Hi,
thanks for providing the GPU crash dump and call stack. The GPU dump indicates that a fragment shader is trying to access a resource (172 MB in size) which has been destroyed, and thus results in a page fault error: “MMU Fault Error during a GPU memory Read of a destroyed unnamed resource at address 0x0000002590cc4000 in this shader location: Fragment Shader fragment_01 @ 0x00000180. A shader instruction caused an MMU fault when accessing memory. This can be caused by shader bugs and binding setup issues.” ).
It could be helpful to know which shader (with id 7C12D170C21A741D) and shader instruction causes the crash. This debug information can be obtained by configuring the shader debug search paths in Nvidia Nsight Graphics to point at the project’s shader debug info (typically located in the folder [YourProjectRoot]/Saved/ShaderDebugInfo/[ShaderTarget]). Would it be possible to obtain this info?
A few more questions which may help diagnose the crash:
- are all the call stacks similar? If not, is it possible to upload a second call stack?
- given that the GPU was busy processing textures when the crash occurred, the resource that the shader is trying to access is most likely a transient (virtual) texture (which has already been destroyed by the time the shader accesses it) with a size of 176MB. Is it possible to provide a full memory dump (with “memreport -full”) in an area where the game is known to crash?
Thanks,
Sam