[arm64] Enable ARM_CCA_GUEST as module

See merge request !1444

The module is needed to run Debian in a protected VM under the Arm Confidential
Compute Architecture (CCA).

See https://lore.kernel.org/kvm/20240701095505.165383-1-steven.price@arm.com/

Update to 6.13.11

See merge request !1455

Drop patches applied upstream

[arm64,powerpc,ppc64,ppc64el,riscv64,s390x] Enable KALLSYMS_ALL

See merge request !1419

Enable the setting on all architectures where DEBUG_INFO_BTF is on.

See #1100634

md: fix mddev uaf while iterating all_mddevs list

See merge request !1451

Closes: #1086175

Update to 6.13.10

See merge request !1449

Drop patche applied upstream

Cleanup debian/changelog file

ALSA: hda/realtek: Fix built-in mic on another ASUS VivoBook model

See merge request !1447

Closes: #1100928

Merge build profiles into build-depends

See merge request !1443

Use gcc-for-host for depends and build-depends

See merge request !1442

d/b/genpatch-rt: Fix subprocess cleanup with Python 3.13

See merge request !1439

As per flake8 testing reporting now

	debian/bin/genpatch-rt:4:1: F401 'io' imported but unused

Fixes: 55caad2c ("d/b/genpatch-rt: Fix subprocess cleanup with Python 3.13")
Link: https://www.flake8rules.com/rules/F401.html


Signed-off-by: Salvatore Bonaccorso <carnil@debian.org>

Ben Hutchings authored 2 weeks ago and Salvatore Bonaccorso committed 2 weeks ago

Quoting Colin Watson from wireless-regdb@c464cfbb:

> With Python 3.13, reopening the stdout file descriptor [...] causes
> cleanup to fail later.  It's simpler to call `subprocess.Popen` with
> `text=True` so that we get text output directly

Since genpatch-rt is used during source package preparation, this
needs to be fixed on older branches, not just those where the target
release has Python 3.13+.

[powerpc*] Revert "fbdev/offb: Update expected device name"

See merge request !1434

Update to 6.13.9

See merge request !1437

Drop patches applied upstream

drivers/net/wireless/realtek/rtw88: Enable RTW88_8723CS as a module

See merge request !1436

Commit 241d2fb56a18 "of: Make OF framebuffer device names unique",
included in Linux 6.2 and backported to 6.1.12, changed the Linux
platform device names for OpenFirmware "display" device nodes on
powerpc from "of-display" to "of-display.0", "of-display.1", etc.
This caused them to no longer match the offb device driver, though
it's not entirely clear why since they should match through the
compatible string..  This was reported as Debian bug #1033058.

Our patch "fbdev/offb: Update expected device name" was supposed to
fix bug that by changing the driver name of offb to "of-display.0",
matching the new first device name.

Later, commit 0bb8f49cd2cc "of: Preserve "of-display" device name for
compatibility", included in Linux 6.5 and backported to 6.1.42,
reverted the first device name to be "of-display" again.

The combination of both changes means the device and driver names
are mismatched again, resulting in bug #1085949.

Drop our patch.

Closes: #1085949

My manual merge conflict resolution was not quite right.

Fixes: f329eeca ("Merge branch 'ci-lintian-release' into debian/latest")

[cloud, amd64/cloud] Enable virtualisation drivers

See merge request !1433

This reverts commit 9b5d19e4.

This will break at least powerpc and ppc64, but as well needed on Apple
hardware.

This reverts commit 180f39f0.

This will break at least powerpc and ppc64, but as well needed on Apple
hardware.

Enable virtualisation drivers (VIRT_DRIVERS) on cloud flavours,
by removing the constraint in config.cloud file, and do not allow
VBOXGUEST on config.cloud-amd64.

Closes: #1101378

Recover arm64 support for rt flavour

See merge request !1430

d/salsa-ci.yml: Run lintian from the target release, not always unstable

See merge request !1290

Remove rt patchset and replace with a simple flavour

See merge request !1416