Update to vim-9.1.1934.
Update to sqlite-autoconf-3510100 (3.51.1).
Update to ninja-1.13.2.
Update to linux-6.17.9.
Update to libxcrypt-4.5.2.
Update to iana-etc-20251120.
This commit makes ml_all the ARCH default, and `default` got renamed to
`none`. However, `default` is still need for profiling or else the
normal sections refuse to render.
Thus, what happens is ARCH gets set by the Makefile, or the user, the
options available are:
- ml_all
- ml_32
- ml_x32
- none
If `none` is set, R_ARCH is set as `default`, otherwise, R_ARCH is set
as ARCH.
This is the name for all cases of rendering. I think ideally, the
Makefile should reflect by making the default render path include "mlfs"
in some form. I think that would trip up Thomas' rendering scripts which
I don't want to do.
Update to coreutils-9.9.
Update to binutils-2.45.1. This now requires adding
glibc-2.42-upstream_fixes-1.patch.
Update to libxcrypt-4.5.1.
Update to linux-6.17.8.
Update to sqlite-autoconf-3510000 (3.51.0).
Update to vim-9.1.1888.
Update to iana-etc-20251022.
Update to pcre2-10.47.
Update to man-pages-6.16.
Update to linux-6.17.6.
Update to libcap-2.77.
Update to elfutils-0.194.
1
This issue of mismatched test count for other installations of
architectures that are not part of the normal installation probably
exists in several XML files throughout the book due to how merges were
done in this branch. This was reported by Visceral on Discord. Though I
am unsure if I should make a point to go through another pass and make
each arch section fall in line with the normal sections.
An alternative is killing the tests for the other architectures. This is
something I've done in GLFS to reduce maintainence burden. But the goal
for many users following this book is to simply get a working system
with capabilities that allow it to run 32-bit software, no hassle. Tests
are thus very important.
Another issue is that the amount of tests can and will differ for
different architectures which I'm pretty sure is not actually considered
presently.
I haven't ran tests in the past but it will probably be a good idea to
start doing it, or cut the losses. I'll be thinking on this.
