Files
binutils-gdb/gdb/buildsym.h
Lancelot SIX cc96ae7f88 gdb: Add support for DW_LNS_set_prologue_end in line-table
Add support for DW_LNS_set_prologue_end when building line-tables.  This
attribute can be set by the compiler to indicate that an instruction is
an adequate place to set a breakpoint just after the prologue of a
function.

The compiler might set multiple prologue_end, but considering how
current skip_prologue_using_sal works, this commit modifies it to accept
the first instruction with this marker (if any) to be the place where a
breakpoint should be placed to be at the end of the prologue.

The need for this support came from a problematic usecase generated by
hipcc (i.e. clang).  The problem is as follows:  There's a function
(lets call it foo) which covers PC from 0xa800 to 0xa950.  The body of
foo begins with a call to an inlined function, covering from 0xa800 to
0xa94c.   The issue is that when placing a breakpoint at 'foo', GDB
inserts the breakpoint at 0xa818.  The 0x18 offset is what GDB thinks is
foo's first address past the prologue.

Later, when hitting the breakpoint, GDB reports the stop within the
inlined function because the PC falls in its range while the user
expects to stop in FOO.

Looking at the line-table for this location, we have:

    INDEX  LINE   ADDRESS            IS-STMT
    [...]
    14     293    0x000000000000a66c Y
    15     END    0x000000000000a6e0 Y
    16     287    0x000000000000a800 Y
    17     END    0x000000000000a818 Y
    18     287    0x000000000000a824 Y
    [...]

For comparison, let's look at llvm-dwarfdump's output for this CU:

    Address            Line   Column File   ISA Discriminator Flags
    ------------------ ------ ------ ------ --- ------------- -------------
    [...]
    0x000000000000a66c    293     12      2   0             0  is_stmt
    0x000000000000a6e0     96     43     82   0             0  is_stmt
    0x000000000000a6f8    102     18     82   0             0  is_stmt
    0x000000000000a70c    102     24     82   0             0
    0x000000000000a710    102     18     82   0             0
    0x000000000000a72c    101     16     82   0             0  is_stmt
    0x000000000000a73c   2915     50     83   0             0  is_stmt
    0x000000000000a74c    110      1      1   0             0  is_stmt
    0x000000000000a750    110      1      1   0             0  is_stmt end_sequence
    0x000000000000a800    107      0      1   0             0  is_stmt
    0x000000000000a800    287     12      2   0             0  is_stmt prologue_end
    0x000000000000a818    114     59     81   0             0  is_stmt
    0x000000000000a824    287     12      2   0             0  is_stmt
    0x000000000000a828    100     58     82   0             0  is_stmt
    [...]

The main difference we are interested in here is that llvm-dwarfdump's
output tells us that 0xa800 is an adequate place to place a breakpoint
past a function prologue.  Since we know that foo covers from 0xa800 to
0xa94c, 0xa800 is the address at which the breakpoint should be placed
if the user wants to break in foo.

This commit proposes to add support for the prologue_end flag in the
line-program processing.

The processing of this prologue_end flag is made in skip_prologue_sal,
before it calls gdbarch_skip_prologue_noexcept.  The intent is that if
the compiler gave information on where the prologue ends, we should use
this information and not try to rely on architecture dependent logic to
guess it.

The testsuite have been executed using this patch on GNU/Linux x86_64.
Testcases have been compiled with both gcc/g++ (verison 9.4.0) and
clang/clang++ (version 10.0.0) since at the time of writing GCC does not
set the prologue_end marker.  Tests done with GCC 11.2.0 (not over the
entire testsuite) show that it does not emit this flag either.

No regression have been observed with GCC or Clang.  Note that when
using Clang, this patch fixes a failure in
gdb.opt/inline-small-func.exp.

