As a result of this commit,
9b4c5f878f
(Remove support for thread events without PTRACE_EVENT_CLONE in GDBServer.)
the last usage of td_ta_event_addr td_ta_set_event and
td_ta_event_getmsg were removed. They are no longer used. This patch
is to remove them.
gdb/gdbserver:
2016-08-02 Yao Qi <yao.qi@linaro.org>
* thread-db.c (struct thread_db) <td_ta_event_getmsg_p>: Remove.
<td_ta_set_event_p, td_ta_event_addr_p>: Remove.
(thread_db_load_search): Update.
(try_thread_db_load_1): Don't look for td_ta_event_addr,
td_ta_set_event and td_ta_event_getmsg.
Since we now rely on PTRACE_EVENT_CLONE being available (added in
Linux 2.5.46), we're relying on NPTL.
This commit removes the support for older LinuxThreads, as well as the
workarounds for vendor 2.4 kernels with NPTL backported.
- Rely on tkill being available.
- Assume gdb doesn't get cancel signals.
- Remove code that checks the LinuxThreads restart and cancel signals
in the inferior.
- Assume that __WALL is available.
- Assume that non-leader threads report WIFEXITED.
- Thus, no longer need to send signal 0 to check whether threads are
still alive.
- Update comments throughout.
Tested on x86_64 Fedora 20, native and gdbserver.
gdb/ChangeLog:
* configure.ac: Remove tkill checks.
* configure, config.in: Regenerate.
* linux-nat.c: Remove HAVE_TKILL_SYSCALL check. Update top level
comments.
(linux_nat_post_attach_wait): Remove 'cloned' parameter. Use
__WALL.
(attach_proc_task_lwp_callback): Don't set the cloned flag.
(linux_nat_attach): Adjust.
(kill_lwp): Remove HAVE_TKILL_SYSCALL check. No longer fall back
to 'kill'.
(linux_handle_extended_wait): Use __WALL. Don't set the cloned
flag.
(wait_lwp): Use __WALL. Update comments.
(running_callback, stop_and_resume_callback): Delete.
(linux_nat_filter_event): Don't stop and resume all lwps. Don't
check if the event LWP has previously exited.
(check_zombie_leaders): Update comments.
(linux_nat_wait_1): Use __WALL.
(kill_wait_callback): Don't handle clone processes separately.
Use __WALL instead.
(linux_thread_alive): Delete.
(linux_nat_thread_alive): Return true as long as the LWP is in the
LWP list.
(linux_nat_update_thread_list): Assume the kernel supports
PTRACE_EVENT_CLONE.
(get_signo): Delete.
(lin_thread_get_thread_signals): Remove LinuxThreads references.
No longer check __pthread_sig_restart / __pthread_sig_cancel in
the inferior.
* linux-nat.h (struct lwp_info) <cloned>: Delete field.
* linux-thread-db.c: Update comments.
(_initialize_thread_db): Remove LinuxThreads references.
* nat/linux-waitpid.c (my_waitpid): No longer emulate __WALL.
Pass down flags unmodified.
* linux-waitpid.h (my_waitpid): Update documentation.
gdb/gdbserver/ChangeLog:
* linux-low.c (linux_kill_one_lwp): Remove references to
LinuxThreads.
(kill_lwp): Remove HAVE_TKILL_SYSCALL check. No longer fall back
to 'kill'.
(linux_init_signals): Delete.
(initialize_low): Adjust.
* thread-db.c (thread_db_init): Remove LinuxThreads reference.
This patch removes support for thread events if PTRACE_EVENT_CLONE is not
supported in GDBServer.
Before, on systems that did not support PTRACE_EVENT_CLONE, both GDB and
GDBServer coordinated with libthread_db.so to insert breakpoints at magic
locations in libpthread.so, in order to break at thread creation and thread
death.
Simple software single stepping support was implemented to step over these
breakpoints in case there was no hardware single stepping support. However,
these simple software single stepping implementations were not fit for any other
use as discussed in :
https://sourceware.org/ml/gdb-patches/2015-04/msg01110.html
These too simple implementations conflict with ongoing work to make proper
implementations of software single stepping in GDBServer.
The problem is that if some implementations are correct and others are not and
only there for the thread magic breakpoint, we can't enable features based
solely software single step support since some would be broken.
To keep the incorrect implementations and allow the new proper ones at the same
time we would need to implement fallback code and it quickly becomes ugly and
confusing with multiple checks for legacy software single step or proper
software single step.
However, PTRACE_EVENT_CLONE was first introduced in Linux 2.5.46,
released in November 2002.
So I think it's reasonable to just remove support for kernels that don't support
PTRACE_EVENT_CLONE, and sidestep the libthread_db breakpoints issues entirely.
This thread on the mailling list discusses the issue :
https://sourceware.org/ml/gdb/2015-10/msg00078.html
No regressions, tested on ubuntu 14.04 ARMv7 and x86.
With gdbserver-{native,extended} / { -marm -mthumb }
gdb/gdbserver/ChangeLog:
* linux-low.c (linux_look_up_symbols): Don't call
linux_supports_traceclone.
* linux-low.h (thread_db_init): Remove use_events argument.
* thread-db.c (thread_db_use_event): Remove global variable.
(struct thread_db) <td_thr_event_enable_p>: Remove field.
(struct thread_db) <td_create_bp>: Remove field.
(thread_db_create_event): Remove function.
(thread_db_enable_reporting): Likewise.
(find_one_thread): Don't check for thread_db_use_events.
(attach_thread): Likewise.
(thread_db_load_search): Remove td_thr_event_enable_p initialization.
(try_thread_db_load_1): Don't check for thread_db_use_events.
(thread_db_init): Remove use_events argument and thread events
handling.
(remove_thread_event_breakpoints): Remove function.
(thread_db_detach): Remove call to remove_thred_event_breakpoints.
gdb/ChangeLog:
* linux-thread-db.c (find_new_threads_callback): Cast ti.ti_tid to
unsigned long for debug_printf.
(thread_db_pid_to_str): Ditto.
gdb/gdbserver/ChangeLog:
* thread-db.c (find_one_thread): Cast ti.ti_tid to unsigned long
for debug_printf.
(attach_thread, find_new_threads_callback): Ditto.
Implicit void * -> function pointer conversion doesn't work in C++, so
in C++, we need to cast the result of dlsym. This adds a few typedefs
and macros that make this easy. GDBserver's version already had the
CHK macro, so I added it to GDB too.
Tested on x86_64 Fedora 20, native and gdbserver.
gdb/gdbserver/ChangeLog:
2015-08-04 Pedro Alves <palves@redhat.com>
* thread-db.c (struct thread_db): Use new typedefs.
(try_thread_db_load_1): Define local TDB_DLSYM macro and use it in
CHK calls.
(disable_thread_event_reporting): Cast result of dlsym to
destination function pointer type.
(thread_db_mourn): Use td_ta_delete_ftype.
gdb/ChangeLog:
2015-08-04 Pedro Alves <palves@redhat.com>
* nat/gdb_thread_db.h (td_init_ftype, td_ta_new_ftype)
(td_ta_map_lwp2thr_ftype, td_ta_thr_iter_ftype)
(td_ta_event_addr_ftype, td_ta_set_event_ftype)
(td_ta_clear_event_ftype, td_ta_event_getmsg_ftype)
(td_thr_validate_ftype, td_thr_get_info_ftype)
(td_thr_event_enable_ftype, td_thr_tls_get_addr_ftype)
(td_thr_tlsbase_ftype, td_symbol_list_ftype, td_ta_delete_ftype):
New typedefs.
* linux-thread-db.c (struct thread_db_info): Use new typedefs.
(try_thread_db_load_1): Define TDB_VERBOSE_DLSYM, TDB_DLSYM , CHK
local macros and use them instead of verbose_dlsym and dlsym
calls.
stdint.h was added to common-defs.h some months ago and should
no longer be included directly by any file.
gdb_assert.h was added to common-defs.h nearly a year ago, but
three includes have crept in since then.
This commit removes all such redundant include directives.
gdb/ChangeLog:
* common/buffer.c (stdint.h): Do not include.
* common/print-utils.c (stdint.h): Likewise.
* compile/compile-c-symbols.c (gdb_assert.h): Likewise.
* compile/compile-c-types.c (gdb_assert.h): Likewise.
* ft32-tdep.c (gdb_assert.h): Likewise.
* guile/scm-utils.c (stdint.h): Likewise.
* i386-linux-tdep.c (stdint.h): Likewise.
* i386-tdep.c (stdint.h): Likewise.
* nat/linux-btrace.c (stdint.h): Likewise.
* nat/linux-btrace.h (stdint.h): Likewise.
* nat/linux-ptrace.c (stdint.h): Likewise.
* nat/mips-linux-watch.h (stdint.h): Likewise.
* ppc-linux-nat.c (stdint.h): Likewise.
* python/python-internal.h (stdint.h): Likewise.
* stub-termcap.c (stdlib.h): Likewise.
* target/target.h (stdint.h): Likewise.
* xtensa-linux-nat.c (stdint.h): Likewise.
gdb/gdbserver/ChangeLog:
* linux-i386-ipa.c (stdint.h): Do not include.
* lynx-i386-low.c (stdint.h): Likewise.
* lynx-ppc-low.c (stdint.h): Likewise.
* mem-break.c (stdint.h): Likewise.
* thread-db.c (stdint.h): Likewise.
* tracepoint.c (stdint.h): Likewise.
* win32-low.c (stdint.h): Likewise.
This patch renames symbols that happen to have names which are
reserved keywords in C++.
Most of this was generated with Tromey's cxx-conversion.el script.
Some places where later hand massaged a bit, to fix formatting, etc.
And this was rebased several times meanwhile, along with re-running
the script, so re-running the script from scratch probably does not
result in the exact same output. I don't think that matters anyway.
gdb/
2015-02-27 Tom Tromey <tromey@redhat.com>
Pedro Alves <palves@redhat.com>
Rename symbols whose names are reserved C++ keywords throughout.
gdb/gdbserver/
2015-02-27 Tom Tromey <tromey@redhat.com>
Pedro Alves <palves@redhat.com>
Rename symbols whose names are reserved C++ keywords throughout.
TL;DR - GDB can hang if something refreshes the thread list out of the
target while the target is running. GDB hangs inside td_ta_thr_iter.
The fix is to not use that libthread_db function anymore.
Long version:
Running the testsuite against my all-stop-on-top-of-non-stop series is
still exposing latent non-stop bugs.
I was originally seeing this with the multi-create.exp test, back when
we were still using libthread_db thread event breakpoints. The
all-stop-on-top-of-non-stop series forces a thread list refresh each
time GDB needs to start stepping over a breakpoint (to pause all
threads). That test hits the thread event breakpoint often, resulting
in a bunch of step-over operations, thus a bunch of thread list
refreshes while some threads in the target are running.
