Merge ~bdrung/ubuntu/+source/initramfs-tools:swap into ubuntu/+source/initramfs-tools:ubuntu/devel

[Benjamin Drung]

Unmerged commits

5678695... by Benjamin Drung on 2024-07-02

resume: always write valid resume device to /sys/power/resume

Users report that hibernation does not work on Ubuntu 24.04 (noble). The
initrd only sets the major:minor in `/sys/power/resume` in case of
resuming from hibernation, but systemd v255 will not allow hibernate to
start if the resume device has not be set in the kernel.

`blkid -p -o value -s TYPE $device` reports `swap` for swap partitions
and for swap files on Ubuntu 24.04 (noble). The type for resume images
is `swsuspend`.

So also set the major:minor in `/sys/power/resume` when the swap type is
`swap`.

Fixes: b3a78e4f8057 (" - resume: only resume when the partition contains a resume image - Fix resuming a hibernate session from a swapfile")
Fixes: https://bugs.launchpad.net/ubuntu/+source/systemd/+bug/2057687

8fd3b04... by Benjamin Drung on 2024-06-10

fsck: Mention file system name in failed identification warning

`update-initramfs` might be unable to identify the file system type of
the root file system:

```
$ update-initramfs -u
update-initramfs: Generating /boot/initrd.img-5.15.0-1059-kvm
W: Couldn't identify type of root file system for fsck hook
```

This error message can be confusing and misleading for the user. The
Ubuntu bug reported initially assumed it to be a bug in initramfs-tools
but later found out that the file system missed the label.

Mention the file system name in failed identification warning to help
the user finding the reason.

Bug-Ubuntu: https://launchpad.net/bugs/2068077
Forwarded: https://salsa.debian.org/kernel-team/initramfs-tools/-/merge_requests/118
Signed-off-by: Benjamin Drung <email address hidden>

68dae92... by Jeffery To on 2024-06-06

Fix resume failure due to resume=UUID=... in certain cases

If resume= has not been set on the kernel command line, and either:
* RESUME=UUID=... is set in /etc/initramfs-tools/conf.d/resume
* RESUME=auto is set in /etc/initramfs-tool/conf.d/resume and swap is a
  traditional partition (not in LVM)

Then the resume variable will be set to a UUID, which causes the resume
local-premount script to fail as it expects the resume variable to be a
device path.

This moves the UUID handling further down, outside of the parse kernel
command line options loop, to catch these cases.

Fixes: https://bugs.launchpad.net/ubuntu/+source/initramfs-tools/+bug/1769297
Fixes: https://bugs.launchpad.net/ubuntu/+source/initramfs-tools/+bug/1853982
Fixes: https://bugs.launchpad.net/ubuntu/+source/initramfs-tools/+bug/1957194
Signed-off-by: Jeffery To <email address hidden>

326e5a7... by Benjamin Drung on 2024-06-04

Release initramfs-tools 0.142ubuntu28

Signed-off-by: Benjamin Drung <email address hidden>

0a8e905... by Benjamin Drung on 2024-06-04

auto_add_modules: group dracut-install calls for block modules

Group calls to `dracut-install` that use the same `-s` parameter.

Benchmark results
=================

Benchmark on a Raspberry Pi Zero 2W on Ubuntu 24.04 (noble) on
2024-06-04. Previous commit:

```
bdrung@zero2w:~$ sudo hyperfine --warmup 1 -r 10 "update-initramfs -u"
Benchmark 1: update-initramfs -u
  Time (mean ± σ): 228.483 s ± 5.953 s [User: 57.447 s, System: 162.737 s]
  Range (min … max): 219.897 s … 240.041 s 10 runs
```

This commit:

```
bdrung@zero2w:~$ sudo hyperfine --warmup 1 -r 10 "update-initramfs -u"
Benchmark 1: update-initramfs -u
  Time (mean ± σ): 226.658 s ± 5.861 s [User: 54.878 s, System: 160.566 s]
  Range (min … max): 220.091 s … 237.076 s 10 runs
```

bb20916... by Benjamin Drung on 2024-06-04

Use copy_modules_dir instead of dracut-install calls

dracut-install has some startup costs and reducing the number of
dracut-install will increase the execution time. So use
`copy_modules_dir` again for calls that do not use `-s`.

