app-office/libreoffice: Backport upstream patch to allow compilation with GCC 14

[0xr0bert]

Bug: https://bugs.gentoo.org/928152

[gentoo-bot]

Pull Request assignment

Submitter: @0xr0bert
Areas affected: ebuilds
Packages affected: app-office/libreoffice

app-office/libreoffice: @gentoo/office

Linked bugs

Bugs linked: 928152

---

In order to force reassignment and/or bug reference scan, please append [please reassign] to the pull request title.

Docs: Code of Conduct ● Copyright policy (expl.) ● Devmanual ● GitHub PRs ● Proxy-maint guide

[gentoo-repo-qa-bot]

Pull request CI report

Report generated at: 2024-04-01 19:22 UTC
Newest commit scanned: b419abe
Status: ✅ good

There are existing issues already. Please look into the report to make sure none of them affect the packages in question:
https://qa-reports.gentoo.org/output/gentoo-ci/003db9de11/output.html

[thesamesam]

Thanks. No need for new revisions for this as it's a build-time only fix and it's not a correctness issue at all (I mean, it is, but if it builds, it's because the transitive include got picked up, so no runtime behaviour diff. before/after).

Could you fix that and then lgtm?

[0xr0bert]

All done @thesamesam

[gentoo-repo-qa-bot]

Pull request CI report

Report generated at: 2024-04-02 16:40 UTC
Newest commit scanned: 6cfc834
Status: ✅ good

There are existing issues already. Please look into the report to make sure none of them affect the packages in question:
https://qa-reports.gentoo.org/output/gentoo-ci/cb079ee164/output.html

[a17r]

@thesamesam, @0xr0bert: As far as I can see this is just duplicating what we already have as libreoffice-7.6.5.2-gcc14.patch but got dropped from 7.6.6.3 because it was not applied to 7.6.9999 and 7.6.5.2 ebuild had said # upstream only, instead of which branch this was coming from.

So, neither do we have to add the patch to FILESDIR in the first place nor should we touch the stable 7.6.4.1 as a matter of principle. We only need to re-add the patch to 7.6.6.3, there's probably not going to be another release spun from 7.6.9999.

See also commit 321f0ca

[a17r]

Also, since 7.6.5.2 already has the patch applied by ebuild, the change here makes it fail in src_prepare() if you really test this.

[0xr0bert]

Thanks @a17r

Should it not be applied to the stable 7.6.4.1? If not, could we add specify as a dependency that gcc < 14.

Should it not also be added to 7.6.9999 as well as 7.6.6.3?

[a17r]

7.6.4.1 does not need any change as long as GCC-14 does not even have any keywords. No one should try this with arch-only and if they do, they may apply user patches. It is just busywork for us and we do not touch stable ebuilds unless really really necessary. For 7.6.9999 it is likely to be dropped soon, or if upstream decide to backport, it was again just busywork for us when it fails to apply the patch for everyone, not just GCC-14 users.

[0xr0bert]

I have done as you've suggested @a17r and it compiles with GCC 14

[gentoo-repo-qa-bot]

Pull request CI report

Report generated at: 2024-04-02 21:00 UTC
Newest commit scanned: 2c36b9e
Status: ✅ good

There are existing issues already. Please look into the report to make sure none of them affect the packages in question:
https://qa-reports.gentoo.org/output/gentoo-ci/d5c6634055/output.html

[thesamesam]

Thank you! Yeah, when I looked at it before, I was convinced it'd been backported everywhere upstream. Apparently not.