search

NVlabs / curobo

CUDA Accelerated Robot Library
https://curobo.org
Other
798 stars 125 forks source link

Nvidia WARP does not build on Jetson orin #61

Closed nivetaiyer closed 10 months ago

nivetaiyer commented 12 months ago
  1. cuRobo installation mode (choose from [python, isaac sim, docker python, docker isaac sim]): Python
  2. python version: 3.8
  3. Isaac Sim version (if using):

Issue Details Curobo builds fine, but unit tests fail because it can't find warp.so. I'm trying to build WARP on arm64 Ubuntu (jetson orin) but it fails because llvm is not installed. Can curobo work without warp? If not, what are the steps to successfully compile warp on arm64?

balakumar-s commented 12 months ago

cuRobo requires warp. We have only been running cuRobo within a docker on the jetson. You can follow instructions to build a docker for jetson here: https://curobo.org/source/getting_started/5_docker_development.html

Compiling warp on jetson can be done following: https://curobo.org/source/getting_started/5_docker_development.html#build-warp-for-nvidia-jetson

nivetaiyer commented 12 months ago

I tried that but it still fails. Doesn't the base image used in the docker file come with jetpack 5.1 which contains cuda 11.4, while warp apparently requires cuda 11.5?

nivetaiyer commented 12 months ago

Did you run the x86 docker image on orin?

nivetaiyer commented 12 months ago

When I try to build warp, it only builds for x86 on a Linux platform - https://github.com/NVIDIA/warp/blob/main/build_llvm.py#L30

The instructions for building warp here https://curobo.org/source/getting_started/5_docker_development.html#build-warp-for-nvidia-jetson don't actually seem to work.

balakumar-s commented 12 months ago

You need to run all docker build commands on a jetson to build an image for the jetson.

The instructions for building warp here https://curobo.org/source/getting_started/5_docker_development.html#build-warp-for-nvidia-jetson don't actually seem to work.

What issue are you facing in building warp on a jetson? You should install cuda 11.5+ on a jetson device (outside docker) and then compile warp (outside docker).

nivetaiyer commented 12 months ago

Are the instructions here https://curobo.org/source/getting_started/5_docker_development.html#build-warp-for-nvidia-jetson for aarch64 or x86? It was a linker error, trying to link with an x86 llvm library. I unfortunately don't have the exact log right now since I'm trying to install warp by passing the --build_llvm flag.

balakumar-s commented 12 months ago

They are for aarch64.

Note: We do not have instructions for cross compilation of curobo. To install curobo on a jetson, all instructions need to be done on a jetson.

balakumar-s commented 12 months ago

Also, can you check out an older branch of warp. I tested the instructions with v0.10.0 and v0.9.0

In case you want to use the latest version of warp, you can build without llvm by adding this argument --no-standalone to build_lib.py

nivetaiyer commented 11 months ago

Thanks! I managed to build warp but the tests fail with this error -

============================= test session starts ==============================
 platform linux -- Python 3.8.10, pytest-7.4.3, pluggy-1.3.0
 rootdir: /home/user/curobo
 configfile: pyproject.toml
 plugins: typeguard-4.1.5, torchtyping-0.1.4
 collected 121 items / 2 errors / 2 skipped
 ==================================== ERRORS ====================================
 ___________________ ERROR collecting tests/ik_config_test.py ___________________
 ../.local/lib/python3.8/site-packages/warp/codegen.py:620: in build
     adj.eval(adj.tree.body[0])
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1158: in emit_FunctionDef
     adj.eval(f)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1786: in emit_Assign
     value = adj.eval(node.value)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1674: in emit_Call
     out = adj.add_call(func=func, args=args, kwds=kwds, templates=templates, min_outputs=min_outputs)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:895: in add_call
     func = adj.resolve_func(func, args, min_outputs, templates, kwds)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:890: in resolve_func
     raise WarpCodegenError(
 E   warp.codegen.WarpCodegenError: Couldn't find function overload for 'index' that matched inputs with types: [warp.types.quatf, int32]
 During handling of the above exception, another exception occurred:
 ../.local/lib/python3.8/site-packages/_pytest/runner.py:341: in from_call
     result: Optional[TResult] = func()
 ../.local/lib/python3.8/site-packages/_pytest/runner.py:372: in <lambda>
     call = CallInfo.from_call(lambda: list(collector.collect()), "collect")
 ../.local/lib/python3.8/site-packages/_pytest/python.py:531: in collect
     self._inject_setup_module_fixture()
 ../.local/lib/python3.8/site-packages/_pytest/python.py:545: in _inject_setup_module_fixture
     self.obj, ("setUpModule", "setup_module")
 ../.local/lib/python3.8/site-packages/_pytest/python.py:310: in obj
     self._obj = obj = self._getobj()
 ../.local/lib/python3.8/site-packages/_pytest/python.py:528: in _getobj
     return self._importtestmodule()
 ../.local/lib/python3.8/site-packages/_pytest/python.py:617: in _importtestmodule
     mod = import_path(self.path, mode=importmode, root=self.config.rootpath)
 ../.local/lib/python3.8/site-packages/_pytest/pathlib.py:567: in import_path
     importlib.import_module(module_name)
 /usr/lib/python3.8/importlib/__init__.py:127: in import_module
     return _bootstrap._gcd_import(name[level:], package, level)
 <frozen importlib._bootstrap>:1014: in _gcd_import
     ???
 <frozen importlib._bootstrap>:991: in _find_and_load
     ???
 <frozen importlib._bootstrap>:975: in _find_and_load_unlocked
     ???
 <frozen importlib._bootstrap>:671: in _load_unlocked
     ???
 ../.local/lib/python3.8/site-packages/_pytest/assertion/rewrite.py:186: in exec_module
     exec(co, module.__dict__)
 tests/ik_config_test.py:134: in <module>
     (ik_base_config(), True),
 tests/ik_config_test.py:35: in ik_base_config
     robot_cfg = RobotConfig.from_dict(
 src/curobo/types/robot.py:49: in from_dict
     CudaRobotGeneratorConfig(**data_dict_in["kinematics"], tensor_args=tensor_args)
 <string>:31: in __init__
     ???
 src/curobo/cuda_robot_model/cuda_robot_generator.py:203: in __post_init__
     self.extra_links[k] = LinkParams.from_dict(self.extra_links[k])
 src/curobo/cuda_robot_model/kinematics_parser.py:43: in from_dict
     Pose.from_list(dict_data["fixed_transform"], tensor_args=TensorDeviceType())
 src/curobo/types/math.py:281: in get_numpy_matrix
     return self.get_matrix().cpu().numpy()
 /usr/lib/python3.8/contextlib.py:75: in inner
     return func(*args, **kwds)
 src/curobo/types/math.py:277: in get_matrix
     full_mat = pose_to_matrix(self.position, self.quaternion, out_matrix)
 src/curobo/geom/transform.py:163: in pose_to_matrix
     out_matrix[..., :3, :3] = quaternion_to_matrix(quaternion)
 src/curobo/geom/transform.py:110: in quaternion_to_matrix
     out_mat = QuatToMatrix.apply(quaternions, out_mat, adj_quaternion)
 ../.local/lib/python3.8/site-packages/torch/autograd/function.py:506: in apply
     return super().apply(*args, **kwargs)  # type: ignore[misc]
 src/curobo/geom/transform.py:1067: in forward
     init_warp()
 src/curobo/util/warp.py:27: in init_warp
     wp.force_load(wp.device_from_torch(tensor_args.device))
 ../.local/lib/python3.8/site-packages/warp/context.py:3613: in force_load
     m.load(d)
 ../.local/lib/python3.8/site-packages/warp/context.py:1449: in load
     builder = ModuleBuilder(self, self.options)
 ../.local/lib/python3.8/site-packages/warp/context.py:1081: in __init__
     self.build_kernel(kernel)
 ../.local/lib/python3.8/site-packages/warp/context.py:1109: in build_kernel
     kernel.adj.build(self)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:634: in build
     raise e
 ../.local/lib/python3.8/site-packages/warp/codegen.py:620: in build
     adj.eval(adj.tree.body[0])
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1158: in emit_FunctionDef
     adj.eval(f)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1786: in emit_Assign
     value = adj.eval(node.value)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1674: in emit_Call
     out = adj.add_call(func=func, args=args, kwds=kwds, templates=templates, min_outputs=min_outputs)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:895: in add_call
     func = adj.resolve_func(func, args, min_outputs, templates, kwds)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:890: in resolve_func
     raise WarpCodegenError(
 E   warp.codegen.WarpCodegenError: Error while parsing function "compute_pose_inverse" at /home/user/curobo/src/curobo/geom/transform.py:356:
 E       out_v[0] = wp.index(out_q, 3)
 E   ;Couldn't find function overload for 'index' that matched inputs with types: [warp.types.quatf, int32]
 ________________ ERROR collecting tests/trajopt_config_test.py _________________
 ../.local/lib/python3.8/site-packages/warp/codegen.py:620: in build
     adj.eval(adj.tree.body[0])
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1158: in emit_FunctionDef
     adj.eval(f)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1786: in emit_Assign
     value = adj.eval(node.value)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1674: in emit_Call
     out = adj.add_call(func=func, args=args, kwds=kwds, templates=templates, min_outputs=min_outputs)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:895: in add_call
     func = adj.resolve_func(func, args, min_outputs, templates, kwds)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:890: in resolve_func
     raise WarpCodegenError(
 E   warp.codegen.WarpCodegenError: Couldn't find function overload for 'index' that matched inputs with types: [warp.types.quatf, int32]
 During handling of the above exception, another exception occurred:
 ../.local/lib/python3.8/site-packages/_pytest/runner.py:341: in from_call
     result: Optional[TResult] = func()
 ../.local/lib/python3.8/site-packages/_pytest/runner.py:372: in <lambda>
     call = CallInfo.from_call(lambda: list(collector.collect()), "collect")
 ../.local/lib/python3.8/site-packages/_pytest/python.py:531: in collect
     self._inject_setup_module_fixture()
 ../.local/lib/python3.8/site-packages/_pytest/python.py:545: in _inject_setup_module_fixture
     self.obj, ("setUpModule", "setup_module")
 ../.local/lib/python3.8/site-packages/_pytest/python.py:310: in obj
     self._obj = obj = self._getobj()
 ../.local/lib/python3.8/site-packages/_pytest/python.py:528: in _getobj
     return self._importtestmodule()
 ../.local/lib/python3.8/site-packages/_pytest/python.py:617: in _importtestmodule
     mod = import_path(self.path, mode=importmode, root=self.config.rootpath)
 ../.local/lib/python3.8/site-packages/_pytest/pathlib.py:567: in import_path
     importlib.import_module(module_name)
 /usr/lib/python3.8/importlib/__init__.py:127: in import_module
     return _bootstrap._gcd_import(name[level:], package, level)
 <frozen importlib._bootstrap>:1014: in _gcd_import
