This is an archive of the discontinued LLVM Phabricator instance.

Differential D132422

[BOLT][DWARF] Fix updating CU that has no entry in .debug_addr
ClosedPublic

Authored by ayermolo on Aug 22 2022, 5:09 PM.

Details

Reviewers
rafauler
Amir
maksfb
jdoerfert
Commits
rG1ee74064e083: [BOLT][DWARF] Fix updating CU that has no entry in .debug_addr
Summary

We were trying to process .debug_addr for CU that doesn't have it. This resulted
in assert. Example came from GCC that also doesn't use DW_OP_addrx in
DW_FORM_exprloc.

Diff Detail

Event Timeline

ayermolo created this revision.Aug 22 2022, 5:09 PM
Herald added a reviewer: rafauler. · View Herald TranscriptAug 22 2022, 5:09 PM
Herald added a reviewer: Amir. · View Herald Transcript
Herald added a reviewer: maksfb. · View Herald Transcript
Herald added a project: Restricted Project. · View Herald Transcript
Herald added subscribers: hoy, modimo, wenlei. · View Herald Transcript
ayermolo requested review of this revision.Aug 22 2022, 5:09 PM
Herald added a reviewer: jdoerfert. · View Herald TranscriptAug 22 2022, 5:09 PM
Herald added a project: Restricted Project. · View Herald Transcript
Herald added subscribers: llvm-commits, yota9, sstefan1. · View Herald Transcript
Harbormaster completed remote builds in B182714: Diff 454656.Aug 22 2022, 5:18 PM
ayermolo updated this revision to Diff 455708.Aug 25 2022, 1:50 PM

type fix

Harbormaster completed remote builds in B183469: Diff 455708.Aug 25 2022, 1:56 PM
maksfb accepted this revision.Aug 25 2022, 2:01 PM

LGTM

This revision is now accepted and ready to land.Aug 25 2022, 2:01 PM
Closed by commit rG1ee74064e083: [BOLT][DWARF] Fix updating CU that has no entry in .debug_addr (authored by ayermolo). · Explain WhyAug 25 2022, 5:03 PM
This revision was automatically updated to reflect the committed changes.
ayermolo added a commit: rG1ee74064e083: [BOLT][DWARF] Fix updating CU that has no entry in .debug_addr.
dblaikie added a subscriber: dblaikie.Sep 8 2022, 12:23 PM
Herald added a subscriber: treapster. · View Herald TranscriptSep 8 2022, 12:23 PM

Revision Contents

PathSize
bolt/
include/
bolt/
Core/
5 lines
lib/
Rewrite/
3 lines
test/
X86/
Inputs/
177 lines
161 lines
20 lines

Diff 455764

bolt/include/bolt/Core/DebugData.h

Show First 20 Lines • Show All 313 Lines • ▼ Show 20 Linespublic:
/// Creates consolidated .debug_addr section, and builds DWOID to offset map. /// Creates consolidated .debug_addr section, and builds DWOID to offset map.
virtual AddressSectionBuffer finalize(); virtual AddressSectionBuffer finalize();
/// Given DWARFUnit \p Unit returns offset of this CU in to .debug_addr /// Given DWARFUnit \p Unit returns offset of this CU in to .debug_addr
/// section. /// section.
virtual uint64_t getOffset(DWARFUnit &Unit); virtual uint64_t getOffset(DWARFUnit &Unit);
/// Returns True if CU exists in the DebugAddrWriter.
bool doesCUExist(DWARFUnit &Unit) {
return DWOIdToOffsetMap.count(getCUID(Unit)) > 0;
}
/// Returns False if .debug_addr section was created.. /// Returns False if .debug_addr section was created..
bool isInitialized() const { return !AddressMaps.empty(); } bool isInitialized() const { return !AddressMaps.empty(); }
protected: protected:
class AddressForDWOCU { class AddressForDWOCU {
public: public:
AddressToIndexMap::iterator find(uint64_t Adddress) { AddressToIndexMap::iterator find(uint64_t Adddress) {
return AddressToIndex.find(Adddress); return AddressToIndex.find(Adddress);
▲ Show 20 Lines • Show All 855 Lines • Show Last 20 Lines

