From: Andrew Burgess <aburgess@redhat.com>
To: gdb-patches@sourceware.org
Cc: Andrew Burgess <aburgess@redhat.com>,
Eli Zaretskii <eliz@gnu.org>,
Simon Marchi <simon.marchi@efficios.com>
Subject: [PATCHv7 2/4] gdb: introduce program_space::get_entry_point_info function
Date: Sat, 18 Jul 2026 14:11:25 +0100 [thread overview]
Message-ID: <90ad3b9d2c136a19decfdd7e1966c41a6c3b779d.1784380038.git.aburgess@redhat.com> (raw)
In-Reply-To: <cover.1784380038.git.aburgess@redhat.com>
I noticed that when debugging a dynamically linked executable, if I
started the inferior with 'starti' then used 'bt' I would see some
bogus frames:
(gdb) starti
Starting program: /tmp/hello
Program stopped.
0x00007ffff7fd3110 in _start () from /lib64/ld-linux-x86-64.so.2
(gdb) bt
#0 0x00007ffff7fd3110 in _start () from /lib64/ld-linux-x86-64.so.2
#1 0x0000000000000001 in ?? ()
#2 0x00007fffffffac13 in ?? ()
#3 0x0000000000000000 in ?? ()
(gdb)
This surprised me as 'backtrace past-entry' was off:
(gdb) show backtrace past-entry
Whether backtraces should continue past the entry point of a program is off.
I was expecting GDB to stop the backtrace at the inferior's entry
address.
Frame unwinding starts in get_prev_frame, and in here we find this
block:
if (this_frame->level >= 0
&& get_frame_type (this_frame) == NORMAL_FRAME
&& !user_set_backtrace_options.backtrace_past_entry
&& frame_pc.has_value ()
&& inside_entry_func (this_frame))
{
frame_debug_got_null_frame (this_frame, "inside entry func");
return NULL;
}
Which uses inside_entry_func to terminate the backtrace when we reach
the entry frame. The inside_entry_func function calls
current_program_space->exec_entry_point_address_if_available and uses
the result to figure out if we are in the entry frame.
And here the problem becomes obvious, we are only checking if we are
in the "entry frame" for the main executable, not for the inferior as
a whole. And indeed, if I compile the same test program as a static
binary, where there will be no run-time linker, and the entry point of
the executable is the entry point for the inferior, then the 'bt'
problem I saw above goes away.
This suggests, I think, that we need to track two different entry
addresses, the entry address for the executable file, and the entry
address for the entire inferior. Then, for dynamically linked
executables, these two addresses can be different, the former will
still be the same address within the main executable file, while the
latter will be the address of the entry point within the run-time
linker.
If we had this information then we could extend inside_entry_func to
check both addresses, and the 'bt' problem seen above will be
resolved.
To make this information available I added a new solib_ops method,
solib_ops::inferior_entry_point_address, for svr4 targets this figures
out if the main executable is dynamically linked, and if it is, uses
the AT_BASE auxv entry and the entry address pulled from the ELF
header to compute the inferior entry address.
I then added program_space::get_entry_point_info, which returns a
struct containing the two entry point addresses, one comes from the
new solib_ops method, and one comes from the existing method
program_space::exec_entry_point_address_if_available.
With the infrastructure in place I can then update inside_entry_func
to check against both entry addresses.
To aid in debugging GDB, I added a new maintenance command:
maintenance info entry-address
which just calls program_space::get_entry_point_info and then prints
the two addresses. I left this as a maintenance command as I don't
see much user utility in this right now, but it made it easier for me
to see what GDB was doing, so I left the command in this commit.
Reviewed-By: Eli Zaretskii <eliz@gnu.org>
Approved-By: Simon Marchi <simon.marchi@efficios.com>
---
gdb/NEWS | 7 +
gdb/doc/gdb.texinfo | 22 ++
gdb/frame.c | 10 +-
gdb/progspace.c | 60 +++++
gdb/progspace.h | 49 ++++
gdb/solib-svr4.c | 85 +++++++
gdb/solib-svr4.h | 1 +
gdb/solib.h | 14 ++
gdb/testsuite/gdb.base/bt-after-starti.exp | 253 +++++++++++++++++++++
gdb/testsuite/lib/gdb.exp | 39 ++++
10 files changed, 535 insertions(+), 5 deletions(-)
create mode 100644 gdb/testsuite/gdb.base/bt-after-starti.exp
diff --git a/gdb/NEWS b/gdb/NEWS
index 8f40ca5cb11..9c51f476c84 100644
--- a/gdb/NEWS
+++ b/gdb/NEWS
@@ -179,6 +179,13 @@ disable skip
These are new aliases for 'skip delete', 'skip enable', and 'skip
disable' respectively.
