Those willing to reuse archdep-filter elsewhere
will likely have to figure it out from Makefiles
anyway, so probably not worth overdocumenting...

Duplicating the architecture tag handling logic was afwul
indeed; one-liner (well, two lines for readability) is not
that bad at least, factoring it out wouldn't gain a lot.

Fixes: 75735473

This code started breeding within pkg.in/*/Makefile,
and it was lacking both E2K (as a generic alias,
not just "e2k or e2kv4") _and_ the ability to negate
selection, as in "everywhere but not on any e2k").

Let's fix all of that at once, and parallelize sed
execution as well; my tests with 1000-line file
containing "a b c d@!E2K e@e2k f g@!X86 h i@IA32"
lines show roughly 3x higher CPU load and lower
execution time (~0.3 sec vs ~0.9 sec on 801-PC).

And turn that code snippet into a proper filter
with inline edit capability (I've dropped the exit
trap as failing to mv signals a disaster anyway).

- make bails out on single image build error, see also make -k (iv@)
- mipsel support (iv@, antohami@)
- regular-vm.mk, refactoring, cleanups (antohami@)

Conflicts with systemd-utils.

See-also: http://bugzilla.altlinux.org/37352

See-also:
http://git.altlinux.org/gears/s/systemd.git?p=systemd.git;a=commitdiff;h=30c28ce9e6d8d006c81b6bbecaf6423a08d8f3f5

Suggested-by: Anton Farygin <rider@altlinux.org>

...it's no more in Sisyphus too:
See-also: http://git.altlinux.org/tasks/archive/done/_232/238525

...so that bits required for PXE boot are handy in the image.

Suggested-by: glebfm@altlinux.org, klark@altlinux.org

It's required to set BASE_BOOTLOADER variable,
otherwise grub will get in *too*.

Not used by systemd and looks obsolete generally
as sysvinit-based disk images aren't really targeting
low-resource systems these days _by default_ but rather
_can_ target those as well; feel free to reconstruct
these "RAM saving" bits as a part of e.g. lowmem patch.

bfk3 needs more than 10 seconds to bring the interface up
and get the lease.

TODO: make this configurable.

sm750 is seen as framebuffer-only device, so it's not marked with
master-of-seat tag with current udev (which does this mostly to
dri cards). But without master-of-seat device logind does not
report that the seat "CanGraphical", and thus sddm does not
work for it; there might be other problems, too.

This change adds udev rule that marks sm705's fb device as
master-of-seat.

We only plan to distribute XZ-compressed rootfs tarballs.

(mike@: merging cleaned-up version)

Tavolga seldom has a serial port connected, so it makes
more sense to write kernel and systemd's messages
to tty1.

We don't know the root UUID at this point; recovery.rc
will deal with it later, but we'd like to have
even the tarball in a somwhat working state.

...with xorg-drv-btsm750 and packaged xorg.conf.

aarch64 repo has got std-def/un-def flavours by now.

This commit fixes duplication of root device entry in /etc/fstab
and changes LABEL=ROOT entry with the proper one if found.

Apparently this didn't hurt qemu images much though...

Before this change, we could have 0 exit code from the main
make process even when some targets failed to build on some
architectures:

$ make ARCH=mipsel [...] vm/tavolga-regular-mate-ex.recovery.tar
** ARCH: mipsel
10:15:06 cleaning up
10:15:06 initializing BUILDDIR: build/
10:15:06 preparing distro config: build/distcfg.mk
make[1]: *** No rule to make target 'use/systemd-coredump', needed by 'profile/.extra'.  Stop.
[nightly-mipsel@loongson5 mkimage-profiles]$ echo $?
0

Use `make -k' to get the old behaviour.

- recovery.tar support (iv@)
- mixin/mixin deps, aarch64 EFI, mcom02 & jetson nano support,
  multikernel & portability fixes, refactoring & cleanups (antohami@)
- document STAGE1_MODLISTS (me)

This fixes warning at enlightenment startup.

...to extend the supported protocols list as suggested.

See-also: b869d365f195a7070f1f6c425795b4c058b18c6c

...to make it easier for all builds including nm-applet
to add this target.