Control: forwarded -1 https://github.com/GeorgeFilipkin/pulsemixer/issues/86
On 2023-12-13 19:53 +0100, Sven Joachim wrote:
> Package: pulsemixer
> Version: 1.5.1-1.1
> Severity: normal
>
> Sending SIGINT to pulsemixer (e.g. by pressing Ctrl-C) prints a
> traceback on stdout:
>
> ,----
> | Traceback (most recent call last):
> | File "/usr/lib/python3.11/curses/__init__.py", line 94, in wrapper
> | return func(stdscr, *args, **kwds)
> | ^^^^^^^^^^^^^^^^^^^^^^^^^^^
> | File "/usr/bin/pulsemixer", line 1275, in run
> | c = self.getch()
> | ^^^^^^^^^^^^
> | File "/usr/bin/pulsemixer", line 1103, in getch
> | self.ev.wait()
> | File "/usr/lib/python3.11/threading.py", line 629, in wait
> | signaled = self._cond.wait(timeout)
> | ^^^^^^^^^^^^^^^^^^^^^^^^
> | File "/usr/lib/python3.11/threading.py", line 327, in wait
> | waiter.acquire()
> | File "/usr/bin/pulsemixer", line 1243, in <lambda>
> | signal.signal(signal.SIGINT, lambda s, f: self.terminate())
> | ^^^^^^^^^^^^^^^^
> | File "/usr/bin/pulsemixer", line 1228, in terminate
> | sys.exit()
> | SystemExit
> |
> | During handling of the above exception, another exception occurred:
> |
> | Traceback (most recent call last):
> | File "/usr/bin/pulsemixer", line 2054, in <module>
> | main()
> | File "/usr/bin/pulsemixer", line 1948, in main
> | curses.wrapper(Screen(CFG.ui.color, CFG.ui.mouse).run)
> | File "/usr/lib/python3.11/curses/__init__.py", line 101, in wrapper
> | endwin()
> | _curses.error: endwin() returned ERR
> `----
>
> The reason is that curses.endwin() is run twice, by the signal handler
> pulsemixer sets up and by curses.wrapper. In ncurses patchlevel
> 20231111 and later, calling endwin() twice in a row is considered an
> error. See bug #1058041 for more information.
>
> This is the signal handler which pulsemixer sets up:
>
> ,----
> | def terminate(self):
> | # if ^C pressed while sleeping in reconnect wrapper.restore won't
> be called
> | # so have to restore it manually here
> | self.screen.keypad(0)
> | curses.echo()
> | curses.nocbreak()
> | curses.endwin()
> | sys.exit()
> `----
>
> I don't know if the comment is actually true, this should best be
> discussed with upstream. Unfortunately they have been inactive for the
> last 3.5 years. :-(
Somebody else also discovered the problem and reported it upstream, but
did not receive a reply so far.
Cheers,
Sven