bolt/lib/Rewrite/DWARFRewriter.cpp

Show First 20 Lines • Show All 1,041 Lines • ▼ Show 20 Lines for (auto &CU : BC.DwCtx->compile_units()) {
findAttributeInfo(DIE, dwarf::DW_AT_GNU_addr_base); findAttributeInfo(DIE, dwarf::DW_AT_GNU_addr_base);
Optional<AttrInfo> AttrVal = Optional<AttrInfo> AttrVal =
findAttributeInfo(DIE, dwarf::DW_AT_addr_base); findAttributeInfo(DIE, dwarf::DW_AT_addr_base);
// For cases where Skeleton CU does not have DW_AT_GNU_addr_base // For cases where Skeleton CU does not have DW_AT_GNU_addr_base
if (!AttrValGnu && CU->getVersion() < 5) if (!AttrValGnu && CU->getVersion() < 5)
continue; continue;
if (!AttrVal && CU->getVersion() >= 5 && !AddrWriter->doesCUExist(*CU))
continue;
Offset = AddrWriter->getOffset(*CU); Offset = AddrWriter->getOffset(*CU);
if (AttrValGnu) { if (AttrValGnu) {
AttrOffset = AttrValGnu->Offset; AttrOffset = AttrValGnu->Offset;
Size = AttrValGnu->Size; Size = AttrValGnu->Size;
} }
if (AttrVal) { if (AttrVal) {
▲ Show 20 Lines • Show All 882 Lines • Show Last 20 Lines