+maint info entry-address
+ Display the inferior and main executable entry addresses for the
+ current inferior. The inferior entry address is the address of the
+ first instruction in the inferior that was executed. The main
+ executable entry address is the address of the first instruction in
+ the main executable that will be executed.
+
* MI changes
** The "-trace-save" command no longer supports the "-ctf" flag.
diff --git a/gdb/doc/gdb.texinfo b/gdb/doc/gdb.texinfo
index 0030698dcee..35fb9f78b08 100644
--- a/gdb/doc/gdb.texinfo
+++ b/gdb/doc/gdb.texinfo
@@ -43260,6 +43260,28 @@ Maintenance Commands
Ignoring SystemTap probe libc longjmp in /lib64/libc.so.6.^M
Ignoring SystemTap probe libc longjmp in /lib64/libc.so.6.^M
@end smallexample
+
+@kindex maint info entry-address
+@item maint info entry-address
+Display the entry addresses for the currently selected inferior. Two
+addresses are displayed, the inferior entry address and the executable
+entry address.
+
+Neither address is the address of @code{main}. When a program is
+started, low-level startup code (typically a function called
+@code{_start}) runs before @code{main} is called.
+
+The executable entry address is the address of this startup code
+within the main executable (@pxref{Files, ,Commands to Specify
+Files}).
+
+The inferior entry address is the address of the very first
+instruction executed when the inferior is started. For statically
+linked programs this is the same as the executable entry address. For
+dynamically linked programs the run-time linker must execute first in
+order to load shared libraries, so the inferior entry address will be
+an address within the run-time linker rather than within the main
+executable.
@end table
The following command is useful for non-interactive invocations of
diff --git a/gdb/frame.c b/gdb/frame.c
index a83f0b20686..e61bfd8d1f7 100644
--- a/gdb/frame.c
+++ b/gdb/frame.c
@@ -2703,12 +2703,12 @@ inside_main_func (const frame_info_ptr &this_frame)
static bool
inside_entry_func (const frame_info_ptr &this_frame)
{
- std::optional<CORE_ADDR> entry_point
- = current_program_space->exec_entry_point_address_if_available ();
- if (!entry_point.has_value ())
- return false;
+ const entry_point_info &ep_info
+ = current_program_space->get_entry_point_info ();
- return get_frame_func (this_frame) == *entry_point;
+ CORE_ADDR frame_func_addr = get_frame_func (this_frame);
+ return (ep_info.exec_entry_address () == frame_func_addr
+ || ep_info.inferior_entry_address () == frame_func_addr);
}
/* Return a structure containing various interesting information about
diff --git a/gdb/progspace.c b/gdb/progspace.c
index 0e8516f4640..4de32bc522f 100644
--- a/gdb/progspace.c
+++ b/gdb/progspace.c
@@ -26,6 +26,7 @@
#include <algorithm>
#include "cli/cli-style.h"
#include "observable.h"
+#include "arch-utils.h"
/* The last program space number assigned. */
static int last_program_space_num = 0;
@@ -288,6 +289,55 @@ program_space::exec_entry_point_address () const
/* See progspace.h. */
+entry_point_info
+program_space::get_entry_point_info () const
+{
+ std::optional<CORE_ADDR> exec_entry_address
+ = this->exec_entry_point_address_if_available ();
+
+ std::optional<CORE_ADDR> inferior_entry_address;
+ if (m_solib_ops != nullptr)
+ inferior_entry_address = m_solib_ops->inferior_entry_point_address ();
+
+ return entry_point_info (std::move (inferior_entry_address),
+ std::move (exec_entry_address));
+}
+
+/* Implement the 'maint info entry-address' command. */
+
+static void
+maintenance_info_entry_address (const char *args, int from_tty)
+{
+ if (args != nullptr && *args != '\0')
