Bug#878496: python-escript FTBFS on 32bit with gcc 7

Juhani Numminen juhaninumminen0 at gmail.com
Fri Dec 8 21:52:01 UTC 2017


Control: tags -1 pending

On Sat, 14 Oct 2017 04:42:54 +0300 Adrian Bunk <bunk at debian.org> wrote:
> Source: python-escript
> Version: 5.0-4
> Severity: serious
> 
> https://buildd.debian.org/status/package.php?p=python-escript&suite=sid
> 
> ...
> escriptcore/src/DataFactory.cpp: In function 'escript::Data escript::VectorFromObj(boost::python::api::object, const escript::FunctionSpace&, bool)':
> escriptcore/src/DataFactory.cpp:49:41: error: '*((void*)&<anonymous> +12)' may be used uninitialized in this function [-Werror=maybe-uninitialized]
>          double v = bp::extract<double>(o);
>                                          ^
> escriptcore/src/DataFactory.cpp: In function 'escript::Data escript::TensorFromObj(boost::python::api::object, const escript::FunctionSpace&, bool)':
> escriptcore/src/DataFactory.cpp:75:41: error: '*((void*)&<anonymous> +12)' may be used uninitialized in this function [-Werror=maybe-uninitialized]
>          double v = bp::extract<double>(o);
>                                          ^
> escriptcore/src/DataFactory.cpp: In function 'escript::Data escript::Tensor3FromObj(boost::python::api::object, const escript::FunctionSpace&, bool)':
> escriptcore/src/DataFactory.cpp:99:41: error: '*((void*)&<anonymous> +12)' may be used uninitialized in this function [-Werror=maybe-uninitialized]
>          double v = bp::extract<double>(o);
>                                          ^
> escriptcore/src/DataFactory.cpp: In function 'escript::Data escript::Tensor4FromObj(boost::python::api::object, const escript::FunctionSpace&, bool)':
> escriptcore/src/DataFactory.cpp:123:41: error: '*((void*)&<anonymous> +12)' may be used uninitialized in this function [-Werror=maybe-uninitialized]
>          double v = bp::extract<double>(o);
>                                          ^

I'm adding the "pending" tag since there seems to be a fix committed 
already, at
https://anonscm.debian.org/git/debian-science/packages/python-escript.git

Joel, if I am not mistaken and your commit fixes the build failure, please
remember to close this bug (#878496) through debian/changelog.


Regards,
Juhani Numminen



More information about the debian-science-maintainers mailing list