bolt/test/X86/Inputs/dwarf5-cu-no-debug-addr-helper.s

  • This file was added.
# clang++ -g2 -gdwarf-5 helper.cpp -S
# __attribute__ ((used))
# static struct {
# int i;
# } Foo;
# Manually modifed to remove DW_AT_addr_base and DW_AT_location.
.text
.file "helper.cpp"
.file 0 "." "helper.cpp" md5 0x18c63f39ce5af555b264ad09798e35d5
.type _ZL3Foo,@object # @_ZL3Foo
.local _ZL3Foo
.comm _ZL3Foo,4,4
.section .debug_abbrev,"",@progbits
.byte 1 # Abbreviation Code
.byte 17 # DW_TAG_compile_unit
.byte 1 # DW_CHILDREN_yes
.byte 37 # DW_AT_producer
.byte 37 # DW_FORM_strx1
.byte 19 # DW_AT_language
.byte 5 # DW_FORM_data2
.byte 3 # DW_AT_name
.byte 37 # DW_FORM_strx1
.byte 114 # DW_AT_str_offsets_base
.byte 23 # DW_FORM_sec_offset
.byte 16 # DW_AT_stmt_list
.byte 23 # DW_FORM_sec_offset
.byte 27 # DW_AT_comp_dir
.byte 37 # DW_FORM_strx1
#.byte 115 # DW_AT_addr_base manually removed
#.byte 23 # DW_FORM_sec_offset
.byte 0 # EOM(1)
.byte 0 # EOM(2)
.byte 2 # Abbreviation Code
.byte 52 # DW_TAG_variable
.byte 0 # DW_CHILDREN_no
.byte 3 # DW_AT_name
.byte 37 # DW_FORM_strx1
.byte 73 # DW_AT_type
.byte 19 # DW_FORM_ref4
.byte 58 # DW_AT_decl_file
.byte 11 # DW_FORM_data1
.byte 59 # DW_AT_decl_line
.byte 11 # DW_FORM_data1
#.byte 2 # DW_AT_location manually removed
#.byte 24 # DW_FORM_exprloc
.byte 110 # DW_AT_linkage_name
.byte 37 # DW_FORM_strx1
.byte 0 # EOM(1)
.byte 0 # EOM(2)
.byte 3 # Abbreviation Code
.byte 19 # DW_TAG_structure_type
.byte 1 # DW_CHILDREN_yes
.byte 54 # DW_AT_calling_convention
.byte 11 # DW_FORM_data1
.byte 11 # DW_AT_byte_size
.byte 11 # DW_FORM_data1
.byte 58 # DW_AT_decl_file
.byte 11 # DW_FORM_data1
.byte 59 # DW_AT_decl_line
.byte 11 # DW_FORM_data1
.byte 0 # EOM(1)
.byte 0 # EOM(2)
.byte 4 # Abbreviation Code
.byte 13 # DW_TAG_member
.byte 0 # DW_CHILDREN_no
.byte 3 # DW_AT_name
.byte 37 # DW_FORM_strx1
.byte 73 # DW_AT_type
.byte 19 # DW_FORM_ref4
.byte 58 # DW_AT_decl_file
.byte 11 # DW_FORM_data1
.byte 59 # DW_AT_decl_line
.byte 11 # DW_FORM_data1
.byte 56 # DW_AT_data_member_location
.byte 11 # DW_FORM_data1
.byte 0 # EOM(1)
.byte 0 # EOM(2)
.byte 5 # Abbreviation Code
.byte 36 # DW_TAG_base_type
.byte 0 # DW_CHILDREN_no
.byte 3 # DW_AT_name
.byte 37 # DW_FORM_strx1
.byte 62 # DW_AT_encoding
.byte 11 # DW_FORM_data1
.byte 11 # DW_AT_byte_size
.byte 11 # DW_FORM_data1
.byte 0 # EOM(1)
.byte 0 # EOM(2)
.byte 0 # EOM(3)
.section .debug_info,"",@progbits
.Lcu_begin0:
.long .Ldebug_info_end0-.Ldebug_info_start0 # Length of Unit
.Ldebug_info_start0:
.short 5 # DWARF version number
.byte 1 # DWARF Unit Type
.byte 8 # Address Size (in bytes)
.long .debug_abbrev # Offset Into Abbrev. Section
.byte 1 # Abbrev [1] 0xc:0x32 DW_TAG_compile_unit
.byte 0 # DW_AT_producer
.short 33 # DW_AT_language
.byte 1 # DW_AT_name
.long .Lstr_offsets_base0 # DW_AT_str_offsets_base
.long .Lline_table_start0 # DW_AT_stmt_list
.byte 2 # DW_AT_comp_dir
#.long .Laddr_table_base0 # DW_AT_addr_base mannually removed
.byte 2 # Abbrev [2] 0x1e:0xc DW_TAG_variable
.byte 3 # DW_AT_name
.long 42 # DW_AT_type
.byte 0 # DW_AT_decl_file
.byte 4 # DW_AT_decl_line
#.byte 2 # DW_AT_location manually removed
#.byte 161
#.byte 0
.byte 6 # DW_AT_linkage_name
.byte 3 # Abbrev [3] 0x2a:0xf DW_TAG_structure_type
.byte 5 # DW_AT_calling_convention
.byte 4 # DW_AT_byte_size
.byte 0 # DW_AT_decl_file
.byte 2 # DW_AT_decl_line
.byte 4 # Abbrev [4] 0x2f:0x9 DW_TAG_member
.byte 4 # DW_AT_name
.long 57 # DW_AT_type
.byte 0 # DW_AT_decl_file
.byte 3 # DW_AT_decl_line
.byte 0 # DW_AT_data_member_location
.byte 0 # End Of Children Mark
.byte 5 # Abbrev [5] 0x39:0x4 DW_TAG_base_type
.byte 5 # DW_AT_name
.byte 5 # DW_AT_encoding
.byte 4 # DW_AT_byte_size
.byte 0 # End Of Children Mark
.Ldebug_info_end0:
.section .debug_str_offsets,"",@progbits
.long 32 # Length of String Offsets Set
.short 5
.short 0
.Lstr_offsets_base0:
.section .debug_str,"MS",@progbits,1
.Linfo_string0:
.asciz "clang version 16.0.0" # string offset=0
.Linfo_string1:
.asciz "helper.cpp" # string offset=134
.Linfo_string2:
.asciz "." # string offset=145
.Linfo_string3:
.asciz "Foo" # string offset=183
.Linfo_string4:
.asciz "i" # string offset=187
.Linfo_string5:
.asciz "int" # string offset=189
.Linfo_string6:
.asciz "_ZL3Foo" # string offset=193
.section .debug_str_offsets,"",@progbits
.long .Linfo_string0
.long .Linfo_string1
.long .Linfo_string2
.long .Linfo_string3
.long .Linfo_string4
.long .Linfo_string5
.long .Linfo_string6
.section .debug_addr,"",@progbits
.long .Ldebug_addr_end0-.Ldebug_addr_start0 # Length of contribution
.Ldebug_addr_start0:
.short 5 # DWARF version number
.byte 8 # Address size
.byte 0 # Segment selector size
.Laddr_table_base0:
.quad _ZL3Foo
.Ldebug_addr_end0:
.ident "clang version 16.0.0"
.section ".note.GNU-stack","",@progbits
.addrsig
.addrsig_sym _ZL3Foo
.section .debug_line,"",@progbits
.Lline_table_start0:

bolt/test/X86/Inputs/dwarf5-cu-no-debug-addr-main.s

  • This file was added.
# clang++ -g2 -gdwarf-5 main.cpp -S
# int main(){
# return 0;
# }
.text
.file "main.cpp"
.globl main # -- Begin function main
.p2align 4, 0x90
.type main,@function
main: # @main
.Lfunc_begin0:
.file 0 "." "main.cpp" md5 0x1dc672a3317ca2ae06986e1275dcc6e7
.loc 0 1 0 # main.cpp:1:0
.cfi_startproc
# %bb.0: # %entry
pushq %rbp
.cfi_def_cfa_offset 16
.cfi_offset %rbp, -16
movq %rsp, %rbp
.cfi_def_cfa_register %rbp
movl $0, -4(%rbp)
.Ltmp0:
.loc 0 2 3 prologue_end # main.cpp:2:3
xorl %eax, %eax
popq %rbp
.cfi_def_cfa %rsp, 8
retq
.Ltmp1:
.Lfunc_end0:
.size main, .Lfunc_end0-main
.cfi_endproc
# -- End function
.section .debug_abbrev,"",@progbits
.byte 1 # Abbreviation Code
.byte 17 # DW_TAG_compile_unit
.byte 1 # DW_CHILDREN_yes
.byte 37 # DW_AT_producer
.byte 37 # DW_FORM_strx1
.byte 19 # DW_AT_language
.byte 5 # DW_FORM_data2
.byte 3 # DW_AT_name
.byte 37 # DW_FORM_strx1
.byte 114 # DW_AT_str_offsets_base
.byte 23 # DW_FORM_sec_offset
.byte 16 # DW_AT_stmt_list
.byte 23 # DW_FORM_sec_offset
.byte 27 # DW_AT_comp_dir
.byte 37 # DW_FORM_strx1
.byte 17 # DW_AT_low_pc
.byte 27 # DW_FORM_addrx
.byte 18 # DW_AT_high_pc
.byte 6 # DW_FORM_data4
.byte 115 # DW_AT_addr_base
.byte 23 # DW_FORM_sec_offset
.byte 0 # EOM(1)
.byte 0 # EOM(2)
.byte 2 # Abbreviation Code
.byte 46 # DW_TAG_subprogram
.byte 0 # DW_CHILDREN_no
.byte 17 # DW_AT_low_pc
.byte 27 # DW_FORM_addrx
.byte 18 # DW_AT_high_pc
.byte 6 # DW_FORM_data4
.byte 64 # DW_AT_frame_base
.byte 24 # DW_FORM_exprloc
.byte 3 # DW_AT_name
.byte 37 # DW_FORM_strx1
.byte 58 # DW_AT_decl_file
.byte 11 # DW_FORM_data1
.byte 59 # DW_AT_decl_line
.byte 11 # DW_FORM_data1
.byte 73 # DW_AT_type
.byte 19 # DW_FORM_ref4
.byte 63 # DW_AT_external
.byte 25 # DW_FORM_flag_present
.byte 0 # EOM(1)
.byte 0 # EOM(2)
.byte 3 # Abbreviation Code
.byte 36 # DW_TAG_base_type
.byte 0 # DW_CHILDREN_no
.byte 3 # DW_AT_name
.byte 37 # DW_FORM_strx1
.byte 62 # DW_AT_encoding
.byte 11 # DW_FORM_data1
.byte 11 # DW_AT_byte_size
.byte 11 # DW_FORM_data1
.byte 0 # EOM(1)
.byte 0 # EOM(2)
.byte 0 # EOM(3)
.section .debug_info,"",@progbits
.Lcu_begin0:
.long .Ldebug_info_end0-.Ldebug_info_start0 # Length of Unit
.Ldebug_info_start0:
.short 5 # DWARF version number
.byte 1 # DWARF Unit Type
.byte 8 # Address Size (in bytes)
.long .debug_abbrev # Offset Into Abbrev. Section
.byte 1 # Abbrev [1] 0xc:0x2b DW_TAG_compile_unit
.byte 0 # DW_AT_producer
.short 33 # DW_AT_language
.byte 1 # DW_AT_name
.long .Lstr_offsets_base0 # DW_AT_str_offsets_base
.long .Lline_table_start0 # DW_AT_stmt_list
.byte 2 # DW_AT_comp_dir
.byte 0 # DW_AT_low_pc
.long .Lfunc_end0-.Lfunc_begin0 # DW_AT_high_pc
.long .Laddr_table_base0 # DW_AT_addr_base
.byte 2 # Abbrev [2] 0x23:0xf DW_TAG_subprogram
.byte 0 # DW_AT_low_pc
.long .Lfunc_end0-.Lfunc_begin0 # DW_AT_high_pc
.byte 1 # DW_AT_frame_base
.byte 86
.byte 3 # DW_AT_name
.byte 0 # DW_AT_decl_file
.byte 1 # DW_AT_decl_line
.long 50 # DW_AT_type
# DW_AT_external
.byte 3 # Abbrev [3] 0x32:0x4 DW_TAG_base_type
.byte 4 # DW_AT_name
.byte 5 # DW_AT_encoding
.byte 4 # DW_AT_byte_size
.byte 0 # End Of Children Mark
.Ldebug_info_end0:
.section .debug_str_offsets,"",@progbits
.long 24 # Length of String Offsets Set
.short 5
.short 0
.Lstr_offsets_base0:
.section .debug_str,"MS",@progbits,1
.Linfo_string0:
.asciz "clang version 16.0.0" # string offset=0
.Linfo_string1:
.asciz "main.cpp" # string offset=134
.Linfo_string2:
.asciz "." # string offset=143
.Linfo_string3:
.asciz "main" # string offset=181
.Linfo_string4:
.asciz "int" # string offset=186
.section .debug_str_offsets,"",@progbits
.long .Linfo_string0
.long .Linfo_string1
.long .Linfo_string2
.long .Linfo_string3
.long .Linfo_string4
.section .debug_addr,"",@progbits
.long .Ldebug_addr_end0-.Ldebug_addr_start0 # Length of contribution
.Ldebug_addr_start0:
.short 5 # DWARF version number
.byte 8 # Address size
.byte 0 # Segment selector size
.Laddr_table_base0:
.quad .Lfunc_begin0
.Ldebug_addr_end0:
.ident "clang version 16.0.0"
.section ".note.GNU-stack","",@progbits
.addrsig
.section .debug_line,"",@progbits
.Lline_table_start0:

