On Sat, 31 Dec 2005 17:09:35 -0500 Christopher Michael
<[EMAIL PROTECTED]> wrote:

> Hi All,
> 
>       Lately it seems there's been issues with exe's exiting from
> within E. In efforts to track this down, I managed to grab this
> backtrace from E by attaching to the pid. Hope it helps !
> 
> devilhorns
> 
> 
> #5  0xb7ce069b in _ecore_main_select (timeout=-nan(0xfffdfe0dfe3bf))
>      at ecore_main.c:348
> #6  0xb7ce0b8c in _ecore_main_loop_iterate_internal (once_only=0)
>      at ecore_main.c:585
> #7  0xb7ce0be7 in ecore_main_loop_begin () at ecore_main.c:79
> #8  0x080604f8 in main ()
> 
> Timeout = NaN ??

NaN means "Not A Number", timeout is a double (big floating point
number) and NaN is a valid value, usually the result of a divide by
zero or some other impossible maths. The problem is that it's not a
valid timeout. We do some maths on it to construct a timeval structure
to pass to the select() call that is causing the crash.

In this case, the timeout comes from _ecore_timer_next_get().  This is
code I haven't looked at yet, so I currently have no idea.  Since it is
an issue with exe's exiting, it is likely my fault coz I've been
playing with that code recently.  I do set up a timer under some
circumstances on exe exit, but it is a fixed, small value.

I'll investigate this more in an hour or two.

Attachment: pgp0wgj5aW5mX.pgp
Description: PGP signature

Reply via email to