Ken Sakai (sakai++at++sgi600.msd.lmsc.lockheed.com)
Sat, 22 Feb 1997 10:03:11 -0800
Like Elwin Lee, I too have been plagued by a persistant _pfDirtCheck() problem.
This problem arises after pfDelete'ing a pfGroup. (Note: I was careful to
do a pfRemoveChild() of the pfGroup from it's parent before pfDelete'ing
it.) I don't call pfInitState() at all. Normally, I start getting _pfDirtCheck
warnings after I do the following:
pfNodeTravMask(sceneRoot,PFTRAV_ISECT,0x0,PFTRAV_IS_CACHE|PFTRAV_DESCEND,PF_SET);
My only way of getting around this is to NOT pfDelete'ing the pfGroups after
pfRemove'ing them and simply let the pfGroup float away resulting in a
serious memory leak but at least I don't get _pfDirtCheck'ed.
I surmise that somehow something that has been pfDelete'd is still being
referenced in this traversal. Is there something else that I have to do to the
scene graph after pfRemove'ing and pfDelete'ing a group (such as re-running
pfNodePickSetup() on the scene graph) ?. This problem only seems to appear
on our Impact R10000 running either Performer 2.01 or Performer 2.1.
I can't seem to duplicate the problem on our ONYX running Performer 2.01.
I would be grateful for any suggestions.
Ken
Kenneth N. Sakai Lockheed Martin Missiles & Space
Research Specialist/3-D Computer Graphics P.O. Box 3504
Email: sakai++at++lmsc.lockheed.com Sunnyvale, CA 94088-3504
Phone: (408) 756-0682
=======================================================================
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 - 17:54:41 PDT