[cctbxbb] Boost Python 1.56
Marcin Wojdyr
wojdyr at gmail.com
Thu Feb 11 10:50:09 PST 2016
Hi Aaron,
yes, I didn't run test yesterday. I did run it today and at first 20 failed.
It's mostly this problem:
https://github.com/boostorg/python/issues/56
Other failures I got were probably not related to the boost update.
I've committed 3 fixes.
I started looking at 4th, which was:
File "/home/marcin/dials/modules/cctbx_project/iotbx/mtz/tst_ext.py",
line 1093, in exercise_modifiers
assert approx_equal(v, [-1]*4)
but I must leave now. Does this test also fail for you?
Marcin
On Wed, Feb 10, 2016 at 11:20 PM, Aaron Brewster <asbrewster at lbl.gov> wrote:
> Hi Marcin, thanks for the detective work. I've reproduced your results on
> Centos 7 using the cctbx bootstrap script (not sure from your email if
> that's what you used):
>
> svn export
> svn://svn.code.sf.net/p/cctbx/code/trunk/libtbx/auto_build/bootstrap.py
> python bootstrap.py --builder=dials hot update
> cd modules
> mv boost oldboost
> wget
> https://sourceforge.net/projects/boost/files/latest/download?source=files -O
> ./boost_1_60_0.tar.gz
> mv boost_1_60_0 boost
> vi boost/boost/rational.hpp (revert
> https://github.com/boostorg/rational/commit/5fddb3f889cd2a2fe59cdcae182f8b3f78ec31e3,
> this is necessary to compile)
> cd ..
> python bootstrap.py --builder=dials base build
>
> I then exercised the cctbx regression tests:
>
> source build/setpaths.sh
> mkdir tests
> cd tests
> cctbx_regression.test_nightly
>
> 17 tests failed, but I think they are fixable, most of them being interface
> changes in boost. I ran the tests again after restoring the original boost
> and recompiling. There were still 3 errors, but I need to verify those
> errors from a completely clean system that hadn't seen boost 1.60 before I'd
> think they were real.
>
> Anyhow the next step is to track down those 17 errors, fix them, then try
> the whole process again with the phenix builder and the full set of dials
> and phenix tests.
>
> -Aaron
>
>
>
>
>
> On Wed, Feb 10, 2016 at 10:45 AM, Marcin Wojdyr <wojdyr at gmail.com> wrote:
>>
>> To help a bit I went through compilation of cctbx/Dials with Boost 1.60.
>>
>> There is a lot of warnings from Boost.Python,
>> about deprecated header:
>>
>> https://github.com/boostorg/python/commit/0a4c76b9ac16974d7d4f164cf6179095217a981a
>> and unused typedefs. After adding -Wno-unused-local-typedefs and
>> removing the deprecated #include almost all warnings are gone.
>>
>> Then there is one error, copy&pasted below.
>> I tracked it down to this change in boost.rational:
>>
>> https://github.com/boostorg/rational/commit/5fddb3f889cd2a2fe59cdcae182f8b3f78ec31e3
>> I just reverted this change to finish compilation, but I suppose a
>> proper fix is also not difficult.
>>
>> I don't know if this is relevant, but I'm using Boost compiled outside of
>> cctbx.
>> I'm attaching a patch that we use in CCP4 for this. It's based on
>> Debian/Gentoo patches.
>>
>> cheers,
>> Marcin
>>
>>
>> g++ -o boost_adaptbx/rational_ext.o -c
>> -I/home/marcin/dials/boost_160/include -Wno-unused-local-typedefs
>> -DBOOST_PYTHON_MAX_BASES=2 -fPIC -fno-strict-aliasing -w -DNDEBUG -O3
>> -ffast-math -DBOOST_ALL_NO_LIB -I/home/marcin/dials/build/include
>> -I/home/marcin/miniconda2/include/python2.7
>> /home/marcin/dials/modules/cctbx_project/boost_adaptbx/rational_ext.cpp
>> In file included from
>>
>> /home/marcin/dials/boost_160/include/boost/preprocessor/iteration/detail/iter/forward1.hpp:47:0,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/detail/invoke.hpp:63,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/detail/caller.hpp:16,
>> from
>>
>> /home/marcin/dials/boost_160/include/boost/python/object/function_handle.hpp:8,
>> from
>>
>> /home/marcin/dials/boost_160/include/boost/python/converter/arg_to_python.hpp:19,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/call.hpp:15,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/object_core.hpp:14,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/args.hpp:25,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/make_function.hpp:11,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/def.hpp:11,
>> from
>> /home/marcin/dials/modules/cctbx_project/boost_adaptbx/rational_ext.cpp:2:
>> /home/marcin/dials/boost_160/include/boost/python/detail/invoke.hpp:
>> In instantiation of 'PyObject*
>> boost::python::detail::invoke(boost::python::detail::invoke_tag_<false,
>> true>, const RC&, F&, TC&) [with RC =
>>
>> boost::python::detail::specify_a_return_value_policy_to_wrap_functions_returning<const
>> int&>; F = const int& (boost::rational<int>::*)()const; TC =
>> boost::python::arg_from_python<boost::rational<int>&>; PyObject =
>> _object]':
>>
>> /home/marcin/dials/boost_160/include/boost/python/detail/caller.hpp:223:13:
>> required from 'PyObject*
>> boost::python::detail::caller_arity<1u>::impl<F, Policies,
>> Sig>::operator()(PyObject*, PyObject*) [with F = const int&
>> (boost::rational<int>::*)()const; Policies =
>> boost::python::default_call_policies; Sig = boost::mpl::vector2<const
>> int&, boost::rational<int>&>; PyObject = _object]'
>>
>> /home/marcin/dials/boost_160/include/boost/python/object/py_function.hpp:38:33:
>> required from 'PyObject*
>>
>> boost::python::objects::caller_py_function_impl<Caller>::operator()(PyObject*,
>> PyObject*) [with Caller = boost::python::detail::caller<const int&
>> (boost::rational<int>::*)()const,
>> boost::python::default_call_policies, boost::mpl::vector2<const int&,
>> boost::rational<int>&> >; PyObject = _object]'
>>
>> /home/marcin/dials/modules/cctbx_project/boost_adaptbx/rational_ext.cpp:231:1:
>> required from here
>> /home/marcin/dials/boost_160/include/boost/python/detail/invoke.hpp:88:90:
>> error: no match for call to '(const
>>
>> boost::python::detail::specify_a_return_value_policy_to_wrap_functions_returning<const
>> int&>) (const int&)'
>> return rc( (tc().*f)(BOOST_PP_ENUM_BINARY_PARAMS_Z(1, N, ac, ()
>> BOOST_PP_INTERCEPT)) );
>>
>> ^
>> In file included from
>>
>> /home/marcin/dials/boost_160/include/boost/python/object/function_handle.hpp:8:0,
>> from
>>
>> /home/marcin/dials/boost_160/include/boost/python/converter/arg_to_python.hpp:19,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/call.hpp:15,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/object_core.hpp:14,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/args.hpp:25,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/make_function.hpp:11,
>> from
>> /home/marcin/dials/boost_160/include/boost/python/def.hpp:11,
>> from
>> /home/marcin/dials/modules/cctbx_project/boost_adaptbx/rational_ext.cpp:2:
>> /home/marcin/dials/boost_160/include/boost/python/detail/caller.hpp:
>> In instantiation of 'static const PyTypeObject*
>>
>> boost::python::detail::converter_target_type<ResultConverter>::get_pytype()
>> [with ResultConverter =
>>
>> boost::python::detail::specify_a_return_value_policy_to_wrap_functions_returning<const
>> int&>; PyTypeObject = _typeobject]':
>>
>> /home/marcin/dials/boost_160/include/boost/python/detail/caller.hpp:240:19:
>> required from 'static boost::python::detail::py_func_sig_info
>> boost::python::detail::caller_arity<1u>::impl<F, Policies,
>> Sig>::signature() [with F = const int&
>> (boost::rational<int>::*)()const; Policies =
>> boost::python::default_call_policies; Sig = boost::mpl::vector2<const
>> int&, boost::rational<int>&>]'
>>
>> /home/marcin/dials/boost_160/include/boost/python/object/py_function.hpp:48:35:
>> required from 'boost::python::detail::py_func_sig_info
>> boost::python::objects::caller_py_function_impl<Caller>::signature()
>> const [with Caller = boost::python::detail::caller<const int&
>> (boost::rational<int>::*)()const,
>> boost::python::default_call_policies, boost::mpl::vector2<const int&,
>> boost::rational<int>&> >]'
>>
>> /home/marcin/dials/modules/cctbx_project/boost_adaptbx/rational_ext.cpp:231:1:
>> required from here
>>
>> /home/marcin/dials/boost_160/include/boost/python/detail/caller.hpp:102:109:
>> error: 'struct
>> boost::python::detail::specify_a_return_value_policy_to_wrap_functions_returning<const
>> int&>' has no member named 'get_pytype'
>> return create_result_converter((PyObject*)0, (ResultConverter
>> *)0, (ResultConverter *)0).get_pytype();
>>
>> ^
>> scons: *** [boost_adaptbx/rational_ext.o] Error 1
>> scons: building terminated because of errors.
>>
>>
>>
>>
>>
>> On Wed, Feb 10, 2016 at 4:44 PM, Aaron Brewster <asbrewster at lbl.gov>
>> wrote:
>> > Hi Nick, Billy and I will work on this today,
>> >
>> > -Aaron
>> >
>> > On Wed, Feb 10, 2016 at 6:40 AM, Marcin Wojdyr <wojdyr at gmail.com> wrote:
>> >>
>> >> FWIW the latest version of Boost is 1.60.
>> >> CCP4 6.5 was using cctbx compiled with Boost 1.56
>> >> and CCP4 7.0 uses Boost 1.58.
>> >>
>> >> Marcin
>> >>
>> >> On Wed, Feb 10, 2016 at 2:17 PM, Nicholas Sauter <nksauter at lbl.gov>
>> >> wrote:
>> >> > Nigel, Billy & Aaron,
>> >> >
>> >> > I completely endorse this move to Boost 1.56. Can we update our
>> >> > build?
>> >> >
>> >> > Nick
>> >> >
>> >> > Nicholas K. Sauter, Ph. D.
>> >> > Computer Staff Scientist, Molecular Biophysics and Integrated
>> >> > Bioimaging
>> >> > Division
>> >> > Lawrence Berkeley National Laboratory
>> >> > 1 Cyclotron Rd., Bldg. 33R0345
>> >> > Berkeley, CA 94720
>> >> > (510) 486-5713
>> >> >
>> >> > On Wed, Feb 10, 2016 at 2:41 PM, Luc Bourhis <luc_j_bourhis at mac.com>
>> >> > wrote:
>> >> >>
>> >> >> Hi,
>> >> >>
>> >> >> I have improvements to the smtbx on their way to be committed which
>> >> >> require Boost version 1.56. This is related to Boost.Threads, whose
>> >> >> support
>> >> >> I re-activated a few months ago on Nick’s request. I need the
>> >> >> function
>> >> >> boost::thread::physical_concurrency which returns the number of
>> >> >> physical
>> >> >> cores on the machine, as opposed to virtual cores when
>> >> >> hyperthreading
>> >> >> is
>> >> >> enabled (which it is by default on any Intel machine). That function
>> >> >> is
>> >> >> not
>> >> >> available in Boost 1.55 which is the version currently used in the
>> >> >> nightly
>> >> >> tests: it appeared in 1.56.
>> >> >>
>> >> >> So, would it be possible to move to Boost 1.56? Otherwise, I will
>> >> >> need
>> >> >> to
>> >> >> backport that function. Not too difficult but not thrilling.
>> >> >>
>> >> >> Best wishes,
>> >> >>
>> >> >> Luc
>> >> >>
>> >> >>
>> >> >> _______________________________________________
>> >> >> cctbxbb mailing list
>> >> >> cctbxbb at phenix-online.org
>> >> >> http://phenix-online.org/mailman/listinfo/cctbxbb
>> >> >
>> >> >
>> >> >
>> >> > _______________________________________________
>> >> > cctbxbb mailing list
>> >> > cctbxbb at phenix-online.org
>> >> > http://phenix-online.org/mailman/listinfo/cctbxbb
>> >> >
>> >>
>> >> _______________________________________________
>> >> cctbxbb mailing list
>> >> cctbxbb at phenix-online.org
>> >> http://phenix-online.org/mailman/listinfo/cctbxbb
>> >
>> >
>> >
>> > _______________________________________________
>> > cctbxbb mailing list
>> > cctbxbb at phenix-online.org
>> > http://phenix-online.org/mailman/listinfo/cctbxbb
>> >
>>
>> _______________________________________________
>> cctbxbb mailing list
>> cctbxbb at phenix-online.org
>> http://phenix-online.org/mailman/listinfo/cctbxbb
>>
>
>
> _______________________________________________
> cctbxbb mailing list
> cctbxbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/cctbxbb
>
More information about the cctbxbb
mailing list