2018-04-12 00:03:07 +08:00
; RUN: llc -fast-isel-sink-local-values -O0 -mtriple=x86_64-unknown-linux-gnu < %s | FileCheck %s
; RUN: llc -fast-isel-sink-local-values -O0 -mtriple=x86_64-unknown-linux-gnu -filetype=obj < %s \
2017-09-12 06:59:45 +08:00
; RUN: | llvm-dwarfdump -debug-info - | FileCheck %s --check-prefix=DWARF
2014-06-10 05:53:47 +08:00
; Verify that we have correct debug info for local variables in code
; instrumented with AddressSanitizer.
; Generated from the source file test.cc:
; int bar(int y) {
; return y + 2;
; }
Clean up the processing of dbg.value in various places
Summary:
First up is instcombine, where in the dbg.declare -> dbg.value conversion,
the llvm.dbg.value needs to be called on the actual loaded value, rather
than the address (since the whole point of this transformation is to be
able to get rid of the alloca). Further, now that that's cleaned up, we
can remove a hack in the backend, that would add an implicit OP_deref if
the argument to dbg.value was an alloca. This stems from before the
existence of DIExpression and is no longer necessary since the deref can
be expressed explicitly.
Now, in order to make sure that the tests pass with this change, we need to
correct the printing of DEBUG_VALUE comments to take into account the
expression, which wasn't taken into account before.
Unfortunately, for both these changes, there were a number of incorrect
test cases (mostly the wrong number of DW_OP_derefs, but also a couple
where the test itself was broken more badly). aprantl and I have gone
through and adjusted these test case in order to make them pass with
these fixes and in some cases to make sure they're actually testing
what they are meant to test.
Reviewers: aprantl
Subscribers: dsanders
Differential Revision: http://reviews.llvm.org/D14186
llvm-svn: 256077
2015-12-19 10:02:44 +08:00
; with "clang++ -S -emit-llvm -mllvm -asan-skip-promotable-allocas=0 -fsanitize=address -O0 -g test.cc"
2014-06-10 05:53:47 +08:00
Clean up the processing of dbg.value in various places
Summary:
First up is instcombine, where in the dbg.declare -> dbg.value conversion,
the llvm.dbg.value needs to be called on the actual loaded value, rather
than the address (since the whole point of this transformation is to be
able to get rid of the alloca). Further, now that that's cleaned up, we
can remove a hack in the backend, that would add an implicit OP_deref if
the argument to dbg.value was an alloca. This stems from before the
existence of DIExpression and is no longer necessary since the deref can
be expressed explicitly.
Now, in order to make sure that the tests pass with this change, we need to
correct the printing of DEBUG_VALUE comments to take into account the
expression, which wasn't taken into account before.
Unfortunately, for both these changes, there were a number of incorrect
test cases (mostly the wrong number of DW_OP_derefs, but also a couple
where the test itself was broken more badly). aprantl and I have gone
through and adjusted these test case in order to make them pass with
these fixes and in some cases to make sure they're actually testing
what they are meant to test.
Reviewers: aprantl
Subscribers: dsanders
Differential Revision: http://reviews.llvm.org/D14186
llvm-svn: 256077
2015-12-19 10:02:44 +08:00
; The address of the (potentially now malloc'ed) alloca ends up
2017-11-29 01:15:09 +08:00
; in rdi, after which it is spilled to the stack. We record the
Clean up the processing of dbg.value in various places
Summary:
First up is instcombine, where in the dbg.declare -> dbg.value conversion,
the llvm.dbg.value needs to be called on the actual loaded value, rather
than the address (since the whole point of this transformation is to be
able to get rid of the alloca). Further, now that that's cleaned up, we
can remove a hack in the backend, that would add an implicit OP_deref if
the argument to dbg.value was an alloca. This stems from before the
existence of DIExpression and is no longer necessary since the deref can
be expressed explicitly.
Now, in order to make sure that the tests pass with this change, we need to
correct the printing of DEBUG_VALUE comments to take into account the
expression, which wasn't taken into account before.
Unfortunately, for both these changes, there were a number of incorrect
test cases (mostly the wrong number of DW_OP_derefs, but also a couple
where the test itself was broken more badly). aprantl and I have gone
through and adjusted these test case in order to make them pass with
these fixes and in some cases to make sure they're actually testing
what they are meant to test.
Reviewers: aprantl
Subscribers: dsanders
Differential Revision: http://reviews.llvm.org/D14186
llvm-svn: 256077
2015-12-19 10:02:44 +08:00
; spill OFFSET on the stack for checking the debug info below.
