[Bug c++/16997] New: Stepping hangs on gettimeofday with recording enabled

[Bug c++/16997] New: Stepping hangs on gettimeofday with recording enabled

[konradsa at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=16997

            Bug ID: 16997
           Summary: Stepping hangs on gettimeofday with recording enabled
           Product: gdb
           Version: 7.7
            Status: NEW
          Severity: normal
          Priority: P2
         Component: c++
          Assignee: unassigned at sourceware dot org
          Reporter: konradsa at gmail dot com

Consider this simple program:

#include <iostream>

#include <sys/time.h>
#include <time.h>

using namespace std;

int main() {
    timeval now;
    gettimeofday( &now, 0);

    cout << "!!!Hello World!!!" << endl; // prints !!!Hello World!!!
    return 0;
}


When stepping through the code with reverse debugging enabled (recording
enabled), gdb seems to hang on the gettimeofday call. Here some sample output:

(gdb) run
Starting program: /home12/skonrad/workspace-luna/gdbTest/Debug/gdbTest

Breakpoint 1, main () at ../src/gdbTest.cpp:18
18          gettimeofday( &now, 0);
(gdb) record
(gdb) n
Do you want to auto delete previous execution log entries when record/replay
buffer becomes full (record full stop-at-limit)?([y] or n) y
...

Waited for several minutes, step never completed. Without recording, no issue.
This was tested on RedHat Linux with g++ (GCC) 4.4.7 20120313 (Red Hat
4.4.7-3).

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug c++/16997] Stepping hangs on gettimeofday with recording enabled

[konradsa at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=16997

konradsa <konradsa at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |konradsa at gmail dot com

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug c++/16997] Stepping hangs on gettimeofday with recording enabled

[palves at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=16997

Pedro Alves <palves at redhat dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |WAITING
                 CC|                            |palves at redhat dot com

--- Comment #1 from Pedro Alves <palves at redhat dot com> ---
Try with "set debug lin-lwp 1" and "set debug infrun 1", which enables relevant
logging.

gettimeofday lives in the vDSO nowadays.  Maybe it's something to do with
single-stepping in the vDSO.  Recording behind the scenes single-steps the
programs one instruction at a time.

Can you do that manually, with "si", and confirm whether you can single-step
all the way across the vDSO (in and out) ?

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug c++/16997] Stepping hangs on gettimeofday with recording enabled

[konradsa at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=16997

--- Comment #2 from konradsa <konradsa at gmail dot com> ---
It seems to get stuck in some sort of loop:

Here the tail end of the output I see when stepping with si:

...
(gdb) si
infrun: clear_proceed_status_thread (process 45441)
infrun: proceed (addr=0xffffffffffffffff, signal=GDB_SIGNAL_DEFAULT, step=1)
infrun: resume (step=1, signal=GDB_SIGNAL_0), trap_expected=0, current thread
[process 45441] at 0x3181609d56
LLR: Preparing to step process 45441, 0, inferior_ptid process 45441
RC: Not resuming sibling process 45441 (not stopped)
LLR: PTRACE_SINGLESTEP process 45441, 0 (resume event thread)
sigchld
infrun: wait_for_inferior ()
linux_nat_wait: [process -1], []
LLW: enter
LNW: waitpid(-1, ...) returned 45441, No child processes
LLW: waitpid 45441 received Trace/breakpoint trap (stopped)
LLW: Candidate event Trace/breakpoint trap (stopped) in process 45441.
SEL: Select single-step process 45441
LLW: trap ptid is process 45441.
LLW: exit
infrun: target_wait (-1, status) =
infrun:   45441 [process 45441],
infrun:   status->kind = stopped, signal = GDB_SIGNAL_TRAP
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x3181609d59
infrun: stepi/nexti
infrun: stop_stepping
0x0000003181609d59 in _dl_lookup_symbol_x () from /lib64/ld-linux-x86-64.so.2
(gdb) si
infrun: clear_proceed_status_thread (process 45441)
infrun: proceed (addr=0xffffffffffffffff, signal=GDB_SIGNAL_DEFAULT, step=1)
infrun: resume (step=1, signal=GDB_SIGNAL_0), trap_expected=0, current thread
[process 45441] at 0x3181609d59
LLR: Preparing to step process 45441, 0, inferior_ptid process 45441
RC: Not resuming sibling process 45441 (not stopped)
LLR: PTRACE_SINGLESTEP process 45441, 0 (resume event thread)
sigchld
infrun: wait_for_inferior ()
linux_nat_wait: [process -1], []
LLW: enter
LNW: waitpid(-1, ...) returned 45441, No child processes
LLW: waitpid 45441 received Trace/breakpoint trap (stopped)
LLW: Candidate event Trace/breakpoint trap (stopped) in process 45441.
SEL: Select single-step process 45441
LLW: trap ptid is process 45441.
LLW: exit
infrun: target_wait (-1, status) =
infrun:   45441 [process 45441],
infrun:   status->kind = stopped, signal = GDB_SIGNAL_TRAP
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x3181609d5b
infrun: stepi/nexti
infrun: stop_stepping
0x0000003181609d5b in _dl_lookup_symbol_x () from /lib64/ld-linux-x86-64.so.2
(gdb) si
infrun: clear_proceed_status_thread (process 45441)
infrun: proceed (addr=0xffffffffffffffff, signal=GDB_SIGNAL_DEFAULT, step=1)
infrun: resume (step=1, signal=GDB_SIGNAL_0), trap_expected=0, current thread
[process 45441] at 0x3181609d5b
LLR: Preparing to step process 45441, 0, inferior_ptid process 45441
RC: Not resuming sibling process 45441 (not stopped)
LLR: PTRACE_SINGLESTEP process 45441, 0 (resume event thread)
sigchld
infrun: wait_for_inferior ()
linux_nat_wait: [process -1], []
LLW: enter
LNW: waitpid(-1, ...) returned 45441, No child processes
LLW: waitpid 45441 received Trace/breakpoint trap (stopped)
LLW: Candidate event Trace/breakpoint trap (stopped) in process 45441.
SEL: Select single-step process 45441
LLW: trap ptid is process 45441.
LLW: exit
infrun: target_wait (-1, status) =
infrun:   45441 [process 45441],
infrun:   status->kind = stopped, signal = GDB_SIGNAL_TRAP
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0x3181609d40
infrun: stepi/nexti
infrun: stop_stepping
0x0000003181609d40 in _dl_lookup_symbol_x () from /lib64/ld-linux-x86-64.so.2

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug c++/16997] Stepping hangs on gettimeofday with recording enabled

[konradsa at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=16997

--- Comment #3 from konradsa <konradsa at gmail dot com> ---
If I just continue, here is what I continuously see:

...

