[Bf-blender-cvs] CVS commit: blender/source/blender/makesdna DNA_actuator_types.h DNA_scene_types.h DNA_sound_types.h blender/source/blender/makesdna/intern makesdna.c
Ton Roosendaal
ton at blender.org
Sun Nov 26 13:23:21 CET 2006
ton (Ton Roosendaal) 2006/11/26 13:23:21 CET
Modified files:
blender/source/blender/makesdna DNA_actuator_types.h
DNA_scene_types.h
DNA_sound_types.h
blender/source/blender/makesdna/intern makesdna.c
Log:
SculptMode fix:
DNA definition of Sculpt structs in Scene were not properly aligned,
causing memory errors on quit ("Memoryblock reconstruct: end corrupt").
More testing reveiled padding errors in two other DNA_ includes, for
sound and gamelogic. Both potentially crashers... and caused by
commenting out struct members with a C++ comment, that seems to not
work...
I've revived the DNA padding test method, which saves out a simple C
file you can compile to see where padding issues are. This now works
as follows:
- change line 991 in makesdna.c to become (1) (true).
- recompile makesdna.c
- you now have a padding.c in the same dir as makesdna.c
- compile it, command line: "gcc -o padding padding.c"
- now run it (./padding), and it will print out errors, if there are.
For me, the DNA files are now 100% padding free. Might be interesting to
check it in 64 bits though!
Revision Changes Path
1.8 +1 -3 blender/source/blender/makesdna/DNA_actuator_types.h
<http://projects.blender.org/viewcvs/viewcvs.cgi/blender/source/blender/makesdna/DNA_actuator_types.h.diff?r1=1.7&r2=1.8&cvsroot=bf-blender>
1.86 +8 -6 blender/source/blender/makesdna/DNA_scene_types.h
<http://projects.blender.org/viewcvs/viewcvs.cgi/blender/source/blender/makesdna/DNA_scene_types.h.diff?r1=1.85&r2=1.86&cvsroot=bf-blender>
1.11 +1 -3 blender/source/blender/makesdna/DNA_sound_types.h
<http://projects.blender.org/viewcvs/viewcvs.cgi/blender/source/blender/makesdna/DNA_sound_types.h.diff?r1=1.10&r2=1.11&cvsroot=bf-blender>
1.24 +10 -9 blender/source/blender/makesdna/intern/makesdna.c
<http://projects.blender.org/viewcvs/viewcvs.cgi/blender/source/blender/makesdna/intern/makesdna.c.diff?r1=1.23&r2=1.24&cvsroot=bf-blender>
More information about the Bf-blender-cvs
mailing list