You are right, the issues seem different. I'm confused how a couple of you got past the activation error described in the ticket.

On 12/27/24 17:36, Alan Bram wrote:
I don't see the connection between this issue and the other ticket you mentioned; they seem like separate issues to me. (The only relation I see between them is that de-selecting nativecomp avoids the issue in both cases.) But perhaps I'm not fully understanding?

The run-time libgccjit failure gets more puzzling the more I look at it. Apparently it has been over three years since some version of this issue has existed:

https://github.com/emacs-mirror/emacs/commit/3d575534d43802552f59eb1b990939cb3412d4e9

But for some reason some people (myself included) encountered it only recently, upon upgrading to Sequoia. And even when it was discussed a few years ago, only some, but not everyone, were affected by it, and no one seems to have put a finger on what the difference was.



On Tue, Dec 24, 2024 at 6:23 AM Jonathan Stickel <jjstic...@gmail.com> wrote:

    There is a macports ticket open for emacs/emacs-app on Sequoia:

    https://trac.macports.org/ticket/71047

    It is interesting that you were able to install and activate
    emacs-app +nativecomp, but later experienced warnings and errors,
    whereas the reported behavior in the ticket is failing at the
    activate step. In any case, disabling nativecomp provided a
    successful workaround.

    Alan, thanks for the reddit link. That might be a solution for a
    patch to the port. I suggest sharing that link on the ticket.

    Regards,
    Jonathan


    On 12/24/24 05:00, macports-users-requ...@lists.macports.org wrote:
    Date: Mon, 23 Dec 2024 11:36:07 -0800
    From: Alan Bram<alan.b...@cornell.edu> <mailto:alan.b...@cornell.edu>
    To:rco...@rcobbe.net
    Cc:macports-users@lists.macports.org
    Subject: Re: emacs-app: error invoking gcc driver
    Message-ID:
        <cachqos8ptqznt4kcj2qfs3bneneq7w0axscerrdh_cqghdc...@mail.gmail.com> 
<mailto:cachqos8ptqznt4kcj2qfs3bneneq7w0axscerrdh_cqghdc...@mail.gmail.com>
    Content-Type: text/plain; charset="utf-8"

    Have a look at this comment from user 0dBgain on Reddit. I had the same
    problem a couple of months ago, but didn't find this Reddit thread until
    yesterday, so I haven't tried it yet. (Instead, I reinstalled Sonoma,
    because the machine was new and there wasn't yet much new data on it.)

    https://www.reddit.com/r/emacs/comments/u28dji/comment/lv0au2o/

    If you do try this I would love to hear whether it works.

    On Mon, Dec 23, 2024 at 8:02?AM Richard Cobbe<rco...@rcobbe.net> 
<mailto:rco...@rcobbe.net> wrote:

    I recently upgraded one of my machines to Sequoia 15.0 (there's an update
    pending to 15.1.1 that I haven't yet installed) and am in the process of
    reinstalling macports.  (I originally tried 'macports migrate' but I ran
    into the Apple CLI installer bug preventing a lot of C++ compilation and
    deleted my previous macports installation before I discovered the fix for
    that.)

    I just reinstalled emacs-app @29.4_2+nativecomp+rsvg+treesitter.  It built
    correctly, and the resulting executable starts, but after a brief time, the
    **Warnings** buffer pops up, and it's full of a bunch of messages like the
    following:

    ? Warning (comp): libgccjit.so: error: error invoking gcc driver
    ? Warning (comp)://Users/r.cobbe//.emacs.d/elpa/auctex-14.0.8/tex-site.el:
    Error: Internal native compiler error failed to compile
    ? Warning (comp): libgccjit.so: error: error invoking gcc driver
    ? Warning (comp)://Users/r.cobbe//.emacs.d/elpa/auctex-14.0.8/texmathp.el:
    Error: Internal native compiler error failed to compile

    Has anyone seen this?  Is there a workaround, or should I file a bug?
    Rebuild without the nativecomp variant?

    MacOS 15.0, Apple M2 chip.

    Thanks,

    Richard

Reply via email to