIScope so the common prefix of all DIScopes cannot be a DIFile - instead it's the raw filename/directory name pair. llvm-svn: 177239
2013-03-18 05:13:55 +08:00
!63 = metadata !{metadata !"foo.m", metadata !"/Users/echristo"}
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
!64 = metadata !{i32 1, metadata !"Debug Info Version", i32 2}
!65 = metadata !{metadata !"0x102\0034\0032"} ; [ DW_TAG_expression ] [DW_OP_plus 32]