Compiling python on OS X gives warning about compact unwind

[vajrasky]

BPO	22093
Nosy	@ronaldoussoren, @amauryfa, @abalkin, @ned-deily, @meadori, @vajrasky

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = None
closed_at = None
created_at = <Date 2014-07-28.04:35:02.834>
labels = []
title = 'Compiling python on OS X gives warning about compact unwind'
updated_at = <Date 2014-07-28.06:07:53.301>
user = 'https://github.com/vajrasky'

bugs.python.org fields:

activity = <Date 2014-07-28.06:07:53.301>
actor = 'ned.deily'
assignee = 'none'
closed = False
closed_date = None
closer = None
components = []
creation = <Date 2014-07-28.04:35:02.834>
creator = 'vajrasky'
dependencies = []
files = []
hgrepos = []
issue_num = 22093
keywords = []
message_count = 2.0
messages = ['224151', '224152']
nosy_count = 6.0
nosy_names = ['ronaldoussoren', 'amaury.forgeotdarc', 'belopolsky', 'ned.deily', 'meador.inge', 'vajrasky']
pr_nums = []
priority = 'normal'
resolution = None
stage = None
status = 'open'
superseder = None
type = None
url = 'https://bugs.python.org/issue22093'
versions = ['Python 3.5']

[vajrasky]

OS X version 10.9.4.

$ gcc --version
Configured with: --prefix=/Applications/Xcode.app/Contents/Developer/usr --with-gxx-include-dir=/usr/include/c++/4.2.1
Apple LLVM version 5.1 (clang-503.0.40) (based on LLVM 3.4svn)
Target: x86_64-apple-darwin13.3.0
Thread model: posix

This is the warning:

...
gcc -bundle -undefined dynamic_lookup build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/_ctypes.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/callbacks.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/callproc.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/stgdict.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/cfield.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/malloc_closure.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/darwin/dlfcn_simple.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/libffi_osx/ffi.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/libffi_osx/x86/darwin64.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/libffi_osx/x86/x86-darwin.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/libffi_osx/x86/x86-ffi_darwin.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/libffi_osx/x86/x86-ffi64.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/libffi_osx/powerpc/ppc-darwin.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/libffi_osx/powerpc/ppc-darwin_closure.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/libffi_osx/powerpc/ppc-ffi_darwin.o build/temp.macosx-10.9-x86_64-3.5-pydebug/Users/sky/Code/python/cpython/Modules/_ctypes/libffi_osx/powerpc/ppc64-darwin_closure.o -L/usr/local/lib -o build/lib.macosx-10.9-x86_64-3.5-pydebug/_ctypes.so
ld: warning: could not create compact unwind for _ffi_call_unix64: does not use RBP or RSP based frame

Python build finished successfully!

[ned-deily]

That warning has been around for some time. The question is: is it something to be concerned about? I don't know of any ctypes test failures on OS X. There have been various discussions of the warning in other (non-Python) contexts, for example, https://sourceware.org/ml/libffi-discuss/2011/msg00109.html.

[iritkatriel]

@ned-deily IIUC this is not something we can fix.

[ronaldoussoren]

For recent versions of Python (IIRC 3.10 and later) we've dropped the vendored copy of libffi and use the system install instead. That makes this issue out of date, it is not relevant for any version receiving regular bug fixes and this is not a security issue.

I propose closing the issue.

[iritkatriel]

Thanks @ronaldoussoren .