[Bug gdb/28942] New: Problem with breakpoint condition calling a function in multi-threaded program
simon.marchi at polymtl dot ca
sourceware-bugzilla@sourceware.org
Thu Mar 3 19:40:58 GMT 2022
https://sourceware.org/bugzilla/show_bug.cgi?id=28942
Bug ID: 28942
Summary: Problem with breakpoint condition calling a function
in multi-threaded program
Product: gdb
Version: HEAD
Status: UNCONFIRMED
Severity: normal
Priority: P2
Component: gdb
Assignee: unassigned at sourceware dot org
Reporter: simon.marchi at polymtl dot ca
Target Milestone: ---
This program:
---8<---
#include <pthread.h>
#include <unistd.h>
static void
function_that_segfaults (void)
{
int *p = 0;
*p = 1;
}
static void
break_here (void)
{}
static void *
thread_func (void *p)
{
for (;;)
sleep (1);
return NULL;
}
static void *
thread_func2 (void *p)
{
sleep (1);
break_here ();
return NULL;
}
int
main (void)
{
pthread_t threads[10];
pthread_create (&threads[0], NULL, thread_func, NULL);
pthread_create (&threads[1], NULL, thread_func, NULL);
pthread_create (&threads[2], NULL, thread_func, NULL);
pthread_create (&threads[3], NULL, thread_func, NULL);
pthread_create (&threads[5], NULL, thread_func, NULL);
pthread_create (&threads[6], NULL, thread_func, NULL);
pthread_create (&threads[4], NULL, thread_func2, NULL);
sleep (60);
return function_that_segfaults != 0;
}
--->8---
$ gcc test.c -g3 -O0 -pthread
$ ./gdb -q -nx --data-directory=data-directory a.out -ex "b break_here if
function_that_segfaults()"
Reading symbols from a.out...
Breakpoint 1 at 0x11ae: file test.c, line 13.
(gdb) r
Starting program: /home/smarchi/build/binutils-gdb/gdb/a.out
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/x86_64-linux-gnu/libthread_db.so.1".
[New Thread 0x7ffff7d99700 (LWP 3567019)]
[New Thread 0x7ffff7598700 (LWP 3567020)]
[New Thread 0x7ffff6d97700 (LWP 3567021)]
[New Thread 0x7ffff6596700 (LWP 3567022)]
[New Thread 0x7ffff5d95700 (LWP 3567023)]
[New Thread 0x7ffff5594700 (LWP 3567024)]
[New Thread 0x7ffff4d93700 (LWP 3567025)]
Error in testing breakpoint condition:
Couldn't get registers: No such process.
An error occurred while in a function called from GDB.
Evaluation of the expression containing the function
(function_that_segfaults) will be abandoned.
When the function is done executing, GDB will silently stop.
Selected thread is running.
(gdb)
The "Couldn't get registers: No such process." is very strange. We expect GDB
to say that the thread received a signal (SIGSEGV) while running the
hand-called function.
And then if you continue with:
(gdb) kill
Kill the program being debugged? (y or n) y
[Inferior 1 (process 3567034) killed]
(gdb) r
Starting program: /home/smarchi/build/binutils-gdb/gdb/a.out
/home/smarchi/src/binutils-gdb/gdb/target.c:2607: internal-error: target_wait:
Assertion `!proc_target->commit_resumed_state' failed.
A problem internal to GDB has been detected,
further debugging may prove unreliable.
Looking at the proceed call here:
(top-gdb) bt
#0 proceed (addr=0x555555555189, siggnal=GDB_SIGNAL_0) at
/home/smarchi/src/binutils-gdb/gdb/infrun.c:3046
#1 0x0000558e5d95a128 in run_inferior_call
(sm=std::unique_ptr<call_thread_fsm> = {...}, call_thread=0x61700009e680,
real_pc=0x555555555189) at /home/smarchi/src/binutils-gdb/gdb/infcall.c:610
#2 0x0000558e5d95ff6e in call_function_by_hand_dummy (function=0x611000489d00,
default_return_type=0x0, args=..., dummy_dtor=0x0, dummy_dtor_data=0x0) at
/home/smarchi/src/binutils-gdb/gdb/infcall.c:1279
#3 0x0000558e5d95b4be in call_function_by_hand (function=0x611000489d00,
default_return_type=0x0, args=...) at
/home/smarchi/src/binutils-gdb/gdb/infcall.c:741
#4 0x0000558e5d609a2e in evaluate_subexp_do_call (exp=0x6030001579f0,
noside=EVAL_NORMAL, callee=0x611000489d00, argvec=..., function_name=0x0,
