[ci] increase valgrind timeout to 6 hours

[jameslamb]

Blocks #5952.

• upgrades the timeout on the r-valgrind CI job to 6 hours, to try to fix Release v4.0.0 #5952 (comment)

Notes for Reviewers

As described in #5414, this has to be merged to master to be tested, since the valgrind job's configuration is only sourced from master.

6 hours is the absolute maximum GitHub Actions allows: https://docs.github.com/en/actions/learn-github-actions/usage-limits-billing-and-administration#usage-limits.

So if this doesn't work, we'll have to come up with another strategy for running the valgrind tests faster (like maybe finding a source of Linux precompiled binaries for {lightgbm}'s dependencies), or will have to rely on running them somewhere else (even maybe just locally on a maintainer's laptop 😬 ).

[jameslamb]

check-docs is failing with this error that seems to be occuring while rendering the sphinx docs to HTML

Warning, treated as error:
/home/runner/work/LightGBM/LightGBM/python-package/lightgbm/dask.py:docstring of lightgbm.dask.DaskLGBMClassifier:1:undefined label: 'metadata_routing'

(build link)

I see that on master too (build link).

Since that blocks merging, I'll try to fix it in this PR too. Moving this back to draft until that's fixed.

[jameslamb]

@guolinke @shiyu1994 could you please review this? So we can continue testing v4.0.0 on #5952.

[jameslamb]

thanks @jmoralez !

hopefully 6 hours is enough 🤞🏻

[github-actions]

This pull request has been automatically locked since there has not been any recent activity since it was closed. To start a new related discussion, open a new issue at https://github.com/microsoft/LightGBM/issues including a reference to this.