+ error (_("unknown argument: %s"), args);
+
+ struct gdbarch *gdbarch = get_current_arch ();
+
+ const entry_point_info &ep_info
+ = current_program_space->get_entry_point_info ();
+
+ /* Display a single entry address ADDR with TITLE as a description. */
+ auto display_entry_address
+ = [&gdbarch] (const char *title,
+ const std::optional<CORE_ADDR> &addr) -> void
+ {
+ gdb_puts (title);
+ if (addr.has_value ())
+ fputs_styled (paddress (gdbarch, addr.value ()),
+ address_style.style (), gdb_stdout);
+ else
+ fputs_styled ("<unknown>", metadata_style.style (), gdb_stdout);
+ gdb_puts ("\n");
+ };
+
+ display_entry_address (_("Inferior entry address: "),
+ ep_info.inferior_entry_address ());
+ display_entry_address (_("Executable entry address: "),
+ ep_info.exec_entry_address ());
+}
+
+/* See progspace.h. */
+
bool
program_space::has_partial_symbols ()
{
@@ -498,6 +548,16 @@ initialize_progspace ()
_("Info about currently known program spaces."),
&maintenanceinfolist);
+ add_cmd ("entry-address", class_maintenance,
+ maintenance_info_entry_address,
+ _("Information about the current inferior's entry addresses.\n\
+Display the address of the first instruction executed within the\n\
+inferior, and the first instruction executed within the main executable.\n\
+The entry address of the whole inferior will be <unknown> prior to the\n\
+inferior starting, and either or both addresses can be <unknown> if\n\
+GDB is unable to find the required information."),
+ &maintenanceinfolist);
+
/* There's always one program space. Note that this function isn't
an automatic _initialize_foo function, since other
_initialize_foo routines may need to install their per-pspace
diff --git a/gdb/progspace.h b/gdb/progspace.h
index 1ae1e42f3bb..be2d3f7e88c 100644
--- a/gdb/progspace.h
+++ b/gdb/progspace.h
@@ -67,6 +67,50 @@ new_address_space ()
return address_space_ref_ptr::new_reference (new address_space);
}
+/* Class for tracking two possible entry points that an inferior might
+ have, the entry point for the entire inferior, and the entry point
+ within the main executable. */
+
+struct entry_point_info
+{
+ entry_point_info (std::optional<CORE_ADDR> inferior_entry_address,
+ std::optional<CORE_ADDR> exec_entry_address)
+ : m_inferior_entry_address (inferior_entry_address),
+ m_exec_entry_address (exec_entry_address)
+ { /* Nothing. */ }
+
+ DISABLE_COPY_AND_ASSIGN (entry_point_info);
+
+ /* The entry address within the inferior as a whole. See the member
+ variable definition below for more details. */
+ std::optional<CORE_ADDR> inferior_entry_address () const
+ {
+ return m_inferior_entry_address;
+ }
+
+ /* The entry address within the main executable. See the member
+ variable definition below for more details. */
+ std::optional<CORE_ADDR> exec_entry_address () const
+ {
+ return m_exec_entry_address;
+ }
+
+private:
+ /* The entry address within the inferior. This can be outside of the
+ main executable, e.g. for dynamically linked executables this could
+ be the entry address for the run-time linker. For statically linked
+ executables this will be the entry address of the main executable.
+ This can be empty if GDB doesn't know how to figure out the correct
+ entry address for any reason. */
+ std::optional<CORE_ADDR> m_inferior_entry_address;
+
+ /* The entry address within the main executable. This can be empty if
+ the main executable is not set yet, or GDB doesn't have an objfile
+ associated with the main executable, e.g. in some attach, or remote
+ debug cases. */
+ std::optional<CORE_ADDR> m_exec_entry_address;
+};
+
/* A program space represents a symbolic view of an address space.
Roughly speaking, it holds all the data associated with a
non-running-yet program (main executable, main symbols), and when
@@ -323,6 +367,11 @@ struct program_space
return m_target_sections;
}
+ /* Return information about the entry point in the main executable, and
+ the entry point for the inferior, which might be different from the
+ main executable. */
+ entry_point_info get_entry_point_info () const;
+
/* If there is a valid and known entry point in the main executable of
this program space, return it. Otherwise return an empty optional. */
std::optional<CORE_ADDR> exec_entry_point_address_if_available () const;
diff --git a/gdb/solib-svr4.c b/gdb/solib-svr4.c
index 8e3de4d3ea1..12f3e783744 100644
--- a/gdb/solib-svr4.c
+++ b/gdb/solib-svr4.c
@@ -3777,6 +3777,91 @@ svr4_solib_ops::get_solibs_in_ns (int nsid) const
return ns_solibs;
}
+/* See solib.h. */
+
+std::optional<CORE_ADDR>
+svr4_solib_ops::inferior_entry_point_address () const
+{
+ std::optional<gdb::byte_vector> interp_name_holder
+ = svr4_find_program_interpreter ();
+
+ /* No interpreter means this is a static executable. Ask the
+ program_space for the entry address within the main executable. */
+ if (!interp_name_holder.has_value ())
+ return m_pspace->exec_entry_point_address_if_available ();
+
+ /* For a dynamically linked executable the inferior's true entry point
+ is the entry point of the dynamic linker. We find this using the
+ AT_BASE auxiliary vector entry, which gives the dynamic linker's
+ load address, combined with e_entry address pulled from the
+ inferior. We assume that the ELF header can be read from AT_BASE. */
+ CORE_ADDR at_base_addr;
+ if (target_auxv_search (AT_BASE, &at_base_addr) <= 0)
+ return {};
+
+ /* Determine ELF architecture type. Use the size of a program header
+ entry to determine which ELF header we can expect to find. */
+ size_t e_entry_offset = 0;
+ size_t e_entry_size = 0;
+ CORE_ADDR at_phent;
+ if (target_auxv_search (AT_PHENT, &at_phent) <= 0)
+ return {};
+ if (at_phent == sizeof (Elf32_External_Phdr))
+ {
+ e_entry_offset = offsetof (Elf32_External_Ehdr, e_entry);
+ e_entry_size = 4;
+ }
+ else if (at_phent == sizeof (Elf64_External_Phdr))
+ {
+ e_entry_offset = offsetof (Elf64_External_Ehdr, e_entry);
+ e_entry_size = 8;
+ }
+ else
+ return {};
+
+ /* Architecture of the current inferior. */
+ gdbarch *gdbarch = current_inferior ()->arch ();
+
+ /* Read the entry address from the ELF header at AT_BASE_ADDR. */
+ CORE_ADDR e_entry;
+ gdb_byte buffer[sizeof (CORE_ADDR)];
+ if (target_read_memory (at_base_addr + e_entry_offset, buffer, e_entry_size))
+ return {};
+ bfd_endian byte_order = gdbarch_byte_order (gdbarch);
+ e_entry = extract_unsigned_integer (buffer, e_entry_size, byte_order);
+
+ /* Ensure AT_BASE_ADDR has proper sign in its possible upper bits so
+ that `+ at_base_addr' will overflow CORE_ADDR width not creating
+ invalid addresses like 0x101234567 for 32bit inferiors on 64bit
+ GDB. */
+ int addr_bit = gdbarch_addr_bit (gdbarch);
+ if (addr_bit < (sizeof (CORE_ADDR) * HOST_CHAR_BIT))
+ {
+ CORE_ADDR space_size = (CORE_ADDR) 1 << addr_bit;
+
+ gdb_assert (at_base_addr < space_size);
+
+ /* E_ENTRY exceeding SPACE_SIZE would be for prelinked
+ 64bit ld.so with 32bit executable, it should not happen. */
+ if (e_entry < space_size
+ && e_entry + at_base_addr >= space_size)
+ at_base_addr -= space_size;
+ }
+
+ /* Compute the entry address. */
+ e_entry = at_base_addr + e_entry;
+
+ /* Handle the case where E_ENTRY is a function descriptor. Also remove
+ any non-address (e.g. tag) bits from E_ENTRY. */
+ e_entry
+ = gdbarch_convert_from_func_ptr_addr (gdbarch, e_entry,
+ current_inferior ()->top_target ());
+ e_entry
+ = gdbarch_addr_bits_remove (gdbarch, e_entry);
+
+ return e_entry;
+}
+
INIT_GDB_FILE (svr4_solib)
{
gdb::observers::free_objfile.attach (svr4_free_objfile_observer,
diff --git a/gdb/solib-svr4.h b/gdb/solib-svr4.h
index 3078a092778..f76f350d951 100644
--- a/gdb/solib-svr4.h
+++ b/gdb/solib-svr4.h
@@ -115,6 +115,7 @@ struct svr4_solib_ops : public solib_ops
void iterate_over_objfiles_in_search_order
(iterate_over_objfiles_in_search_order_cb_ftype cb,
objfile *current_objfile) const override;
+ std::optional<CORE_ADDR> inferior_entry_point_address () const override;
/* Return the appropriate link map offsets table for the architecture. */
virtual link_map_offsets *fetch_link_map_offsets () const = 0;
diff --git a/gdb/solib.h b/gdb/solib.h
index 662d467b0cc..2db836dc005 100644
--- a/gdb/solib.h
+++ b/gdb/solib.h
@@ -283,6 +283,20 @@ struct solib_ops
(iterate_over_objfiles_in_search_order_cb_ftype cb,
objfile *current_objfile) const;
+ /* Return the inferior entry point address. This isn't always the entry
+ point of the main executable, but will be the actual entry point where
+ the inferior starts (or started) executing.
+
+ For example on SVR4 targets, for dynamically linked executables, this
+ will be the entry address of the dynamic linker. But for statically
+ linked executables, this will be the entry point of the main
+ executable.
+
+ If the entry address cannot be found then an empty optional is
+ returned. */
+ virtual std::optional<CORE_ADDR> inferior_entry_point_address () const
+ { return {}; }
+
protected:
/* The program space for which this solib_ops was created. */
program_space *m_pspace;
diff --git a/gdb/testsuite/gdb.base/bt-after-starti.exp b/gdb/testsuite/gdb.base/bt-after-starti.exp
new file mode 100644
index 00000000000..778c8030d26
--- /dev/null
+++ b/gdb/testsuite/gdb.base/bt-after-starti.exp
@@ -0,0 +1,253 @@
+# Copyright 2026 Free Software Foundation, Inc.
+#
+# 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/>.
+
+# Check that 'bt' from the very first instruction (where 'starti'
+# stops GDB) doesn't display any unexpected frames.
+
+standard_testfile main.c
+
+# Build default executable, hopefully dynamically linked, but we check
+# this below.
+foreach pie_mode { pie nopie } {
+ if { [build_executable "failed to build" $testfile-$pie_mode $srcfile \
+ [list debug $pie_mode]] } {
+ return
+ }
+}
+
+# Build a statically linked executable.
+set testfile_static ${testfile}-static
+if { [build_executable "failed to build" $testfile_static $srcfile \
+ {debug additional_flags=-static}] } {
+ return
+}
+
+# Run the 'maint info entry-address' command. Expect INF_RE to match
+# the inferior entry address and EXE_RE to match the executable entry
+# address. TESTNAME is used as the test name.
+proc check_maint_info_entry_addr { inf_re exe_re testname } {
+ gdb_test "maint info entry-address" \
+ [multi_line \
+ "Inferior entry address: ${inf_re}" \
+ "Executable entry address: ${exe_re}"] \
+ $testname
+}
+
+# Use 'info proc mappings' to find the address of the first mapping of
+# TESTFILE, return this address. If the address cannot be found then
+# the empty string is returned.
+#
+# This emits a PASS or FAIL depending on whether the address is found
+# or not.
+proc get_mapped_file_base_addr { testfile } {
+ set base_addr ""
+ gdb_test_multiple "info proc mappings" "" -lbl {
+ -re "\r\n($::hex) \[^\r\n\]+/${testfile}\\s*(?=\r\n)" {
+ if { $base_addr eq "" } {
+ set base_addr $expect_out(1,string)
+ }
+ exp_continue
+ }
+ -re "\r\n$::gdb_prompt $" {
+ gdb_assert { $base_addr ne "" } \
+ "find base address of $testfile"
+ }
+ }
+ return $base_addr
+}
+
+# Use 'objdump' to extract the entry point address from TESTFILE.
+proc get_exec_entry_address { testfile } {
+ set objdump_program [gdb_find_objdump]
+ set command "exec $objdump_program -f [standard_output_file $testfile]"
+ verbose -log "command is $command"
+ set result [catch {{*}$command} output]
+ verbose -log "result is $result"
+ verbose -log "output is $output"
+
+ set testname "get_exec_entry_address"
+
+ if {$result != 0} {
+ fail $testname
+ return "UNKNOWN"
+ }
+
+ if {![regexp "\nstart address ($::hex)\n" $output trash addr]} {
+ fail "$testname (no entry point address header)"
+ return "UNKNOWN"
+ }
+
+ return 0x[format %lx $addr]
+}
+
+# Start TESTFILE using 'starti'. Check if the 'backtrace' command
+# prints any additional frames; we don't expect any. Create a core
+# file from the running inferior. Restart GDB and load the core file,
+# check that GDB can still figure out the entry addresses.
+#
+# When IS_DYNAMIC is true then TESTFILE is a dynamically linked
+# executable, otherwise TESTFILE is statically linked.
+proc run_test { testfile is_dynamic } {
+ clean_restart $testfile
+
+ set exe_entry_addr [get_exec_entry_address $testfile]
+
+ # The inferior entry address is always unknown before starting the
+ # inferior because GDB has no solib_ops yet, and it is only
+ # through that that we figure out the inferior entry address.
+ check_maint_info_entry_addr "<unknown>" $exe_entry_addr \
+ "check 'maint info entry-address' before inferior starts"
+
+ if {$::use_gdb_stub && [target_info exists gdb,do_reload_on_run]} {
+ # This is the path taken by the 'native-gdbserver' board.
+ # Start gdbserver and connect, this should leave the inferior
+ # at the first instruction.
+ if { [gdb_reload] != 0 } {
+ return -1
+ }
+ pass "connected to gdbserver"
+ } else {
+ # Start inferior with 'starti' and then wait for a prompt.
+ if { [gdb_starti_cmd] < 0 } {
+ fail "could not issue starti command"
+ return -1
+ }
+ gdb_test "" "Program stopped.*" "prompt after starti"
+ }
+
+ # Not every target supports finding the process entry point.
+ # Targets that don't support this report '<unknown>', we check for
+ # this too on those targets.
+ if {[supports_process_entry_point]} {
+ set pc [get_hexadecimal_valueof "\$pc" "UNKNOWN" \
+ "get current program counter"]
+
+ if { !$is_dynamic } {
+ # In static binaries, the first address should be the first
+ # address in the executable.
+ gdb_assert { $pc == $exe_entry_addr } \
+ "stopped at executable entry address"
+ }
+
+ set inf_entry_addr $pc
+ } else {
+ set inf_entry_addr "<unknown>"
+ }
+
+ # If using PIE then the executable entry address we looked up from
+ # the objfile will have been relocated. Attempt to compute the
+ # relocated address and updated EXE_ENTRY_ADDR, otherwise set
+ # EXE_ENTRY_ADDR to a generic pattern that matches any address.
+ if {[exec_is_pie [standard_output_file $testfile]]} {
+ set base_addr [get_mapped_file_base_addr $testfile]
+ if { $base_addr ne "" } {
+ set exe_entry_addr \
+ 0x[format %lx [expr {$base_addr + $exe_entry_addr}]]
+ } else {
+ # We failed to find the base address. The
+ # get_mapped_file_base_addr will have already emitted a
+ # FAIL. Set the expected pattern to something generic and
+ # push on, we can still test a bunch of things, we just
+ # don't know the address is correct.
+ set exe_entry_addr $::hex
+ }
+ }
+
+ check_maint_info_entry_addr $inf_entry_addr $exe_entry_addr \
+ "check 'maint info entry-address' after starti"
+
+ # Allow backtrace past the entry frame, count how many frames GDB
+ # finds. If on this target there are no additional frames then
+ # the following test isn't going to tell us much, so skip it.
+ gdb_test_no_output "set backtrace past-entry on"
+ set frame_count 0
+ set first_frame_is_unnamed false
+ gdb_test_multiple "bt" "count possible frames" -lbl {
+ -re "\r\n#0\\s+$::hex in \\?\\? \\(\\) \[^\r\n\]+(?=\r\n)" {
+ incr frame_count
+ set first_frame_is_unnamed true
+ exp_continue
+ }
+ -re "\r\n#($::decimal)\\s+\[^\r\n\]+(?=\r\n)" {
+ incr frame_count
+ exp_continue
+ }
+ -re "\r\n$::gdb_prompt $" {
+ gdb_assert { $frame_count > 0 } $gdb_test_name
+ }
+ }
+
+ if { $frame_count <= 1 } {
+ unsupported "no additional frames that GDB can hide"
+ } elseif { $first_frame_is_unnamed } {
+ # GDB's entry frame detection relies on figuring out the $pc
+ # for the start of the frame. We know that frame #0 should be
+ # the entry frame, but if it is nameless then GDB will have
+ # failed to find the start of the frame, and so will not be
+ # able to perform entry frame detection.