Change-Id: I720449a8a9b2e1fb45b54c6095d3b1e9da9152f8
2022-04-04 23:03:32 +01:00

425 lines
12 KiB
C++
Raw Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

/* Build symbol tables in GDB's internal format.
Copyright (C) 1986-2022 Free Software Foundation, Inc.
This file is part of GDB.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 3 of the License, or
(at your option) any later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program. If not, see <http://www.gnu.org/licenses/>. */
#if !defined (BUILDSYM_H)
#define BUILDSYM_H 1
#include "gdbsupport/gdb_obstack.h"
struct objfile;
struct symbol;
struct addrmap;
struct compunit_symtab;
enum language;
/* This module provides definitions used for creating and adding to
the symbol table. These routines are called from various symbol-
file-reading routines.
They originated in dbxread.c of gdb-4.2, and were split out to
make xcoffread.c more maintainable by sharing code. */
struct block;
struct pending_block;
struct dynamic_prop;
/* The list of sub-source-files within the current individual
compilation. Each file gets its own symtab with its own linetable
and associated info, but they all share one blockvector. */
struct subfile
{
struct subfile *next;
/* Space for this is malloc'd. */
char *name;
/* Space for this is malloc'd. */
struct linetable *line_vector;
int line_vector_length;
/* The "containing" compunit. */
struct buildsym_compunit *buildsym_compunit;
enum language language;
struct symtab *symtab;
};
/* Record the symbols defined for each context in a list. We don't
create a struct block for the context until we know how long to
make it. */
#define PENDINGSIZE 100
struct pending
{
struct pending *next;
int nsyms;
struct symbol *symbol[PENDINGSIZE];
};
/* Stack representing unclosed lexical contexts (that will become
blocks, eventually). */
struct context_stack
{
/* Outer locals at the time we entered */
struct pending *locals;
/* Pending using directives at the time we entered. */
struct using_direct *local_using_directives;
/* Pointer into blocklist as of entry */
struct pending_block *old_blocks;
/* Name of function, if any, defining context */
struct symbol *name;
/* Expression that computes the frame base of the lexically enclosing
function, if any. NULL otherwise. */
struct dynamic_prop *static_link;
/* PC where this context starts */
CORE_ADDR start_addr;
/* Temp slot for exception handling. */
CORE_ADDR end_addr;
/* For error-checking matching push/pop */
int depth;
};
/* Flags associated with a linetable entry. */
enum linetable_entry_flag : unsigned
{
/* Indicates this PC is a good location to place a breakpoint at LINE. */
LEF_IS_STMT = 1 << 1,
/* Indicates this PC is a good location to place a breakpoint at the first
instruction past a function prologue. */
LEF_PROLOGUE_END = 1 << 2,
};
DEF_ENUM_FLAGS_TYPE (enum linetable_entry_flag, linetable_entry_flags);
/* Buildsym's counterpart to struct compunit_symtab. */
struct buildsym_compunit
{
/* Start recording information about a primary source file (IOW, not an
included source file).
COMP_DIR is the directory in which the compilation unit was compiled
(or NULL if not known). */
buildsym_compunit (struct objfile *objfile_, const char *name,
const char *comp_dir_, enum language language_,
CORE_ADDR last_addr);
/* Reopen an existing compunit_symtab so that additional symbols can
be added to it. Arguments are as for the main constructor. CUST
is the expandable compunit_symtab to be reopened. */
buildsym_compunit (struct objfile *objfile_, const char *name,
const char *comp_dir_, enum language language_,
CORE_ADDR last_addr, struct compunit_symtab *cust)
: m_objfile (objfile_),
m_last_source_file (name == nullptr ? nullptr : xstrdup (name)),
m_comp_dir (comp_dir_ == nullptr ? nullptr : xstrdup (comp_dir_)),
m_compunit_symtab (cust),
m_language (language_),
m_last_source_start_addr (last_addr)
{
}
~buildsym_compunit ();
DISABLE_COPY_AND_ASSIGN (buildsym_compunit);
void set_last_source_file (const char *name)
{
char *new_name = name == NULL ? NULL : xstrdup (name);
m_last_source_file.reset (new_name);
}
const char *get_last_source_file ()
{
return m_last_source_file.get ();
}
struct macro_table *get_macro_table ();
struct macro_table *release_macros ()
{
struct macro_table *result = m_pending_macros;
m_pending_macros = nullptr;
return result;
}
/* This function is called to discard any pending blocks. */
void free_pending_blocks ()
{
m_pending_block_obstack.clear ();
m_pending_blocks = nullptr;
}
struct block *finish_block (struct symbol *symbol,
struct pending_block *old_blocks,
const struct dynamic_prop *static_link,
CORE_ADDR start, CORE_ADDR end);
void record_block_range (struct block *block,
CORE_ADDR start, CORE_ADDR end_inclusive);
void start_subfile (const char *name);
void patch_subfile_names (struct subfile *subfile, const char *name);
void push_subfile ();
const char *pop_subfile ();
void record_line (struct subfile *subfile, int line, CORE_ADDR pc,
linetable_entry_flags flags);
struct compunit_symtab *get_compunit_symtab ()
{
return m_compunit_symtab;
}
void set_last_source_start_addr (CORE_ADDR addr)
{
m_last_source_start_addr = addr;
}
CORE_ADDR get_last_source_start_addr ()
{
return m_last_source_start_addr;
}
struct using_direct **get_local_using_directives ()
{
return &m_local_using_directives;
}
void set_local_using_directives (struct using_direct *new_local)
{
m_local_using_directives = new_local;
}
struct using_direct **get_global_using_directives ()
