emacs-{,app-}devel: fix build on arm64

[catap]

Description

gcc11 can't be build on arm64, use gcc-devel instead

Type(s)

• bugfix
• enhancement
• security fix

Tested on

macOS 12.0.1 21A559 arm64
Xcode 13.1 13A1030d

Verification

Have you

• followed our Commit Message Guidelines?
• squashed and minimized your commits?
• checked that there aren't other open pull requests for the same change?
• referenced existing tickets on Trac with full URL?
• checked your Portfile with port lint?
• tried existing tests with sudo port test?
• tried a full install with sudo port -vst install?
• tested basic functionality of all binary files?

[macportsbot]

Notifying maintainers:
@drkp for port emacs.

[catap]

@cjones051073 I'd like to ask you to review it. Maybe you know an easy way to use libgccjit that my hack.

[pmetzger]

@mascguy Are you actually deleting comments from fellow project members that say you committed something in error? Really?

[mascguy]

Perhaps we need to have a frank real-time discussion with the Managers, about appropriate conduct and decorum. Between the utter inappropriateness of these comments - they absolutely should NOT be here in a contributor's PR, as they're related to internal MacPorts-related matters - combined with your continual adversarial comments toward folks (including @kencu, among others) - I've had enough.

CC: @ryandesign
CC: @jmroot
CC: @raimue

[pmetzger]

I already sent email to the mailing list about this earlier. And my comments weren't inappropriate, they noted that you had been violating the long-standing rules we have about how long to wait for maintainers to comment, and they were not unfriendly. Deleting my comments, however, was an escalation.

(BTW, making this about me isn't very reasonable. If someone points out that you've done something that violates policy, the polite thing to do is to check if that's true, and to not do it going forward if it is. I didn't make any personal comments about you; I'd ask that you avoid that in return.)

[pmetzger]

On the technical merits: generally, one would probably want a solution here that would continue to work if gcc11 is later made available for the platform.

(Note also that the lack of availability of a -devel port isn't usually an emergency, and Dan is quite responsive in general.)

[drkp]

FWIW, I don't mind merging this to fix the build on arm64 (though I can't test it myself since I only just got an arm64 system and haven't even set it up yet).

That said, I'm wondering if a better solution here is to patch emacs's autoconf script, which already checks which gcc ports are installed (!) to set the right libgccjit path, to detect gcc-devel if installed and prefer it, perhaps just on arm64. That's something that we could upstream, as I did earlier with a patch to handle multiple installed gcc ports.

[jmroot]

Perhaps we need to have a frank real-time discussion with the Managers, about appropriate conduct and decorum.

TBH I didn't even know non-admins had the ability to delete comments. This was undoubtedly an abuse of that ability in any case. Off-topic comments should at most be hidden. AFAICT there's no way to undelete comments, so I and anyone else who didn't see them at the time will never know what they said.

@mascguy If inappropriate posts are made (and I'm not saying they were here), report them to portmgr. If someone points out a project policy that may have been overlooked in a PR, that is not in itself inappropriate.

@pmetzger If you were using an adversarial tone (and I'm not saying you were), please try to moderate it for the comfort of your fellow project members. It's easy to lose nuance in a text-based medium, so keep in mind the different ways in which a message could be interpreted.

[mascguy]

I already sent email to the mailing list about this earlier. And my comments weren't inappropriate, they noted that you had been violating the long-standing rules we have about how long to wait for maintainers to comment, and they were not unfriendly. Deleting my comments, however, was an escalation.

(BTW, making this about me isn't very reasonable. If someone points out that you've done something that violates policy, the polite thing to do is to check if that's true, and to not do it going forward if it is. I didn't make any personal comments about you; I'd ask that you avoid that in return.)

You're right Perry, and I sincerely apologize.

Ditto for policy reminders, as they're certainly there for a reason. Indeed, I made the (dubious) decision to prematurely merge a PR for one of Josh's ports - py-docutils - only to have it come back and bite us/me in the keister.

I'll certainly try my best to not let my feathers get ruffled by future policy reminders. Ideally, though, I'll try not to violate them, period, to help ensure folks don't need to remind me at all!

Thanks for keeping me honest, folks!

[pmetzger]

@drkp Whether or not we patch autoconf, I think it's best if the port continues to work correctly if gcc11 becomes available for the M1 macs.

(As an aside, Emacs 28 seems to now be in the midst of its release process, so the jit functionality will soon be migrating to the main port.)

[catap]

@pmetzger the issue that the only gcc which is available for M1 is gcc-devel and from specified place: https://github.com/iains/gcc-darwin-arm64

I doubt that this code will be merge into gcc mainline and released as gcc12 before emacs.

Anyway, the bad news that my hacks doesn't work for buildbot which is strange because I've used the same port to have this one: