doonny / PipeCNN

An OpenCL-based FPGA Accelerator for Convolutional Neural Networks
Apache License 2.0
1.26k stars 370 forks source link

./run.exe conv.xclbin: terminate called recursively #179

Open chu-chiao opened 1 year ago

chu-chiao commented 1 year ago

=======================================LOG======================================= *** PipeCNN: An OpenCL-Based FPGA Accelerator for CNNs *** XRT build version: 2.13.466 Build hash: f5505e402c2ca1ffe45eb6d3a9399b23a0dc8776 Build date: 2022-04-14 17:43:11 Git branch: 2022.1 PID: 160167 UID: 605465 [Sat Jun 17 15:32:30 2023 GMT] HOST: alveo-u50d-02.inf.ethz.ch EXE: /home/xxx/examples/PipeCNN/project_xilinx/run.exe [XRT] WARNING: The xrt.ini flag "data_transfer_trace" is deprecated and will be removed in a future release. Please use the equivalent flag "device_trace."

Platform: Xilinx Totally 1 device(s) are found Using Device 0: xilinx_u50_gen3x16_xdma_base_5 Device OpenCL Version: OpenCL 1.0 Device Max Compute Units: 0 Device Max WorkGroup Size: -1 Device Max WorkItem Size: -1 Device Global Memory Size: 0 MBytes Device Local Memory Size: 16 KBytes Device Max Clock Freq: 500 Mhz [XRT] WARNING: Trace buffer size for 0th. TS2MM is too big for memory resource. Using 268435456 instead.

138455872 total weights read 1024 total output reference read

terminate called after throwing an instance of '__gnu_cxx::recursive_init_error' what(): std::exception Aborted =======================================LOG=======================================

The exception happened around Line #446 and #450 that when execute clEnqueueMigrateMemObjects(). If I commend Line #446 and #450, the program can continue executing. (But there will be another recursive_init_error exception caused by other line)

Please help me solve this problem. Many thanks.