54 Commits

Author SHA1 Message Date
sthen
d858b7cc2a Ports built with setuptools on python 3.10 don't need a RUN_DEPENDS on
setuptools (it's used as a package locator but importlib.metadata in
newer Python core or the external importlib_metadata are preferred).
So drop the RDEP in that case (it's still kept for py27) and bump
revisions.
2022-11-26 23:28:11 +00:00
sthen
dae9f91e31 bump for MODPY_DEFAULT_VERSION_3 change 2022-11-13 15:28:39 +00:00
naddy
ab45f39af6 drop RCS Ids 2022-03-11 19:36:11 +00:00
sthen
3fb7cdbcf6 bump REVISION for switch from Python 3.8 -> 3.9 2021-11-02 00:01:35 +00:00
pascal
7d63a0f078 Remove USE_LLD=No hack on i386. A fix for the R_386_GOTOFF bug has been
backported.

ok sthen@
2021-07-31 16:37:56 +00:00
sthen
d80c418015 regen PLISTs for python ports with .so files that end up renamed due to
the EXT_SUFFIX change in Python 3.8.7
2021-01-04 14:06:26 +00:00
daniel
31b0ea8cdb update to 1.5.3; a bug-fix release 2020-10-28 01:24:38 +00:00
daniel
9b8e165000 Update py-scipy to 1.5.2 which is now python3-only.
tweaks and ok bket@
2020-10-12 22:04:32 +00:00
sthen
d9cfe4113e bump REVISION; python 3 default changed to 3.8 2020-07-03 21:12:24 +00:00
jca
c9fa63c46c Use math/cblas instead of a bundled copy
scipy properly detects cblas now that libcblas.so has a dep on
libblas.so.  Register the dep on math/cblas, along with math/lapack,
because scipy directly links against those.  (py-numpy, a BUILD
and RUN_DEPENDS, also links against those but better be explicit).

Spotted by naddy@

While here, bytecompile all the installed .py files
2020-04-03 15:12:07 +00:00
daniel
add32f09f8 drop maintainer 2020-03-10 03:31:45 +00:00
daniel
fef0569685 Update to py-scipy 1.2.3. 2020-02-16 17:53:36 +00:00
daniel
f393e581f7 Update to py-scipy 1.2.2. 2019-08-18 01:19:36 +00:00
daniel
81e6976bfa PERMIT_PACKAGE_CDROM -> PERMIT_PACKAGE 2019-07-07 14:58:04 +00:00
sthen
42dc443195 Use ld.bfd to link math/py-scipy and telephony/asterisk on i386, like gcc/8
itself they fail with "has non-ABS relocation R_386_GOTOFF against symbol XXX"
with LLD.

Mention these ports in the comment in lang/gcc/8/Makefile.
2019-05-01 12:12:24 +00:00
naddy
3e042ba62d Bump all ports that depend on ports-gcc on the clang archs.
SYSTEM_VERSION didn't quite work out how we expected and it's
easier|safer to do it this way than fiddle with pkg_add now.
2019-04-28 21:08:26 +00:00
sthen
d7f0752227 bump all the py3 things, _SYSTEM_VERSION didn't quite work out how
we expected and it's easier|safer to do it this way than fiddle with
pkg_add now. thanks aja for update tests with a quick bulk.
2019-04-28 20:51:26 +00:00
daniel
0f2051a4d3 Repair PLIST. 2019-04-20 23:55:21 +00:00
daniel
8a1596e617 Update py-scipy to 1.2.1. 2019-04-20 21:35:01 +00:00
sthen
25f0e460f2 Add COMPILER lines to c++ ports which currently use the default. Adjust
some existing COMPILER lines with arch restrictions etc. In the usual
case this is now using "COMPILER = base-clang ports-gcc base-gcc" on
ports with c++ libraries in WANTLIB.

This is basically intended to be a noop on architectures using clang
as the system compiler, but help with other architectures where we
currently have many ports knocked out due to building with an unsuitable
compiler -

- some ports require c++11/newer so the GCC version in base that is used
on these archirtectures is too old.

- some ports have conflicts where an executable is built with one compiler
(e.g. gcc from base) but a library dependency is built with a different
one (e.g. gcc from ports), resulted in mixing incompatible libraries in the
same address space.

devel/gmp is intentionally skipped as it's on the path to building gcc -
the c++ library there is unused in ports (and not built by default upstream)
so intending to disable building gmpcxx in a future commit.
2018-10-24 14:27:57 +00:00
ajacoutot
09d389a6c2 Drop CPPFLAGS="-I${LOCALBASE}/include/amd", it's not needed since rev. 1.24,
3 years ago.

pointed by daniel@ (maintainer)
2018-08-15 14:55:22 +00:00
daniel
9298dadb3e Update py-scipy to 1.1.0. 2018-07-01 23:14:19 +00:00
daniel
45d5ea1b40 Prefer gfortran over g77 when both are present.
Build breakage pointed out by aja@
2018-06-26 11:25:41 +00:00
daniel
1d4a9acd9a Update scipy to 0.17.1 since newer lapack has removed certain symbols
which caused scipy 0.16.1 to stop working in certain situations.

