-
-
Notifications
You must be signed in to change notification settings - Fork 479
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
cysignals, cvxopt fail to build in conda #32087
Comments
Attachment: conda-list.txt Attachment: config.status.gz |
Attachment: config.log |
This comment has been minimized.
This comment has been minimized.
comment:1
Attachment: pynac-0.7.27.p8.log |
comment:3
Any idea where the |
comment:4
From the build log:
|
comment:5
pynac gets its flags from python, see https://github.com/pynac/pynac/blob/master/configure.ac#L56 |
comment:6
Note that for this build I explicitely used conda's Python with
|
comment:7
A good question for @isuruf then. Not recommended for use unless you really know what you are doing or you are compiling for a BlueGene [now completely dead, I personally shutdown two of those in 2016]. |
comment:8
|
comment:9
Got
And the output of |
comment:10
Attachment: conda-list-p.txt Thanks. Can you post the output of |
comment:11
|
comment:12
Sorry, I can't reproduce this. All I know is that the flag Can you try printing that env variable |
comment:13
Nvm, I can reproduce this. I'll fix this soon. |
comment:14
Upstream issue conda-forge/python-feedstock#474. You can downgrade your setuptools version as a workaround. |
comment:15
Replying to @isuruf:
Downgrading to setuptools 49.5.0 was only possible together with downgrading Python to 3.8. I am trying to build sage with that new setup now. |
comment:16
Same problem
Note that sage installs its own version of setuptools (version 56.2.0). |
comment:17
Right. Until upstream issue is solved, we have to downgrade setuptools to 49.* in sage itself. |
comment:18
Perhaps it's good enough to just change the value of |
comment:19
Yes, that seems reasonable until we get the upstream issue solved. |
comment:20
Replying to @mkoeppe:
How could I do that? |
comment:22
Try this New commits:
|
Commit: |
comment:23
Thanks |
comment:25
Can you attach the pari log as well? |
comment:26
Attachment: pari-2.13.1.log available as pari-2.13.1.log. Thanks for your help. |
comment:27
This may be a good moment to mention sagemath/cysignals#125, which needs help |
Changed branch from u/mkoeppe/pynac_fails_to_build_in_conda to u/isuruf/pari-conda |
comment:29
The issue was what @mkoeppe said about LDFLAGS in python overriding sage built version. For now, I've updated conda in pari to use pthread version for pari>=2.13. Can you check that this works? |
comment:32
It looks like this may be breaking |
comment:33
@isuruf: with your branch, sage did build in conda. Thanks. Though I got into trouble when compiling the documentation
|
comment:37
pynac issue will disappear with package removal in #32387 |
Dependencies: #32387 |
comment:38
Outdated; in 9.5.beta8 (with #32113), the packages build OK. |
comment:39
Replying to @videlec:
}}} I've opened #33056 for this. |
Sage sources 9.4.beta3
Depends on #32387
CC: @isuruf
Component: packages: standard
Branch/Commit: u/isuruf/pari-conda @
3592860
Issue created by migration from https://trac.sagemath.org/ticket/32087
The text was updated successfully, but these errors were encountered: