a problem with pfBuffers in 2.0?

New Message Reply Date view Thread view Subject view Author view

Randall Barker (barker++at++cs.nps.navy.mil)
Tue, 19 Dec 1995 17:08:05 -0800

I have been working with pfBuffers in Performer 2.0 beta (a151). I have
noticed some unusual behavior and was hoping some one could tell me why the
following code segment is valid. I'm running it on an Indigo2 and I have
verified that the pfDBase process in getting forked and also that this code is
being called in that forked process:

            tempPage->node = new pfSCS (theMat);
            pfNode *tile = pfdLoadFile (buf);
            tempPage->node->addChild (tile);
            [...]
            scene->bufferAddChild (tempPage->node);
            pfBuffer::merge ();

While this one:

            pfNode *tile = pfdLoadFile (buf);
            tempPage->node = new pfSCS (theMat);
            tempPage->node->addChild (tile);
            [...]
            scene->bufferAddChild (tempPage->node);
            pfBuffer::merge ();

causes a core dump with dbx having the following to say:

(dbx) where
> 0 pfSCS::nb_clean(int)(0x180e7c60, 0x5100, 0x18041370, 0x180e811c)
["../../../lib/libpf/pfSCS.C":176, 0x5d132a38]
   1 pfBuffer::pf_prepareNewUpdatables(int)(0x180e1340, 0x1, 0x18041370,
0x180e811c) ["../../../lib/libpf/pfBuffer.C":752, 0x5d1114a4]
   2 pfBuffer::merge(void)(0x180e7cb4, 0x0, 0x18041370, 0x180e811c)
["../../../lib/libpf/pfBuffer.C":556, 0x5d1107cc]
   3 dbStorage::index_hit(int,int)(0x10003720, 0xffffffff, 0xffffffff,
0x180e811c) ["dbStorageClass.cc":113, 0x410a30]
   4 dbPaging::update(float,float)(0x10005698, 0x0, 0x0, 0x180e811c)
["dbPagingClass.cc":225, 0x40f83c]
   5 dbPaging::dbPaging(float,float,char*,pfScene*)(0x10005698, 0x0, 0x0,
0x100000fc) ["dbPagingClass.cc":137, 0x40f350]
   6 db_paging_process(void*)(0x0, 0x0, 0x18041370, 0x180e811c) ["cig.cc":307,
0x405fc0]
   7 mpDBase(void)(0x180e7cb4, 0x0, 0x18041370, 0x180e811c)
["../../../lib/libpf/pfProcess.C":3087, 0x5d15b714]
   8 pfConfig(0x180e7cb4, 0x18040830, 0x18041370, 0x180e811c)
["../../../lib/libpf/pfProcess.C":1699, 0x5d0e9cfc]
   9 main(0x1, 0x7fffaf44, 0x18041370, 0x180e811c) ["cig.cc":78, 0x404c18]
   10 __start() ["crt1text.s":133, 0x40493c]
(dbx)

  Is this a bug, and if so, has it been fixed? If it is not a bug, what am I
doing wrong? Does order matter when creating geometry in pfBuffers?

-Randall

-- 
Randall E. Barker                 Spanagel Hall, Room 525
Naval Postgraduate School         e-mail: barker++at++cs.nps.navy.mil
Computer Science, Code CS/Barker  Phone:  408.656.2251
Monterey, California 93943        Fax:    408.656.2814
URL: http://www-npsnet.cs.nps.navy.mil/barker/

New Message Reply Date view Thread view Subject view Author view

This archive was generated by hypermail 2.0b2 on Mon Aug 10 1998 - 17:52:10 PDT

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