[FastISel] Sink local value materializations to first use
Summary:
Local values are constants, global addresses, and stack addresses that
can't be folded into the instruction that uses them. For example, when
storing the address of a global variable into memory, we need to
materialize that address into a register.
FastISel doesn't want to materialize any given local value more than
once, so it generates all local value materialization code at
EmitStartPt, which always dominates the current insertion point. This
allows it to maintain a map of local value registers, and it knows that
the local value area will always dominate the current insertion point.
The downside is that local value instructions are always emitted without
a source location. This is done to prevent jumpy line tables, but it
means that the local value area will be considered part of the previous
statement. Consider this C code:
call1(); // line 1
++global; // line 2
++global; // line 3
call2(&global, &local); // line 4
Today we end up with assembly and line tables like this:
.loc 1 1
callq call1
leaq global(%rip), %rdi
leaq local(%rsp), %rsi
.loc 1 2
addq $1, global(%rip)
.loc 1 3
addq $1, global(%rip)
.loc 1 4
callq call2
The LEA instructions in the local value area have no source location and
are treated as being on line 1. Stepping through the code in a debugger
and correlating it with the assembly won't make much sense, because
these materializations are only required for line 4.
This is actually problematic for the VS debugger "set next statement"
feature, which effectively assumes that there are no registers live
across statement boundaries. By sinking the local value code into the
statement and fixing up the source location, we can make that feature
work. This was filed as https://bugs.llvm.org/show_bug.cgi?id=35975 and
https://crbug.com/793819.
This change is obviously not enough to make this feature work reliably
in all cases, but I felt that it was worth doing anyway because it
usually generates smaller, more comprehensible -O0 code. I measured a
0.12% regression in code generation time with LLC on the sqlite3
amalgamation, so I think this is worth doing.
There are some special cases worth calling out in the commit message:
1. local values materialized for phis
2. local values used by no-op casts
3. dead local value code
Local values can be materialized for phis, and this does not show up as
a vreg use in MachineRegisterInfo. In this case, if there are no other
uses, this patch sinks the value to the first terminator, EH label, or
the end of the BB if nothing else exists.
Local values may also be used by no-op casts, which adds the register to
the RegFixups table. Without reversing the RegFixups map direction, we
don't have enough information to sink these instructions.
Lastly, if the local value register has no other uses, we can delete it.
This comes up when fastisel tries two instruction selection approaches
and the first materializes the value but fails and the second succeeds
without using the local value.
Reviewers: aprantl, dblaikie, qcolombet, MatzeB, vsk, echristo
Subscribers: dotdash, chandlerc, hans, sdardis, amccarth, javed.absar, zturner, llvm-commits, hiraditya
Differential Revision: https://reviews.llvm.org/D43093
llvm-svn: 327581
2018-03-15 05:54:21 +08:00
; CHECK: #DEBUG_VALUE: bar:y <- [DW_OP_deref] [$rcx+0]
; CHECK: movq %rcx, [[OFFSET:[0-9]+]](%rsp)
2014-06-10 05:53:47 +08:00
; CHECK-NEXT: [[START_LABEL:.Ltmp[0-9]+]]
2017-08-02 05:45:24 +08:00
; CHECK-NEXT: #DEBUG_VALUE: bar:y <- [DW_OP_plus_uconst [[OFFSET]], DW_OP_deref, DW_OP_deref]
2014-06-10 05:53:47 +08:00
; This location should be valid until the end of the function.
2016-03-26 01:54:46 +08:00
; CHECK: movq %rbp, %rsp
; CHECK-NEXT: [[END_LABEL:.Ltmp[0-9]+]]:
2014-06-10 05:53:47 +08:00
; CHECK: .Ldebug_loc{{[0-9]+}}:
; We expect two location ranges for the variable.
[FastISel] Sink local value materializations to first use
Summary:
Local values are constants, global addresses, and stack addresses that
can't be folded into the instruction that uses them. For example, when
storing the address of a global variable into memory, we need to
materialize that address into a register.
FastISel doesn't want to materialize any given local value more than
once, so it generates all local value materialization code at
EmitStartPt, which always dominates the current insertion point. This
allows it to maintain a map of local value registers, and it knows that
the local value area will always dominate the current insertion point.
The downside is that local value instructions are always emitted without
a source location. This is done to prevent jumpy line tables, but it
means that the local value area will be considered part of the previous
statement. Consider this C code:
call1(); // line 1
++global; // line 2
++global; // line 3
call2(&global, &local); // line 4
Today we end up with assembly and line tables like this:
.loc 1 1
callq call1
leaq global(%rip), %rdi
leaq local(%rsp), %rsi
.loc 1 2
addq $1, global(%rip)
.loc 1 3
addq $1, global(%rip)
.loc 1 4
callq call2
The LEA instructions in the local value area have no source location and
are treated as being on line 1. Stepping through the code in a debugger
and correlating it with the assembly won't make much sense, because
these materializations are only required for line 4.
