uzh-rpg / agile_autonomy

Repository Containing the Code associated with the Paper: "Learning High-Speed Flight in the Wild"
GNU General Public License v3.0
607 stars 164 forks source link

Further instruction needed for someone new to this. #47

Closed psun-autel closed 2 years ago

psun-autel commented 2 years ago

After running roslaunch agile_autonomy simulation.launch

I am not sure what I should expect or what I should do next in order to see this demo. In the Unity window, the environment is being rendered in real time, and you can observe the grass is moving. The view angle is fixed.

The partial console output is shown as below.

roslaunch agile_autonomy simulation.launch

[ INFO] [1649622819.898319698]: [/hummingbird/agile_autonomy] camera/baseline = 0.1 [ INFO] [1649622819.898993112]: [/hummingbird/agile_autonomy] camera/pitch_angle_deg = 0 Vertical Fov is 74.7021 Initializing ZMQ connections... Initializing ZMQ done... [ INFO] [1649622819.901140818]: Connecting to unity... Ready ? : 0 [ INFO] [1649622820.083847360]: Finished loading Gazebo ROS API Plugin. [ INFO] [1649622820.084971604]: waitForService: Service [/gazebo/set_physics_properties] has not been advertised, waiting... Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 [ INFO] [1649622821.892760684]: waitForService: Service [/gazebo/set_physics_properties] is now available. Ready ? : 0 [ INFO] [1649622821.910235945]: Physics dynamic reconfigure ready. [INFO] [1649622821.992587, 0.000000]: Calling service /gazebo/spawn_urdf_model [INFO] [1649622822.084647, 0.000000]: Spawn status: SpawnModel: Successfully spawned entity Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 [hummingbird/spawn_hummingbird-3] process has finished cleanly log file: /home/psun/.ros/log/808eb694-b90d-11ec-b32c-11a93648a3ed/hummingbird-spawn_hummingbird-3*.log Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 0 Ready ? : 1 [ INFO] [1649622827.802552360, 2.876000000]: Unity is ready! [ERROR] [1649622828.212042755, 3.080000000]: no kernel image is available for execution on the device 209

[ERROR] [1649622828.321543409, 3.135000000]: no kernel image is available for execution on the device 209

[ERROR] [1649622828.459425264, 3.203000000]: no kernel image is available for execution on the device 209

[ERROR] [1649622828.582797969, 3.265000000]: no kernel image is available for execution on the device 209

[ERROR] [1649622828.721430274, 3.334000000]: no kernel image is available for execution on the device 209

^C[ERROR] [1649622828.859060628, 3.403000000]: no kernel image is available for execution on the device 209

psun-autel commented 2 years ago

I figured out myself. Closing this one.

blu666 commented 1 year ago

How did you fix this? I'm running into the same problem. Would be great if you could share your solution!! Thank you!

blu666 commented 1 year ago

I resolved the "no kernel image is available" issue by following #18 and editing line 14 in the agile_autonomy/data_generation/agile_autonomy/CMakeLists.txt according to GPU architecture

random-user-in-space commented 1 year ago

@blu666 Can you tell me what did you use? I am using NVIDIA GeForce RTX 3060 Laptop GPU with cuda 11.2

blu666 commented 1 year ago

@random-user-in-space I am using NVIDIA RTX A2000 ti laptop GPU which I believe is equivalent to 3050. I also installed cuda 11.2. After changing the CMakeList, sometimes it will throw a lot of warning messages about sm_86, but according to other sources these messages can be ignored. Hope this helps!

random-user-in-space commented 1 year ago

@blu666 Thank you for quick response. Do you mind sharing the exact config that you used?

I used the below config but I get build error if I use that. set(CUDA_NVCC_FLAGS ${CUDA_NVCC_FLAGS}; -O3 -gencode=arch=compute_52,code=sm_52 -gencode=arch=compute_61,code=sm_61 -gencode=arch=compute_62,code=sm_62 -gencode=arch=compute_75,code=sm_75 -gencode=arch=compute_86,code=sm_86

    )

error while building-->

Errors << agile_autonomy:make /home/sumit/agile_autonomy_ws/catkin_aa/logs/agile_autonomy/build.make.041.log nvcc fatal : Unsupported gpu architecture 'compute_86' CMake Error at sgm_gpu_generated_median_filter.cu.o.Release.cmake:220 (message): Error generating /home/sumit/agile_autonomy_ws/catkin_aa/build/agile_autonomy/CMakeFiles/sgm_gpu.dir/src/sgm_gpu/./sgm_gpu_generated_median_filter.cu.o

make[2]: [CMakeFiles/sgm_gpu.dir/build.make:2289: CMakeFiles/sgm_gpu.dir/src/sgm_gpu/sgm_gpu_generated_median_filter.cu.o] Error 1 make[2]: Waiting for unfinished jobs.... nvcc fatal : Unsupported gpu architecture 'compute_86' CMake Error at sgm_gpu_generated_costs.cu.o.Release.cmake:220 (message): Error generating /home/sumit/agile_autonomy_ws/catkin_aa/build/agile_autonomy/CMakeFiles/sgm_gpu.dir/src/sgm_gpu/./sgm_gpu_generated_costs.cu.o

nvcc fatal : Unsupported gpu architecture 'compute_86' CMake Error at sgm_gpu_generated_sgm_gpu.cu.o.Release.cmake:220 (message): Error generating /home/sumit/agile_autonomy_ws/catkin_aa/build/agile_autonomy/CMakeFiles/sgm_gpu.dir/src/sgm_gpu/./sgm_gpu_generated_sgm_gpu.cu.o

make[2]: [CMakeFiles/sgm_gpu.dir/build.make:1673: CMakeFiles/sgm_gpu.dir/src/sgm_gpu/sgm_gpu_generated_costs.cu.o] Error 1 make[2]: [CMakeFiles/sgm_gpu.dir/build.make:1498: CMakeFiles/sgm_gpu.dir/src/sgm_gpu/sgm_gpu_generated_sgm_gpu.cu.o] Error 1 nvcc fatal : Unsupported gpu architecture 'compute_86' CMake Error at sgm_gpu_generated_hamming_cost.cu.o.Release.cmake:220 (message): Error generating /home/sumit/agile_autonomy_ws/catkin_aa/build/agile_autonomy/CMakeFiles/sgm_gpu.dir/src/sgm_gpu/./sgm_gpu_generated_hamming_cost.cu.o

nvcc fatal : Unsupported gpu architecture 'compute_86' CMake Error at sgm_gpu_generated_left_right_consistency.cu.o.Release.cmake:220 (message): Error generating /home/sumit/agile_autonomy_ws/catkin_aa/build/agile_autonomy/CMakeFiles/sgm_gpu.dir/src/sgm_gpu/./sgm_gpu_generated_left_right_consistency.cu.o

make[2]: [CMakeFiles/sgm_gpu.dir/build.make:1940: CMakeFiles/sgm_gpu.dir/src/sgm_gpu/sgm_gpu_generated_hamming_cost.cu.o] Error 1 make[2]: [CMakeFiles/sgm_gpu.dir/build.make:2115: CMakeFiles/sgm_gpu.dir/src/sgm_gpu/sgm_gpu_generated_left_right_consistency.cu.o] Error 1 make[1]: [CMakeFiles/Makefile2:1557: CMakeFiles/sgm_gpu.dir/all] Error 2 make: [Makefile:141: all] Error 2

As suggested by nvcc --help command, compute_86 seems to be valid. --gpu-code ,...
Allowed values for this option: 'compute_35','compute_37','compute_50', 'compute_52','compute_53','compute_60','compute_61','compute_62','compute_70', 'compute_72','compute_75','compute_80','compute_86','lto_35','lto_37','lto_50', 'lto_52','lto_53','lto_60','lto_61','lto_62','lto_70','lto_72','lto_75', 'lto_80','lto_86','sm_35','sm_37','sm_50','sm_52','sm_53','sm_60','sm_61', 'sm_62','sm_70','sm_72','sm_75','sm_80','sm_86'.

blu666 commented 1 year ago

@random-user-in-space Have you resolved this issue?

I've encountered this error before, but it went away. It is a weird error.

"nvcc fatal : Unsupported gpu architecture 'compute_86'" but 'compute_86' is a allowed value according to the last line. Try restarting the computer (: Or might be related to something in your Cuda and driver.