Steve Baker (sbaker++at++link.com)
Thu, 28 May 1998 13:19:13 -0500 (CDT)
> Greetings All,
>
> Our Performer apps have all worked very well on the Impact and the Octane,
> both single processor. Now that we have moved these apps to the 4 processor
> Onyx RE2 they stop somewhere and only get through our Performer loop once or
> twice.
>
> I tried using pfMultiprocess(PFMP_APP_CULL_DRAW) and
> pfMultiprocess(PFMP_DEFAULT) to try to setup the program for multiple
> processors but the only thing that works is if I put everything on one
> processor using pfMultiprocess(PFMP_APPCULLDRAW).
>
> Is there something else I must do to utilize multiple processors?
No, that call is enough - you are probably suffering from the classic
problem that 99% of people converting from single CPU to multiproc
have.
Make sure that all Performer data structures are allocated out of
shared memory. This especially includes stuff you send to pfGeoSet
API: vertex arrays, normal arrays, index arrays, colour tables, etc.
Search you code for calls to malloc/calloc/new/delete/amalloc etc, and
convince yourself whether these should in fact be pfMalloc/pfCalloc.
Also, beware of global variables that are written to after the other
processes are split off.
Steve Baker (817)619-8776 (Vox/Vox-Mail)
Raytheon Systems Inc. (817)619-4028 (Fax)
Work: SBaker++at++link.com http://www.hti.com
Home: SJBaker1++at++airmail.net http://web2.airmail.net/sjbaker1
=======================================================================
List Archives, FAQ, FTP: http://www.sgi.com/Technology/Performer/
Submissions: info-performer++at++sgi.com
Admin. requests: info-performer-request++at++sgi.com
This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:57:26 PDT