This is actually problematic for the VS debugger "set next statement"
feature, which effectively assumes that there are no registers live
across statement boundaries. By sinking the local value code into the
statement and fixing up the source location, we can make that feature
work. This was filed as https://bugs.llvm.org/show_bug.cgi?id=35975 and
https://crbug.com/793819.
This change is obviously not enough to make this feature work reliably
in all cases, but I felt that it was worth doing anyway because it
usually generates smaller, more comprehensible -O0 code. I measured a
0.12% regression in code generation time with LLC on the sqlite3
amalgamation, so I think this is worth doing.
There are some special cases worth calling out in the commit message:
1. local values materialized for phis
2. local values used by no-op casts
3. dead local value code
Local values can be materialized for phis, and this does not show up as
a vreg use in MachineRegisterInfo. In this case, if there are no other
uses, this patch sinks the value to the first terminator, EH label, or
the end of the BB if nothing else exists.
Local values may also be used by no-op casts, which adds the register to
the RegFixups table. Without reversing the RegFixups map direction, we
don't have enough information to sink these instructions.
Lastly, if the local value register has no other uses, we can delete it.
This comes up when fastisel tries two instruction selection approaches
and the first materializes the value but fails and the second succeeds
without using the local value.
Reviewers: aprantl, dblaikie, qcolombet, MatzeB, vsk, echristo
Subscribers: dotdash, chandlerc, hans, sdardis, amccarth, javed.absar, zturner, llvm-commits, hiraditya
Differential Revision: https://reviews.llvm.org/D43093
llvm-svn: 327581
2018-03-15 05:54:21 +08:00
; First, its address is stored in %rcx:
2014-10-21 09:17:30 +08:00
; CHECK: .quad .Lfunc_begin0-.Lfunc_begin0
; CHECK-NEXT: .quad [[START_LABEL]]-.Lfunc_begin0
[FastISel] Sink local value materializations to first use
Summary:
Local values are constants, global addresses, and stack addresses that
can't be folded into the instruction that uses them. For example, when
storing the address of a global variable into memory, we need to
materialize that address into a register.
FastISel doesn't want to materialize any given local value more than
once, so it generates all local value materialization code at
EmitStartPt, which always dominates the current insertion point. This
allows it to maintain a map of local value registers, and it knows that
the local value area will always dominate the current insertion point.
The downside is that local value instructions are always emitted without
a source location. This is done to prevent jumpy line tables, but it
means that the local value area will be considered part of the previous
statement. Consider this C code:
call1(); // line 1
++global; // line 2
++global; // line 3
call2(&global, &local); // line 4
Today we end up with assembly and line tables like this:
.loc 1 1
callq call1
leaq global(%rip), %rdi
leaq local(%rsp), %rsi
.loc 1 2
addq $1, global(%rip)
.loc 1 3
addq $1, global(%rip)
.loc 1 4
callq call2
The LEA instructions in the local value area have no source location and
are treated as being on line 1. Stepping through the code in a debugger
and correlating it with the assembly won't make much sense, because
these materializations are only required for line 4.
This is actually problematic for the VS debugger "set next statement"
feature, which effectively assumes that there are no registers live
across statement boundaries. By sinking the local value code into the
statement and fixing up the source location, we can make that feature
work. This was filed as https://bugs.llvm.org/show_bug.cgi?id=35975 and
https://crbug.com/793819.
This change is obviously not enough to make this feature work reliably
in all cases, but I felt that it was worth doing anyway because it
usually generates smaller, more comprehensible -O0 code. I measured a
0.12% regression in code generation time with LLC on the sqlite3
amalgamation, so I think this is worth doing.
There are some special cases worth calling out in the commit message:
1. local values materialized for phis
2. local values used by no-op casts
3. dead local value code
Local values can be materialized for phis, and this does not show up as
a vreg use in MachineRegisterInfo. In this case, if there are no other
uses, this patch sinks the value to the first terminator, EH label, or
the end of the BB if nothing else exists.
Local values may also be used by no-op casts, which adds the register to
the RegFixups table. Without reversing the RegFixups map direction, we
don't have enough information to sink these instructions.
Lastly, if the local value register has no other uses, we can delete it.
This comes up when fastisel tries two instruction selection approaches
and the first materializes the value but fails and the second succeeds
without using the local value.
Reviewers: aprantl, dblaikie, qcolombet, MatzeB, vsk, echristo
Subscribers: dotdash, chandlerc, hans, sdardis, amccarth, javed.absar, zturner, llvm-commits, hiraditya
Differential Revision: https://reviews.llvm.org/D43093
llvm-svn: 327581
2018-03-15 05:54:21 +08:00
; CHECK: DW_OP_breg2
[dwarfdump] Pretty print location expressions and location lists
Summary:
Based on Fred's patch here: https://reviews.llvm.org/D6771
I can't seem to commandeer the old review, so I'm creating a new one.
With that change the locations exrpessions are pretty printed inline in the
DIE tree. The output looks like this for debug_loc entries:
DW_AT_location [DW_FORM_data4] (0x00000000
0x0000000000000001 - 0x000000000000000b: DW_OP_consts +3
0x000000000000000b - 0x0000000000000012: DW_OP_consts +7
0x0000000000000012 - 0x000000000000001b: DW_OP_reg0 RAX, DW_OP_piece 0x4
0x000000000000001b - 0x0000000000000024: DW_OP_breg5 RDI+0)
And like this for debug_loc.dwo entries:
DW_AT_location [DW_FORM_sec_offset] (0x00000000
Addr idx 2 (w/ length 190): DW_OP_consts +0, DW_OP_stack_value
Addr idx 3 (w/ length 23): DW_OP_reg0 RAX, DW_OP_piece 0x4)
Simple locations without ranges are printed inline:
DW_AT_location [DW_FORM_block1] (DW_OP_reg4 RSI, DW_OP_piece 0x4, DW_OP_bit_piece 0x20 0x0)
The debug_loc(.dwo) dumping in changed accordingly to factor the code.
Reviewers: dblaikie, aprantl, friss
Subscribers: mgorny, javed.absar, hiraditya, llvm-commits, JDevlieghere
Differential Revision: https://reviews.llvm.org/D37123
llvm-svn: 312042
2017-08-30 05:41:21 +08:00
; DWARF: DW_TAG_formal_parameter
; DWARF: DW_AT_location
[FastISel] Sink local value materializations to first use
Summary:
Local values are constants, global addresses, and stack addresses that
can't be folded into the instruction that uses them. For example, when
storing the address of a global variable into memory, we need to
materialize that address into a register.
FastISel doesn't want to materialize any given local value more than
once, so it generates all local value materialization code at
EmitStartPt, which always dominates the current insertion point. This
allows it to maintain a map of local value registers, and it knows that
the local value area will always dominate the current insertion point.
The downside is that local value instructions are always emitted without
a source location. This is done to prevent jumpy line tables, but it
means that the local value area will be considered part of the previous
statement. Consider this C code:
call1(); // line 1
++global; // line 2
++global; // line 3
call2(&global, &local); // line 4
Today we end up with assembly and line tables like this:
.loc 1 1
callq call1
leaq global(%rip), %rdi
leaq local(%rsp), %rsi
.loc 1 2
addq $1, global(%rip)
.loc 1 3
addq $1, global(%rip)
.loc 1 4
callq call2
The LEA instructions in the local value area have no source location and
are treated as being on line 1. Stepping through the code in a debugger
and correlating it with the assembly won't make much sense, because
these materializations are only required for line 4.
This is actually problematic for the VS debugger "set next statement"
feature, which effectively assumes that there are no registers live
across statement boundaries. By sinking the local value code into the
statement and fixing up the source location, we can make that feature
work. This was filed as https://bugs.llvm.org/show_bug.cgi?id=35975 and
https://crbug.com/793819.
This change is obviously not enough to make this feature work reliably
in all cases, but I felt that it was worth doing anyway because it
usually generates smaller, more comprehensible -O0 code. I measured a
0.12% regression in code generation time with LLC on the sqlite3
amalgamation, so I think this is worth doing.
There are some special cases worth calling out in the commit message:
1. local values materialized for phis
2. local values used by no-op casts
3. dead local value code
Local values can be materialized for phis, and this does not show up as
a vreg use in MachineRegisterInfo. In this case, if there are no other
uses, this patch sinks the value to the first terminator, EH label, or
the end of the BB if nothing else exists.
Local values may also be used by no-op casts, which adds the register to
the RegFixups table. Without reversing the RegFixups map direction, we
don't have enough information to sink these instructions.
Lastly, if the local value register has no other uses, we can delete it.
This comes up when fastisel tries two instruction selection approaches
and the first materializes the value but fails and the second succeeds
without using the local value.
Reviewers: aprantl, dblaikie, qcolombet, MatzeB, vsk, echristo
Subscribers: dotdash, chandlerc, hans, sdardis, amccarth, javed.absar, zturner, llvm-commits, hiraditya
Differential Revision: https://reviews.llvm.org/D43093
llvm-svn: 327581
2018-03-15 05:54:21 +08:00
; DWARF-NEXT: [{{.*}}, {{.*}}): DW_OP_breg2 RCX+0, DW_OP_deref
2014-06-10 05:53:47 +08:00
; Then it's addressed via %rsp:
2014-10-21 09:17:30 +08:00
; CHECK: .quad [[START_LABEL]]-.Lfunc_begin0
2016-03-26 01:54:46 +08:00
; CHECK-NEXT: .quad [[END_LABEL]]-.Lfunc_begin0
2014-06-10 05:53:47 +08:00
; CHECK: DW_OP_breg7
; CHECK-NEXT: [[OFFSET]]
; CHECK: DW_OP_deref
2018-01-16 19:17:57 +08:00
; DWARF-NEXT: [{{.*}}, {{.*}}): DW_OP_breg7 RSP+{{[0-9]+}}, DW_OP_deref, DW_OP_deref)
2014-06-10 05:53:47 +08:00
; ModuleID = 'test.cc'
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
target triple = "x86_64-unknown-linux-gnu"
@llvm.global_ctors = appending global [ 1 x { i32 , void ( ) * } ] [ { i32 , void ( ) * } { i32 1 , void ( ) * @asan.module_ctor } ]
@__asan_option_detect_stack_use_after_return = external global i32
@__asan_gen_ = private unnamed_addr constant [ 16 x i8 ] c "1 32 4 6 y.addr\00" , align 1
; Function Attrs: nounwind sanitize_address uwtable
2015-11-06 06:03:56 +08:00
define i32 @_Z3bari ( i32 %y ) #0 !dbg !4 {
2014-06-10 05:53:47 +08:00
entry:
%MyAlloca = alloca [ 64 x i8 ] , align 32
%0 = ptrtoint [ 64 x i8 ] * %MyAlloca to i64
2015-02-28 05:17:42 +08:00
%1 = load i32 , i32 * @__asan_option_detect_stack_use_after_return
2014-06-10 05:53:47 +08:00
%2 = icmp ne i32 %1 , 0
br i1 %2 , label %3 , label %5
; <label>:3 ; preds = %entry
%4 = call i64 @__asan_stack_malloc_0 ( i64 64 , i64 %0 )
br label %5
; <label>:5 ; preds = %entry, %3
%6 = phi i64 [ %0 , %entry ] , [ %4 , %3 ]
%7 = add i64 %6 , 32
%8 = inttoptr i64 %7 to i32 *
%9 = inttoptr i64 %6 to i64 *
store i64 1102416563 , i64 * %9
%10 = add i64 %6 , 8
%11 = inttoptr i64 %10 to i64 *
store i64 ptrtoint ( [ 16 x i8 ] * @__asan_gen_ to i64 ) , i64 * %11
%12 = add i64 %6 , 16
%13 = inttoptr i64 %12 to i64 *
store i64 ptrtoint ( i32 ( i32 ) * @_Z3bari to i64 ) , i64 * %13
%14 = lshr i64 %6 , 3
%15 = add i64 %14 , 2147450880
%16 = add i64 %15 , 0
%17 = inttoptr i64 %16 to i64 *
store i64 -868083100587789839 , i64 * %17
%18 = ptrtoint i32 * %8 to i64
%19 = lshr i64 %18 , 3
%20 = add i64 %19 , 2147450880
%21 = inttoptr i64 %20 to i8 *
2015-02-28 05:17:42 +08:00
%22 = load i8 , i8 * %21
2014-06-10 05:53:47 +08:00
%23 = icmp ne i8 %22 , 0
2015-04-30 00:38:44 +08:00
call void @llvm.dbg.declare ( metadata i32 * %8 , metadata !12 , metadata !14 ) , !dbg !DILocation ( scope: !4 )
2014-06-10 05:53:47 +08:00
br i1 %23 , label %24 , label %30
; <label>:24 ; preds = %5
%25 = and i64 %18 , 7
%26 = add i64 %25 , 3
%27 = trunc i64 %26 to i8
%28 = icmp sge i8 %27 , %22
br i1 %28 , label %29 , label %30
; <label>:29 ; preds = %24
call void @__asan_report_store4 ( i64 %18 )
call void asm sideeffect "" , "" ( )
unreachable
; <label>:30 ; preds = %24, %5
store i32 %y , i32 * %8 , align 4
%31 = ptrtoint i32 * %8 to i64 , !dbg !13