The commit adds a real non-stop mode test that triggers the issue,
based on multi-create.exp, that does an explicit "info threads" when a
breakpoint is hit. IOW, it does the same things the as-ns series was
doing when testing multi-create.exp.
The bug is a race, so it unfortunately takes several runs for the test
to trigger it. In fact, even when setting the test running in a loop,
it sometimes takes several minutes for it to trigger for me.
The race is related to libthread_db's td_ta_thr_iter. This is
libthread_db's entry point for walking the thread list of the
inferior.
Sometimes, when GDB refreshes the thread list from the target,
libthread_db's td_ta_thr_iter can somehow see glibc's thread list as a
cycle, and get stuck in an infinite loop.
The issue is that when a thread exits, its thread control structure in
glibc is moved from a "used" list to a "cache" list. These lists are
simply circular linked lists where the "next/prev" pointers are
embedded in the thread control structure itself. The "next" pointer
of the last element of the list points back to the list's sentinel
"head". There's only one set of "next/prev" pointers for both lists;
thus a thread can only be in one of the lists at a time, not in both
simultaneously.
So when thread C exits, simplifying, the following happens. A-C are
threads. stack_used and stack_cache are the list's heads.
Before:
stack_used -> A -> B -> C -> (&stack_used)
stack_cache -> (&stack_cache)
After:
stack_used -> A -> B -> (&stack_used)
stack_cache -> C -> (&stack_cache)
td_ta_thr_iter starts by iterating at the list's head's next, and
iterates until it sees a thread whose next pointer points to the
list's head again. Thus in the before case above, C's next points to
stack_used, indicating end of list. In the same case, the stack_cache
list is empty.
For each thread being iterated, td_ta_thr_iter reads the whole thread
object out of the inferior. This includes the thread's "next"
pointer.
In the scenario above, it may happen that td_ta_thr_iter is iterating
thread B and has already read B's thread structure just before thread
C exits and its control structure moves to the cached list.
Now, recall that td_ta_thr_iter is running in the context of GDB, and
there's no locking between GDB and the inferior. From it's local copy
of B, td_ta_thr_iter believes that the next thread after B is thread
C, so it happilly continues iterating to C, a thread that has already
exited, and is now in the stack cache list.
After iterating C, td_ta_thr_iter finds the stack_cache head, which
because it is not stack_used, td_ta_thr_iter assumes it's just another
thread. After this, unless the reverse race triggers, GDB gets stuck
in td_ta_thr_iter forever walking the stack_cache list, as no thread
in thatlist has a next pointer that points back to stack_used (the
terminating condition).
Before fully understanding the issue, I tried adding cycle detection
to GDB's td_ta_thr_iter callback. However, td_ta_thr_iter skips
calling the callback in some cases, which means that it's possible
that the callback isn't called at all, making it impossible for GDB to
break the loop. I did manage to get GDB stuck in that state more than
once.
Fortunately, we can avoid the issue altogether. We don't really need
td_ta_thr_iter for live debugging nowadays, given PTRACE_EVENT_CLONE.
We already know how to map and lwp id to a thread id without iterating
(thread_from_lwp), so use that more.
gdb/ChangeLog:
2015-02-20 Pedro Alves <palves@redhat.com>
* linux-nat.c (linux_handle_extended_wait): Call
thread_db_notice_clone whenever a new clone LWP is detected.
(linux_stop_and_wait_all_lwps, linux_unstop_all_lwps): New
functions.
* linux-nat.h (thread_db_attach_lwp): Delete declaration.
(thread_db_notice_clone, linux_stop_and_wait_all_lwps)
(linux_unstop_all_lwps): Declare.
* linux-thread-db.c (struct thread_get_info_inout): Delete.
(thread_get_info_callback): Delete.
(thread_from_lwp): Use td_thr_get_info and record_thread.
(thread_db_attach_lwp): Delete.
(thread_db_notice_clone): New function.
(try_thread_db_load_1): If /proc is mounted and shows the
process'es task list, walk over all LWPs and call thread_from_lwp
instead of relying on td_ta_thr_iter.
(attach_thread): Don't call check_thread_signals here. Split the
tail part of the function (which adds the thread to the core GDB
thread list) to ...
(record_thread): ... this function. Call check_thread_signals
here.
(thread_db_wait): Don't call thread_db_find_new_threads_1. Always
call thread_from_lwp.
(thread_db_update_thread_list): Rename to ...
(thread_db_update_thread_list_org): ... this.
(thread_db_update_thread_list): New function.
(thread_db_find_thread_from_tid): Delete.
(thread_db_get_ada_task_ptid): Simplify.
* nat/linux-procfs.c: Include <sys/stat.h>.
(linux_proc_task_list_dir_exists): New function.
* nat/linux-procfs.h (linux_proc_task_list_dir_exists): Declare.
gdb/gdbserver/ChangeLog:
2015-02-20 Pedro Alves <palves@redhat.com>
* thread-db.c: Include "nat/linux-procfs.h".
(thread_db_init): Skip listing new threads if the kernel supports
PTRACE_EVENT_CLONE and /proc/PID/task/ is accessible.
gdb/testsuite/ChangeLog:
2015-02-20 Pedro Alves <palves@redhat.com>
* gdb.threads/multi-create-ns-info-thr.exp: New file.
Add a bit of debug output that made things a bit easier for me before.
gdb/
2015-02-06 Pedro Alves <palves@redhat.com>
* linux-thread-db.c (find_new_threads_callback): Add debug output.
gdb/gdbserver/
2015-02-06 Pedro Alves <palves@redhat.com>
* thread-db.c (find_new_threads_callback): Add debug output.