Benchmark results
=================

Benchmark on a Raspberry Pi Zero 2W on Ubuntu 24.04 (noble) on
2024-06-04. Previous commit:

```
bdrung@zero2w:~$ sudo hyperfine --warmup 1 -r 10 "update-initramfs -u"
Benchmark 1: update-initramfs -u
  Time (mean ± σ): 236.326 s ± 6.033 s [User: 62.455 s, System: 165.245 s]
  Range (min … max): 229.254 s … 244.094 s 10 runs
```

This commit:

```
bdrung@zero2w:~$ sudo hyperfine --warmup 1 -r 10 "update-initramfs -u"
Benchmark 1: update-initramfs -u
  Time (mean ± σ): 228.483 s ± 5.953 s [User: 57.447 s, System: 162.737 s]
  Range (min … max): 219.897 s … 240.041 s 10 runs
```

The expected difference (for two dracut-install calls less) is a few
seconds which fits within the margin of error. The changes of the range
is a good indicator.

3121df3... by Benjamin Drung on 2024-06-03

reduce number of dracut-install calls

dracut-install has a slow start-up time on ARM devices. Alper Nebi Yasak
wrote: "On my ARM64 Chromebooks dracut-install takes about 2-3 seconds
to start up even for a single module." [1]

Instead of `manual_add_modules` calling `dracut-install` directly,
collect the wanted kernel modules in a temporary file. Call
dracut-install at the end on all modules in one go.

Since the `vmxnet` hook from `open-vm-tools` and `bootcdinitramfshook`
from `bootcd` look at the copied kernel modules, call
`apply_add_modules` before calling the hooks to not break those hooks.

`hidden_dep_add_modules` looks at the installed kernel modules. So call
`apply_add_modules` before it. Since it is the last function that
installs kernel modules, let it call dracut-install directly.

This can be a breaking change. To restore the previous behavior, call
`apply_add_modules` without arguments after a `manual_add_modules` call.

Benchmark results
=================

Benchmark on a Raspberry Pi Zero 2W on Ubuntu 24.04 (noble) on
2024-06-03. initramfs-tools 0.142ubuntu25.1:

```
bdrung@zero2w:~$ sudo hyperfine --warmup 1 -r 10 "update-initramfs -u"
Benchmark 1: update-initramfs -u
  Time (mean ± σ): 402.751 s ± 5.592 s [User: 166.316 s, System: 228.909 s]
  Range (min … max): 394.380 s … 411.445 s 10 runs
```

This commit:

```
bdrung@zero2w:~$ sudo hyperfine --warmup 1 -r 10 "update-initramfs -u"
Benchmark 1: update-initramfs -u
  Time (mean ± σ): 236.326 s ± 6.033 s [User: 62.455 s, System: 165.245 s]
  Range (min … max): 229.254 s … 244.094 s 10 runs
```

LP: #2065180
Forwarded: https://salsa.debian.org/kernel-team/initramfs-tools/-/merge_requests/114
[1] https://salsa.debian.org/kernel-team/initramfs-tools/-/merge_requests/102

949a2ce... by Benjamin Drung on 2024-05-31

mkinitramfs: Resolve hidden dependencies after all modules were copied

Move the `hidden_dep_add_modules` call after all kernel modules were
copied. Hooks might add or remove kernel modules that will infuence what
kernel modules `hidden_dep_add_modules` will add.

c2dfc45... by Benjamin Drung on 2024-05-28

hook-functions: Use firmware search order from kernel

`add_firmware` should behave like the kernel searches for the firmware
files in `_request_firmware`. This functions calls
`fw_get_filesystem_firmware` (that iterates over the paths) for
uncompressed files, then `.zstd`, and then `.xz`.