+ unsupported "first frame is unnamed"
+ } else {
+ # Turn off backtrace past the entry frame. Use the 'bt' command,
+ # and check we see only a single frame.
+ gdb_test_no_output "set backtrace past-entry off"
+ gdb_test "bt" "^#0 \[^\r\n\]+" \
+ "single frame when backtrace past-entry is off"
+
+ # Check 'bt -past-entry'.
+ set frame_count_with_past_entry_flag 0
+ gdb_test_multiple "bt -past-entry" "count frames again" -lbl {
+ -re "\r\n#($::decimal)\\s+\[^\r\n\]+(?=\r\n)" {
+ incr frame_count_with_past_entry_flag
+ exp_continue
+ }
+ -re "\r\n$::gdb_prompt $" {
+ gdb_assert \
+ { $frame_count == $frame_count_with_past_entry_flag } \
+ $gdb_test_name
+ }
+ }
+ }
+
+ # Create a core file. Restart GDB. Load the core file. Check
+ # that 'maint info entry-address' gives the correct output.
+ set corefile [host_standard_output_file $testfile].core
+ if {![gdb_gcore_cmd $corefile "generate corefile"]} {
+ return
+ }
+
+ clean_restart $testfile
+
+ if {[gdb_core_cmd $corefile "load corefile"] != 1} {
+ return
+ }
+
+ check_maint_info_entry_addr $inf_entry_addr $exe_entry_addr \
+ "check 'maint info entry-address' after loading core file"
+}
+
+with_test_prefix "dynamic" {
+ foreach_with_prefix pie_mode { pie nopie } {
+ set dyln_name [section_get $binfile-$pie_mode .interp]
+ if { $dyln_name eq "" } {
+ unsupported "couldn't find dynamic linker name"
+ } else {
+ run_test $testfile-$pie_mode true
+ }
+ }
+}
+
+with_test_prefix "static" {
+ run_test $testfile_static false
+}
diff --git a/gdb/testsuite/lib/gdb.exp b/gdb/testsuite/lib/gdb.exp
index a40c87c6727..11aa6bfbf35 100644
--- a/gdb/testsuite/lib/gdb.exp
+++ b/gdb/testsuite/lib/gdb.exp
@@ -4217,6 +4217,45 @@ proc is_aarch64_target {} {
return [expr {![is_aarch32_target]}]
}
+# Return true for svr4 targets, otherwise, return false.
+
+proc is_svr4_target {} {
+ return [expr {[istarget *-linux*] || [istarget *-freebsd*]
+ || [istarget *-netbsd*] || [istarget *-openbsd*]
+ || [istarget *-solaris*] || [istarget *-gnu]}]
+}
+
+# Return true for targets that support finding the whole process entry
+# address, otherwise, return false.
+
+proc supports_process_entry_point {} {
+ # SVR4 targets support finding the entry point. This is done
+ # within GDB so will work even for remote targets.
+ if {[is_svr4_target]} {
+ return true
+ }
+
+ # Windows and Darwin don't currently support this.
+ if {[istarget *-*-mingw*]
+ || [istarget *-*-cygwin*]
+ || [istarget *-*-pe*]
+ || [istarget *-*-darwin*]} {
+ return false
+ }
+
+ # For remote targets there is no RSP packet to retrieve the entry
+ # point, so this won't work unless the solib code can handle this
+ # within GDB, see the svr4 check above.
+ if {[gdb_protocol_is_remote]} {
+ return false
+ }
+
+ # Assume everything else supports this by default. If a test
+ # fails because we get here then either fix GDB to support this
+ # feature, or add a new deny list entry above.
+ return true
+}
+
# Return 1 if displaced stepping is supported on target, otherwise, return 0.
proc support_displaced_stepping {} {
--
2.25.4
next prev parent reply other threads:[~2026-07-18 13:13 UTC|newest]
Thread overview: 54+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-06-08 18:14 [PATCH] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-06-11 19:24 ` Guinevere Larsen
2026-06-11 20:40 ` Andrew Burgess
2026-06-12 12:49 ` Guinevere Larsen
2026-06-11 21:59 ` [PATCHv2 0/3] " Andrew Burgess
2026-06-11 21:59 ` [PATCHv2 1/3] gdb: rename program_space::entry_point_address* functions Andrew Burgess
2026-06-12 15:41 ` Pedro Alves
2026-06-11 21:59 ` [PATCHv2 2/3] gdb: introduce program_space::get_entry_point_info function Andrew Burgess
2026-06-12 6:07 ` Eli Zaretskii
2026-06-15 10:14 ` Andrew Burgess
2026-06-15 12:01 ` Eli Zaretskii
2026-06-12 15:41 ` Pedro Alves
2026-06-15 10:29 ` Andrew Burgess
2026-06-16 18:36 ` Pedro Alves
2026-06-23 9:46 ` Andrew Burgess
2026-06-23 10:20 ` Pedro Alves
2026-06-11 21:59 ` [PATCHv2 3/3] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-06-12 15:57 ` Pedro Alves
2026-06-16 19:47 ` [PATCHv3 0/4] " Andrew Burgess
2026-06-16 19:47 ` [PATCHv3 1/4] gdb: rename program_space::entry_point_address* functions Andrew Burgess
2026-06-16 19:47 ` [PATCHv3 2/4] gdb: introduce program_space::get_entry_point_info function Andrew Burgess
2026-06-17 11:52 ` Eli Zaretskii
2026-06-16 19:47 ` [PATCHv3 3/4] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-06-16 19:47 ` [PATCHv3 4/4] gdb: cache program space entry point information Andrew Burgess
2026-06-23 10:47 ` [PATCHv4 0/4] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-06-23 10:47 ` [PATCHv4 1/4] gdb: rename program_space::entry_point_address* functions Andrew Burgess
2026-06-23 10:47 ` [PATCHv4 2/4] gdb: introduce program_space::get_entry_point_info function Andrew Burgess
2026-06-23 10:47 ` [PATCHv4 3/4] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-06-23 10:47 ` [PATCHv4 4/4] gdb: cache program space entry point information Andrew Burgess
2026-06-25 15:26 ` [PATCHv5 0/4] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-06-25 15:26 ` [PATCHv5 1/4] gdb: rename program_space::entry_point_address* functions Andrew Burgess
2026-06-25 15:26 ` [PATCHv5 2/4] gdb: introduce program_space::get_entry_point_info function Andrew Burgess
2026-06-25 15:26 ` [PATCHv5 3/4] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-06-25 15:26 ` [PATCHv5 4/4] gdb: cache program space entry point information Andrew Burgess
2026-07-10 14:24 ` [PATCHv6 0/4] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-07-10 14:24 ` [PATCHv6 1/4] gdb: rename program_space::entry_point_address* functions Andrew Burgess
2026-07-10 14:24 ` [PATCHv6 2/4] gdb: introduce program_space::get_entry_point_info function Andrew Burgess
2026-07-10 15:39 ` Simon Marchi
2026-07-10 14:24 ` [PATCHv6 3/4] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-07-10 17:00 ` Simon Marchi
2026-07-10 17:01 ` Simon Marchi
2026-07-16 15:06 ` Andrew Burgess
2026-07-10 14:24 ` [PATCHv6 4/4] gdb: cache program space entry point information Andrew Burgess
2026-07-10 17:07 ` Simon Marchi
2026-07-18 13:11 ` [PATCHv7 0/4] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-07-18 13:11 ` [PATCHv7 1/4] gdb: rename program_space::entry_point_address* functions Andrew Burgess
2026-07-18 13:11 ` Andrew Burgess [this message]
2026-07-18 13:11 ` [PATCHv7 3/4] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-07-18 13:11 ` [PATCHv7 4/4] gdb: cache program space entry point information Andrew Burgess
2026-07-20 9:52 ` [PATCHv8 0/4] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-07-20 9:52 ` [PATCHv8 1/4] gdb: rename program_space::entry_point_address* functions Andrew Burgess
2026-07-20 9:52 ` [PATCHv8 2/4] gdb: introduce program_space::get_entry_point_info function Andrew Burgess
2026-07-20 9:52 ` [PATCHv8 3/4] gdb: allow 'until' to work in outermost frame Andrew Burgess
2026-07-20 9:52 ` [PATCHv8 4/4] gdb: cache program space entry point information Andrew Burgess
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=90ad3b9d2c136a19decfdd7e1966c41a6c3b779d.1784380038.git.aburgess@redhat.com \
--to=aburgess@redhat.com \
--cc=eliz@gnu.org \
--cc=gdb-patches@sourceware.org \
--cc=simon.marchi@efficios.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox