ev_timer_again() causing segmentation fault

2013-10-29 Thread utku genç
​Hello All, Following code doesn't yield any problems: static void udp_read_cb(EV_P_ ev_io *watcher, int revents) { ev_timer_again(EV_A_ communication_timeout_watcher); do_something(); } void do_something() { ... } But if I move the ev_timer_again() one call deeper, as in: static void

Re: ev_timer_again() causing segmentation fault

2013-10-29 Thread Dan White
On 10/29/13 22:30 +0200, utku genç wrote: ​Hello All, Following code doesn't yield any problems: static void udp_read_cb(EV_P_ ev_io *watcher, int revents) { ev_timer_again(EV_A_ communication_timeout_watcher); do_something(); } void do_something() { ... } But if I move the ev_timer_again()

Re: ev_timer_again() causing segmentation fault

2013-10-29 Thread Marc Lehmann
On Tue, Oct 29, 2013 at 10:30:20PM +0200, utku genç utkug...@gmail.com wrote: I get a segmentation fault. Am I missing something here?​ You probably have some memory corruption bug elsewhere in your program. A tool such as valgrind may be able to help you find out whats going on. This is

Re: ev_timer_again() causing segmentation fault

2013-10-29 Thread Marc Lehmann
On Tue, Oct 29, 2013 at 10:30:20PM +0200, utku genç utkug...@gmail.com wrote: But if I move the ev_timer_again() one call deeper, as in: I see now, you are probably passing in an illegal loop pointer. void do_something() { ev_timer_again(EV_A_ communication_timeout_watcher); Where do you

Re: ev_timer_again() causing segmentation fault

2013-10-29 Thread Rodrigo Campos
On Tue, Oct 29, 2013 at 11:20:14PM +0200, utku genç wrote: Probably you are right. But I will post this anyway, maybe you will catch something else. Compiling with -fstack-protector-all might help (depends on what is actually happening, maybe it doesn't help at all) to see where the problem is.

Re: ev_timer_again() causing segmentation fault

2013-10-29 Thread utku genç
Thank you so much Marc, you were right, an EV_DEFAULT_ instead of EV_A_ (in ev_timer_again(EV_A_ communication_timeout_watcher); ) solved it for me. Thank you, Utku On Tue, Oct 29, 2013 at 11:31 PM, Marc Lehmann schm...@schmorp.de wrote: On Tue, Oct 29, 2013 at 10:30:20PM +0200, utku genç