CI builds fail due to runner upgrade to CMake 4.0.0

[rafbiels]

Summary

GitHub Ubuntu runner upgraded CMake to 4.0.0 recently and now oneMath no longer builds in CI:

CMake Error at deps/googletest/CMakeLists.txt:53 (cmake_minimum_required):
  Compatibility with CMake < 3.5 has been removed from CMake.

  Update the VERSION argument <min> value.  Or, use the <min>...<max> syntax
  to tell CMake that the project requires at least <min> but has been updated
  to work with policies introduced by <max> or earlier.

  Or, add -DCMAKE_POLICY_VERSION_MINIMUM=3.5 to try configuring anyway.

Example CI failure:
https://github.com/uxlfoundation/oneMath/actions/runs/14200577688/job/39903482981?pr=655
Runner change:
actions/runner-images@83ff3a6#diff-cc12e5c5d005932feaacf280af949d83e2cb6020dff8203d9ead2e00c30e8b7fR70

Version

Head of develop

Environment

oneMath CI

Steps to reproduce

Run CI on any PR

Observed behavior

CMake configure fails in CI

Expected behavior

CMake configure succeeds in CI

[rafbiels]

@Rbiessy this affects not only the googletest dependency:

oneMath/deps/googletest/CMakeLists.txt

Line 53 in 8517a20

cmake_minimum_required(VERSION 2.6.4)

but also the generic SYCL BLAS:
https://github.com/uxlfoundation/generic-sycl-components/blob/badb37970edb6f96115fedd3752a7f2bfda25f81/onemath/sycl/blas/CMakeLists.txt#L23
as seen in this job:
https://github.com/uxlfoundation/oneMath/actions/runs/14200577688/job/39903480796?pr=655

Would you suggest updating both projects to cmake_minimum_required(VERSION 3.5.0) or do you prefer to go with the workaround suggested in the error message, i.e. using -DCMAKE_POLICY_VERSION_MINIMUM=3.5 in the CI cmake commands?

I can submit fixes on Monday unless someone does it sooner (I'm not available today I'm afraid).

[Rbiessy]

The issue you mentioned should be fixed by uxlfoundation/generic-sycl-components#9. I suggest that we update the minimum CMake version of generic-sycl-components to 3.13 to match with oneMath.

No worries if you're not available today, this can wait a bit more if a fix is not merged by then.

[mkrainiuk]

#657 with the fix was merged, closing the issue.

[rafbiels]

Many thanks @mkrainiuk for the fixes!