Robert X
Robert X

Reputation: 21

Issues with pipenv install on M1 Mac for python dependencies

I'm trying to set up a local repository that was originally created on intel-based macs. The repository uses pyenv and pipenv to manage python dependencies. Whenever I try to run pipenv install, I get the following error:

ERROR:pip.subprocessor:[present-rich] pip subprocess to install build dependencies exited with 1
[ResolutionFailure]:   File "/Users/xur19/.pyenv/versions/3.8.10/lib/python3.8/site-packages/pipenv/resolver.py", line 782, in _main
[ResolutionFailure]:       resolve_packages(
[ResolutionFailure]:   File "/Users/xur19/.pyenv/versions/3.8.10/lib/python3.8/site-packages/pipenv/resolver.py", line 730, in resolve_packages
[ResolutionFailure]:       results, resolver = resolve(
[ResolutionFailure]:   File "/Users/xur19/.pyenv/versions/3.8.10/lib/python3.8/site-packages/pipenv/resolver.py", line 709, in resolve
[ResolutionFailure]:       return resolve_deps(
[ResolutionFailure]:   File "/Users/xur19/.pyenv/versions/3.8.10/lib/python3.8/site-packages/pipenv/utils/resolver.py", line 1105, in resolve_deps
[ResolutionFailure]:       results, hashes, markers_lookup, resolver, skipped = actually_resolve_deps(
[ResolutionFailure]:   File "/Users/xur19/.pyenv/versions/3.8.10/lib/python3.8/site-packages/pipenv/utils/resolver.py", line 894, in actually_resolve_deps
[ResolutionFailure]:       resolver.resolve()
[ResolutionFailure]:   File "/Users/xur19/.pyenv/versions/3.8.10/lib/python3.8/site-packages/pipenv/utils/resolver.py", line 689, in resolve
[ResolutionFailure]:       raise ResolutionFailure(message=str(e))
[pipenv.exceptions.ResolutionFailure]: Warning: Your dependencies could not be resolved. You likely have a mismatch in your sub-dependencies.
  You can use $ pipenv install --skip-lock to bypass this mechanism, then run $ pipenv graph to inspect the situation.
  Hint: try $ pipenv lock --pre if it is a pre-release dependency.
ERROR: pip subprocess to install build dependencies exited with 1

The repository has an existing Pipfile and Pipfile.lock

My steps for setup are:

brew install pyenv
pyenv install 3.8.7
pyenv global 3.8.7
echo -e 'if command -v pyenv 1>/dev/null 2>&1; then\n  eval "$(pyenv init)"\nfi' >> ~/.zshrc
brew install python
pip install pipenv
pipenv shell
pipenv install

My Mac is an M1 mac and I've tried the following:

I'm not sure what else to try - was wondering if anyone had advice on how to tackle this. For context, here is my Pipfile.

[[source]]
name = "pypi"
url = "https://pypi.org/simple"
verify_ssl = true

[dev-packages]
autopep8 = "*"
flake8 = "*"
black = "==20.8b1"

[packages]
celery = "==4.4.7"
channels = "==3.0.3"
channels-redis = "==3.2.0"
#cx_Oracle = "==8.2.1" # Django DATABASE
Django = "==2.2.24"
django-allauth = "==0.44.0"
django-celery-beat = "==2.2.0"
django-cors-headers = "==3.7.0"
django-enumfield = "==1.5"
django-environ = "==0.4.5"
django-extensions = "==3.1.2"
django-filter = "==2.1.0"
django-graphql-jwt = "==0.3.1"
django-graphql-persist = "==0.1.1"
django-jet = "==1.0.8"
django-redis = "==4.12.1"
django-reversion = "==3.0.9"
django-reversion-compare = "==0.14.0"
django-measurement = "==3.2.3"
django-jsoneditor = "==0.1.6"
django-storages = {extras = ["google"],version = "==1.11.1"}
flower = "==0.9.7"
future = "==0.18.2"
google-auth = "==1.29.0"
google-auth-oauthlib = "==0.4.4"
google-cloud-storage = "==1.37.1"
google-cloud-pubsub = "==2.4.0"
graphene = "==2.1.7"
graphql-core = "==2.2"
graphene-django = "==2.7"
graphene-file-upload = "==1.3.0"
graphql-relay = "==2"
graphql-server-core = "==1.1.1"
gunicorn = "==19.9.0"
ijson = "==3.1.4"
ldap3 = "==2.9"
matplotlib = "==3.4.1"
munch = "==2.5.0"
numpy = "==1.20.2"
openpyxl = "==3.0.7"
pandas = "==1.2.4"
pillow = "==8.2.0"
psycopg2-binary = "==2.8.6"
pyhumps = "==1.6.1"
pyjwt = "==1.7.1"
pymongo = "==3.11.3"
pyteomics = "==4.4.2"
python-pptx = "==0.6.18"
pytz = "==2018.9"
redis = "==3.2.0" # Doesn't get installed by django-redis
requests = "==2.21.0"
tqdm = "==4.60.0"
xlrd = "==2.0.1"
jinja2 = "==3.0.2"
premailer = "==3.10.0"
docxtpl = "==0.14.2"
inflection = "==0.4.0"
sympy = "==1.9"

[requires]
python_version = "3.7"

Upvotes: 2

Views: 5230

Answers (2)

Osama Remlawi
Osama Remlawi

Reputation: 2990

In my case I did the following:

  1. arch -arm64 brew install libffi
  2. export LDFLAGS="-L/opt/homebrew/opt/libffi/lib"
  3. export CPPFLAGS="-I/opt/homebrew/opt/libffi/include"
  4. CONFIGURE_OPTS='--with-system-ffi' pyenv install 3.7.10 pyenv local 3.7.10
  5. pipenv install

And then it worked!

Upvotes: 0

theherk
theherk

Reputation: 7546

I'm not 100% sure, since I was dealing with 3.7.10, but my problem looked suspiciously similar. Based on this issue on pipenv, I was led to this article, which led me to a solution.

The summarized solution from that article is as follows.

I'm not sure you need to install libffi from brew and make these exports, but I was in a morass so it may help.

brew install libffi
export LDFLAGS="-L/opt/homebrew/opt/libffi/lib"
export CPPFLAGS="-I/opt/homebrew/opt/libffi/include"

This part you definitely need to do.

CONFIGURE_OPTS='--with-system-ffi' pyenv install 3.7.10
pyenv local 3.7.10

Then edit the file that causes the MemoryError. For 3.7.10 ~/.pyenv/versions/3.7.10/lib/python3.7/ctypes/__init__.py. Comment or remove the last line of this function.

def _reset_cache():
    _pointer_type_cache.clear()
    _c_functype_cache.clear()
    if _os.name == "nt":
        _win_functype_cache.clear()
    # _SimpleCData.c_wchar_p_from_param
    POINTER(c_wchar).from_param = c_wchar_p.from_param
    # _SimpleCData.c_char_p_from_param
    POINTER(c_char).from_param = c_char_p.from_param
    _pointer_type_cache[None] = c_void_p
    # XXX for whatever reasons, creating the first instance of a callback
    # function is needed for the unittests on Win64 to succeed.  This MAY
    # be a compiler bug, since the problem occurs only when _ctypes is
    # compiled with the MS SDK compiler.  Or an uninitialized variable?
    CFUNCTYPE(c_int)(lambda: None) # comment or remove me

Then it worked for me. pipenv installed many things without issue.

Upvotes: -2

Related Questions