Remove experimental dmraid support

See merge request !6

Was behind a boot flag anyway.

Remove unnecessary constraints

See merge request !5

* Build-Depends: Drop versioned constraint on dpkg-dev, libdebian-installer4-dev and po-debconf.
* hw-detect: Drop versioned constraint on cdebconf-udeb, di-utils, rootskel and udpkg in Depends.
* ethdetect: Drop versioned constraint on cdebconf-udeb, di-utils and rootskel in Depends.
* disk-detect: Drop versioned constraint on cdebconf-udeb and di-utils in Depends.
* driver-injection-disk-detect: Drop versioned constraint on cdebconf-udeb in Depends.

Changes-By: deb-scrub-obsolete

Fix some issues reported by lintian

See merge request !4

Changes-By: lintian-brush
Fixes: lintian: uses-debhelper-compat-file
See-also: https://lintian.debian.org/tags/uses-debhelper-compat-file.html

Changes-By: lintian-brush
Fixes: lintian: package-uses-deprecated-debhelper-compat-version
See-also: https://lintian.debian.org/tags/package-uses-deprecated-debhelper-compat-version.html

Add missing ${misc:Depends} to Depends for hw-detect, ethdetect, disk-detect, driver-injection-disk-detect, archdetect.

Changes-By: lintian-brush
Fixes: lintian: debhelper-but-no-misc-depends
See-also: https://lintian.debian.org/tags/debhelper-but-no-misc-depends.html

Install opal-prd package on OpenPOWER machines.

See merge request !1

This makes it possible for some callers (e.g. cdrom-detect, maybe
iso-scan too) to avoid triggering firmware-related prompts when it's too
early in the installation process to do anything about it (See: #991587,
#991590).

See check-missing-firmware:
 - /firmware is used when booting from PXE;
 - /cdrom/firmware is used when booting from an ISO.

It seems to be about debugging, it doesn't seem actually required, and
it's not packaged anyway (see: #969264, 966218).

Appease lintian (udev-rule-in-etc).

They were obsoleted upstream (udev) in 2013. This only generates noise
in the installer's syslog.

Avoid repeating n times the same module if it requests n firmware files
(e.g. iwlwifi).

Check the actual module behind a given driver, since there might be a
slight mismatch between the name coming up in dmesg (e.g. rtw_8821ce)
and the actual module name (e.g. rtw88_8821ce).

With many thanks to Daniel Lewart for the original patch.

Note: We could move the initial code block below the loop and check
whether $actual_module was ever set, but it seems safer to keep
reloading $module in any cases. The only downside is two log lines
(modprobe errors) when there's no actual module by that name.

We would alternate between base-installer lines (grep hits) and
install-firmware (log messages). Opt for install-firmware all the way.