sal/typesconfig/typeconfig.c

[Jung-uk Kim]

On Tuesday 23 May 2006 12:56 pm, Jung-uk Kim wrote:
> On Tuesday 23 May 2006 12:16 pm, Carlos F. A. Paniago wrote:
> > I`m compiling the amd64 OO-2.0-devel and I have to change this:
> > in sal/typesconfig/typeconfig.c line 41:
> >
> > #ifndef __FreeBSD__
> > #define USE_FORK_TO_CHECK 1
> > #endif
> >
> > then the typesconfig program stops exiting as
> >
> > pid 10359 (typesconfig), uid 0: exited on signal 11 (core dumped)
> >
> > But this does not affect the compilation proccess.
>
> It does *intentionally* causes signal 11 to find out certain things
> are supported or not but I believe FreeBSD doesn't have to use fork
> for this test because it was Linux 2.6 kernel workaround.

POSIX says:
http://www.opengroup.org/onlinepubs/009695399/functions/signal.html

'If and when the function returns, if the value of sig was SIGFPE, 
SIGILL, or SIGSEGV or any other implementation-defined value 
corresponding to a computational exception, the behavior is 
undefined. Otherwise, the program shall resume execution at the point 
it was interrupted.'

I guess you should not do that even if that works in FreeBSD. :-)

Jung-uk Kim