{
return &m_global_using_directives;
}
bool outermost_context_p () const
{
return m_context_stack.empty ();
}
struct context_stack *get_current_context_stack ()
{
if (m_context_stack.empty ())
return nullptr;
return &m_context_stack.back ();
}
int get_context_stack_depth () const
{
return m_context_stack.size ();
}
struct subfile *get_current_subfile ()
{
return m_current_subfile;
}
struct pending **get_local_symbols ()
{
return &m_local_symbols;
}
struct pending **get_file_symbols ()
{
return &m_file_symbols;
}
struct pending **get_global_symbols ()
{
return &m_global_symbols;
}
void record_debugformat (const char *format)
{
m_debugformat = format;
}
void record_producer (const char *producer)
{
m_producer = producer;
}
struct context_stack *push_context (int desc, CORE_ADDR valu);
struct context_stack pop_context ();
struct block *end_compunit_symtab_get_static_block
(CORE_ADDR end_addr, int expandable, int required);
struct compunit_symtab *end_compunit_symtab_from_static_block
(struct block *static_block, int section, int expandable);
struct compunit_symtab *end_compunit_symtab (CORE_ADDR end_addr, int section);
struct compunit_symtab *end_expandable_symtab (CORE_ADDR end_addr,
int section);
void augment_type_symtab ();
private:
void record_pending_block (struct block *block, struct pending_block *opblock);
struct block *finish_block_internal (struct symbol *symbol,
struct pending **listhead,
struct pending_block *old_blocks,
const struct dynamic_prop *static_link,
CORE_ADDR start, CORE_ADDR end,
int is_global, int expandable);
struct blockvector *make_blockvector ();
void watch_main_source_file_lossage ();
struct compunit_symtab *end_compunit_symtab_with_blockvector
(struct block *static_block, int section, int expandable);
/* The objfile we're reading debug info from. */
struct objfile *m_objfile;
/* List of subfiles (source files).
Files are added to the front of the list.
This is important mostly for the language determination hacks we use,
which iterate over previously added files. */
struct subfile *m_subfiles = nullptr;
/* The subfile of the main source file. */
struct subfile *m_main_subfile = nullptr;
/* Name of source file whose symbol data we are now processing. This
comes from a symbol of type N_SO for stabs. For DWARF it comes
from the DW_AT_name attribute of a DW_TAG_compile_unit DIE. */
gdb::unique_xmalloc_ptr<char> m_last_source_file;
/* E.g., DW_AT_comp_dir if DWARF. Space for this is malloc'd. */
gdb::unique_xmalloc_ptr<char> m_comp_dir;
/* Space for this is not malloc'd, and is assumed to have at least
the same lifetime as objfile. */
const char *m_producer = nullptr;
/* Space for this is not malloc'd, and is assumed to have at least
the same lifetime as objfile. */
const char *m_debugformat = nullptr;
/* The compunit we are building. */
struct compunit_symtab *m_compunit_symtab = nullptr;
/* Language of this compunit_symtab. */
enum language m_language;
/* The macro table for the compilation unit whose symbols we're
currently reading. */
struct macro_table *m_pending_macros = nullptr;
/* True if symtab has line number info. This prevents an otherwise
empty symtab from being tossed. */
bool m_have_line_numbers = false;
/* Core address of start of text of current source file. This too
comes from the N_SO symbol. For Dwarf it typically comes from the
DW_AT_low_pc attribute of a DW_TAG_compile_unit DIE. */
CORE_ADDR m_last_source_start_addr;
/* Stack of subfile names. */
std::vector<const char *> m_subfile_stack;
/* The "using" directives local to lexical context. */
struct using_direct *m_local_using_directives = nullptr;
/* Global "using" directives. */
struct using_direct *m_global_using_directives = nullptr;
/* The stack of contexts that are pushed by push_context and popped
by pop_context. */
std::vector<struct context_stack> m_context_stack;
struct subfile *m_current_subfile = nullptr;
/* The mutable address map for the compilation unit whose symbols
we're currently reading. The symtabs' shared blockvector will
point to a fixed copy of this. */
struct addrmap *m_pending_addrmap = nullptr;
/* The obstack on which we allocate pending_addrmap.
If pending_addrmap is NULL, this is uninitialized; otherwise, it is
initialized (and holds pending_addrmap). */
auto_obstack m_pending_addrmap_obstack;
/* True if we recorded any ranges in the addrmap that are different
from those in the blockvector already. We set this to false when
we start processing a symfile, and if it's still false at the
end, then we just toss the addrmap. */
bool m_pending_addrmap_interesting = false;
/* An obstack used for allocating pending blocks. */
auto_obstack m_pending_block_obstack;
/* Pointer to the head of a linked list of symbol blocks which have
already been finalized (lexical contexts already closed) and which
are just waiting to be built into a blockvector when finalizing the
associated symtab. */
struct pending_block *m_pending_blocks = nullptr;
/* Pending static symbols and types at the top level. */
struct pending *m_file_symbols = nullptr;
/* Pending global functions and variables. */
struct pending *m_global_symbols = nullptr;
/* Pending symbols that are local to the lexical context. */
struct pending *m_local_symbols = nullptr;
};
extern void add_symbol_to_list (struct symbol *symbol,
struct pending **listhead);
extern struct symbol *find_symbol_in_list (struct pending *list,
char *name, int length);
#endif /* defined (BUILDSYM_H) */