Page 1 of 1
luxcore standalone annoying infinite loop
Posted: Tue Mar 16, 2021 9:07 am
by bartek_zgo
Hi,
I want to report a small bug. If I start render with standalone version, set batch.haltspp = 3, the render stops at 3 Samples but there is some strange infinite loop that will apply OIDN. As you can see on attached image, render stops at Pass 3, but in console you can see OIDN that is repeated. The CPU load is very high (one core 100%) and the interface is terribly slow.
If I remove OIDN from pipeline there is nothing in Console but CPU remains very high. So it looks that the the loop is still running but without any output
Re: luxcore standalone annoying infinite loop
Posted: Wed Mar 17, 2021 8:28 am
by kintuX
I can confirm the issue.
After render hits Halt then just switching to Image Pipeline 1 (OIDN denoiser) LuxCore starts an infinite loop.
Code: Select all
[19.703] [MergeSampleBuffersOCL] Kernels compilation time: 16ms
[LuxRays][19.734] [LinearToneMap] Compiler options: -D LUXRAYS_OPENCL_KERNEL -D SLG_OPENCL_KERNEL -D LUXRAYS_CUDA_DEVICE -D LUXRAYS_OS_WINDOWS --use_fast_math
[LuxRays][19.734] [LinearToneMap] Compiling kernels
[LuxRays][19.734] [LinearToneMap] Program cached
[LuxCore][19.734] [AutoLinearToneMap] Compiling LinearToneMap_Apply Kernel
[LuxCore][19.734] [LinearToneMap] Kernels compilation time: 0ms
[LuxRays][19.734] [Device GeForce GTX 1070 CUDAIntersect] Gamma table buffer size: 16Kbytes
[LuxRays][19.734] [GammaCorrectionPlugin] Compiler options: -D LUXRAYS_OPENCL_KERNEL -D SLG_OPENCL_KERNEL -D LUXRAYS_CUDA_DEVICE -D LUXRAYS_OS_WINDOWS --use_fast_math
[LuxRays][19.734] [GammaCorrectionPlugin] Compiling kernels
[LuxRays][19.750] [GammaCorrectionPlugin] Program cached
[LuxCore][19.750] [GammaCorrectionPlugin] Compiling GammaCorrectionPlugin_Apply Kernel
[LuxCore][19.750] [GammaCorrectionPlugin] Kernels compilation time: 16ms
[LuxCore][20.375] Samples per pixel 100%, rendering done.
[LuxCore][65.969] [IntelOIDNPlugin] Applying single OIDN
[LuxCore][66.000] IntelOIDNPlugin executing filter
[LuxCore][66.453] IntelOIDNPlugin apply took: 0.5secs
[LuxCore][66.453] IntelOIDNPlugin copying output buffer
[LuxCore][66.453] IntelOIDNPlugin single execution took a total of 0.484secs
[LuxRays][66.547] [LinearToneMap] Compiler options: -D LUXRAYS_OPENCL_KERNEL -D SLG_OPENCL_KERNEL -D LUXRAYS_CUDA_DEVICE -D LUXRAYS_OS_WINDOWS --use_fast_math
[LuxRays][66.547] [LinearToneMap] Compiling kernels
[LuxRays][66.547] [LinearToneMap] Program cached
[LuxCore][66.547] [AutoLinearToneMap] Compiling LinearToneMap_Apply Kernel
[LuxCore][66.547] [LinearToneMap] Kernels compilation time: 0ms
[LuxRays][66.547] [Device GeForce GTX 1070 CUDAIntersect] Gamma table buffer size: 16Kbytes
[LuxRays][66.547] [GammaCorrectionPlugin] Compiler options: -D LUXRAYS_OPENCL_KERNEL -D SLG_OPENCL_KERNEL -D LUXRAYS_CUDA_DEVICE -D LUXRAYS_OS_WINDOWS --use_fast_math
[LuxRays][66.547] [GammaCorrectionPlugin] Compiling kernels
[LuxRays][66.547] [GammaCorrectionPlugin] Program cached
[LuxCore][66.547] [GammaCorrectionPlugin] Compiling GammaCorrectionPlugin_Apply Kernel
[LuxCore][66.547] [GammaCorrectionPlugin] Kernels compilation time: 0ms
[LuxCore][66.594] [IntelOIDNPlugin] Applying single OIDN
[LuxCore][66.625] IntelOIDNPlugin executing filter
[LuxCore][67.094] IntelOIDNPlugin apply took: 0.5secs
[LuxCore][67.094] IntelOIDNPlugin copying output buffer
[LuxCore][67.094] IntelOIDNPlugin single execution took a total of 0.500secs
[LuxCore][67.281] [IntelOIDNPlugin] Applying single OIDN
[LuxCore][67.297] IntelOIDNPlugin executing filter
[LuxCore][67.766] IntelOIDNPlugin apply took: 0.5secs
[LuxCore][67.766] IntelOIDNPlugin copying output buffer
[LuxCore][67.766] IntelOIDNPlugin single execution took a total of 0.485secs
[LuxCore][67.906] [IntelOIDNPlugin] Applying single OIDN
[LuxCore][67.922] IntelOIDNPlugin executing filter
[LuxCore][68.391] IntelOIDNPlugin apply took: 0.5secs
[LuxCore][68.391] IntelOIDNPlugin copying output buffer
[LuxCore][68.391] IntelOIDNPlugin single execution took a total of 0.485secs
[LuxCore][68.531] [IntelOIDNPlugin] Applying single OIDN
[LuxCore][68.562] IntelOIDNPlugin executing filter
[LuxCore][69.031] IntelOIDNPlugin apply took: 0.5secs
[LuxCore][69.031] IntelOIDNPlugin copying output buffer
[LuxCore][69.031] IntelOIDNPlugin single execution took a total of 0.500secs
...
PS. After disabling OIDN (Image Pipeline 0) things return to normal - there's no excessive CPU usage (CPU load 0%)