%32 = lshr i64 %31 , 3 , !dbg !13
%33 = add i64 %32 , 2147450880 , !dbg !13
%34 = inttoptr i64 %33 to i8 * , !dbg !13
2015-02-28 05:17:42 +08:00
%35 = load i8 , i8 * %34 , !dbg !13
2014-06-10 05:53:47 +08:00
%36 = icmp ne i8 %35 , 0 , !dbg !13
br i1 %36 , label %37 , label %43 , !dbg !13
; <label>:37 ; preds = %30
%38 = and i64 %31 , 7 , !dbg !13
%39 = add i64 %38 , 3 , !dbg !13
%40 = trunc i64 %39 to i8 , !dbg !13
%41 = icmp sge i8 %40 , %35 , !dbg !13
br i1 %41 , label %42 , label %43
; <label>:42 ; preds = %37
call void @__asan_report_load4 ( i64 %31 ) , !dbg !13
call void asm sideeffect "" , "" ( )
unreachable
; <label>:43 ; preds = %37, %30
2015-02-28 05:17:42 +08:00
%44 = load i32 , i32 * %8 , align 4 , !dbg !13
2014-06-10 05:53:47 +08:00
%add = add nsw i32 %44 , 2 , !dbg !13
store i64 1172321806 , i64 * %9 , !dbg !13
%45 = icmp ne i64 %6 , %0 , !dbg !13
br i1 %45 , label %46 , label %53 , !dbg !13
; <label>:46 ; preds = %43
%47 = add i64 %15 , 0 , !dbg !13
%48 = inttoptr i64 %47 to i64 * , !dbg !13
store i64 -723401728380766731 , i64 * %48 , !dbg !13
%49 = add i64 %6 , 56 , !dbg !13
%50 = inttoptr i64 %49 to i64 * , !dbg !13
2015-02-28 05:17:42 +08:00
%51 = load i64 , i64 * %50 , !dbg !13
2014-06-10 05:53:47 +08:00
%52 = inttoptr i64 %51 to i8 * , !dbg !13
store i8 0 , i8 * %52 , !dbg !13
br label %56 , !dbg !13
; <label>:53 ; preds = %43
%54 = add i64 %15 , 0 , !dbg !13
%55 = inttoptr i64 %54 to i64 * , !dbg !13
store i64 0 , i64 * %55 , !dbg !13
br label %56 , !dbg !13
; <label>:56 ; preds = %53, %46
ret i32 %add , !dbg !13
}
; 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.
llvm-svn: 218787
2014-10-02 02:55:02 +08:00
declare void @llvm.dbg.declare ( metadata , metadata , metadata ) #1
2014-06-10 05:53:47 +08:00
define internal void @asan.module_ctor ( ) {
call void @__asan_init_v3 ( )
ret void
}
declare void @__asan_init_v3 ( )
declare void @__asan_report_load4 ( i64 )
declare void @__asan_report_store4 ( i64 )
declare i64 @__asan_stack_malloc_0 ( i64 , i64 )
attributes #0 = { nounwind sanitize_address uwtable "less-precise-fpmad" = "false" "no-frame-pointer-elim" = "true" "no-frame-pointer-elim-non-leaf" "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 = ! { !9 , !10 }
!llvm.ident = ! { !11 }
2016-04-15 23:57:41 +08:00
!0 = distinct !DICompileUnit ( language: D W _ L A N G _ C _ p l u s _ p l u s , producer: "clang version 3.5.0 (209308)" , isOptimized: false , emissionKind: F u l l D e b u g , file: !1 , enums: !2 , retainedTypes: !2 , globals: !2 , imports: !2 )
2015-04-30 00:38:44 +08:00
!1 = !DIFile ( filename: "test.cc" , directory: "/llvm_cmake_gcc" )
IR: Make metadata typeless in assembly
Now that `Metadata` is typeless, reflect that in the assembly. These
are the matching assembly changes for the metadata/value split in
r223802.
- Only use the `metadata` type when referencing metadata from a call
intrinsic -- i.e., only when it's used as a `Value`.
- Stop pretending that `ValueAsMetadata` is wrapped in an `MDNode`
when referencing it from call intrinsics.