I wrote a test that attaches to a program that constantly spawns
short-lived threads, which exposed several issues. This is one of
them.
On GNU/Linux, attaching to a multi-threaded program sometimes prints
out warnings like:
...
[New LWP 20700]
warning: unable to open /proc file '/proc/-1/status'
[New LWP 20850]
[New LWP 21019]
...
That happens because when a thread exits, and is joined, glibc does:
nptl/pthread_join.c:
pthread_join ()
{
...
if (__glibc_likely (result == 0))
{
/* We mark the thread as terminated and as joined. */
pd->tid = -1;
...
/* Free the TCB. */
__free_tcb (pd);
}
So if we attach or interrupt the program (which does an implicit "info
threads") at just the right (or rather, wrong) time, we can find and
return threads in the libthread_db/pthreads thread list with kernel
thread ID -1. I've filed glibc PR nptl/17707 for this. You'll find
more info there.
This patch handles this as a special case in GDB.
This is actually more than just a cosmetic issue. lin_lwp_attach_lwp
will think that this -1 is an LWP we're not attached to yet, and after
failing to attach will try to check we were already attached to the
process, using a waitpid call, which in this case ends up being
"waitpid (-1, ...", which obviously results in GDB potentially
discarding an event when it shouldn't...
Tested on x86_64 Fedora 20, native and gdbserver.
gdb/gdbserver/
2015-01-09 Pedro Alves <palves@redhat.com>
* thread-db.c (find_new_threads_callback): Ignore thread if the
kernel thread ID is -1.
gdb/
2015-01-09 Pedro Alves <palves@redhat.com>
* linux-nat.c (lin_lwp_attach_lwp): Assert that the lwp id we're
about to wait for is > 0.
* linux-thread-db.c (find_new_threads_callback): Ignore thread if
the kernel thread ID is -1.
... instead of relying on libthread_db.
I wrote a test that attaches to a program that constantly spawns
short-lived threads, which exposed several issues. This is one of
them.
On Linux, we need to attach to all threads of a process (thread group)
individually. We currently rely on libthread_db to list the threads,
but that is problematic, because libthread_db relies on reading data
structures out of the inferior (which may well be corrupted). If
threads are being created or exiting just while we try to attach, we
may trip on inconsistencies in the inferior's thread list. To work
around that, when we see a seemingly corrupt list, we currently retry
a few times:
static void
thread_db_find_new_threads_2 (ptid_t ptid, int until_no_new)
{
...
if (until_no_new)
{
/* Require 4 successive iterations which do not find any new threads.
The 4 is a heuristic: there is an inherent race here, and I have
seen that 2 iterations in a row are not always sufficient to
"capture" all threads. */
...
That heuristic may well fail, and when it does, we end up with threads
in the program that aren't under GDB's control. That's obviously bad
and results in quite mistifying failures, like e.g., the process dying
for seeminly no reason when a thread that wasn't attached trips on a
breakpoint.
There's really no reason to rely on libthread_db for this nowadays
when we have /proc mounted. In that case, which is the usual case, we
can list the LWPs from /proc/PID/task/. In fact, GDBserver is already
doing this. The patch factors out that code that knows to walk the
task/ directory out of GDBserver, and makes GDB use it too.
Like GDBserver, the patch makes GDB attach to LWPs and _not_ wait for
them to stop immediately. Instead, we just tag the LWP as having an
expected stop. Because we can only set the ptrace options when the
thread stops, we need a new flag in the lwp structure to keep track of
whether we've already set the ptrace options, just like in GDBserver.
Note that nothing issues any ptrace command to the threads between the
PTRACE_ATTACH and the stop, so this is safe (unlike one scenario
described in gdbserver's linux-low.c).
When we attach to a program that has threads exiting while we attach,
it's easy to race with a thread just exiting as we try to attach to
it, like:
#1 - get current list of threads
#2 - attach to each listed thread
#3 - ooops, attach failed, thread is already gone
As this is pretty normal, we shouldn't be issuing a scary warning in
step #3.
When #3 happens, PTRACE_ATTACH usually fails with ESRCH, but sometimes
we'll see EPERM as well. That happens when the kernel still has the
thread in its task list, but the thread is marked as dead.
Unfortunately, EPERM is ambiguous and we'll get it also on other
scenarios where the thread isn't dead, and in those cases, it's useful
to get a warning. To distiguish the cases, when we get an EPERM
failure, we open /proc/PID/status, and check the thread's state -- if
the /proc file no longer exists, or the state is "Z (Zombie)" or "X
(Dead)", we ignore the EPERM error silently; otherwise, we'll warn.
Unfortunately, there seems to be a kernel race here. Sometimes I get
EPERM, and then the /proc state still indicates "R (Running)"... If
we wait a bit and retry, we do end up seeing X or Z state, or get an
ESRCH. I thought of making GDB retry the attach a few times, but even
with a 500ms wait and 4 retries, I still see the warning sometimes. I
haven't been able to identify the kernel path that causes this yet,
but in any case, it looks like a kernel bug to me. As this just
results failure to suppress a warning that we've been printing since
about forever anyway, I'm just making the test cope with it, and issue
an XFAIL.
gdb/gdbserver/
2015-01-09 Pedro Alves <palves@redhat.com>
* linux-low.c (linux_attach_fail_reason_string): Move to
nat/linux-ptrace.c, and rename.
