[Lldb-commits] [lldb] f3fab39 - [lldb/DWARF] Don't get confused by line sequences with tombstone values
Pavel Labath via lldb-commits
lldb-commits at lists.llvm.org
Fri Jul 17 02:52:12 PDT 2020
Author: Pavel Labath
Date: 2020-07-17T11:51:52+02:00
New Revision: f3fab392f57421a5bdabfb7e40820257d8f637b2
URL: https://github.com/llvm/llvm-project/commit/f3fab392f57421a5bdabfb7e40820257d8f637b2
DIFF: https://github.com/llvm/llvm-project/commit/f3fab392f57421a5bdabfb7e40820257d8f637b2.diff
LOG: [lldb/DWARF] Don't get confused by line sequences with tombstone values
Summary:
With D81784, lld has started debug info resolving relocations to
garbage-collected symbols as -1 (instead of relocation addend). For an
unaware consumer this generated sequences which seemingly wrap the
address space -- their first entry was 0xfffff, but all other entries
were low numbers.
Lldb stores line sequences concatenated into one large vector, sorted by
the first entry, and searched with std::lower_bound. This resulted in
the low-value entries being placed at the end of the vector, which
utterly confused the lower_bound algorithm, and caused it to not find a
match. (Previously, these sequences would be at the start of the vector,
and normally would contain addresses that are far smaller than any real
address we want to look up, so std::lower_bound was fine.)
This patch makes lldb ignore these kinds of sequences completely. It
does that by changing the construction algorithm from iterating over the
rows (as parsed by llvm), to iterating over the sequences. This is
important because the llvm parsed performs validity checks when
constructing the sequence array, whereas the row array contains raw
data.
Reviewers: JDevlieghere, MaskRay
Differential Revision: https://reviews.llvm.org/D83957
Added:
lldb/test/Shell/SymbolFile/DWARF/debug_line-tombstone.s
Modified:
lldb/source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.cpp
Removed:
################################################################################
diff --git a/lldb/source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.cpp b/lldb/source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.cpp
index 9f64e5255fd5..0b7e31ae2d1d 100644
--- a/lldb/source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.cpp
+++ b/lldb/source/Plugins/SymbolFile/DWARF/SymbolFileDWARF.cpp
@@ -1036,18 +1036,20 @@ bool SymbolFileDWARF::ParseLineTable(CompileUnit &comp_unit) {
// FIXME: Rather than parsing the whole line table and then copying it over
// into LLDB, we should explore using a callback to populate the line table
// while we parse to reduce memory usage.
- std::unique_ptr<LineSequence> sequence =
- LineTable::CreateLineSequenceContainer();
std::vector<std::unique_ptr<LineSequence>> sequences;
- for (auto &row : line_table->Rows) {
- LineTable::AppendLineEntryToSequence(
- sequence.get(), row.Address.Address, row.Line, row.Column, row.File,
- row.IsStmt, row.BasicBlock, row.PrologueEnd, row.EpilogueBegin,
- row.EndSequence);
- if (row.EndSequence) {
- sequences.push_back(std::move(sequence));
- sequence = LineTable::CreateLineSequenceContainer();
+ // The Sequences view contains only valid line sequences. Don't iterate over
+ // the Rows directly.
+ for (const llvm::DWARFDebugLine::Sequence &seq : line_table->Sequences) {
+ std::unique_ptr<LineSequence> sequence =
+ LineTable::CreateLineSequenceContainer();
+ for (unsigned idx = seq.FirstRowIndex; idx < seq.LastRowIndex; ++idx) {
+ const llvm::DWARFDebugLine::Row &row = line_table->Rows[idx];
+ LineTable::AppendLineEntryToSequence(
+ sequence.get(), row.Address.Address, row.Line, row.Column, row.File,
+ row.IsStmt, row.BasicBlock, row.PrologueEnd, row.EpilogueBegin,
+ row.EndSequence);
}
+ sequences.push_back(std::move(sequence));
}
std::unique_ptr<LineTable> line_table_up =
diff --git a/lldb/test/Shell/SymbolFile/DWARF/debug_line-tombstone.s b/lldb/test/Shell/SymbolFile/DWARF/debug_line-tombstone.s
new file mode 100644
index 000000000000..53600ac5f4b1
--- /dev/null
+++ b/lldb/test/Shell/SymbolFile/DWARF/debug_line-tombstone.s
@@ -0,0 +1,106 @@
+# This test that we don't get confused by line tables containing a tombstone
+# (-1) value, as produced by recent lld's. Line sequences with the tombstone
+# value should be completely ignored. The tombstone sequence is deliberately
+# longer so that any attempt at an address binary search will likely land inside
+# the sequence.