default_return_type=0x0) at /home/smarchi/src/binutils-gdb/gdb/eval.c:674
#5 0x0000558e5d60a7c5 in expr::operation::evaluate_funcall
(this=0x603000157ab0, expect_type=0x0, exp=0x6030001579f0, noside=EVAL_NORMAL,
function_name=0x0, args=std::__debug::vector of length 0, capacity 0) at
/home/smarchi/src/binutils-gdb/gdb/eval.c:702
#6 0x0000558e5c4090aa in expr::operation::evaluate_funcall
(this=0x603000157ab0, expect_type=0x0, exp=0x6030001579f0, noside=EVAL_NORMAL,
args=std::__debug::vector of length 0, capacity 0) at
/home/smarchi/src/binutils-gdb/gdb/expression.h:136
#7 0x0000558e5d60ad63 in expr::var_value_operation::evaluate_funcall
(this=0x603000157ab0, expect_type=0x0, exp=0x6030001579f0, noside=EVAL_NORMAL,
args=std::__debug::vector of length 0, capacity 0) at
/home/smarchi/src/binutils-gdb/gdb/eval.c:714
#8 0x0000558e5cb8d2be in expr::funcall_operation::evaluate
(this=0x607000083f80, expect_type=0x0, exp=0x6030001579f0, noside=EVAL_NORMAL)
at /home/smarchi/src/binutils-gdb/gdb/expop.h:2178
#9 0x0000558e5d604e00 in expression::evaluate (During symbol reading: Child
DIE 0x8d876c and its abstract origin 0x8f9b2b have different parents
sthis=0x6030001579f0, expect_type=0x0, noside=EVAL_NORMAL) at
/home/smarchi/src/binutils-gdb/gdb/eval.c:101
#10 0x0000558e5d604f71 in evaluate_expression (exp=0x6030001579f0,
expect_type=0x0) at /home/smarchi/src/binutils-gdb/gdb/eval.c:115
#11 0x0000558e5c8c99b9 in breakpoint_cond_eval (exp=0x6030001579f0) at
/home/smarchi/src/binutils-gdb/gdb/breakpoint.c:4739
#12 0x0000558e5c8d1f11 in bpstat_check_breakpoint_conditions
(bs=0x6060001b29c0, thread=0x61700009e680) at
/home/smarchi/src/binutils-gdb/gdb/breakpoint.c:5303
#13 0x0000558e5c8d4b45 in bpstat_stop_status (aspace=0x603000045a00,
bp_addr=0x5555555551ae, thread=0x61700009e680, ws=...,
stop_chain=0x6060001b29c0) at
/home/smarchi/src/binutils-gdb/gdb/breakpoint.c:5475
#14 0x0000558e5da1f939 in handle_signal_stop (ecs=0x7fff97a4bd50) at
/home/smarchi/src/binutils-gdb/gdb/infrun.c:6200
#15 0x0000558e5da19441 in handle_inferior_event (ecs=0x7fff97a4bd50) at
/home/smarchi/src/binutils-gdb/gdb/infrun.c:5690
#16 0x0000558e5da05206 in fetch_inferior_event () at
/home/smarchi/src/binutils-gdb/gdb/infrun.c:4091
#17 0x0000558e5d94fad4 in inferior_event_handler (event_type=INF_REG_EVENT) at
/home/smarchi/src/binutils-gdb/gdb/inf-loop.c:41
#18 0x0000558e5dc29bdd in handle_target_event (error=0, client_data=0x0) at
/home/smarchi/src/binutils-gdb/gdb/linux-nat.c:4096
#19 0x0000558e5f4e4dd1 in handle_file_event (file_ptr=0x607000016050,
ready_mask=1) at /home/smarchi/src/binutils-gdb/gdbsupport/event-loop.cc:574
#20 0x0000558e5f4e562c in gdb_wait_for_event (block=0) at
/home/smarchi/src/binutils-gdb/gdbsupport/event-loop.cc:700
#21 0x0000558e5f4e343c in gdb_do_one_event () at
/home/smarchi/src/binutils-gdb/gdbsupport/event-loop.cc:212
#22 0x0000558e5dd29d99 in start_event_loop () at
/home/smarchi/src/binutils-gdb/gdb/main.c:421
#23 0x0000558e5dd2a1df in captured_command_loop () at
/home/smarchi/src/binutils-gdb/gdb/main.c:481
#24 0x0000558e5dd2fad9 in captured_main (data=0x7fff97a4c200) at
/home/smarchi/src/binutils-gdb/gdb/main.c:1348
#25 0x0000558e5dd2fbc2 in gdb_main (args=0x7fff97a4c200) at
/home/smarchi/src/binutils-gdb/gdb/main.c:1363
#26 0x0000558e5c3e1ddd in main (argc=7, argv=0x7fff97a4c378) at
/home/smarchi/src/binutils-gdb/gdb/gdb.c:32
We find that GDB tries to resume some other threads than the event thread (for
which we evaluate the breakpoint condition), because it thinks they are not
resumed. Probably because when the linux-nat target added them, they were added
in the non-resumed state and stayed this way.
--
You are receiving this mail because:
You are on the CC list for the bug.
More information about the Gdb-prs
mailing list