Re: M-chip

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Wed, 4 Dec 2024, 21:49 Bert Gold via Gcc-help, <gcc-help@xxxxxxxxxxx>
wrote:

> I am posting my recent interaction with chatGPT.  I hope it sheds light for
> some of you.
> bgold wrote:
> Clang is unable to support pybind11 on a MiniMac M4 computer.  Here is my
> bug report to the pybind11 group:
> https://github.com/pybind/pybind11/issues/5455
> ChatGPT said:
> ChatGPT
> Thank you for sharing the background and your detailed bug report on the
> pybind11 GitHub repository. Given the constraints with both clang and gcc
> on the macOS ARM64 architecture, here are some tailored suggestions to
> proceed effectively:
>
> Key Insights from the Issue
> clang and pybind11 Incompatibilities:
>
> pybind11 relies heavily on standard C++ features, which can have nuanced
> differences in their implementation between compilers. The failure you
> described with clang likely involves macOS-specific quirks in either its
> standard library or runtime support for templates and attributes that
> pybind11 uses.
>


A lot of waffle and no actual information.

gcc on macOS:
>
> While gcc can theoretically work, Homebrew's gcc isn't optimized for native
> macOS toolchains. It is more suited for POSIX-compliant environments and
> can create cross-platform complications like generating ELF binaries
> instead of Mach-O.
>

This is typical chatgpt nonsense. Please don't waste our time with this
mess.


Compilers and macOS Toolchain:
>
> Apple tightly integrates its development stack with clang and Xcode, and
> deviations (like forcing gcc) often run into edge-case issues due to the
> lack of native gcc support.
>
> On Wed, Dec 4, 2024 at 5:09 AM Hans Åberg <hans.aberg_1@xxxxxxxxxx> wrote:
>
> >
> > > On 4 Dec 2024, at 09:52, Simon Wright <simon@xxxxxxxxxxxx> wrote:
> > >
> > > On 3 Dec 2024, at 23:43, Bert Gold via Gcc-help <gcc-help@xxxxxxxxxxx>
> > wrote:
> > >> One might expect that using
> > >>
> > >>
> >
> Cellar/gcc/14.2.0_1/bin/../libexec/gcc/aarch64-apple-darwin24/14/lto-wrapper
> > >>
> > >> on an M-chip Mac, we would get an ARM64 shared library.  But, no!
> > >>
> > >> This is what I get:
> > >>
> > >> ELF 64-bit LSB shared object, x86-64, version 1 (GNU/Linux),
> dynamically
> > >> linked, BuildID[sha1]=abf0c9f9112ea87ddc35d0e7a99d6259bbbd74be, not
> > stripped
> > >
> > > I get
> > > $ file lto-wrapper
> > > lto-wrapper: Mach-O 64-bit executable arm64
> > >
> > > so perhaps you need to reinstall.
> > >
> > >> Of course the x86_64 won't work with the reduced instruction set of
> the
> > >> M-chip.
> > >
> > > An x86_64 executable will run on Apple silicon under Rosetta 2
> >
> > On MacPorts, one gets a warning that the installation is running under
> > Rosetta, and should run 'port migrate', after which GCC14 is arm64.
> >
> >
> >
>
> --
> Bert Gold
> Falmouth, MA
> 302-507-1827
>




[Index of Archives]     [Linux C Programming]     [Linux Kernel]     [eCos]     [Fedora Development]     [Fedora Announce]     [Autoconf]     [The DWARVES Debugging Tools]     [Yosemite Campsites]     [Yosemite News]     [Linux GCC]

  Powered by Linux