Windows devs required to help pointpoint vsgXchange regresion #716
Replies: 2 comments 4 replies
-
Hello Robert, I have tried the latest source code and I also have an issue now. And there is a stack overflow.
Probably this array which is allocated on stack is too big for VC++ compiler without extra compile flags
When I've made data static error was fixed and vsgview and vsgconv started to work fine.
|
Beta Was this translation helpful? Give feedback.
-
@Mikalai I have applied your suggestion to vsgXchange's cpp writer: And have updated the VulkanSceneGraph's built in shaders wit commits : 7924ac9 & 2e1ddb9. I'd appreciate testing of the latest VulkanSceneGraph and vsgXchange master, so we can close this Issue. Thanks. |
Beta Was this translation helpful? Give feedback.
-
Hi All,
@rainergericke has reported an issue he's having with latest VSG/vsgXchange master - it crashes when trying to load models via assimp, the isuee is on the vsgXchange Issue tracker:
vsg-dev/vsgXchange#147
I have tried reproduce the issue under Linux, and reviewed the code but neither Rainer or myself has come up with a possible cause of regression. So I'm looking for Windows users to test out the latest VSG/vsgXciange and see if you can load .obj, .gltg etc. models.
My hope is with more data points of success/failure with different Visual Studio/MinGW/Cygwin build combinations we might be a hint to root cause of the problem.
Loading a model that relies upon vsgXchange::assimp to load it should be sufficient to trigger the crash if it's going to happen, i.e.:
Whether it works or not could you post your findings here, along with the build tools you used.
Thanks.
Robert.
Beta Was this translation helpful? Give feedback.
All reactions