Pthreads?

Pthreads?

Post by Scott Vette » Sat, 24 Mar 2001 22:36:36



   Is there such a thing as pthreads in OS/2?
 
 
 

Pthreads?

Post by Shaun Tanche » Sun, 25 Mar 2001 05:58:13




Quote:>   Is there such a thing as pthreads in OS/2?

Everblue ... netlabs ...

http://ais.gmd.de/~veit/os2/mailinglist2/0189.html

ftp://ftp.netlabs.org/pub/everblue/pthread6.zip

HTH
 http://shaun.tancheff.com/       /?/                 -_-


 
 
 

Pthreads?

Post by Scott Vette » Sun, 25 Mar 2001 08:27:47


Shaun:

    pthreads is supposed to be about multitasking is it not?

Scott

---------




> >   Is there such a thing as pthreads in OS/2?

> Everblue ... netlabs ...

> http://ais.gmd.de/~veit/os2/mailinglist2/0189.html

> ftp://ftp.netlabs.org/pub/everblue/pthread6.zip

> HTH
>  http://shaun.tancheff.com/       /?/                 -_-


 
 
 

Pthreads?

Post by Alexander Ma » Mon, 02 Apr 2001 02:39:20



>Shaun:

>    pthreads is supposed to be about multitasking is it not?

>Scott

Yes
(e.g. http://www.opengroup.org/onlinepubs/007908799/xsh/threads.html)
but was there a contradiction to the previous posting?

---
Alexander Mai


 
 
 

1. solaris 2.5 pthreads vs. pthreads

Is it known whether Sun's pthread implementation adheres to the standard?

Specifically, I am trying to port the public domain DCE RPC stuff to
solaris.  It is based on
an implementation of pthreads that seems to be a little different.  In most
cases it is
merely a matter of passing pointers rather than values, or passing a
parameter rather
than getting a value through the return.

But, there are two points where I'm not sure if I am porting this
correctly.

First, the code has a call to
        int pthread_setasynccancel(int state)

Which doesn't seem to exist under solaris.  There is, though a

        int pthread_setcancelstate(int newsate, int * oldstate)

The problem is that I don't see that the type is set to
PTHREAD_CANCEL_ASYNCHRONOUS.

So, is the solution to take a call that was

        oldstate = pthread_setasynccancel(newstate);

and make it

        retval = pthread_setcancelstate(newstate, &oldstate);

or is it

        retval = pthread_setcanceltype(PTHREAD_CANCEL_ASYNCHRONOUS, &oldtype);
        retval = pthread_setcancelstate(newstate /* like PTHREAD_CANCEL_ENABLED
*/, &oldstate);

??

The second thing is related.  The code has a call to

        int pthread_setcancel(state);   /* which seems to be for non-asynchronous
cancels */

Again, solaris has no direct equivalent.  So, does

        oldstate = pthread_setcancel(state);

become

        retval = pthread_setcancelstate(newstate, &oldstate);

or

        retval = pthread_setcanceltype(PTHREAD_CANCEL_DEFFERED, &oldtype);
        retval = pthread_setcancelstate(newstate, &oldstate);

Any ideas?

Thanks,
-brennan

+===========================================================================
====+
+ Brennan HIldebrand
+ Senior Architect
+ CoSite, Inc

+===========================================================================
====+

2. Scrollbar Not Visible in Panel

3. Problem with pthread and clone() + pthread and context switching question...

4. Irwin tape drive / backup softare

5. solaris threads,pthreads & MIT pthreads

6. Anyone in Royal Oak, MI area with outages??

7. Why doesn't a thread exit in pthreads-win32 version? (C++ / pthreads / pthreads-win32)

8. How can I send the conditional variable from one pthread to the other pthread in Linux C++ codes?

9. pthread.h is missing from SFU V3.0

10. Pthreads?

11. compiling Sendmail-8.12.6 for milters (pthread-safe) on SGI

12. Sendmail 8.12.3 HP-UX 10.20 libmilter: GCC is missing pthread.h