-
-
Notifications
You must be signed in to change notification settings - Fork 5.1k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
TST, MAINT: some tests blocking 1.11.0 on MacOS ARM64 with NumPy 1.25.0
#18732
Comments
I made a few attempts to reproduce on ARM Mac this evening locally with I mean, we may just need some simple tol bumps, but it is still nice to repro locally when reasonable. |
Do you need any help on this one? I have an M1 at home. |
* deals with the 1 of the 2 Apple silicon test failures in scipygh-18732 * I was able to reproduce locally, but because it was flaky I needed `pytest-repeat` for consistent failure: `CC=clang CXX=clang++ FC=gfortran-12 python dev.py test -t scipy/signal/tests/test_filter_design.py::TestSOSFreqz::test_fs_param -- --count=200` * I verified that with the provided patch/`atol` bump, we consistently pass even with `2000` repeats [skip circle]
* deals with the second part of scipygh-18732 * I was able to reproduce locally, but this one was even flakier--analysis of patch success below, and possibly related to subtle low-level MacOS ARM issues upstream in numpy/numpy#23996 (NumPy `1.25.0`) * before this patch, 4 trials of this command with `pytest-repeat`: `CC=clang CXX=clang++ FC=gfortran-12 python dev.py test -t scipy/signal/tests/test_signaltools.py::TestDecimate::test_complex_iir_dlti -- --count=20000` `81 failed, 19919 passed in 43.68s` `81 failed, 19919 passed in 43.79s` `20000 passed in 33.24s` `43 failed, 19957 passed in 35.17s` * after this patch, 6 trials of the same command: `20000 passed in 30.44s` `20000 passed in 33.08s` `20000 passed in 33.46s` `20000 passed in 33.45s` `20000 passed in 33.56s` `20000 passed in 33.30s` [skip circle]
I'm seeing unrelated test fails. Python3.10, numpy 1.25, scipy maintenance/1.11.0
1.11.0 1.25.0 sys.version_info(major=3, minor=10, micro=11, releaselevel='final', serial=0)
Build Dependencies:
blas:
detection method: pkgconfig
found: true
include directory: /opt/arm64-builds/include
lib directory: /opt/arm64-builds/lib
name: openblas
openblas configuration: USE_64BITINT= DYNAMIC_ARCH=1 DYNAMIC_OLDER= NO_CBLAS=
NO_LAPACK= NO_LAPACKE= NO_AFFINITY=1 USE_OPENMP= SANDYBRIDGE MAX_THREADS=3
pc file directory: /opt/arm64-builds/lib/pkgconfig
version: 0.3.21.dev
lapack:
detection method: pkgconfig
found: true
include directory: /opt/arm64-builds/include
lib directory: /opt/arm64-builds/lib
name: openblas
openblas configuration: USE_64BITINT= DYNAMIC_ARCH=1 DYNAMIC_OLDER= NO_CBLAS=
NO_LAPACK= NO_LAPACKE= NO_AFFINITY=1 USE_OPENMP= SANDYBRIDGE MAX_THREADS=3
pc file directory: /opt/arm64-builds/lib/pkgconfig
version: 0.3.21.dev
pybind11:
detection method: config-tool
include directory: unknown
name: pybind11
version: 2.10.4
Compilers:
c:
commands: cc
linker: ld64
name: clang
version: 14.0.3
c++:
commands: c++
linker: ld64
name: clang
version: 14.0.3
cython:
commands: cython
linker: cython
name: cython
version: 0.29.35
fortran:
commands: gfortran
linker: ld64
name: gcc
version: 12.2.0
pythran:
include directory: ../../../../../miniconda3/envs/dev3/lib/python3.10/site-packages/pythran
version: 0.13.1
Machine Information:
build:
cpu: aarch64
endian: little
family: aarch64
system: darwin
cross-compiled: false
host:
cpu: aarch64
endian: little
family: aarch64
system: darwin
Python Information:
path: /Users/andrew/miniconda3/envs/dev3/bin/python3.10
version: '3.10'
|
* deals with the 1 of the 2 Apple silicon test failures in scipygh-18732 * I was able to reproduce locally, but because it was flaky I needed `pytest-repeat` for consistent failure: `CC=clang CXX=clang++ FC=gfortran-12 python dev.py test -t scipy/signal/tests/test_filter_design.py::TestSOSFreqz::test_fs_param -- --count=200` * I verified that with the provided patch/`atol` bump, we consistently pass even with `2000` repeats [skip circle]
* deals with the second part of scipygh-18732 * I was able to reproduce locally, but this one was even flakier--analysis of patch success below, and possibly related to subtle low-level MacOS ARM issues upstream in numpy/numpy#23996 (NumPy `1.25.0`) * before this patch, 4 trials of this command with `pytest-repeat`: `CC=clang CXX=clang++ FC=gfortran-12 python dev.py test -t scipy/signal/tests/test_signaltools.py::TestDecimate::test_complex_iir_dlti -- --count=20000` `81 failed, 19919 passed in 43.68s` `81 failed, 19919 passed in 43.79s` `20000 passed in 33.24s` `43 failed, 19957 passed in 35.17s` * after this patch, 6 trials of the same command: `20000 passed in 30.44s` `20000 passed in 33.08s` `20000 passed in 33.46s` `20000 passed in 33.45s` `20000 passed in 33.56s` `20000 passed in 33.30s` [skip circle]
A few more issues blocked me during the release process today on MacOS ARM64 Cirrus CI: https://cirrus-ci.com/build/6410513341480960
Actual test failures appear to be moderate tolerance issues, presumably just showing up now because of NumPy
1.25.0
being in the mix for the final tests in the release-specific Cirrus CIcibuildwheel
workflows:There are also quite a large number of warnings, that perhaps won't cause failures on their own, mostly like this:
The text was updated successfully, but these errors were encountered: