https://bugzilla.redhat.com/show_bug.cgi?id=2366921



--- Comment #13 from Ben Beasley <[email protected]> ---
(In reply to Jannik Müller from comment #12)
> Should I do write a package review for that now? I didn't think it would
> explode this far with so many packages I have to submit. I just want to fix
> Hyprland right now, so that the Fedora users can enjoy the newest version
> again. I don't know if I have the time to maintain more packages.
> Although it might be necessary...

I hacked together a quick proof of concept, as follows:

  - https://music.fedorapeople.org/openalgz-ut.spec
  - https://music.fedorapeople.org/openalgz-ut-0.0.5-1.fc42.src.rpm

Then I tried to unbundle with

  BuildRequires:  openalgz-ut

and in %prep

  # Adjust dependency name in tests to match Fedora’s openalgz-ut package.
  sed -r -i 's/\but::(ut)\b/\1/' tests/CMakeLists.txt

After that, things still weren’t quite right:

  cd
/builddir/build/BUILD/glaze-5.2.1-build/glaze-5.2.1/redhat-linux-build/tests/compare_test
&& /usr/bin/cmake -E cmake_link_script CMakeFiles/compare_test.dir/link.txt
--verbose=1
  /usr/bin/ld: cannot find -lut: No such file or directory

(At the same time, in a different parallel process, I did see the GCC 15
incompatibility you reported, “static assertion failed: hash function must be
copy constructible.”)

So the CMake configs are trying to have it link a nonexistent libut.so for a
header-only library. I also don’t like that my proposed openalgz-ut package
Provides cmake(ut) rather than cmake(ut::ut), and that this is the same
Provides as ut-devel. I didn’t have the time or interest to investigate any of
this further. You are welcome to take my links above as a starting point and
try to get everything working properly, if you like.

I think unbundling (openalgz-ut) here is less urgent since it’s only used for
tests and isn’t installed as part of the binary RPMs, so if you can get glaze
working with a bundled copy provided as an additional source, I think that
would be reasonable too.

For the time being, I guess it could also be ok to build this without the tests
in Fedora 42+ due to the GCC 15 issue, as long as you are sure it works with
hyprland, although I defer to Benson as the actual reviewer here. It would be
nice to at least prove it out with tests in a Fedora 41 scratch build, where
you have GCC 14, to show that the GCC 15 regression is the only problem.


-- 
You are receiving this mail because:
You are always notified about changes to this product and component
You are on the CC list for the bug.
https://bugzilla.redhat.com/show_bug.cgi?id=2366921

Report this comment as SPAM: 
https://bugzilla.redhat.com/enter_bug.cgi?product=Bugzilla&format=report-spam&short_desc=Report%20of%20Bug%202366921%23c13

-- 
_______________________________________________
package-review mailing list -- [email protected]
To unsubscribe send an email to [email protected]
Fedora Code of Conduct: 
https://docs.fedoraproject.org/en-US/project/code-of-conduct/
List Guidelines: https://fedoraproject.org/wiki/Mailing_list_guidelines
List Archives: 
https://lists.fedoraproject.org/archives/list/[email protected]
Do not reply to spam, report it: 
https://pagure.io/fedora-infrastructure/new_issue

Reply via email to