PR19388: DebugInfo: Emit dead arguments in their originally declared order.
Unused arguments were not being added to the argument list, but instead
treated as arbitrary scope variables. This meant they weren't carefully
added in the original argument order.
In this particular example, though, it turns out the argument is only
/mostly/ unused (well, actually it's entirely used, but in a specific
way). It's a struct that, due to ABI reasons, is decomposed into chunks
(exactly one chunk, since it has one member) and then passed. Since only
one of those chunks is used (SROA, etc, kill the original reconstitution
code) we don't have a location to describe the whole variable.
In this particular case, since the struct consists of just the one int,
once we have partial location information, this should have a location
that describes the entire variable (since the piece is the entirety of
the object).
And at some point we'll need to describe the location of even /entirely/
unused arguments so that they can at least be printed on function entry.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@210231 91177308-0d34-0410-b5e6-96231b3b80d8
2014-06-05 00:51:35 +00:00
|
|
|
; REQUIRES: object-emission
|
|
|
|
|
|
|
|
|
|
; RUN: %llc_dwarf -O0 -filetype=obj < %s | llvm-dwarfdump -debug-dump=info - | FileCheck %s
|
|
|
|
|
|
|
|
|
|
; Built from the following source with clang -O1
|
|
|
|
|
; struct S { int i; };
|
|
|
|
|
; int function(struct S s, int i) { return s.i + i; }
|
|
|
|
|
|
|
|
|
|
; Due to the X86_64 ABI, 's' is passed in registers and once optimized, the
|
|
|
|
|
; entirety of 's' is never reconstituted, since only the int is required, and
|
|
|
|
|
; thus the variable's location is unknown/dead to debug info.
|
|
|
|
|
|
|
|
|
|
; Future/current work should enable us to describe partial variables, which, in
|
|
|
|
|
; this case, happens to be the entire variable.
|
|
|
|
|
|
|
|
|
|
; CHECK: DW_TAG_subprogram
|
|
|
|
|
; CHECK-NOT: DW_TAG
|
|
|
|
|
; CHECK: DW_AT_name {{.*}} "function"
|
|
|
|
|
; CHECK-NOT: {{DW_TAG|NULL}}
|
|
|
|
|
; CHECK: DW_TAG_formal_parameter
|
|
|
|
|
; CHECK-NOT: DW_TAG
|
|
|
|
|
; CHECK: DW_AT_name {{.*}} "s"
|
|
|
|
|
; CHECK-NOT: DW_TAG
|
|
|
|
|
; FIXME: Even though 's' is never reconstituted into a struct, the one member
|
|
|
|
|
; variable is still live and used, and so we should be able to describe 's's
|
|
|
|
|
; location as the location of that int.
|
|
|
|
|
; CHECK-NOT: DW_AT_location
|
|
|
|
|
; CHECK-NOT: {{DW_TAG|NULL}}
|
|
|
|
|
; CHECK: DW_TAG_formal_parameter
|
|
|
|
|
; CHECK-NOT: DW_TAG
|
|
|
|
|
; CHECK: DW_AT_location
|
2014-06-13 22:18:23 +00:00
|
|
|
; CHECK-NOT: DW_TAG
|
|
|
|
|
; CHECK: DW_AT_name {{.*}} "i"
|
PR19388: DebugInfo: Emit dead arguments in their originally declared order.
Unused arguments were not being added to the argument list, but instead
treated as arbitrary scope variables. This meant they weren't carefully
added in the original argument order.
In this particular example, though, it turns out the argument is only
/mostly/ unused (well, actually it's entirely used, but in a specific
way). It's a struct that, due to ABI reasons, is decomposed into chunks
(exactly one chunk, since it has one member) and then passed. Since only
one of those chunks is used (SROA, etc, kill the original reconstitution
code) we don't have a location to describe the whole variable.
In this particular case, since the struct consists of just the one int,
once we have partial location information, this should have a location
that describes the entire variable (since the piece is the entirety of
the object).
And at some point we'll need to describe the location of even /entirely/
unused arguments so that they can at least be printed on function entry.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@210231 91177308-0d34-0410-b5e6-96231b3b80d8
2014-06-05 00:51:35 +00:00
|
|
|
|
|
|
|
|
|
|
|
|
|
%struct.S = type { i32 }
|
|
|
|
|
|
|
|
|
|
; Function Attrs: nounwind readnone uwtable
|
|
|
|
|
define i32 @_Z8function1Si(i32 %s.coerce, i32 %i) #0 {
|
|
|
|
|
entry:
|
2014-10-02 22:15:31 +00:00
|
|
|
tail call void @llvm.dbg.declare(metadata !19, metadata !14, metadata !{i32 786690}), !dbg !20
|
|
|
|
|
tail call void @llvm.dbg.value(metadata !{i32 %i}, i64 0, metadata !15, metadata !{i32 786690}), !dbg !20
|
PR19388: DebugInfo: Emit dead arguments in their originally declared order.
