Restore big endian support

[fitzsim]

Restore big endian support, which worked with the Makefile, to CMake, eliminate C++23 reliance, fix some whitespace. It might make sense to move this logic into ggml at some point.

[fitzsim]

It looks like the linux/arm64 continuous integration machine had a system issue:

Processing triggers for libc-bin (2.35-0ubuntu3.8) ...
qemu: uncaught target signal 11 (Segmentation fault) - core dumped
Segmentation fault (core dumped)
qemu: uncaught target signal 11 (Segmentation fault) - core dumped
Segmentation fault (core dumped)
dpkg: error processing package libc-bin (--configure):
 installed libc-bin package post-installation script subprocess returned error exit status 139
Processing triggers for ca-certificates (20240203~22.04.1) ...
Updating certificates in /etc/ssl/certs...
0 added, 0 removed; done.
Running hooks in /etc/ca-certificates/update.d...
done.
Errors were encountered while processing:
 libc-bin
E: Sub-process /usr/bin/dpkg returned an error code (1)

This seems unrelated to the change in this pull request. I cannot find a way to restart the failed checks though.

[petterreinholdtsen]

The easiest way to restart the CI runs is to do fetch the latest master from ggerganov, followed by 'git rebase ggerganov/master' and 'git push -f fitzsim restore-big-endian-support' assuming your remote git repo is 'fitzsim'.

[fitzsim]

The easiest way to restart the CI runs is to do fetch the latest master from ggerganov, followed by 'git rebase ggerganov/master' and 'git push -f fitzsim restore-big-endian-support' assuming your remote git repo is 'fitzsim'.

Makes sense, thanks, however I will have to wait until a new commit lands on master.

[ggerganov]

To fix the CI, we can apply the same fix that we did in llama.cpp: ggml-org/llama.cpp#11895