     ???
 <frozen importlib._bootstrap>:991: in _find_and_load
     ???
 <frozen importlib._bootstrap>:975: in _find_and_load_unlocked
     ???
 <frozen importlib._bootstrap>:671: in _load_unlocked
     ???
 ../.local/lib/python3.8/site-packages/_pytest/assertion/rewrite.py:186: in exec_module
     exec(co, module.__dict__)
 tests/trajopt_config_test.py:115: in <module>
     (trajopt_base_config(), True),
 tests/trajopt_config_test.py:30: in trajopt_base_config
     robot_cfg = RobotConfig.from_dict(
 src/curobo/types/robot.py:49: in from_dict
     CudaRobotGeneratorConfig(**data_dict_in["kinematics"], tensor_args=tensor_args)
 <string>:31: in __init__
     ???
 src/curobo/cuda_robot_model/cuda_robot_generator.py:203: in __post_init__
     self.extra_links[k] = LinkParams.from_dict(self.extra_links[k])
 src/curobo/cuda_robot_model/kinematics_parser.py:43: in from_dict
     Pose.from_list(dict_data["fixed_transform"], tensor_args=TensorDeviceType())
 src/curobo/types/math.py:281: in get_numpy_matrix
     return self.get_matrix().cpu().numpy()
 /usr/lib/python3.8/contextlib.py:75: in inner
     return func(*args, **kwds)
 src/curobo/types/math.py:277: in get_matrix
     full_mat = pose_to_matrix(self.position, self.quaternion, out_matrix)
 src/curobo/geom/transform.py:163: in pose_to_matrix
     out_matrix[..., :3, :3] = quaternion_to_matrix(quaternion)
 src/curobo/geom/transform.py:110: in quaternion_to_matrix
     out_mat = QuatToMatrix.apply(quaternions, out_mat, adj_quaternion)
 ../.local/lib/python3.8/site-packages/torch/autograd/function.py:506: in apply
     return super().apply(*args, **kwargs)  # type: ignore[misc]
 src/curobo/geom/transform.py:1067: in forward
     init_warp()
 src/curobo/util/warp.py:27: in init_warp
     wp.force_load(wp.device_from_torch(tensor_args.device))
 ../.local/lib/python3.8/site-packages/warp/context.py:3613: in force_load
     m.load(d)
 ../.local/lib/python3.8/site-packages/warp/context.py:1449: in load
     builder = ModuleBuilder(self, self.options)
 ../.local/lib/python3.8/site-packages/warp/context.py:1081: in __init__
     self.build_kernel(kernel)
 ../.local/lib/python3.8/site-packages/warp/context.py:1109: in build_kernel
     kernel.adj.build(self)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:634: in build
     raise e
 ../.local/lib/python3.8/site-packages/warp/codegen.py:620: in build
    adj.eval(adj.tree.body[0])
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1158: in emit_FunctionDef
     adj.eval(f)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1786: in emit_Assign
     value = adj.eval(node.value)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1957: in eval
     return emit_node(adj, node)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:1674: in emit_Call
     out = adj.add_call(func=func, args=args, kwds=kwds, templates=templates, min_outputs=min_outputs)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:895: in add_call
     func = adj.resolve_func(func, args, min_outputs, templates, kwds)
 ../.local/lib/python3.8/site-packages/warp/codegen.py:890: in resolve_func
     raise WarpCodegenError(
 E   warp.codegen.WarpCodegenError: Error while parsing function "compute_matrix_to_quat" at /home/user/curobo/src/curobo/geom/transform.py:401:
 E       out_v[0] = wp.index(out_q, 3)
 E   ;Couldn't find function overload for 'index' that matched inputs with types: [warp.types.quatf, int32]
 =============================== warnings summary ===============================
 <unknown>:3
   <unknown>:3: DeprecationWarning: invalid escape sequence \p
 -- Docs: https://docs.pytest.org/en/stable/how-to/capture-warnings.html
 =========================== short test summary info ============================
 ERROR tests/ik_config_test.py - warp.codegen.WarpCodegenError: Error while pa...
 ERROR tests/trajopt_config_test.py - warp.codegen.WarpCodegenError: Error whi...
 !!!!!!!!!!!!!!!!!!! Interrupted: 2 errors during collection !!!!!!!!!!!!!!!!!!!!
 =================== 2 skipped, 1 warning, 2 errors in 12.56s ===================
balakumar-s commented 11 months ago

What version of warp are you using? Can you install warp 0.10.0 or lower?

waldezjr commented 10 months ago

I was able to compile libwarp in my Orin Nx (not AGX). I used tag v0.9.0 and the option to build llvm from source.

python3 build.py -build_llvm

Without this option, warp uses the already built files for x64, which (maybe obviously in hindsight xD) don't work in aarch64

c0d1f1ed commented 10 months ago

I've been working on this recently. The next release of Warp will include Jetson support, and we'll publish pip packages.

c0d1f1ed commented 10 months ago

Warp release 1.0.0-beta.6 adds Jetson support. It can be installed using pip install warp-lang==1.0.0b6.

Hermanye996 commented 10 months ago

pip install warp-lang==1.0.0b6 works for me, but error occurred when I run curobo fk. What should I do? Thanks. @c0d1f1ed @balakumar-s 

[INFO]  Creating curobo robot config instance...
Traceback (most recent call last):
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 620, in build
    adj.eval(adj.tree.body[0])
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1162, in emit_FunctionDef
    adj.eval(f)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1803, in emit_Assign
    value = adj.eval(node.value)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1691, in emit_Call
    out = adj.add_call(func=func, args=args, kwds=kwds, templates=templates, min_outputs=min_outputs)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 896, in add_call
    func = adj.resolve_func(func, args, min_outputs, templates, kwds)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 891, in resolve_func
    raise WarpCodegenError(
warp.codegen.WarpCodegenError: Couldn't find function overload for 'index' that matched inputs with types: [warp.types.quatf, int32]

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "herman_curobo_warpper.py", line 1482, in <module>
    main()
  File "herman_curobo_warpper.py", line 1468, in main
    demo_compute_fk(command_line_args)
  File "herman_curobo_warpper.py", line 1297, in demo_compute_fk
    herman_curobo_warpper = hermanCuroboWarpper(
  File "herman_curobo_warpper.py", line 52, in __init__
    self.curobo_init()
  File "herman_curobo_warpper.py", line 91, in curobo_init
    self.curobo_kinemetics_init()
  File "herman_curobo_warpper.py", line 239, in curobo_kinemetics_init
    self.curobo_robot_config_instance = RobotConfig.from_dict(
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/types/robot.py", line 49, in from_dict
    CudaRobotGeneratorConfig(**data_dict_in["kinematics"], tensor_args=tensor_args)
  File "<string>", line 33, in __init__
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/cuda_robot_model/cuda_robot_generator.py", line 208, in __post_init__
    self.extra_links[k] = LinkParams.from_dict(self.extra_links[k])
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/cuda_robot_model/kinematics_parser.py", line 43, in from_dict
    Pose.from_list(dict_data["fixed_transform"], tensor_args=TensorDeviceType())
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/types/math.py", line 281, in get_numpy_matrix
    return self.get_matrix().cpu().numpy()
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/contextlib.py", line 75, in inner
    return func(*args, **kwds)
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/types/math.py", line 277, in get_matrix
    full_mat = pose_to_matrix(self.position, self.quaternion, out_matrix)
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/geom/transform.py", line 163, in pose_to_matrix
    out_matrix[..., :3, :3] = quaternion_to_matrix(quaternion)
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/geom/transform.py", line 110, in quaternion_to_matrix
    out_mat = QuatToMatrix.apply(quaternions, out_mat, adj_quaternion)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/torch/autograd/function.py", line 506, in apply
    return super().apply(*args, **kwargs)  # type: ignore[misc]
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/geom/transform.py", line 1067, in forward
    init_warp()
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/util/warp.py", line 27, in init_warp
    wp.force_load(wp.device_from_torch(tensor_args.device))
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/context.py", line 3698, in force_load
    m.load(d)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/context.py", line 1534, in load
    builder = ModuleBuilder(self, self.options)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/context.py", line 1166, in __init__
    self.build_kernel(kernel)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/context.py", line 1194, in build_kernel
    kernel.adj.build(self)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 634, in build
    raise e
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 620, in build
    adj.eval(adj.tree.body[0])
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1162, in emit_FunctionDef
    adj.eval(f)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1803, in emit_Assign
    value = adj.eval(node.value)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1691, in emit_Call
    out = adj.add_call(func=func, args=args, kwds=kwds, templates=templates, min_outputs=min_outputs)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 896, in add_call
    func = adj.resolve_func(func, args, min_outputs, templates, kwds)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 891, in resolve_func
    raise WarpCodegenError(
warp.codegen.WarpCodegenError: Error while parsing function "compute_pose_inverse" at /home/herman_orin_240118/gmp_ws/curobo/src/curobo/geom/transform.py:356:
    out_v[0] = wp.index(out_q, 3)
;Couldn't find function overload for 'index' that matched inputs with types: [warp.types.quatf, int32]
Hermanye996 commented 10 months ago

Official Kinematics_example.py also:

(gmp) herman_orin_240118@ubuntu:~/gmp_ws/curobo/examples$ python kinematics_example.py 
Traceback (most recent call last):
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 620, in build
    adj.eval(adj.tree.body[0])
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1162, in emit_FunctionDef
    adj.eval(f)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1803, in emit_Assign
    value = adj.eval(node.value)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1691, in emit_Call
    out = adj.add_call(func=func, args=args, kwds=kwds, templates=templates, min_outputs=min_outputs)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 896, in add_call
    func = adj.resolve_func(func, args, min_outputs, templates, kwds)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 891, in resolve_func
    raise WarpCodegenError(
warp.codegen.WarpCodegenError: Couldn't find function overload for 'index' that matched inputs with types: [warp.types.quatf, int32]

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "kinematics_example.py", line 66, in <module>
    demo_full_config_robot()
  File "kinematics_example.py", line 53, in demo_full_config_robot
    robot_cfg = RobotConfig.from_dict(config_file, tensor_args)
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/types/robot.py", line 49, in from_dict
    CudaRobotGeneratorConfig(**data_dict_in["kinematics"], tensor_args=tensor_args)
  File "<string>", line 33, in __init__
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/cuda_robot_model/cuda_robot_generator.py", line 208, in __post_init__
    self.extra_links[k] = LinkParams.from_dict(self.extra_links[k])
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/cuda_robot_model/kinematics_parser.py", line 43, in from_dict
    Pose.from_list(dict_data["fixed_transform"], tensor_args=TensorDeviceType())
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/types/math.py", line 281, in get_numpy_matrix
    return self.get_matrix().cpu().numpy()
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/contextlib.py", line 75, in inner
    return func(*args, **kwds)
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/types/math.py", line 277, in get_matrix
    full_mat = pose_to_matrix(self.position, self.quaternion, out_matrix)
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/geom/transform.py", line 163, in pose_to_matrix
    out_matrix[..., :3, :3] = quaternion_to_matrix(quaternion)
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/geom/transform.py", line 110, in quaternion_to_matrix
    out_mat = QuatToMatrix.apply(quaternions, out_mat, adj_quaternion)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/torch/autograd/function.py", line 506, in apply
    return super().apply(*args, **kwargs)  # type: ignore[misc]
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/geom/transform.py", line 1067, in forward
    init_warp()
  File "/home/herman_orin_240118/gmp_ws/curobo/src/curobo/util/warp.py", line 27, in init_warp
    wp.force_load(wp.device_from_torch(tensor_args.device))
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/context.py", line 3698, in force_load
    m.load(d)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/context.py", line 1534, in load
    builder = ModuleBuilder(self, self.options)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/context.py", line 1166, in __init__
    self.build_kernel(kernel)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/context.py", line 1194, in build_kernel
    kernel.adj.build(self)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 634, in build
    raise e
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 620, in build
    adj.eval(adj.tree.body[0])
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1162, in emit_FunctionDef
    adj.eval(f)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1803, in emit_Assign
    value = adj.eval(node.value)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1975, in eval
    return emit_node(adj, node)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 1691, in emit_Call
    out = adj.add_call(func=func, args=args, kwds=kwds, templates=templates, min_outputs=min_outputs)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 896, in add_call
    func = adj.resolve_func(func, args, min_outputs, templates, kwds)
  File "/home/herman_orin_240118/anaconda3/envs/gmp/lib/python3.8/site-packages/warp/codegen.py", line 891, in resolve_func
    raise WarpCodegenError(
warp.codegen.WarpCodegenError: Error while parsing function "compute_pose_inverse" at /home/herman_orin_240118/gmp_ws/curobo/src/curobo/geom/transform.py:356:
    out_v[0] = wp.index(out_q, 3)
;Couldn't find function overload for 'index' that matched inputs with types: [warp.types.quatf, int32]
c0d1f1ed commented 10 months ago

Sorry about that. I made a breaking change by adding wp.extract() to replace wp.index() when used to extract elements from a vector or quaternion. wp.index() now only handles array indexing. I should have documented that change better.

balakumar-s commented 10 months ago

We will push a fix on the curobo repo to remove use of wp.index().

balakumar-s commented 10 months ago

We pushed a fix to curobo, there are now no instances of wp.index and it should work with the latest warp.