+
+# RUN: llvm-mc --filetype=obj --triple=x86_64-pc-linux %s -o %t
+# RUN: %lldb -o "image lookup -n main -v" -o "image dump line-table main.cpp" \
+# RUN: -o exit %t | FileCheck %s
+
+# CHECK-LABEL: image lookup -n main -v
+# CHECK: LineEntry: [0x0000000000001000-0x0000000000001001): main.cpp:1
+# CHECK-LABEL: image dump line-table main.cpp
+# CHECK-NEXT: Line table for main.cpp
+# CHECK-NEXT: 0x0000000000001000: main.cpp:1
+# CHECK-NEXT: 0x0000000000001001: main.cpp:1
+# CHECK-EMPTY:
+# CHECK-NEXT: exit
+
+ .text
+.space 0x1000
+main:
+ nop
+.Lmain_end:
+
+ .section .debug_abbrev,"", at progbits
+ .byte 1 # Abbreviation Code
+ .byte 17 # DW_TAG_compile_unit
+ .byte 0 # DW_CHILDREN_no
+ .byte 37 # DW_AT_producer
+ .byte 8 # DW_FORM_string
+ .byte 3 # DW_AT_name
+ .byte 8 # DW_FORM_string
+ .byte 16 # DW_AT_stmt_list
+ .byte 23 # DW_FORM_sec_offset
+ .byte 17 # DW_AT_low_pc
+ .byte 1 # DW_FORM_addr
+ .byte 18 # DW_AT_high_pc
+ .byte 6 # DW_FORM_data4
+ .byte 0 # EOM(1)
+ .byte 0 # EOM(2)
+ .byte 0 # EOM(3)
+
+ .section .debug_info,"", at progbits
+.Lcu_begin0:
+ .long .Ldebug_info_end0-.Ldebug_info_start0 # Length of Unit
+.Ldebug_info_start0:
+ .short 4 # DWARF version number
+ .long 0 # Offset Into Abbrev. Section
+ .byte 8 # Address Size (in bytes)
+ .byte 1 # Abbrev [1] 0xb:0xc4 DW_TAG_compile_unit
+ .asciz "Hand-written DWARF" # DW_AT_producer
+ .asciz "main.cpp" # DW_AT_name
+ .long 0 # DW_AT_stmt_list
+ .quad main-.text # DW_AT_low_pc
+ .long .Lmain_end-main # DW_AT_high_pc
+.Ldebug_info_end0:
+
+.section .debug_line,"", at progbits
+ .long .Llt1_end - .Llt1_start # Length of Unit (DWARF-32 format)
+.Llt1_start:
+ .short 4 # DWARF version number
+ .long .Lprologue1_end-.Lprologue1_start # Length of Prologue
+.Lprologue1_start:
+ .byte 1 # Minimum Instruction Length
+ .byte 1 # Maximum Operations per Instruction
+ .byte 1 # Default is_stmt
+ .byte -5 # Line Base
+ .byte 14 # Line Range
+ .byte 13 # Opcode Base
+ .byte 0, 1, 1, 1, 1, 0, 0, 0, 1, 0, 0, 1 # Standard Opcode Lengths
+ .byte 0
+ .asciz "main.cpp" # File table
+ .byte 0, 0, 0
+ .byte 0
+.Lprologue1_end:
+ .byte 0, 9, 2 # DW_LNE_set_address
+ .quad -1
+ .byte 1 # DW_LNS_copy
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 33 # address += 1, line += 1
+ .byte 2 # DW_LNS_advance_pc
+ .uleb128 1
+ .byte 0, 1, 1 # DW_LNE_end_sequence
+
+ .byte 0, 9, 2 # DW_LNE_set_address
+ .quad main-.text
+ .byte 18 # address += 0, line += 0
+ .byte 2 # DW_LNS_advance_pc
+ .uleb128 1
+ .byte 0, 1, 1 # DW_LNE_end_sequence
+.Llt1_end:
+
More information about the lldb-commits
mailing list