I’ve reported that back a week ago or so.
Yesterday the fix was added!
Blender 2.5 Bug Tracker item #22823, was opened at 2010-07-12 08:32
You can respond by visiting:
http://projects.blender.org/tracker/?func=detail&atid=498&aid=22823&group_id=9
Or by replying to this e-mail entering your response between the following markers:
#+#+#+#+#+#+#+#+#+#+#+#+#+#+#+#+#+
(enter your response here)
#+#+#+#+#+#+#+#+#+#+#+#+#+#+#+#+#+
>Status: Closed
Priority: 3
Submitted By: Andreas Esau (ndee)
Assigned to: Janne Karhu (jhk)
Summary: Children Particle Rendering is broken
>Resolution: Fixed
Category: Physics
Initial Comment:
I’ve downloaded an actual blender build (Revision 30209) and tried to render my scene. It always chrashed. Tried to find the bug, and it is particle children.
When you create a new scene with a cube. Add particle hairs everything works fine when rendering. Now enabling particle children with a renderamount of 1 extra particle, blender crashes if I try to render the scene.
The Revision 30094 works fine concerning the the children rendering for particles.
All test were done under windows xp 32 bit.
>Comment By: Diego Borghetti (bdiego)
Date: 2010-07-12 16:52
Message:
Hi,
Confirmed, Linux Ubuntu 64 bits. I do a backtrace and I found that the problem is at convertblender.c L1791, psys->childcache is NULL (a and totpart are fine).
Brecht, can you take a look at this ?
Here is the backtrace:
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7fffeded2910 (LWP 28042)]
0x00000000007741c0 in render_new_particle_system (re=<value optimized out>, obr=0x385c428, psys=<value optimized out>, timeoffset=<value optimized out>)
at convertblender.c:1791
1791 cache = psys->childcache[a-totpart];
(gdb) bt
#0 0x00000000007741c0 in render_new_particle_system (re=<value optimized out>, obr=0x385c428, psys=<value optimized out>, timeoffset=<value optimized out>)
at convertblender.c:1791
#1 0x0000000000775989 in init_render_object_data (re=0x39ac158, obr=0x385c428, timeoffset=<value optimized out>) at convertblender.c:4423
#2 0x00000000007783a9 in add_render_object (re=0x39ac158, ob=0x2063858, par=<value optimized out>, dob=0x0, timeoffset=<value optimized out>,
vectorlay=<value optimized out>) at convertblender.c:4504
#3 init_render_object (re=0x39ac158, ob=0x2063858, par=<value optimized out>, dob=0x0, timeoffset=<value optimized out>, vectorlay=<value optimized out>)
at convertblender.c:4529
#4 0x0000000000778b7a in database_init_objects (re=0x39ac158, renderlay=<value optimized out>, nolamps=<value optimized out>,
onlyselected=<value optimized out>, actob=<value optimized out>, timeoffset=<value optimized out>) at convertblender.c:4902
#5 0x00000000007795fe in RE_Database_FromScene (re=0x39ac158, scene=0x2061498, lay=1, use_camera_view=1) at convertblender.c:4989
#6 0x000000000072dadf in do_render_3d (re=0x39ac158) at pipeline.c:1773
#7 0x000000000072dfd3 in do_render_fields_blur_3d (re=0x39ac158) at pipeline.c:2080
#8 0x000000000072e6c7 in do_render_composite_fields_blur_3d (re=0x39ac158) at pipeline.c:2362
#9 do_render_all_options (re=0x39ac158) at pipeline.c:2539
#10 0x000000000072eee4 in RE_BlenderFrame (re=0x39ac158, scene=<value optimized out>, srl=<value optimized out>, lay=<value optimized out>,
frame=<value optimized out>) at pipeline.c:2756
#11 0x000000000070dbea in do_job_thread (job_v=0x393e4c8) at wm_jobs.c:238
#12 0x00007ffff706fa04 in start_thread () from /lib/libpthread.so.0
#13 0x00007ffff4566d4d in clone () from /lib/libc.so.6
Thanks,
Comment By: Brecht Van Lommel (blendix)
Date: 2010-07-13 13:47
Message:
Tracked down to this commit:
http://projects.blender.org/plugins/scmsvn/viewcvs.php?view=rev&root=bf-blender&revision=30187
Comment By: Diego Borghetti (bdiego)
Date: 2010-07-14 16:43
Message:
Hi,
I see, reassign to Campbell.
Comment By: Matt Ebb (broken)
Date: 2010-07-16 02:29
Message:
Found and confirmed here too, this is pretty bad stuff - I’ve attached a test file here, just try rendering it. I wonder how this is not affecting durian? 
You can also see problems by adding a new hair particle system, and enabling children - nothing updates in the 3D View, and changing the number of children for display does nothing.
Comment By: Brecht Van Lommel (blendix)
Date: 2010-07-16 11:37
Message:
I left out a commit done in trunk in the render branch a while ago:
http://projects.blender.org/plugins/scmsvn/viewcvs.php?view=rev&root=bf-blender&revision=29245
That commit broke all kinds of things, and I suspect is responsible for this as well. What changed there was that instead of computing paths in advance as part of particle_system_update, this was moved to do it just before drawing or rendering.
When the new particle system was just committed, it also worked this way but gave various problems, and I had to remove that behavior to get things working stable. Basically it meant that you could get have a path cache that was outdated in various situations, and it doesn’t fit well with dependency graph evaluation order. Particle paths should just be calculated in object_handle_update/particle_system_update like modifiers, animation, etc.
I’ll just revert that commit in trunk as well, since it seems to be causing a few other bugs as well, and the beta is coming up.
Comment By: Brecht Van Lommel (blendix)
Date: 2010-07-17 19:08
Message:
Fix in svn, thanks for the report.
You can respond by visiting:
http://projects.blender.org/tracker/?func=detail&atid=498&aid=22823&group_id=9
