Also, remove the catching of signals in the signal handler, so the next similar signal will kill Pike. Just in case it hangs when making the backtrace...
In the current case, the printing of the backtrace also calls exit(). I don't think "stealing" SIGQUIT is a good idea. Rather this could be implemented on the Pike-level using the Pike-level C backtrace method.
/ David Hedbor
Previous text:
2003-10-02 17:45: Subject: Some tasks from the conference
Also, remove the catching of signals in the signal handler, so the next similar signal will kill Pike. Just in case it hangs when making the backtrace...
/ Mirar
SIGINT is "stolen" by Hilfe already...
/ Mirar
Previous text:
2003-10-02 17:47: Subject: Some tasks from the conference
In the current case, the printing of the backtrace also calls exit(). I don't think "stealing" SIGQUIT is a good idea. Rather this could be implemented on the Pike-level using the Pike-level C backtrace method.
/ David Hedbor
The default behaviour of SIGQUIT is also to end the program (with a core dump). So the only thing that could be considered "stolen" is that you don't get the core dump. But that could be fixed by calling abort() instead of exit() (after removing the SIGABRT handler of course).
/ Marcus Comstedt (ACROSS) (Hail Ilpalazzo!)
Previous text:
2003-10-02 17:47: Subject: Some tasks from the conference
In the current case, the printing of the backtrace also calls exit(). I don't think "stealing" SIGQUIT is a good idea. Rather this could be implemented on the Pike-level using the Pike-level C backtrace method.
/ David Hedbor
Indeed, that's an idea. I'll implement that. I assume calling signal() from pike will disconnect the previous signal handler correctly?
/ David Hedbor
Previous text:
2003-10-02 18:22: Subject: Some tasks from the conference
The default behaviour of SIGQUIT is also to end the program (with a core dump). So the only thing that could be considered "stolen" is that you don't get the core dump. But that could be fixed by calling abort() instead of exit() (after removing the SIGABRT handler of course).
/ Marcus Comstedt (ACROSS) (Hail Ilpalazzo!)
pike-devel@lists.lysator.liu.se
-
David Hedbor ＠ Pike developers forum -
Marcus Comstedt (ACROSS) (Hail Ilpalazzo!) ＠ Pike (-) developers forum
-
Martin Stjernholm, Roxen IS ＠ Pike developers forum
-
Mirar ＠ Pike developers forum