CI: osx and osx_arm64 #5
Open
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
1 task
|
Are you happy to take a look at migrating the PR CI from boa to pixi/rattler-build? |
|
@Tobias-Fischer #8 migrates the PR CI to |
oursland
added a commit
to oursland/ros-jazzy
that referenced
this pull request
Sep 20, 2024
Current processes:
rebuild:
reset ; rm -rf output/bld && pixi run generate-recipes && time pixi run -v rattler-build build --recipe recipes/ros-jazzy-rosidl-generator-py --recipe recipes/ros-jazzy-rclpy --recipe recipes/ros-jazzy-rcl-interfaces -m conda_build_config.yaml -c robostack-jazzy -c conda-forge --keep-build
test:
mamba create -n ros-jazzy -c conda-forge -c file:///Users/jso/code/ros-jazzy/output python=3.11\* ros-jazzy-rclpy
mamba run --live-stream -n ros-jazzy lldb python -- -c 'import rclpy; rclpy.init(); node = rclpy.create_node("test"); rclpy.spin(node)'
Then run with the 'r' command:
The second line will run a simple python script that exhibits the crash. The log looks like this:
(lldb) target create "python"
Current executable set to '/Users/jso/code/FreeCAD/FreeCAD/.conda/ros-jazzy/bin/python' (arm64).
(lldb) settings set -- target.run-args "-c" "import rclpy; rclpy.init(); node = rclpy.create_node(\"test\"); rclpy.spin(node)"
(lldb) r
Process 78922 launched: '/Users/jso/code/FreeCAD/FreeCAD/.conda/ros-jazzy/bin/python' (arm64)
Process 78922 stopped
* thread #1, queue = 'com.apple.main-thread', stop reason = EXC_BAD_ACCESS (code=1, address=0x60)
frame #0: 0x00000001027d6314 libpython3.11.dylib`set_attribute_error_context + 60
libpython3.11.dylib`set_attribute_error_context:
-> 0x1027d6314 <+60>: ldr x0, [x8, #0x60]
0x1027d6318 <+64>: bl 0x1028a0b0c ; PyErr_GivenExceptionMatches
0x1027d631c <+68>: cbz w0, 0x1027d6384 ; <+172>
0x1027d6320 <+72>: ldr x8, [x21, #0x358]
Target 0: (python) stopped.
(lldb) bt
* thread #1, queue = 'com.apple.main-thread', stop reason = EXC_BAD_ACCESS (code=1, address=0x60)
* frame #0: 0x00000001027d6314 libpython3.11.dylib`set_attribute_error_context + 60
frame #1: 0x00000001027d68f0 libpython3.11.dylib`PyObject_GetAttr + 192
frame #2: 0x00000001027d67e0 libpython3.11.dylib`PyObject_GetAttrString + 108
frame #3: 0x0000000101a87b04 libbuiltin_interfaces__rosidl_generator_py.dylib`builtin_interfaces__msg__time__convert_from_py + 60
frame #4: 0x000000010219c0a8 librcl_interfaces__rosidl_generator_py.dylib`rcl_interfaces__msg__parameter_event__convert_from_py + 356
frame RoboStack#5: 0x0000000100e48250 _rclpy_pybind11.cpython-311-darwin.so`___lldb_unnamed_symbol2190 + 360
frame RoboStack#6: 0x0000000100e2b844 _rclpy_pybind11.cpython-311-darwin.so`___lldb_unnamed_symbol1795 + 76
frame RoboStack#7: 0x0000000100e2c1a0 _rclpy_pybind11.cpython-311-darwin.so`___lldb_unnamed_symbol1802 + 208
frame RoboStack#8: 0x0000000100dc17c0 _rclpy_pybind11.cpython-311-darwin.so`___lldb_unnamed_symbol708 + 4556
frame RoboStack#9: 0x00000001000b7f4c python`cfunction_call + 124
frame RoboStack#10: 0x00000001000606f0 python`_PyObject_MakeTpCall + 332
frame RoboStack#11: 0x0000000100162e3c python`_PyEval_EvalFrameDefault + 46484
frame RoboStack#12: 0x00000001001674c8 python`_PyEval_Vector + 184
frame #13: 0x00000001000608b8 python`_PyObject_FastCallDictTstate + 156
frame #14: 0x00000001000617f0 python`_PyObject_Call_Prepend + 176
frame #15: 0x00000001000dc888 python`slot_tp_init + 196
frame #16: 0x00000001000d4de4 python`type_call + 464
frame #17: 0x00000001000606f0 python`_PyObject_MakeTpCall + 332
frame #18: 0x0000000100162e3c python`_PyEval_EvalFrameDefault + 46484
frame #19: 0x00000001001674c8 python`_PyEval_Vector + 184
frame #20: 0x00000001000608b8 python`_PyObject_FastCallDictTstate + 156
frame #21: 0x00000001000617f0 python`_PyObject_Call_Prepend + 176
frame #22: 0x00000001000dc888 python`slot_tp_init + 196
frame #23: 0x00000001000d4de4 python`type_call + 464
frame #24: 0x00000001000606f0 python`_PyObject_MakeTpCall + 332
frame #25: 0x0000000100162e3c python`_PyEval_EvalFrameDefault + 46484
frame #26: 0x00000001001568f4 python`PyEval_EvalCode + 220
frame #27: 0x00000001001bc970 python`run_mod + 144
frame #28: 0x00000001001c04a4 python`PyRun_SimpleStringFlags + 272
frame #29: 0x00000001001e1b3c python`Py_RunMain + 1396
frame #30: 0x00000001001e3050 python`pymain_main + 1252
frame #31: 0x0000000100003398 python`main + 56
frame #32: 0x00000001921f0274 dyld`start + 2840
|
Ok, I see the problem, on conda-forge we have gz-gui 8.1.1 (https://github.com/conda-forge/gz-gui-feedstock/blob/414bd0e904ebdcb2abd6e0dc8acd6babe4ccd031/recipe/meta.yaml#L3) while the vendor expects 8.3.0 (see https://github.com/gazebo-release/gz_gui_vendor/blob/jazzy/CMakeLists.txt#L6-L8). I will update the conda-forge version, but this is an example of why we need to set |
5 tasks
|
I'm a little confused as this appears to be doing what you suggest, installing a |
|
Here's what the output from |
|
@traversaro The change to set |
Ah, good point. This is okish, but explains what you saw. All gz-* projects generation version config files with |
|
I opened ament/ament_cmake#552 that hopefully should be helpful if backported to jazzy. With that and a way to specify different CMake options for each project in vinca, we can simply pass |
|
Looks like I've brought over your patch to force de-vendoring and set it for the gz-*-vendor packages. |
|
Unfortunately still some conflicts that do not seem easily resolved: |
|
I think we need to bump versions in https://github.com/oursland/ros-jazzy/blob/osx-ci/.ci_support/conda_forge_pinnings.yaml - based on https://github.com/conda-forge/conda-forge-pinning-feedstock/blob/main/recipe/conda_build_config.yaml and ongoing migrations (https://conda-forge.org/status/ - relevant are probably assimp543, fmt11, icu75, libboost186, libabseil/grpc/libprotobuf spdlog114 and vtk931) |
|
That was generated from the local build, which I believe uses I did a build locally where I only specified |
|
You should be using both, https://github.com/oursland/ros-jazzy/blob/osx-ci/.ci_support/conda_forge_pinnings.yaml (that needs probably be converted in a format that is understood by rattler-build) and https://github.com/oursland/ros-jazzy/blob/osx-ci/conda_build_config.yaml. If we don't pin against the versions specified in conda_forge_pinnings.yaml, we will inevitably run into issues like you have reported .. |
|
My attempts to find an authoritative source for Where can I find the current |
|
That’s the current authorative one: https://github.com/conda-forge/conda-forge-pinning-feedstock/blob/main/recipe/conda_build_config.yaml The migrations can be found in that repository as well, and the status is linked above. |
|
The authoritative |
…pstream. The file .ci_support/conda_forge_pinnings.yaml does not dictate the which packages are used by RoboStack. vinca uses the file conda_build_config.yaml. The .ci_support/conda_forge_pinnings.yaml serves as a reference for compatible packages.
|
I have updated It looks like the authoritative
|
|
De-vendoring is nearly completed. The following packages are not being detected via |
|
I've forced de-vendoring, but found that the |
Thanks! I am afraid this is a macos-specific issue, can you open an issue at https://github.com/conda-forge/gz-sim-feedstock following the template? I will not have access to a macos machine for a week unfortunately, but at least we track this problem in an issue. |
|
@traversaro Unfortunately, I cannot create an issue. Attempts to do so result in the following message:
|
|
@oursland you mentioned that homebrew installed GZ would not crash this way. Do you have logs from that run as well? I am wondering if the shader file that is corrupted(?) is the problem... |
|
Here's the output from the Homebrew-installed |
|
I wonder if this is the same of conda-forge/gz-sim-feedstock#24 ? In case somebody has access to a macOS machine, it would be good to try also with |
Frustratingly |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.