RE: pthreads and Performer

New Message Reply Date view Thread view Subject view Author view

Michael Baldwin (mbaldwin++at++nvl.army.mil)
Thu, 21 Oct 1999 17:52:16 -0400 (EDT)


I have a similar question about Performer and pthreads. I get the
following error message:

PF Notice(22): MUSTRUN of DBASE process 78677 on CPU 0 failed.

I also get the error message in a short test program (see below) if the pthread
library is linked in. However, it goes away if the -delay_load switch is used
on -lpthread.

In the future I would like to be able to lock down processes to cpus. Is this
a sign that I won't be able to do this in my current app, since I have to link
in libphread for my app? Are there any other concerns other than sysmp
failing that I should be aware of? I'm running Performer 2.2.4 and IRIX 6.5.3m
on an ONYX2.

Thanks,
Mike

Example:

/* CC -o foo foo.C -lpf -lpthread ==> Get error message */
/* CC -o foo foo.C -lpf -delay_load -lpthread ==> OK */

#include <Performer/pf.h>

void main (void)
{
   pfInit();
   pfConfig();
   while (1);
}

CC -o dbase2 dbase2.cc -lpf -lpthread

On 21-Oct-99 Anthony Bavuso wrote:
> Greetings,
>
> Is IRIX Performer compatible with pthreads? More specifically does
> Performer use sprocs?
>
> I know that pthreads and sprocs can not both live in the same process space
> so if Performer uses sprocs then it is incompatible with pthreads.
>
> I ask this because I have tried to compile a program using both Performer
> and pthreads and the -D_POSIX_C_SOURCE=199506L compiler flag required by
> posix threads causes linker errors.
>
> Thanks.
>
>
> -----------------------------------------------------------------------
> List Archives, FAQ, FTP: http://www.sgi.com/software/performer/
> Submissions: info-performer++at++sgi.com
> Admin. requests: info-performer-request++at++sgi.com


New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Thu Oct 21 1999 - 14:52:38 PDT

This message has been cleansed for anti-spam protection. Replace '++at++' in any mail addresses with the '@' symbol.