Game Development Community

[RESOLVED] AdvancedLightmapSupport crash

by Steve Acaster · in Torque 3D Professional · 09/05/2009 (2:32 pm) · 10 replies

T3D crashes due to advancedLightmapSupport = "1"; in the following circumstances:

1. No lightmapped objects present and it is set from 0 to 1 in the editor.
2. On T3D startup, trying to load a advancedLightmapSupport = "1"; mission in Basic Lighting.
3. Loading from a advancedLightmapSupport = "1"; mission to one which is not OR vicea-versa.

Having a lightmapped mesh in level and I can usually flick between the 2 settings without a crash, though sometimes artifacts appear ... and occaisionally it does crash.

7900gs

edit
Still Present in Release.
RESOLVED in 101

NextNext Thread
PrevPrevious Thread
More from Torque 3D Professional


Torque 3D Owner Kenneth Holst
Default Studio Name
#1
09/09/2009 (2:56 pm)
Logged as THREED-739

Associate Steve Acaster
[YorkshireRifles.com]
#2
09/30/2009 (11:05 pm)
Just on a little cleanup.
Resovled in Release

edit: oh no it isn't

Richard Marrevee
R.G.S - Richards Game Studio
#3
10/01/2009 (9:42 am)
@Steve:

T3D 1.0 still crash when switching to advancedLightmapsupport when there are no meshes with lightmaps present.

At first I thought it was a problem with my GFX-card (NVIDIA Geforce8400) but after a change to the NVIDIA Geforce 9800 the problem still remains.

I will try it whit a lightmapped mesh when I have some time.

Associate Steve Acaster
[YorkshireRifles.com]
#4
10/01/2009 (11:01 am)
@Richard
Yeah, you're right.

Silly old me forgot to check unlightmapped mission files.

... okay ... unresolved then ...

Associate Matt Fairfax
PopCap
#5
10/02/2009 (6:43 pm)
Hrm...I can't reproduce this on my development box...

I tried toggling on and off advancedLightmapSupport in the Blank Room level of FPS Example and never got a crash.

I have also loaded the Deathball Desert level a number of times after loading the Burg level (Desert has no advancedLightmapSupport and Burg does).

Any chance you can capture a callstack of the crash?

Associate Steve Acaster
[YorkshireRifles.com]
#6
10/02/2009 (7:24 pm)
Quote:
Any chance you can capture a callstack of the crash?

If you give me idiot proof instructions, then maybe.

Associate Steve Acaster
[YorkshireRifles.com]
#7
10/18/2009 (12:34 pm)
Line endings are not consistent source\gfx\D3D9\gfxD3D9Shader.cpp

Unhandled exception at 0x76c642eb in YorkshireRifles_DEBUG

> YorkshireRifles_DEBUG.dll!GFXD3D9ShaderConstBuffer::set(GFXShaderConstHandle * handle=0x09c22180, const MatrixF & mat={...}, GFXShaderConstType matrixType=GFXSCT_Float4x4) Line 334 + 0x17 bytes C++


- this 0x09c110b8 {mShader={...} mVertexConstBufferLayoutF=0x09c0fb80 mVertexConstBufferF=0x09c1be60 ...} GFXD3D9ShaderConstBuffer * const



- handle 0x09c22180 {mValid=true } GFXShaderConstHandle *

[0] CXX0030: Error: expression cannot be evaluated


incorrect const buffer type (334)

-----------------------------

Hopefully this is the info required.

Associate Steve Acaster
[YorkshireRifles.com]
#8
10/18/2009 (9:12 pm)
Rolfcopter ... a few hours later 101 comes out and it's fixed.

Associate Matt Fairfax
PopCap
#9
10/20/2009 (5:03 pm)
Sweet =)

Associate Scott Burns
GG Alumni
#10
06/10/2010 (1:41 pm)
Confirmed as fixed for 1.1 Beta2.