Hello,
I think I have working OpenCL with RPC example using Docker (like in this instruction: Deploy the Pretrained Model on Android — tvm 0.9.dev182+ge718f5a8a documentation ). But when I run tuning using this Auto-tuning a Convolutional Network for Mobile GPU — tvm 0.9.dev182+ge718f5a8a documentation I’m getting sometimes strange stacktrace in android_rpc app:
/org.apache.tvm.tvmrpc D/TVM_RUNTIME: /workspace/apps/android_rpc/app/src/main/jni/../../../../../../include/../src/runtime/opencl/opencl_module.cc:251: OpenCL build error for device=0x7de805d598
Pass
/org.apache.tvm.tvmrpc D/TVM_RUNTIME: /workspace/apps/android_rpc/app/src/main/jni/../../../../../../include/../src/runtime/library_module.cc:80:
---------------------------------------------------------------
An error occurred during the execution of TVM.
For more information, please see: https://tvm.apache.org/docs/errors.html
---------------------------------------------------------------
Check failed: ret == 0 (-1 vs. 0) : TVMError: OpenCL build error for device=0x7de805d598
Pass
Stack trace:
File "/workspace/apps/android_rpc/app/src/main/jni/../../../../../../include/../src/runtime/opencl/opencl_module.cc", line 251