-
-
Notifications
You must be signed in to change notification settings - Fork 44
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[do not merge] meson catanh sign issue #304
[do not merge] meson catanh sign issue #304
Conversation
Signed-off-by: Marcel Bargull <marcel.bargull@udo.edu>
Signed-off-by: Marcel Bargull <marcel.bargull@udo.edu>
Signed-off-by: Marcel Bargull <marcel.bargull@udo.edu>
…nda-forge-pinning 2023.11.07.08.59.34 Signed-off-by: Marcel Bargull <marcel.bargull@udo.edu>
Hi! This is the friendly automated conda-forge-linting service. I just wanted to let you know that I linted all conda-recipes in your PR ( |
f821db1
to
4ae5464
Compare
Signed-off-by: Marcel Bargull <marcel.bargull@udo.edu>
4ae5464
to
2620333
Compare
Upstream fix at numpy/numpy#25093 . |
We can backport the patch if you want? Though I think a 1.26.2 is on the way reasonably soon, because of issues with #302 |
Yep, saw that too.
No strong opinion from my side. (TBH, I just fixed, but did not assess the impact of the issue.) |
Checklist
0
(if the version changed)conda-smithy
(Use the phrase@conda-forge-admin, please rerender
in a comment in this PR for automated rerendering)On the
numba
feedstock we encountered a regression for glibc=2.17 vianumba
's test suite.See conda-forge/numba-feedstock#128 (comment) .
Essentially, for
1.26.0
, arctanh(-i) gives a "positive zero" real part -- consistent with whatcatanh
from glibc=2.17, but different from NumPy's own implementation andcatanh
from glibc>2.17.I tracked that down to be related to builds via
meson
, i.e., the current1.26
but also1.25
show the same behavior if built viameson
.I'll open an issue upstream laterand let this build here with downloadable artifacts for inspection/comparison.[EDIT]: Upstream issue: numpy/numpy#25087
I don't know why the
meson
-based builds show that behavior.As expected by the observed behavior, one can see that the extension modules link to glibc's versions (not just
catanh
, but that is our example here):readelf --dyn-syms 'numpy/linalg/_umath_linalg.cpython-311-x86_64-linux-gnu.so' | grep -o '[^ ]*catanh[^ ]*'
:with_meson == "yes"
:with_meson == "no"
: