Re: broken python packages?

Daniel,
     Unfortunately, it doesn't seem that the BLDSHARED environmental
is the cause as reverting that change doesn't eliminate the build
failures of numpy 1.13.3 as fink-bld...

Could not locate executable pgfortran
don't know how to compile Fortran code on platform 'posix'
C compiler: gcc -Wno-unused-result -Wsign-compare -Wunreachable-code
-DNDEBUG -g -fwrapv -fwrapv -O3 -Wall -Wstrict-prototypes
-I/sw/include

compile options: '-Inumpy/core/src/private -Inumpy/core/src
-Inumpy/core -Inumpy/core/src/npymath -Inumpy/core/src/multiarray
-Inumpy/core/src/umath -Inumpy/core/src/npysort
-I/sw/include/python3.6m -c'
gcc: _configtest.c

failure.
removing: _configtest.c _configtest.o
Traceback (most recent call last):
  File "setup.py", line 392, in <module>
    setup_package()
  File "setup.py", line 384, in setup_package
    setup(**metadata)
  File "/sw/src/fink.build/numpy-py36-1.13.3-1/numpy-1.13.3/numpy/distutils/core.py",
line 169, in setup
    return old_setup(**new_attr)
  File "/sw/lib/python3.6/distutils/core.py", line 148, in setup
    dist.run_commands()
  File "/sw/lib/python3.6/distutils/dist.py", line 955, in run_commands
    self.run_command(cmd)
  File "/sw/lib/python3.6/distutils/dist.py", line 974, in run_command
    cmd_obj.run()
  File "/sw/src/fink.build/numpy-py36-1.13.3-1/numpy-1.13.3/numpy/distutils/command/build.py",
line 47, in run
    old_build.run(self)
  File "/sw/lib/python3.6/distutils/command/build.py", line 135, in run
    self.run_command(cmd_name)
  File "/sw/lib/python3.6/distutils/cmd.py", line 313, in run_command
    self.distribution.run_command(command)
  File "/sw/lib/python3.6/distutils/dist.py", line 974, in run_command
    cmd_obj.run()
  File "/sw/src/fink.build/numpy-py36-1.13.3-1/numpy-1.13.3/numpy/distutils/command/build_src.py",
line 148, in run
    self.build_sources()
  File "/sw/src/fink.build/numpy-py36-1.13.3-1/numpy-1.13.3/numpy/distutils/command/build_src.py",
line 159, in build_sources
    self.build_library_sources(*libname_info)
  File "/sw/src/fink.build/numpy-py36-1.13.3-1/numpy-1.13.3/numpy/distutils/command/build_src.py",
line 294, in build_library_sources
    sources = self.generate_sources(sources, (lib_name, build_info))
  File "/sw/src/fink.build/numpy-py36-1.13.3-1/numpy-1.13.3/numpy/distutils/command/build_src.py",
line 377, in generate_sources
    source = func(extension, build_dir)
  File "numpy/core/setup.py", line 674, in get_mathlib_info
    raise RuntimeError("Broken toolchain: cannot link a simple C program")
RuntimeError: Broken toolchain: cannot link a simple C program
### execution of /tmp/fink.JpadZ failed, exit code 1
### execution of /tmp/fink.kTHyz failed, exit code 1
Removing runtime build-lock...
Removing build-lock package...

Any suggestions on how to debug this as upstream seems disinterested.
      Jack

On Sun, Oct 29, 2017 at 10:09 AM, Jack Howarth
<[email protected]> wrote:
> Daniel,
>      Both current upstream numpy 1.13.1 and scipy 1.0.0 are no longer
> buildable as fink-bld...
>
> https://github.com/scipy/scipy/issues/7979
>
> Upstream insists this is breakage in how we are building our python
> compiler packages on Darwin. While I was originally dubious about
> their argument, I suspect now that they are correct. The failures can
> be reproduced using the packaging that I have placed on fink tracker
> at...
>
> https://sourceforge.net/p/fink/package-submissions/5028/
>
> and
>
> https://sourceforge.net/p/fink/package-submissions/5056/
>
> by changing the line...
>
> BuildAsNobody: false
>
> to
>
> BuildAsNobody: true
>
> The failures seem to be in the linkage tests of the clang compiler as
> now done in recent numpy and scipy. This issue doesn't occur on
> MacPorts which is already on the newest upstream numpy and scipy
> releases (although they may be building as root since their builds
> default to using sandbox).
>      Comparing our python compiler package patches to those on
> MacPorts, the one that jumps out is...
>
> diff -ru Python-3.6.1.orig/Lib/sysconfig.py Python-3.6.1/Lib/sysconfig.py
> --- Python-3.6.1.orig/Lib/sysconfig.py  2017-03-21 02:32:38.000000000 -0400
> +++ Python-3.6.1/Lib/sysconfig.py       2017-04-29 11:40:11.000000000 -0400
> @@ -374,11 +374,6 @@
>          if hasattr(e, "strerror"):
>              msg = msg + " (%s)" % e.strerror
>          raise OSError(msg)
> -    # On AIX, there are wrong paths to the linker scripts in the Makefile
> -    # -- these paths are relative to the Python source, but when installed
> -    # the scripts are in another directory.
> -    if _PYTHON_BUILD:
> -        vars['BLDSHARED'] = vars['LDSHARED']
>
>      # There's a chicken-and-egg situation on OS X with regards to the
>      # _sysconfigdata module after the changes introduced by #15298:
>
> Do we really have to eliminate setting BLDSHARED to LDSHARED as this
> could potentially be something the numpy and scipy developers are
> relying on being set?
>              Jack

------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Fink-devel mailing list
[email protected]
List archive:
https://sourceforge.net/p/fink/mailman/fink-devel
Subscription management:
https://lists.sourceforge.net/lists/listinfo/fink-devel