Unused arguments were not being added to the argument list, but instead
treated as arbitrary scope variables. This meant they weren't carefully
added in the original argument order.
In this particular example, though, it turns out the argument is only
/mostly/ unused (well, actually it's entirely used, but in a specific
way). It's a struct that, due to ABI reasons, is decomposed into chunks
(exactly one chunk, since it has one member) and then passed. Since only
one of those chunks is used (SROA, etc, kill the original reconstitution
code) we don't have a location to describe the whole variable.
In this particular case, since the struct consists of just the one int,
once we have partial location information, this should have a location
that describes the entire variable (since the piece is the entirety of
the object).
And at some point we'll need to describe the location of even /entirely/
unused arguments so that they can at least be printed on function entry.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@210231 91177308-0d34-0410-b5e6-96231b3b80d8
2014-06-05 00:51:35 +00:00
|
|
|
%add = add nsw i32 %i, %s.coerce, !dbg !20
|
|
|
|
|
ret i32 %add, !dbg !20
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; Function Attrs: nounwind readnone
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@218787 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-01 18:55:02 +00:00
|
|
|
declare void @llvm.dbg.declare(metadata, metadata, metadata) #1
|
PR19388: DebugInfo: Emit dead arguments in their originally declared order.
Unused arguments were not being added to the argument list, but instead
treated as arbitrary scope variables. This meant they weren't carefully
added in the original argument order.
In this particular example, though, it turns out the argument is only
/mostly/ unused (well, actually it's entirely used, but in a specific
way). It's a struct that, due to ABI reasons, is decomposed into chunks
(exactly one chunk, since it has one member) and then passed. Since only
one of those chunks is used (SROA, etc, kill the original reconstitution
code) we don't have a location to describe the whole variable.
In this particular case, since the struct consists of just the one int,
once we have partial location information, this should have a location
that describes the entire variable (since the piece is the entirety of
the object).
And at some point we'll need to describe the location of even /entirely/
unused arguments so that they can at least be printed on function entry.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@210231 91177308-0d34-0410-b5e6-96231b3b80d8
2014-06-05 00:51:35 +00:00
|
|
|
|
|
|
|
|
; Function Attrs: nounwind readnone
|
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.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@218787 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-01 18:55:02 +00:00
|
|
|
declare void @llvm.dbg.value(metadata, i64, metadata, metadata) #1
|
PR19388: DebugInfo: Emit dead arguments in their originally declared order.
Unused arguments were not being added to the argument list, but instead
treated as arbitrary scope variables. This meant they weren't carefully
added in the original argument order.
In this particular example, though, it turns out the argument is only
/mostly/ unused (well, actually it's entirely used, but in a specific
way). It's a struct that, due to ABI reasons, is decomposed into chunks
(exactly one chunk, since it has one member) and then passed. Since only
one of those chunks is used (SROA, etc, kill the original reconstitution
code) we don't have a location to describe the whole variable.
In this particular case, since the struct consists of just the one int,
once we have partial location information, this should have a location
that describes the entire variable (since the piece is the entirety of
the object).
And at some point we'll need to describe the location of even /entirely/
unused arguments so that they can at least be printed on function entry.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@210231 91177308-0d34-0410-b5e6-96231b3b80d8
2014-06-05 00:51:35 +00:00
|
|
|
|
|
|
|
|
attributes #0 = { nounwind readnone uwtable "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
|
|
|
|
|
attributes #1 = { nounwind readnone }
|
|
|
|
|
|
|
|
|
|
!llvm.dbg.cu = !{!0}
|
|
|
|
|
!llvm.module.flags = !{!16, !17}
|
|
|
|
|
!llvm.ident = !{!18}
|
|
|
|
|
|
2014-10-02 22:15:31 +00:00
|
|
|
!0 = metadata !{i32 786449, metadata !1, i32 4, metadata !"clang version 3.5.0 ", i1 true, metadata !"", i32 0, metadata !2, metadata !3, metadata !8, metadata !2, metadata !2, metadata !"", i32 1} ; [ DW_TAG_compile_unit ] [/tmp/dbginfo/dead-argument-order.cpp] [DW_LANG_C_plus_plus]
|
PR19388: DebugInfo: Emit dead arguments in their originally declared order.
Unused arguments were not being added to the argument list, but instead
treated as arbitrary scope variables. This meant they weren't carefully
added in the original argument order.
In this particular example, though, it turns out the argument is only
/mostly/ unused (well, actually it's entirely used, but in a specific
way). It's a struct that, due to ABI reasons, is decomposed into chunks
(exactly one chunk, since it has one member) and then passed. Since only
one of those chunks is used (SROA, etc, kill the original reconstitution
code) we don't have a location to describe the whole variable.
In this particular case, since the struct consists of just the one int,
once we have partial location information, this should have a location
that describes the entire variable (since the piece is the entirety of
the object).
And at some point we'll need to describe the location of even /entirely/
unused arguments so that they can at least be printed on function entry.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@210231 91177308-0d34-0410-b5e6-96231b3b80d8
2014-06-05 00:51:35 +00:00
|
|
|
!1 = metadata !{metadata !"dead-argument-order.cpp", metadata !"/tmp/dbginfo"}
|
|
|
|
|
!2 = metadata !{}
|
|
|
|
|
!3 = metadata !{metadata !4}
|
2014-10-02 22:15:31 +00:00
|
|
|
!4 = metadata !{i32 786451, metadata !1, null, metadata !"S", i32 1, i64 32, i64 32, i32 0, i32 0, null, metadata !5, i32 0, null, null, metadata !"_ZTS1S"} ; [ DW_TAG_structure_type ] [S] [line 1, size 32, align 32, offset 0] [def] [from ]
|
PR19388: DebugInfo: Emit dead arguments in their originally declared order.
Unused arguments were not being added to the argument list, but instead
treated as arbitrary scope variables. This meant they weren't carefully
added in the original argument order.
In this particular example, though, it turns out the argument is only
/mostly/ unused (well, actually it's entirely used, but in a specific
way). It's a struct that, due to ABI reasons, is decomposed into chunks
(exactly one chunk, since it has one member) and then passed. Since only
one of those chunks is used (SROA, etc, kill the original reconstitution
code) we don't have a location to describe the whole variable.
In this particular case, since the struct consists of just the one int,
once we have partial location information, this should have a location
that describes the entire variable (since the piece is the entirety of
the object).
And at some point we'll need to describe the location of even /entirely/
unused arguments so that they can at least be printed on function entry.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@210231 91177308-0d34-0410-b5e6-96231b3b80d8
2014-06-05 00:51:35 +00:00
|
|
|
!5 = metadata !{metadata !6}
|
2014-10-02 22:15:31 +00:00
|
|
|
!6 = metadata !{i32 786445, metadata !1, metadata !"_ZTS1S", metadata !"i", i32 1, i64 32, i64 32, i64 0, i32 0, metadata !7} ; [ DW_TAG_member ] [i] [line 1, size 32, align 32, offset 0] [from int]
|
|
|
|
|
!7 = metadata !{i32 786468, null, null, metadata !"int", i32 0, i64 32, i64 32, i64 0, i32 0, i32 5} ; [ DW_TAG_base_type ] [int] [line 0, size 32, align 32, offset 0, enc DW_ATE_signed]
|
PR19388: DebugInfo: Emit dead arguments in their originally declared order.
Unused arguments were not being added to the argument list, but instead
treated as arbitrary scope variables. This meant they weren't carefully
added in the original argument order.
In this particular example, though, it turns out the argument is only
/mostly/ unused (well, actually it's entirely used, but in a specific
way). It's a struct that, due to ABI reasons, is decomposed into chunks
(exactly one chunk, since it has one member) and then passed. Since only
one of those chunks is used (SROA, etc, kill the original reconstitution
code) we don't have a location to describe the whole variable.
In this particular case, since the struct consists of just the one int,
once we have partial location information, this should have a location
that describes the entire variable (since the piece is the entirety of
the object).
And at some point we'll need to describe the location of even /entirely/
unused arguments so that they can at least be printed on function entry.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@210231 91177308-0d34-0410-b5e6-96231b3b80d8
2014-06-05 00:51:35 +00:00
|
|
|
!8 = metadata !{metadata !9}
|
2014-10-02 22:15:31 +00:00
|
|
|
!9 = metadata !{i32 786478, metadata !1, metadata !10, metadata !"function", metadata !"function", metadata !"_Z8function1Si", i32 2, metadata !11, i1 false, i1 true, i32 0, i32 0, null, i32 256, i1 true, i32 (i32, i32)* @_Z8function1Si, null, null, metadata !13, i32 2} ; [ DW_TAG_subprogram ] [line 2] [def] [function]
|
|
|
|
|
!10 = metadata !{i32 786473, metadata !1} ; [ DW_TAG_file_type ] [/tmp/dbginfo/dead-argument-order.cpp]
|
|
|
|
|
!11 = metadata !{i32 786453, i32 0, null, metadata !"", i32 0, i64 0, i64 0, i64 0, i32 0, null, metadata !12, i32 0, null, null, null} ; [ DW_TAG_subroutine_type ] [line 0, size 0, align 0, offset 0] [from ]
|
PR19388: DebugInfo: Emit dead arguments in their originally declared order.