So, assembly like this:
define @foo(i32 %v) {
call void @llvm.foo(metadata !{i32 %v}, metadata !0)
call void @llvm.foo(metadata !{i32 7}, metadata !0)
call void @llvm.foo(metadata !1, metadata !0)
call void @llvm.foo(metadata !3, metadata !0)
call void @llvm.foo(metadata !{metadata !3}, metadata !0)
ret void, !bar !2
}
!0 = metadata !{metadata !2}
!1 = metadata !{i32* @global}
!2 = metadata !{metadata !3}
!3 = metadata !{}
turns into this:
define @foo(i32 %v) {
call void @llvm.foo(metadata i32 %v, metadata !0)
call void @llvm.foo(metadata i32 7, metadata !0)
call void @llvm.foo(metadata i32* @global, metadata !0)
call void @llvm.foo(metadata !3, metadata !0)
call void @llvm.foo(metadata !{!3}, metadata !0)
ret void, !bar !2
}
!0 = !{!2}
!1 = !{i32* @global}
!2 = !{!3}
!3 = !{}
I wrote an upgrade script that handled almost all of the tests in llvm
and many of the tests in cfe (even handling many `CHECK` lines). I've
attached it (or will attach it in a moment if you're speedy) to PR21532
to help everyone update their out-of-tree testcases.
This is part of PR21532.
llvm-svn: 224257
2014-12-16 03:07:53 +08:00
!2 = ! { }
[DebugInfo] Add DILabel metadata and intrinsic llvm.dbg.label.
In order to set breakpoints on labels and list source code around
labels, we need collect debug information for labels, i.e., label
name, the function label belong, line number in the file, and the
address label located. In order to keep these information in LLVM
IR and to allow backend to generate debug information correctly.
We create a new kind of metadata for labels, DILabel. The format
of DILabel is
!DILabel(scope: !1, name: "foo", file: !2, line: 3)
We hope to keep debug information as much as possible even the
code is optimized. So, we create a new kind of intrinsic for label
metadata to avoid the metadata is eliminated with basic block.
The intrinsic will keep existing if we keep it from optimized out.
The format of the intrinsic is
llvm.dbg.label(metadata !1)
It has only one argument, that is the DILabel metadata. The
intrinsic will follow the label immediately. Backend could get the
label metadata through the intrinsic's parameter.
We also create DIBuilder API for labels to be used by Frontend.
Frontend could use createLabel() to allocate DILabel objects, and use
insertLabel() to insert llvm.dbg.label intrinsic in LLVM IR.
Differential Revision: https://reviews.llvm.org/D45024
Patch by Hsiangkai Wang.
llvm-svn: 331841
2018-05-09 10:40:45 +08:00
!4 = distinct !DISubprogram ( name: "bar" , linkageName: "_Z3bari" , line: 1 , isLocal: false , isDefinition: true , virtualIndex: 6 , flags: D I F l a g P r o t o t y p e d , isOptimized: false , unit: !0 , scopeLine: 1 , file: !1 , scope: !5 , type: !6 , retainedNodes: !2 )
2015-04-30 00:38:44 +08:00
!5 = !DIFile ( filename: "test.cc" , directory: "/llvm_cmake_gcc" )
!6 = !DISubroutineType ( types: !7 )
IR: Make metadata typeless in assembly
Now that `Metadata` is typeless, reflect that in the assembly. These
are the matching assembly changes for the metadata/value split in
r223802.
- Only use the `metadata` type when referencing metadata from a call
intrinsic -- i.e., only when it's used as a `Value`.
- Stop pretending that `ValueAsMetadata` is wrapped in an `MDNode`
when referencing it from call intrinsics.
So, assembly like this:
define @foo(i32 %v) {
call void @llvm.foo(metadata !{i32 %v}, metadata !0)
call void @llvm.foo(metadata !{i32 7}, metadata !0)
call void @llvm.foo(metadata !1, metadata !0)
call void @llvm.foo(metadata !3, metadata !0)
call void @llvm.foo(metadata !{metadata !3}, metadata !0)
ret void, !bar !2
}
!0 = metadata !{metadata !2}
!1 = metadata !{i32* @global}
!2 = metadata !{metadata !3}
!3 = metadata !{}
turns into this:
define @foo(i32 %v) {
call void @llvm.foo(metadata i32 %v, metadata !0)
call void @llvm.foo(metadata i32 7, metadata !0)
call void @llvm.foo(metadata i32* @global, metadata !0)
call void @llvm.foo(metadata !3, metadata !0)
call void @llvm.foo(metadata !{!3}, metadata !0)
ret void, !bar !2
}
!0 = !{!2}
!1 = !{i32* @global}
!2 = !{!3}
!3 = !{}
I wrote an upgrade script that handled almost all of the tests in llvm
and many of the tests in cfe (even handling many `CHECK` lines). I've
attached it (or will attach it in a moment if you're speedy) to PR21532
to help everyone update their out-of-tree testcases.
