Closed strogdon closed 1 year ago
I meant to write something about gap-core
. At least it should have been much easier to deal with than previous gap upgrades. This I haven't seen before. I'll have to check whether I missed something in the last push.
Surrounding gap-core
I had to eventually do an emerge @preserved-rebuild
(or perhaps it was updating world?) and giac
had to be rebuilt which couldn't be done until after I manually rebuilt gmp-ecm
. I'm still not sure what was up.
Did it come with something more than gap-core
being a blocker? I only had to do emerge -C gap-core atlasrep
and it was all good.
I'm sure it was unrelated to gap-core
. I did emerge -C gap-core
and later on updated atlasrep
. After removing gap-core
a preserved-rebuild
was required which I deferred until after I tried to build sage. Perhaps this was my problem.
I cannot reproduce the issue. What version of cython are you using?
# eix -I cython
[I] dev-python/cython
Available versions: 0.29.32^t ~0.29.33^t {doc emacs test PYTHON_TARGETS="pypy3 python3_9 python3_10 python3_11"}
Installed versions: 0.29.32^t(11:28:26 PM 01/29/2023)(-doc -emacs -test PYTHON_TARGETS="python3_10 python3_11 -pypy3 -python3_9")
Homepage: https://cython.org/ https://github.com/cython/cython/ https://pypi.org/project/Cython/
Description: A Python to C compile
Ouch, an issue with giac and gmp-ecm while building documentation. This is new. It looks like the latest update of gmp may have broken some compatibility without a version bump. There is a missing symbol.
For cython, can you try to upgrade, I may have to raise the dependency.
Same problem with cython-0.29.33
?
Are both wrapper_rdf.pxd
and Wrapper_rdf.pxd
to be found under work/sage-9999
? If so, then I don't have them.
They should be generated by a call to sage_setup.
The vbraun
branch fails here as in https://github.com/cschwan/sage-on-gentoo/issues/729#issue-1563022184
Could I have a full log? You can send it to me privately> These seems a bit strange, there is probably a dependency missing or the wrong level that interfere with autogen, but I need to see what happen at that level.
log file sent. Let me know if not received.
Got it. And it is a change that was probably not in the previous beta. It didn't happen to me because it was finding it from a previous install. Which is why I should test install from scratch more often. After https://github.com/sagemath/sage/commit/2f4b0539a17cf2dd9ff53ccb85a3b7f8f218e6a8 autogen_all()
is not run anymore for the build_ext
target, which is what distutils-r1 does during compile. Fix coming.
Ready for testing.
Sage now builds. I didn't see an issue with the documentation. It seems to build just fine.
For the doc I just had to rebuild gmp-ecm (and giac for safety).
After dealing with
gap-core
and significant system updates I finally seeI haven't see this before.