Sage build fails on Ubuntu 25.04

[Jon Skanes]

Sorry if this has been discussed before, if so, kindly point me in the right direction, please.

Installed Sage source from git master branch at version 10.6.

Scrubbed the environment.

Hand installed, where needed, every Ubuntu package mentioned anywhere sensible I could find.

I noticed MANY system packages were not used by configure.

Thanks,

Jon Skanes

[Dima Pasechnik]

./configure without `--enable-system-site-packages` will ignore most Python packages on its list of dependencies.

`./configure --enable-system-site-packages` wil try to use the latter, but it might run into a version conflict.

(because most Linux distros don't have the complete range of these available, so some will have to be built from source, etc)

Yes, I find this state of affairs less than ideal, too.

We now also have an option to only build the main component, sagelib.

(but then you'll have to provide the needed dependencies).

Please see https://doc.sagemath.org/html/en/installation/meson.html

--
You received this message because you are subscribed to the Google Groups "sage-devel" group.
To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/sage-devel/e6afd3dc-9672-48df-a1e0-ab7ba2668d1an%40googlegroups.com.

[Tobia...@gmx.de]

You can try https://github.com/sagemath/sage/pull/40071 by installing all system packages and uv, and then run:

uv pip install meson-python "cysignals >=1.11.2, != 1.12.0" "cython >=3.0, != 3.0.3, != 3.1.0" "gmpy2 ~=2.1.b999" memory_allocator "numpy >=1.25" setuptools jinja2

uv pip install --no-build-isolation "cypari2 >=2.2.1"

uv sync --frozen --no-build-isolation

It's still rough around the edges, especially regarding support of maxima/ecl and expect a few failing doctests due to missing external dependencies. But otherwise it should work and take only a fraction of the time to install via `make`.

[chris wuthrich]

Upgrading to ubuntu 25.04 also broke sage build (both 10.6 and 10.7.beta3) for me. Logs attached. 

Tobias' method via meson worked.

Chris

[chris wuthrich]

The problem is still present with beta4.

(And with the meson approach I didn't get very far either, no notebook for instance.)

Chris

[Dima Pasechnik]

On Sat, May 17, 2025 at 10:43 AM chris wuthrich
<christian...@gmail.com> wrote:
>
> Upgrading to ubuntu 25.04 also broke sage build (both 10.6 and 10.7.beta3) for me. Logs attached.

This is basically the error in
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1093321
for brial/polybori

g++: error: @SIMD_CFLAGS@: linker input file not found: No such file
or directory

> To view this discussion visit https://groups.google.com/d/msgid/sage-devel/4f6bf0c0-bcf7-4184-b230-c41d91470bf2n%40googlegroups.com.

[Grégory Vanuxem]

Hello here, 

Le lun. 19 mai 2025, 19:57, Dima Pasechnik <dim...@gmail.com> a écrit :

On Sat, May 17, 2025 at 10:43 AM chris wuthrich
<christian...@gmail.com> wrote:
>
> Upgrading to ubuntu 25.04 also broke sage build (both 10.6 and 10.7.beta3) for me. Logs attached.


This is basically the error in
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=1093321
for brial/polybori

g++: error: @SIMD_CFLAGS@: linker input file not found: No such file
or directory

Thanks for this information. I have also this error but when building sagelib. 

Greg

[Dima Pasechnik]

I've opened a Brial issue here:
https://github.com/BRiAl/BRiAl/issues/58

Can we get a stand-alone Brial reproducer for this bug?

> --
> You received this message because you are subscribed to the Google Groups "sage-devel" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to sage-devel+...@googlegroups.com.

> To view this discussion visit https://groups.google.com/d/msgid/sage-devel/CAHnU2dagtAdEwL8k-2a51fVEiKeFSo8nf15kzuLT-xwz1nRCgA%40mail.gmail.com.

[Antonio Rojas]

The bug is in m4ri, not in brial. It is fixed in 20250128, sage shouldn't accept anything older.

[François Bissey]

Thanks for that! I was looking at the bug Dima opened in brial as the maintainer and I was, "this is not inside stock brial code" that it comes from m4ri makes sense to me.

François

To view this discussion visit https://groups.google.com/d/msgid/sage-devel/66a44171-cab2-4fd8-a01a-bc350549f634n%40googlegroups.com.

[Dima Pasechnik]

On Mon, May 19, 2025 at 4:00 PM Antonio Rojas <nqn...@gmail.com> wrote:
>
> The bug is in m4ri, not in brial. It is fixed in 20250128, sage shouldn't accept anything older.

OK, I'll open a PR to update, but it would have been great if there
was an issue opened about this.

Please, it's a general request: if you learned about the need for an
spkg update, please open an issue (better, a PR, but at least an
issue). Otherwise someone will waste their time re-discovering it.

> To view this discussion visit https://groups.google.com/d/msgid/sage-devel/66a44171-cab2-4fd8-a01a-bc350549f634n%40googlegroups.com.

[Dima Pasechnik]

Here is the update PR: https://github.com/sagemath/sage/pull/40132

[Tobia...@gmx.de]

What issues did you experience with the meson build? The notebook functionality should still work as usual via `sage -n`.

[Antonio Rojas]

To clarify, this has nothing to do with GCC 15. It's a long known bug in m4ri version 20240729 that some distros still ship [1,2,3]. In particular, sage-the-distro's (very old) version is not affected and, strictly speaking, only system version 20240729 should be disallowed.

[1]  https://groups.google.com/g/sage-release/c/muFP_3aonHc/m/2SxR8t1eAQAJ

[2] https://groups.google.com/g/sage-devel/c/nfStULvihao/m/f-moPx9QEAAJ

[3] https://groups.google.com/u/1/g/sage-support/c/iEfgKWIvzi4/m/OOr2TDSeDQAJ

[chris wuthrich]

Thanks to Dima's quick action here, the changes at https://github.com/sagemath/sage/pull/40132 fix the initial problem and I can now compile sage on the upgraded ubuntu.

Chris