(linux_attach_lwp): Update comment.
(attach_proc_task_lwp_callback): New function.
(linux_attach): Adjust to rename and use
linux_proc_attach_tgid_threads.
(linux_attach_fail_reason_string): Delete declaration.
gdb/
2015-01-09 Pedro Alves <palves@redhat.com>
* linux-nat.c (attach_proc_task_lwp_callback): New function.
(linux_nat_attach): Use linux_proc_attach_tgid_threads.
(wait_lwp, linux_nat_filter_event): If not set yet, set the lwp's
ptrace option flags.
* linux-nat.h (struct lwp_info) <must_set_ptrace_flags>: New
field.
* nat/linux-procfs.c: Include <dirent.h>.
(linux_proc_get_int): New parameter "warn". Handle it.
(linux_proc_get_tgid): Adjust.
(linux_proc_get_tracerpid): Rename to ...
(linux_proc_get_tracerpid_nowarn): ... this.
(linux_proc_pid_get_state): New function, factored out from
(linux_proc_pid_has_state): ... this. Add new parameter "warn"
and handle it.
(linux_proc_pid_is_gone): New function.
(linux_proc_pid_is_stopped): Adjust.
(linux_proc_pid_is_zombie_maybe_warn)
(linux_proc_pid_is_zombie_nowarn): New functions.
(linux_proc_pid_is_zombie): Use
linux_proc_pid_is_zombie_maybe_warn.
(linux_proc_attach_tgid_threads): New function.
* nat/linux-procfs.h (linux_proc_get_tgid): Update comment.
(linux_proc_get_tracerpid): Rename to ...
(linux_proc_get_tracerpid_nowarn): ... this, and update comment.
(linux_proc_pid_is_gone): New declaration.
(linux_proc_pid_is_zombie): Update comment.
(linux_proc_pid_is_zombie_nowarn): New declaration.
(linux_proc_attach_lwp_func): New typedef.
(linux_proc_attach_tgid_threads): New declaration.
* nat/linux-ptrace.c (linux_ptrace_attach_fail_reason): Adjust to
use nowarn functions.
(linux_ptrace_attach_fail_reason_string): Move here from
gdbserver/linux-low.c and rename.
(ptrace_supports_feature): If the current ptrace options are not
known yet, check them now, instead of asserting.
* nat/linux-ptrace.h (linux_ptrace_attach_fail_reason_string):
Declare.
GDB has a function named "current_inferior" and gdbserver has a global
variable named "current_inferior", but the two are not equivalent;
indeed, gdbserver does not have any real equivalent of what GDB calls
an inferior. What gdbserver's "current_inferior" is actually pointing
to is a structure describing the current thread. This commit renames
current_inferior as current_thread in gdbserver to clarify this. It
also renames the function "set_desired_inferior" to "set_desired_thread"
and renames various local variables from foo_inferior to foo_thread.
gdb/gdbserver/ChangeLog:
* inferiors.h (current_inferior): Renamed as...
(current_thread): New variable. All uses updated.
* linux-low.c (get_pc): Renamed saved_inferior as saved_thread.
(maybe_move_out_of_jump_pad): Likewise.
(cancel_breakpoint): Likewise.
(linux_low_filter_event): Likewise.
(wait_for_sigstop): Likewise.
(linux_resume_one_lwp): Likewise.
(need_step_over_p): Likewise.
(start_step_over): Likewise.
(linux_stabilize_threads): Renamed save_inferior as saved_thread.
* linux-x86-low.c (x86_linux_update_xmltarget): Likewise.
* proc-service.c (ps_lgetregs): Renamed reg_inferior as reg_thread
and save_inferior as saved_thread.
* regcache.c (get_thread_regcache): Renamed saved_inferior as
saved_thread.
(regcache_invalidate_thread): Likewise.
* remote-utils.c (prepare_resume_reply): Likewise.
* thread-db.c (thread_db_get_tls_address): Likewise.
(disable_thread_event_reporting): Likewise.
(remove_thread_event_breakpoints): Likewise.
* tracepoint.c (gdb_agent_about_to_close): Renamed save_inferior
as saved_thread.
* target.h (set_desired_inferior): Renamed as...
(set_desired_thread): New declaration. All uses updated.
* server.c (myresume): Updated comment to reference thread instead
of inferior.
(handle_serial_event): Likewise.
(handle_target_event): Likewise.
I have posted:
TLS variables access for -static -lpthread executables
https://sourceware.org/ml/libc-help/2014-03/msg00024.html
and the GDB patch below has been confirmed as OK for current glibcs.
Further work should be done for newer glibcs:
Improve TLS variables glibc compatibility
https://sourceware.org/bugzilla/show_bug.cgi?id=16954
Still the patch below implements the feature in a fully functional way backward
compatible with current glibcs, it depends on the following glibc source line:
csu/libc-tls.c
main_map->l_tls_modid = 1;
gdb/
2014-05-21 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix TLS access for -static -pthread.
* linux-thread-db.c (struct thread_db_info): Add td_thr_tlsbase_p.
(try_thread_db_load_1): Initialize it.
(thread_db_get_thread_local_address): Call it if LM is zero.
* target.c (target_translate_tls_address): Remove LM_ADDR zero check.
* target.h (struct target_ops) (to_get_thread_local_address): Add
load_module_addr comment.
gdb/gdbserver/
2014-05-21 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix TLS access for -static -pthread.
* gdbserver/thread-db.c (struct thread_db): Add td_thr_tlsbase_p.
(thread_db_get_tls_address): Call it if LOAD_MODULE is zero.
(thread_db_load_search, try_thread_db_load_1): Initialize it.
gdb/testsuite/
2014-05-21 Jan Kratochvil <jan.kratochvil@redhat.com>
Fix TLS access for -static -pthread.
* gdb.threads/staticthreads.c <HAVE_TLS> (tlsvar): New.
<HAVE_TLS> (thread_function, main): Initialize it.
* gdb.threads/staticthreads.exp: Try gdb_compile_pthreads for $have_tls.
Add clean_restart.
<$have_tls != "">: Check TLSVAR.
Message-ID: <20140410115204.GB16411@host2.jankratochvil.net>
On Linux, we need to explicitly ptrace attach to all lwps of a
process. Because GDB might not be connected yet when an attach is
requested, and thus it may not be possible to activate thread_db, as
that requires access to symbols (IOW, gdbserver --attach), a while ago
we make linux_attach loop over the lwps as listed by /proc/PID/task to
find the lwps to attach to.
linux_attach_lwp_1 has:
...
if (initial)
/* If lwp is the tgid, we handle adding existing threads later.
Otherwise we just add lwp without bothering about any other
threads. */
ptid = ptid_build (lwpid, lwpid, 0);
else
{
/* Note that extracting the pid from the current inferior is
safe, since we're always called in the context of the same
process as this new thread. */
int pid = pid_of (current_inferior);
ptid = ptid_build (pid, lwpid, 0);
}
That "safe" comment referred to linux_attach_lwp being called by
thread-db.c. But this was clearly missed when a new call to
linux_attach_lwp_1 was added to linux_attach. As a result,
current_inferior will be set to some random process, and non-initial
lwps of the second inferior get assigned the pid of the wrong
inferior. E.g., in the case of attaching to two inferiors, for the
second inferior (and so on), non-initial lwps of the second inferior
get assigned the pid of the first inferior. This doesn't trigger on
the first inferior, when current_inferior is NULL, add_thread switches
the current inferior to the newly added thread.
Rather than making linux_attach switch current_inferior temporarily
(thus avoiding further reliance on global state), or making
linux_attach_lwp_1 get the tgid from /proc, which add extra syscalls,
and will be wrong in case of the user having originally attached
directly to a non-tgid lwp, and then that lwp spawning new clones (the
ptid.pid field of further new clones should be the same as the
original lwp's pid, which is not the tgid), we note that callers of
linux_attach_lwp/linux_attach_lwp_1 always have the right pid handy
already, so they can pass it down along with the lwpid.
The only other reason for the "initial" parameter is to error out
instead of warn in case of attach failure, when we're first attaching
to a process. There are only three callers of
linux_attach_lwp/linux_attach_lwp_1, and each wants to print a
different warn/error string, so we can just move the error/warn out of
linux_attach_lwp_1 to the callers, thus getting rid of the "initial"
parameter.
There really nothing gdbserver-specific about attaching to two
threaded processes, so this adds a new test under gdb.multi/. The
test passes cleanly against the native GNU/Linux target, but
fails/triggers the bug against GDBserver (before the patch), with the
native-extended-remote board (as plain remote doesn't support
multi-process).
Tested on x86_64 Fedora 17, with the native-extended-gdbserver board.
gdb/gdbserver/
2014-04-25 Pedro Alves <palves@redhat.com>
PR server/16255
* linux-low.c (linux_attach_fail_reason_string): New function.
(linux_attach_lwp): Delete.
(linux_attach_lwp_1): Rename to ...
(linux_attach_lwp): ... this. Take a ptid instead of a pid as
argument. Remove "initial" parameter. Return int instead of
void. Don't error or warn here.
(linux_attach): Adjust to call linux_attach_lwp. Call error on
failure to attach to the tgid. Call warning when failing to
attach to an lwp.
* linux-low.h (linux_attach_lwp): Take a ptid instead of a pid as
argument. Remove "initial" parameter. Return int instead of
void. Don't error or warn here.
(linux_attach_fail_reason_string): New declaration.
* thread-db.c (attach_thread): Adjust to linux_attach_lwp's
interface change. Use linux_attach_fail_reason_string.
gdb/
2014-04-25 Pedro Alves <palves@redhat.com>
PR server/16255
* common/linux-ptrace.c (linux_ptrace_attach_warnings): Rename to ...
(linux_ptrace_attach_fail_reason): ... this. Remove "warning: "
and newline from built string.
* common/linux-ptrace.h (linux_ptrace_attach_warnings): Rename to ...
(linux_ptrace_attach_fail_reason): ... this.
* linux-nat.c (linux_nat_attach): Adjust to use
linux_ptrace_attach_fail_reason.
gdb/testsuite/
2014-04-25 Simon Marchi <simon.marchi@ericsson.com>
Pedro Alves <palves@redhat.com>
PR server/16255
* gdb.multi/multi-attach.c: New file.
* gdb.multi/multi-attach.exp: New file.
null ptr check to prevent gdbserver from crashing
Evaluating a thread local storage variable in a remote scenario crashes
gdbserver if libthread-db could not be loaded.
2013-02-12 Sanimir Agovic <sanimir.agovic@intel.com>
gdbserver/
* thread-db.c (thread_db_get_tls_address):
NULL pointer check thread_db.
testsuite/
* gdb.server/no-thread-db.exp: New file.
* gdb.server/no-thread-db.c: New file.
* gdb.server/Makefile.in (EXECUTABLES): Add no-thread-db.
Two modifications:
1. The addition of 2013 to the copyright year range for every file;
2. The use of a single year range, instead of potentially multiple
year ranges, as approved by the FSF.
* common/gdb_string.h: ... here.
* common/vec.h: Remove #ifndef GDBSERVER conditional inclusion of
gdb_string.h and gdb_assert.h.
gdbserver/
* configure.ac: Add check for strstr.
* config.in: Regenerate.
* configure: Regenerate.
* linux-thread-db.c: #include "gdb_vecs.h".
(try_thread_db_load_from_pdir_1): New arg "subdir". All callers
updated.
(try_thread_db_load_from_pdir): New arg "subdir". All callers updated.
(thread_db_load_search): Use a vector to iterate over path elements.
Handle text appearing after "$pdir".
gdbserver/
* Makefile.in (SFILES): Add gdb_vecs.c.
(OBS): Add gdb_vecs.o.
(gdb_vecs_h, host_defs_h): New variables.
(thread-db.o): Add $(gdb_vecs_h) dependency.
(gdb_vecs.o): New rule.
* thread-db.c: #include "gdb_vecs.h".
(thread_db_load_search): Use a vector to iterate over path elements.
Handle text appearing after "$pdir".
* NEWS: Mention $sdir,$pdir.
* gdb_thread_db.h (LIBTHREAD_DB_SEARCH_PATH): Add $sdir:$pdir.
* linux-thread-db.c (try_thread_db_load_from_pdir): New function.
(try_thread_db_load_from_sdir): New function.
(try_thread_db_load_from_dir): New function.
(thread_db_load_search): Handle $pdir, $sdir. Remove trying of
system directories if search of libthread-db-search-path fails,
that is now done via $sdir.
(has_libpthread): New function.
(thread_db_load): Remove search for libthread_db in directory of
libpthread, that is now done via $pdir.
gdbserver/
* thread-db.c (try_thread_db_load_from_sdir): New function.
(try_thread_db_load_from_dir): New function.
(thread_db_load_search): Handle $sdir, ignore $pdir.
Remove trying of system directories if search of
libthread-db-search-path fails, that is now done via $sdir.
doc/
* gdb.texinfo (Threads): Document $sdir,$pdir.
(Server): Document $pdir exception.
(_initialize_thread_db): Add setter for libthread-db-search-path.
gdbserver/
* thread-db.c (thread_db_handle_monitor_command): Handle elided path.
doc/
* gdb.texinfo (Threads): If an empty path is provided for
libthread-db-search-path it is reset to its default value.
(Server): Ditto.
snprintf is not available on LynxOS, so I changed the calls to snprintf
to calls to xsnprintf, which should be strictly equivalent.
gdb/gdbserver/ChangeLog:
* utils.c (xsnprintf): Make non-static.
* server.h: Add xsnprintf declaration.
* linux-low.c, nto-low.c, target.c, thread-db.c, tracepoint.c:
replace calls to snprintf by calls to xsnprintf throughout.
* proc-service.c (ps_pglobal_lookup): Use
thread_db_look_up_one_symbol.
* remote-utils.c (look_up_one_symbol): Add new `may_ask_gdb'
parameter. Use it instead of all_symbols_looked_up.
* server.h (struct process_info) <all_symbols_looked_up>: Delete
field.
(all_symbols_looked_up): Don't declare.
(look_up_one_symbol): Add new `may_ask_gdb' parameter.
* thread-db.c (struct thread_db) <all_symbols_looked_up>: New
field.
(thread_db_look_up_symbols): Adjust call to look_up_one_symbol.
Set all_symbols_looked_up here.
(thread_db_look_up_one_symbol): New.
(thread_db_get_tls_address): Adjust.
(thread_db_load_search, try_thread_db_load_1): Always allocate the
thread_db object on the heap, and tentatively set it in the
process structure.
(thread_db_init): Don't set all_symbols_looked_up here.
* linux-low.h (thread_db_look_up_one_symbol): Declare.
(linux_kill): Stop all lwps here. Don't delete the main lwp here.
(linux_detach_one_lwp): Assume the lwp is stopped.
(any_thread_of): Delete.
(linux_detach): Stop all lwps here. Don't blindly delete all
breakpoints.
(delete_lwp_callback): New.
(linux_mourn): Delete all lwps of the process that is gone.
(linux_wait_1): Don't delete the last lwp of the process here.
* mem-break.h (mark_breakpoints_out): Declare.
* mem-break.c (mark_breakpoints_out): New.
(free_all_breakpoints): Use it.
* server.c (handle_target_event): If the process is gone, mark
breakpoints out.
* thread-db.c (struct thread_db) <create_bp>: New field.
(thread_db_enable_reporting): Fix prototype. Store a thread event
breakpoint reference in the thread_db struct.
(thread_db_load_search): Clear the thread_db object.
(try_thread_db_load_1): Ditto.
(switch_to_process): New.
(disable_thread_event_reporting): Use it.
(remove_thread_event_breakpoints): New.
(thread_db_detach, thread_db_mourn): Use it.
* linux-low.c (linux_remove_process): Delete.
(add_lwp): Don't set last_resume_kind here.
(linux_kill): Use `mourn'.
(linux_detach): Use `thread_db_detach', and `mourn'.
(linux_mourn): New.
(linux_attach_lwp_1): Adjust comment.
(linux_attach): last_resume_kind moved the thread_info; adjust.
(status_pending_p_callback): Adjust.
(linux_wait_for_event_1): Adjust.
(count_events_callback, select_singlestep_lwp_callback)
(select_event_lwp_callback, cancel_breakpoints_callback)
(db_wants_lwp_stopped, linux_wait_1, need_step_over_p)
(proceed_one_lwp): Adjust.
(linux_async): Add debug output.
(linux_thread_stopped): New.
(linux_pause_all): New.
(linux_target_ops): Install linux_mourn, linux_thread_stopped and
linux_pause_all.
* linux-low.h (struct lwp_info): Delete last_resume_kind field.
(thread_db_free): Delete declaration.
(thread_db_detach, thread_db_mourn): Declare.
* thread-db.c (thread_db_init): Use thread_db_mourn.
(thread_db_free): Delete, split in two.
(disable_thread_event_reporting): New.
(thread_db_detach): New.
(thread_db_mourn): New.
* server.h (struct thread_info) <last_resume_kind>: New field.
<attached>: Add comment.
<gdb_detached>: New field.
(handler_func): Change return type to int.
(handle_serial_event, handle_target_event): Ditto.
(gdb_connected): Declare.
(tracing): Delete.
(disconnected_tracing): Declare.
(stop_tracing): Declare.
* server.c (handle_query) <qSupported>: Report support for
disconnected tracing.
(queue_stop_reply_callback): Account for running threads.
(gdb_wants_thread_stopped): New.
(gdb_wants_all_threads_stopped): New.
(gdb_reattached_process): New.
(handle_status): Clear the `gdb_detached' flag of all processes.
In all-stop, stop all threads.
(main): Be sure to leave tfind mode. Handle disconnected tracing.
(process_serial_event): If the remote connection breaks, or if an
exit was forced with "monitor exit", force an event loop exit.
Handle disconnected tracing on detach.
(handle_serial_event): Adjust.
(handle_target_event): If GDB isn't connected, forward events back
to the inferior, unless the last process exited, in which case,
exit gdbserver. Adjust interface.
* remote-utils.c (remote_open): Don't block in accept. Instead
register an event loop source on the listen socket file
descriptor. Refactor bits into ...
(listen_desc): ... this new global.
(gdb_connected): ... this new function.
(enable_async_notification): ... this new function.
(handle_accept_event): ... this new function.
(remote_close): Clear remote_desc.
* inferiors.c (add_thread): Set the new thread's last_resume_kind.
* target.h (struct target_ops) <mourn, thread_stopped, pause_all>:
New fields.
(mourn_inferior): Define.
(target_process_qsupported): Avoid the dangling else problem.
(thread_stopped): Define.
(pause_all): Define.
(target_waitstatus_to_string): Declare.
* target.c (target_waitstatus_to_string): New.
* tracepoint.c (tracing): Make extern.
(disconnected_tracing): New.
(stop_tracing): Make extern. Handle tracing stops due to GDB
disconnecting.
(cmd_qtdisconnected): New.
(cmd_qtstatus): Report disconnected tracing status in trace reply.
(handle_tracepoint_general_set): Handle QTDisconnected.
* event-loop.c (event_handler_func): Change return type to int.
(process_event): Bail out if the event handler wants the event
loop to stop.
(handle_file_event): Ditto.
(start_event_loop): Bail out if the event handler wants the event
loop to stop.
* nto-low.c (nto_target_ops): Adjust.
* spu-low.c (spu_wait): Don't remove the process here.
(spu_target_ops): Adjust.
* win32-low.c (win32_wait): Don't remove the process here.
(win32_target_ops): Adjust.
parameter. Don't release/detach from thread_db here.
(linux_kill): Release/detach from thread_db here, ...
(linux_detach): ... and here, before actually detaching.
(linux_wait_1): ... and here, when a process exits.
* thread-db.c (any_thread_of): New.
(thread_db_free): Switch the current inferior to a thread of the
passed in process.
* gdb_proc_service.h (psaddr_t): Fix type.
* linux-thread-db.c (thread_db_info.td_thr_tls_get_addr_p): Fix
signature to match glibc.
(thread_db_get_thread_local_address): Use psaddr_t for type of
address to match parameter of td_thr_tls_get_addr_p.
Cast through psaddr_t instead of (void*) to match parameter of
td_thr_tls_get_addr_p.
gdbserver/
* gdb_proc_service.h (psaddr_t): Fix type.
* thread-db.c (thread_db_info.td_thr_tls_get_addr_p): Fix
signature to match glibc.
Pass it to thread_db_free.
(linux_kill, linux_detach, linux_wait_1): Adjust to pass the
proper `detaching' argument to linux_remove_process.
* linux-low.h (thread_db_free): Add `detaching' parameter.
* thread-db.c (thread_db_init): Pass false as `detaching' argument
to thread_db_free.
(thread_db_free): Add `detaching' parameter. Only
call td_ta_clear_event if detaching from process.
PR gdb/10757
* thread-db.c (attach_thread): New function.
(maybe_attach_thread): Return success/failure.
(find_new_threads_callback): Adjust.
(thread_db_find_new_threads): Loop until no new threads.
* server.h (get_thread_process): Add prototype.
* thread-db.c (find_one_thread): Use get_thread_process
instead of current_process.
(thread_db_get_tls_address): Do not crash if called when
thread layer is not yet initialized.
Pedro Alves
Yao Qi
Joel Brobecker
Antoine Tremblay
Doug Evans
Simon Marchi
Gary Benson
Jan Kratochvil
Joel Brobecker
Yao Qi
Michael Snyder
gdbadmin
Maxim Kuvyrkov
Paul Pluzhnikov
Ulrich Weigand