This issue was reported by David Kalliecharan and is also documented at:
https://github.com/scipy/scipy/issues/5266

The failure looks like this:

In [9]: import scipy.interpolate as interp
python3.6:/usr/local/lib/python3.6/site-packages/scipy/linalg/_flapack.so: undefined symbol 'sgegv_'
python3.6:/usr/local/lib/python3.6/site-packages/scipy/linalg/_flapack.so: undefined symbol 'dgegv_'
python3.6:/usr/local/lib/python3.6/site-packages/scipy/linalg/_flapack.so: undefined symbol 'cgegv_'
python3.6:/usr/local/lib/python3.6/site-packages/scipy/linalg/_flapack.so: undefined symbol 'zgegv_'

While here also switch MASTER_SITES from sourceforge to pypi and switch from
g77 to gfortran.
2018-06-23 18:45:10 +00:00
sthen
5e964ab0df bump LIBCXX/LIBECXX/COMPILER_LIBCXX ports. 2017-07-26 22:45:14 +00:00
espie
8ac47fd9c6 use COMPILER_LIBCXX where applicable 2017-07-16 19:18:47 +00:00
sthen
bcbf44ab87 use LIBCXX 2017-04-10 11:46:18 +00:00
landry
ffe8bbd592 bump ports depending on python3 in math, misc and multimedia 2017-01-03 19:23:48 +00:00
daniel
55089d2f69 Update py-scipy to 0.16.1. Many thanks to Anton Kasimov for testing.
ok shadchin@
2016-04-25 08:33:23 +00:00
sthen
ee57a78901 bump REVISION for setuptools dep change. 2015-09-29 10:52:28 +00:00
ajacoutot
cc564cd15c umfpack, arpack and amd are not needed.
ok pirofti@
2015-07-20 23:08:19 +00:00
sthen
270db149c4 add some missing pthread to WANTLIB
(using modified portbump, thanks zhuk@)
2015-04-21 18:24:22 +00:00
daniel
aa8a4a5f1a Update py-scipy to 0.14.1 and take over maintainer from eric@ with his
permission.

ok benoit@
2015-01-09 14:40:27 +00:00
dcoppa
a2a0464ede Bye bye MODPY_BADEGGS 2013-10-03 16:37:15 +00:00
landry
3817610542 Mark as BROKEN-alpha: ICE in scipy/optimize/minpack/lmstr.f:462 2013-09-04 18:40:21 +00:00
espie
0662a4e9d6 PERMIT_* / REGRESS->TEST sweep 2013-03-11 11:20:26 +00:00
landry
27a2ddaa6b Mark as BROKEN-hppa, triggers ICE in scipy/special/specfun/specfun.f:5151 2012-09-29 07:23:11 +00:00
landry
9fde09d02b Mark as BROKEN on sparc64 since it never built there in ages; Triggers
an internal compiler error in scipy/optimize/minpack/qrsolv.f:189
2012-09-27 17:42:16 +00:00
rpointel
6dc8c2a4ce update scipy to 0.10.0.
ok eric@ (maintainer).
2011-11-15 14:45:46 +00:00
espie
e820071495 normalize pkgpath 2011-09-16 10:31:20 +00:00
jasper
eec0417831 bump after default python version switch 2011-06-08 17:31:05 +00:00
rpointel
114510e67c Unbreak after recent commit of Python.
ok landry@ jasper@.
2011-06-07 07:33:24 +00:00
eric
c9239c1402 update to 0.9.0
ok rpointel@ landry@
2011-05-27 10:10:19 +00:00
espie
bc8dc9adb1 new depends 2010-11-19 07:23:05 +00:00
steven
83c4da679d replace MODFORTRAN_WANTG77 with MODFORTRAN_COMPILER 2010-10-23 15:09:57 +00:00
sthen
56e01aaf81 - fix LIB_DEPENDS; math/amd should have been a LIB_DEPENDS here.
the scipy package used to work as umfpack also had math/amd as
LIB_DEPENDS but this was removed from umfpack and py-scipy wasn't
changed to match.

- sync WANTLIB, bump REVISION

broken packages noticed by thib@
2010-09-14 12:07:53 +00:00
fgsch
0915f5b894 Bump revision after python's default change. 2010-08-30 17:11:35 +00:00
steven
cedd0e2bb7 REVISION 2010-07-18 10:50:11 +00:00
naddy
f6c04155f3 bump after fortran move to ports and don't put the compiler into run depends 2010-05-28 15:11:51 +00:00
espie
18fe3c3a29 use fortran module
make sure we pick the amd includes (appear to affect gcc3 also)
2010-05-23 15:51:45 +00:00