LLW: enter
LNW: waitpid(-1, ...) returned 45441, No child processes
LLW: waitpid 45441 received Trace/breakpoint trap (stopped)
LLW: Candidate event Trace/breakpoint trap (stopped) in process 45441.
SEL: Select single-step process 45441
LLW: trap ptid is process 45441.
LLW: exit
LLR: Preparing to step process -1, 0, inferior_ptid process -1
RC: Not resuming sibling process 45441 (not stopped)
LLR: PTRACE_SINGLESTEP process 45441, 0 (resume event thread)
linux_nat_wait: [process -1], []
sigchld
LLW: enter
LNW: waitpid(-1, ...) returned 45441, No child processes
LLW: waitpid 45441 received Trace/breakpoint trap (stopped)
LLW: Candidate event Trace/breakpoint trap (stopped) in process 45441.
SEL: Select single-step process 45441
LLW: trap ptid is process 45441.
LLW: exit
LLR: Preparing to step process -1, 0, inferior_ptid process -1
RC: Not resuming sibling process 45441 (not stopped)
LLR: PTRACE_SINGLESTEP process 45441, 0 (resume event thread)
linux_nat_wait: [process -1], []
sigchld
LLW: enter
LNW: waitpid(-1, ...) returned 45441, No child processes
LLW: waitpid 45441 received Trace/breakpoint trap (stopped)
LLW: Candidate event Trace/breakpoint trap (stopped) in process 45441.
SEL: Select single-step process 45441
LLW: trap ptid is process 45441.
LLW: exit
LLR: Preparing to step process -1, 0, inferior_ptid process -1
RC: Not resuming sibling process 45441 (not stopped)
LLR: PTRACE_SINGLESTEP process 45441, 0 (resume event thread)
linux_nat_wait: [process -1], []
sigchld
LLW: enter
LNW: waitpid(-1, ...) returned 45441, No child processes
LLW: waitpid 45441 received Trace/breakpoint trap (stopped)
LLW: Candidate event Trace/breakpoint trap (stopped) in process 45441.
SEL: Select single-step process 45441
LLW: trap ptid is process 45441.
LLW: exit
LLR: Preparing to step process -1, 0, inferior_ptid process -1
RC: Not resuming sibling process 45441 (not stopped)
LLR: PTRACE_SINGLESTEP process 45441, 0 (resume event thread)
linux_nat_wait: [process -1], []
sigchld
LLW: enter
LNW: waitpid(-1, ...) returned 45441, No child processes
LLW: waitpid 45441 received Trace/breakpoint trap (stopped)
LLW: Candidate event Trace/breakpoint trap (stopped) in process 45441.
SEL: Select single-step process 45441
LLW: trap ptid is process 45441.
LLW: exit
LLR: Preparing to step process -1, 0, inferior_ptid process -1
RC: Not resuming sibling process 45441 (not stopped)
LLR: PTRACE_SINGLESTEP process 45441, 0 (resume event thread)
linux_nat_wait: [process -1], []
sigchld
LLW: enter
LNW: waitpid(-1, ...) returned 45441, No child processes
^CLLW: waitpid 45441 received Trace/breakpoint trap (stopped)
LLW: Candidate event Trace/breakpoint trap (stopped) in process 45441.
SEL: Select single-step process 45441
LLW: trap ptid is process 45441.
LLW: exit
LLR: Preparing to step process -1, 0, inferior_ptid process -1
RC: Not resuming sibling process 45441 (not stopped)
LLR: PTRACE_SINGLESTEP process 45441, 0 (resume event thread)
sigchld
linux_nat_wait: [process -1], []
LLW: enter
LNW: waitpid(-1, ...) returned 45441, No child processes
LLW: waitpid 45441 received Interrupt (stopped)
LLW: Candidate event Interrupt (stopped) in process 45441.
SEL: Select single-step process 45441
LLW: exit
infrun: target_wait (-1, status) =
infrun:   45441 [process 45441],
infrun:   status->kind = stopped, signal = GDB_SIGNAL_INT
infrun: infwait_normal_state
infrun: TARGET_WAITKIND_STOPPED
infrun: stop_pc = 0xffffffffff600144
infrun: random signal (GDB_SIGNAL_INT)

Program received signal SIGINT, Interrupt.
infrun: stop_stepping
0xffffffffff600144 in ?? ()
(gdb)

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug c++/16997] Stepping hangs on gettimeofday with recording enabled

[konradsa at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=16997

--- Comment #4 from konradsa <konradsa at gmail dot com> ---
I think vDSO really is the culprit here. When I replace
    gettimeofday( &now, 0);
with a direct system call
    syscall(SYS_gettimeofday, &now, 0);
it works fine, even with recording enabled.

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug c++/16997] vDSO call to gettimeofday hangs with recording enabled

[konradsa at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=16997

konradsa <konradsa at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|Stepping hangs on           |vDSO call to gettimeofday
                   |gettimeofday with recording |hangs with recording
                   |enabled                     |enabled

-- 
You are receiving this mail because:
You are on the CC list for the bug.

[Bug c++/16997] vDSO call to gettimeofday hangs with recording enabled

[guillaume at morinfr dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=16997

Guillaume Morin <guillaume at morinfr dot org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |guillaume at morinfr dot org

-- 
You are receiving this mail because:
You are on the CC list for the bug.