Exiting workmode crashing

[changrilla]

Hi. Not been on for a while but am experiencing crashes with a particular model. I've scrutinized it and can't find anything wrong. Exported to stl and viewed in FreeCad which usually shows up any odd artifacts, even had it printed without any problems but can't switch workmode off. I'm using 1.5.4 but have tried it on 2.1.5 with the same problem.

Hopefully someone can help

tia Gordon

File: http://www.mgcsoft.com/junk/LatchPin_k_tmp.wings

Start of dump:

=erl_crash_dump:0.3
Fri Aug 31 10:41:40 2018
Slogan: binary_alloc: Cannot reallocate 718368715 bytes of memory (of type "binary").
System version: Erlang/OTP 18 [erts-7.0.1] [source-7dfcf74] [smp:2:2] [async-threads:10]
Compiled: Fri Jul 3 08:14:29 2015
Taints:
Atoms: 15737
Calling Thread: scheduler:1
=scheduler:1
Scheduler Sleep Info Flags:
Scheduler Sleep Info Aux Work:
Current Port:
Run Queue Max Length: 0
Run Queue High Length: 0
Run Queue Normal Length: 0
Run Queue Low Length: 0
Run Queue Port Length: 0
Run Queue Flags: NONEMPTY
Current Process: <0.30.0>
Current Process State: Running
Current Process Internal State: ACT_PRIO_NORMAL | USR_PRIO_NORMAL | PRQ_PRIO_NORMAL | ACTIVE | RUNNING
Current Process Program counter: 0x0b5017a8 (wings_draw_setup:dup3/3 + 4)
Current Process CP: 0x0b4fe430 (wings_draw_setupetup_smooth_normals/4 + 1016)
Current Process Limited Stack Trace:
0x05ee6b10:SReturn addr 0xB4FDFC8 (wings_draw_setupetup_smooth_normals/1 + 38
0x05ee6b2c:SReturn addr 0xB467784 (wings_draw:update_fun_2/3 + 1480)
0x05ee6b38:SReturn addr 0xB466FF0 (wings_draw:update_fun/3 + 52)
0x05ee6b40:SReturn addr 0xB45FA64 (wings_dl:map_1/5 + 80)
0x05ee6b4c:SReturn addr 0xB464F3C (wings_draw:refresh_dlists/1 + 52)
0x05ee6b64:SReturn addr 0x4B1EBE8 (wings:main_loop/1 + 44)
0x05ee6b68:SReturn addr 0x552A660 (wings_wm:handle_event/3 + 112)
0x05ee6b70:SReturn addr 0x552A0DC (wings_wmend_event/2 + 572)
0x05ee6b80:SReturn addr 0x5529760 (wings_wm:do_dispatch/2 + 12
0x05ee6bac:SReturn addr 0x55290C0 (wings_wm:get_and_dispatch/0 + 2
0x05ee6bb8:SReturn addr 0x4B1E06C (wings:init/1 + 88
0x05ee6bbc:SReturn addr 0x5AC76F04 (<terminate process normally>)

[micheus]

changrilla, how much memory has your PC? Is it a 32bit system?

The dump seems to to report a memory unavailability issue. I remember to see something like that when using XP 32bits with only 1GB of free memory available and a dense mesh.

[changrilla]

Hi Micheus. I have the same problem on W7 64 bit but it's just this one part. I have lots of parts much bigger that don't give me any problems.

The part itself has 10654 polygons, 21528 edges and 10876 vertices and according to performance monitor on W7 64, 2.3G available?

Sorry didn't answer your question - 4GB. Usually work on W7 32 bit on the same machine.

Gordon

[micheus]

Thanks for the infos.

I just tried it at my PC and got the same.
It's a huge amount of vertex - hard to debug, but it should have something wrong with it since Wings3D is trying to allocate all the memory available.
For me was:
Slogan: binary_alloc: Cannot reallocate 1512964055 bytes of
Windows task manager was showing Wings3D using 11GB and then it crashed.

I'm going to try to break it in parts and check if it's possible to find the problem on it and maybe avoid Wings3D to crash in such situation.

[changrilla]

OK, thanks Micheus. I'll leave it with you it isn't stopping me doing anything.

[micheus]

A fast look at the mesh - when creating regions for cut - it's possible to see some bad topology.

I selected visually the faces of the screw head and after select the edge loop I just found there were unpredictable face selected since I got more than a "circle" selected.

That's something that can makes Wings3D crazy. It looks like you has been using some boolean tool to create that object. They use to do that.

[changrilla]

Found it There's what appears to be an edge spanning vertices. See link.

http://www.mgcsoft.com/junk/latchpin_rogue_edge.png

Sorry if I've wasted your time.

Regards Gordon

---

You're right about the boolean 'cut'. At least I know what to look out for if it happens again. Thanks for your trouble.

Kind regards Gordon

[micheus]

I just saw your post now. Good you found it, but there is another one at the other side.