This is part of PR21532.
llvm-svn: 224257
2014-12-16 03:07:53 +08:00
!7 = ! { !8 , !8 }
2015-04-30 00:38:44 +08:00
!8 = !DIBasicType ( tag: D W _ T A G _ b a s e _ type , name: "int" , size: 32 , align: 32 , encoding: D W _ A T E _ s i g n e d )
IR: Make metadata typeless in assembly
Now that `Metadata` is typeless, reflect that in the assembly. These
are the matching assembly changes for the metadata/value split in
r223802.
- Only use the `metadata` type when referencing metadata from a call
intrinsic -- i.e., only when it's used as a `Value`.
- Stop pretending that `ValueAsMetadata` is wrapped in an `MDNode`
when referencing it from call intrinsics.
So, assembly like this:
define @foo(i32 %v) {
call void @llvm.foo(metadata !{i32 %v}, metadata !0)
call void @llvm.foo(metadata !{i32 7}, metadata !0)
call void @llvm.foo(metadata !1, metadata !0)
call void @llvm.foo(metadata !3, metadata !0)
call void @llvm.foo(metadata !{metadata !3}, metadata !0)
ret void, !bar !2
}
!0 = metadata !{metadata !2}
!1 = metadata !{i32* @global}
!2 = metadata !{metadata !3}
!3 = metadata !{}
turns into this:
define @foo(i32 %v) {
call void @llvm.foo(metadata i32 %v, metadata !0)
call void @llvm.foo(metadata i32 7, metadata !0)
call void @llvm.foo(metadata i32* @global, metadata !0)
call void @llvm.foo(metadata !3, metadata !0)
call void @llvm.foo(metadata !{!3}, metadata !0)
ret void, !bar !2
}
!0 = !{!2}
!1 = !{i32* @global}
!2 = !{!3}
!3 = !{}
I wrote an upgrade script that handled almost all of the tests in llvm
and many of the tests in cfe (even handling many `CHECK` lines). I've
attached it (or will attach it in a moment if you're speedy) to PR21532
to help everyone update their out-of-tree testcases.
This is part of PR21532.
llvm-svn: 224257
2014-12-16 03:07:53 +08:00
!9 = ! { i32 2 , !"Dwarf Version" , i32 4 }
2015-03-04 01:24:31 +08:00
!10 = ! { i32 2 , !"Debug Info Version" , i32 3 }
IR: Make metadata typeless in assembly
Now that `Metadata` is typeless, reflect that in the assembly. These
are the matching assembly changes for the metadata/value split in
r223802.
- Only use the `metadata` type when referencing metadata from a call
intrinsic -- i.e., only when it's used as a `Value`.
- Stop pretending that `ValueAsMetadata` is wrapped in an `MDNode`
when referencing it from call intrinsics.
So, assembly like this:
define @foo(i32 %v) {
call void @llvm.foo(metadata !{i32 %v}, metadata !0)
call void @llvm.foo(metadata !{i32 7}, metadata !0)
call void @llvm.foo(metadata !1, metadata !0)
call void @llvm.foo(metadata !3, metadata !0)
call void @llvm.foo(metadata !{metadata !3}, metadata !0)
ret void, !bar !2
}
!0 = metadata !{metadata !2}
!1 = metadata !{i32* @global}
!2 = metadata !{metadata !3}
!3 = metadata !{}
turns into this:
define @foo(i32 %v) {
call void @llvm.foo(metadata i32 %v, metadata !0)
call void @llvm.foo(metadata i32 7, metadata !0)
call void @llvm.foo(metadata i32* @global, metadata !0)
call void @llvm.foo(metadata !3, metadata !0)
call void @llvm.foo(metadata !{!3}, metadata !0)
ret void, !bar !2
}
!0 = !{!2}
!1 = !{i32* @global}
!2 = !{!3}
!3 = !{}
I wrote an upgrade script that handled almost all of the tests in llvm
and many of the tests in cfe (even handling many `CHECK` lines). I've
attached it (or will attach it in a moment if you're speedy) to PR21532
to help everyone update their out-of-tree testcases.
This is part of PR21532.
llvm-svn: 224257
2014-12-16 03:07:53 +08:00
!11 = ! { !"clang version 3.5.0 (209308)" }
2015-08-01 02:58:39 +08:00
!12 = !DILocalVariable ( name: "y" , line: 1 , arg: 1 , scope: !4 , file: !5 , type: !8 )
2015-04-30 00:38:44 +08:00
!13 = !DILocation ( line: 2 , scope: !4 )
!14 = !DIExpression ( D W _ O P _ d e r e f )
