| Summary: | Unable to quit GDB if remote TCP connection is lost | ||
|---|---|---|---|
| Product: | gdb | Reporter: | jon |
| Component: | remote | Assignee: | Keith Seitz <keiths> |
| Status: | RESOLVED WORKSFORME | ||
| Severity: | normal | CC: | jan.smets, keiths, pedro |
| Priority: | P2 | ||
| Version: | 7.5 | ||
| Target Milestone: | --- | ||
| Host: | windows 7 / cygwin / linux | Target: | |
| Build: | Last reconfirmed: | ||
I cannot reproduce with the 7.3 branch (7.3.1-20111104-cvs) or CVS HEAD. I've tried this on cygwin, too (Windows 2008 Server):
SHELL A:
$ gdbserver :1234 foo
Process foo created; pid = 1936
Listening on port 1234
SHELL B:
$ gdb -nx -q foo
(gdb) b main
Breakpoint 1 at 0x401196: file foo.c, line 6.
(gdb) tar remote :1234
Remote debugging using :1243
warning: Can not parse target XML description; XML support was disabled at compile time
warning: Can not parser XML library list; XML support was disabled at compile time
0x775e0005 in ?? ()
(gdb) c
Breakpoint 1, main () at foo.c:6
6 return 0;
(gdb)
SHELL C:
$ ps waux | grep gdbserver | kill `awk '{print $1}'`
SHELL B:
(gdb) quit
A debugging session is active.
Inferior 1 [Remote target] will be killed.
Quit anyway? (y or n) y
qTStatus: remote connection closed
$
Can you verify any of this or provide me with a test that demonstrates the problem better?
It's almost ten months. This still works, as described, on CVS HEAD (7.5.50.20120828-cvs) and the 7.5 branch. I'm closing this bug. Can someone please re-investigate? I'm having the same issue with 7.5
1) connect to a remote target
2) attach to a process
3) disconnect/reset the remote target
4) watch the timeout
5) try detaching or quit gdb
(gdb) attach 165209617
Attaching to process 165209617
[New Thread 165209617.165209617]
0x07ab29bc in __memPoolInit ()
(gdb) c
Continuing.
^C
[Thread 165209617.165209617] #1 stopped.
0x07ab29bc in __memPoolInit ()
Ignoring packet error, continuing...
Stopping Thread 165209617.165209617 failed: timeout
(gdb) Ignoring packet error, continuing...
Stopping Thread 165209617.165209617 failed: timeout
(gdb) (gdb)
Continuing.
Cannot execute this command while the selected thread is running.
(gdb)
(gdb) exit
Undefined command: "exit". Try "help".
(gdb) quit
A debugging session is active.
Inferior 1 [process 165209617] will be detached.
Quit anyway? (y or n) y
Ignoring packet error, continuing...
Bogus trace status reply from target: timeout
(gdb)
A debugging session is active.
Inferior 1 [process 165209617] will be detached.
.... (can't get out of GDB!!)
(In reply to Jan Smets from comment #3) > Can someone please re-investigate? I'm having the same issue with 7.5 You'll most probably need a more recent GDB. Keith mentioned 7.5.50.20120828-cvs (future 7.6), and the 7.5 branch (what ended up as 7.5.1). In any case 7.5 is old by now. Please reopen if you can reproduce it with a 7.7, or ideally a recent snapshot or git head (we're not that far from 7.8 now). |
If using a TCP remote connection (E.g. tar ext localhost:1234), if the target connection is lost for some reason, it is no longer possible to quit GDB. (gdb) quit A debugging session is active. Inferior 1 [Remote target] will be killed. Quit anyway? (y or n) y putpkt: write failed: Connection reset by peer. (gdb) i.e. you get the putpkt error message, and then get returned to the prompt. Pressing ctrl-c several times just results in: (gdb) Quit (gdb) Quit (gdb) Quit... being printed. This is occuring on a Windows 7 64-bit / Cygwin system.