See also https://salsa.debian.org/kernel-team/initramfs-tools/-/merge_requests/108

4ef7091... by Benjamin Drung on 2024-05-19

Release initramfs-tools 0.142ubuntu27

Signed-off-by: Benjamin Drung <email address hidden>

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 0000000..42eda44
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,2 @@
+*~
+.#*
diff --git a/debian/.gitignore b/debian/.gitignore
new file mode 100644
index 0000000..bc79963
--- /dev/null
+++ b/debian/.gitignore
@@ -0,0 +1,7 @@
+/*.debhelper
+/*.debhelper.log
+/*.substvars
+/.debhelper
+/files
+/initramfs-tools/
+/initramfs-tools-core/
diff --git a/hooks/fsck b/hooks/fsck
index f6348e8..e0cfad0 100755
--- a/hooks/fsck
+++ b/hooks/fsck
@@ -44,6 +44,10 @@ _read_fstab_entry () {
 # Find a specific fstab entry and print its type (if found, and pass != 0)
 # $1=mountpoint
 get_fsck_type_fstab () {
+<<<<<<< hooks/fsck
+=======
+	local device
+>>>>>>> hooks/fsck
 	eval "$(_read_fstab_entry "$1")"
 
 	# Not found by default.
@@ -51,9 +55,15 @@ get_fsck_type_fstab () {
 		# Ignore filesystem type for /, as it is not available and
 		# therefore never used at boot time
 		if [ "${MNT_DIR}" = "/" ] || [ "${MNT_TYPE}" = "auto" ]; then
+<<<<<<< hooks/fsck
 			MNT_FSNAME="$(resolve_device "${MNT_FSNAME}")"
 			if ! get_fstype "${MNT_FSNAME}"; then
 				echo "W: Couldn't identify type of $2 file system for fsck hook" >&2
+=======
+			device="$(resolve_device "${MNT_FSNAME}")"
+			if ! get_fstype "${device}"; then
+				echo "W: Couldn't identify type of $2 file system '$MNT_FSNAME' for fsck hook" >&2
+>>>>>>> hooks/fsck
 			fi
 		else
 			echo "${MNT_TYPE}"
diff --git a/init b/init
index e270969..22fcbe7 100755
--- a/init
+++ b/init
@@ -167,10 +167,13 @@ for x in $(cat /proc/cmdline); do
 		;;
 	resume=*)
 		RESUME="${x#resume=}"
+<<<<<<< init
 		case $RESUME in
 		UUID=*)
 			RESUME="/dev/disk/by-uuid/${RESUME#UUID=}"
 		esac
+=======
+>>>>>>> init
 		;;
 	resume_offset=*)
 		resume_offset="${x#resume_offset=}"
@@ -247,6 +250,13 @@ fi
 if [ -n "${noresume}" ] || [ "$RESUME" = none ]; then
 	noresume=y
 else
+<<<<<<< init
+=======
+	case $RESUME in
+	UUID=*)
+		RESUME="/dev/disk/by-uuid/${RESUME#UUID=}"
+	esac
+>>>>>>> init
 	resume=${RESUME:-}
 fi
 
diff --git a/scripts/local-premount/resume b/scripts/local-premount/resume
index 6bd98e8..b7e975d 100755
--- a/scripts/local-premount/resume
+++ b/scripts/local-premount/resume
@@ -65,7 +65,15 @@ case "${SWAPTYPE}" in
 		# that less likely.
 		sleep 0.1
 	fi
+<<<<<<< scripts/local-premount/resume
 
+=======
+	;;
+esac
+
+case "${SWAPTYPE}" in
+    swap|swsuspend|s1suspend|s2suspend|ulsuspend|tuxonice)
+>>>>>>> scripts/local-premount/resume
 	echo "${MAJMIN}" > /sys/power/resume
 	;;
 esac