[EMAIL PROTECTED] writes:
> On 26 Jun, Sebastien Carpe shouted:
Not that much, hey :)
> Well u'll proly have to mis out on soudn for e 0.14 until ESD can be
> ported to all platfroms for audio.
Well... i don't really care much about sounds (especially at work... i
don't wanna be blamed for boing'ing my collegue :) ). Taht was mainly
to help the port (since it's hard to get all platforms handy :) )
>
> -> E : Met two main problems :
> -> 1) Makefile doesn't honnor includedir. It's pity for people like me
> -> that can't install in "normal" places and have a bunch of goodies
>
> try ./configure --x-includes
> :)
>
> i fixed the --includes now tho.. i was setting INCLUDES n tthe
> Makefile.am to somehting else. thanx for pointing that out.
i did the --x-includes stuff. it's just i have yet another place where
i install all the stuff (and X11R6 is somewhere else... didn't wanted
to mix apps i compiled all around b4 with X11R6 stuff). it's just that
setting --prefix doesn't add the corresponding $(includedir) in the
includes path. But well, i'll look for --includes next time...
>
> but your system does have wait3 - simply configure doesnt pick i up?
> someone esle had probs on solaris. Not much i an do. All I call is:
> AC_FUNC_WAIT3
> in configure.. it does the rest of trying to find wait3... I added a
> dummy struct rusage now for wait3 to fill incase wait3 does fill rusage
> and doesnt check the pointer to the struct before filling it. Still
> nothing I can really do about autoconf not behaving.. check config.log
> to see if it tells u more info on why configure finds what it
> does...
Here is what it exactly do :
*> #include "confdefs.h"
*> #include <sys/types.h>
*> #include <sys/time.h>
*> #include <sys/resource.h>
*> #include <stdio.h>
*> /* HP-UX has wait3 but does not fill in rusage at all. */
*> main() {
*> struct rusage r;
*> int i;
*> /* Use a field that we can force nonzero --
*> voluntary context switches.
*> For systems like NeXT and OSF/1 that don't set it,
*> also use the system CPU time. And page faults (I/O) for Linux. */
*> r.ru_nvcsw = 0;
*> r.ru_stime.tv_sec = 0;
*> r.ru_stime.tv_usec = 0;
*> r.ru_majflt = r.ru_minflt = 0;
*> switch (fork()) {
*> case 0: /* Child. */
*> sleep(1); /* Give up the CPU. */
*> _exit(0);
*> case -1: _exit(0); /* What can we do? */
*> default: /* Parent. */
*> wait3(&i, 0, &r);
*> sleep(2); /* Avoid "text file busy" from rm on fast HP-UX machines. */
*> exit(r.ru_nvcsw == 0 && r.ru_majflt == 0 && r.ru_minflt == 0
*> && r.ru_stime.tv_sec == 0 && r.ru_stime.tv_usec == 0);
*> }
*> }
As you can see, there's a comment about HPUX lacking full support for
rusage. As a matter of fact, the return expression goes wrong if the
system doesn't fully support wait3 (including rusage). Yet, i've seen
the place where wait3 is used, and there's a NULL pointer passed as a
3rd arg. (may be you changed it...), so there's no point on rejecting
system not supporting rusage...
At least, you will to know about that if you really want to use rusage
:)
>
> -> Last but not least, HPUX 9.05 doesn't have inet_aton : dld fails when
> -> loading libesd.sl. I guess i'll comment that part, since it tries to
> -> resolve the IP address from host="0.0.0.0" ?!??! - see esdlib.c -.
> -> May be that would be worth asking someone around if inet_aton is present
> -> on up to date HPUX version (10+). Real last thing : bcopy is obsolet :)
>
> well for now dont use or install esd - if E picks up u have esd
> installed it'll compile to use it.. :)
'Kay :) ... If i find time, i'll kick esd's butt and make it work for
that crappy system ... i'll then let you know about...
Anyway, that wasn't a post complaining it doesn't work or whatever, it
was my .2 cents help for multi-platform support :)
--
Seb C. (mailto:[EMAIL PROTECTED]) | Working for Atos at Lille, France
*********************************************************************
-
To unsubscribe from this list send mail to: [EMAIL PROTECTED]
with the message contents: unsubscribe e-develop