bolt/test/X86/dwarf5-cu-no-debug-addr.test

  • This file was added.
# REQUIRES: system-linux
# RUN: llvm-mc -dwarf-version=5 -filetype=obj -triple x86_64-unknown-linux %p/Inputs/dwarf5-cu-no-debug-addr-main.s -o %t1main.o
# RUN: llvm-mc -dwarf-version=5 -filetype=obj -triple x86_64-unknown-linux %p/Inputs/dwarf5-cu-no-debug-addr-helper.s -o %t1helper.o
# RUN: %clang %cflags -dwarf-5 %t1main.o %t1helper.o -o %t.exe -Wl,-q
# RUN: llvm-bolt %t.exe -o %t.bolt --update-debug-sections
# RUN: llvm-dwarfdump --show-form --verbose --debug-info %t.exe | FileCheck --check-prefix=PRECHECK %s
# RUN: llvm-dwarfdump --show-form --verbose --debug-info %t.bolt | FileCheck --check-prefix=POSTCHECK %s
# This tests checks that we handle correctly, don't crash, DWARF5 CUs that does not access .debug_addr.
# PRECHECK: DW_TAG_compile_unit
# PRECHECK: DW_AT_addr_base
# PRECHECK: DW_TAG_compile_unit
# PRECHECK-NOT: DW_AT_addr_base
# POSTCHECK: DW_TAG_compile_unit
# POSTCHECK: DW_AT_addr_base
# POSTCHECK: DW_TAG_compile_unit
# POSTCHECK-NOT: DW_AT_addr_base