Unused arguments were not being added to the argument list, but instead
treated as arbitrary scope variables. This meant they weren't carefully
added in the original argument order.
In this particular example, though, it turns out the argument is only
/mostly/ unused (well, actually it's entirely used, but in a specific
way). It's a struct that, due to ABI reasons, is decomposed into chunks
(exactly one chunk, since it has one member) and then passed. Since only
one of those chunks is used (SROA, etc, kill the original reconstitution
code) we don't have a location to describe the whole variable.
In this particular case, since the struct consists of just the one int,
once we have partial location information, this should have a location
that describes the entire variable (since the piece is the entirety of
the object).
And at some point we'll need to describe the location of even /entirely/
unused arguments so that they can at least be printed on function entry.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@210231 91177308-0d34-0410-b5e6-96231b3b80d8
2014-06-05 00:51:35 +00:00
|
|
|
!12 = metadata !{metadata !7, metadata !4, metadata !7}
|
|
|
|
|
!13 = metadata !{metadata !14, metadata !15}
|
2014-10-02 22:15:31 +00:00
|
|
|
!14 = metadata !{i32 786689, metadata !9, metadata !"s", metadata !10, i32 16777218, metadata !"_ZTS1S", i32 0, i32 0} ; [ DW_TAG_arg_variable ] [s] [line 2]
|
|
|
|
|
!15 = metadata !{i32 786689, metadata !9, metadata !"i", metadata !10, i32 33554434, metadata !7, i32 0, i32 0} ; [ DW_TAG_arg_variable ] [i] [line 2]
|
PR19388: DebugInfo: Emit dead arguments in their originally declared order.
Unused arguments were not being added to the argument list, but instead
treated as arbitrary scope variables. This meant they weren't carefully
added in the original argument order.
In this particular example, though, it turns out the argument is only
/mostly/ unused (well, actually it's entirely used, but in a specific
way). It's a struct that, due to ABI reasons, is decomposed into chunks
(exactly one chunk, since it has one member) and then passed. Since only
one of those chunks is used (SROA, etc, kill the original reconstitution
code) we don't have a location to describe the whole variable.
In this particular case, since the struct consists of just the one int,
once we have partial location information, this should have a location
that describes the entire variable (since the piece is the entirety of
the object).
And at some point we'll need to describe the location of even /entirely/
unused arguments so that they can at least be printed on function entry.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@210231 91177308-0d34-0410-b5e6-96231b3b80d8
2014-06-05 00:51:35 +00:00
|
|
|
!16 = metadata !{i32 2, metadata !"Dwarf Version", i32 4}
|
2014-10-02 22:15:31 +00:00
|
|
|
!17 = metadata !{i32 2, metadata !"Debug Info Version", i32 1}
|
PR19388: DebugInfo: Emit dead arguments in their originally declared order.
Unused arguments were not being added to the argument list, but instead
treated as arbitrary scope variables. This meant they weren't carefully
added in the original argument order.
In this particular example, though, it turns out the argument is only
/mostly/ unused (well, actually it's entirely used, but in a specific
way). It's a struct that, due to ABI reasons, is decomposed into chunks
(exactly one chunk, since it has one member) and then passed. Since only
one of those chunks is used (SROA, etc, kill the original reconstitution
code) we don't have a location to describe the whole variable.
In this particular case, since the struct consists of just the one int,
once we have partial location information, this should have a location
that describes the entire variable (since the piece is the entirety of
the object).
And at some point we'll need to describe the location of even /entirely/
unused arguments so that they can at least be printed on function entry.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@210231 91177308-0d34-0410-b5e6-96231b3b80d8
2014-06-05 00:51:35 +00:00
|
|
|
!18 = metadata !{metadata !"clang version 3.5.0 "}
|
|
|
|
|
!19 = metadata !{%struct.S* undef}
|
|
|
|
|
!20 = metadata !{i32 2, i32 0, metadata !9, null}
|
|
|
|
|
|
