** Description changed:
[ Impact ]
When compared to Ubuntu 23.10, creating intramfs files with update-
initramfs takes 2 to 5 times more time on ARM devices.
IIUC, dracut-install usage was added to initramfs-tools to speed up the
process. But now its way slower. Even running update
Thanks for the follow-up improvements.
I think, though, that they should be applied under a new bug, instead of
using this one here again. If we have to chase down future regressions,
or even just understand the history of these changes, that is already
becoming hard here, with over 70 comments an
I implemented the change suggested in #2065180-44 in the safest possible
way. Benchmark on a Raspberry Pi Zero 2W on Ubuntu 24.04 (noble) on
2024-09-20. initramfs-tools 0.142ubuntu25.2 and dracut-install
060+5-1ubuntu3.2:
```
bdrung@zero2w:~$ sudo hyperfine --warmup 1 -r 10 "update-initramfs -u"
This bug was fixed in the package cloud-initramfs-tools - 0.49~24.04.1
---
cloud-initramfs-tools (0.49~24.04.1) noble; urgency=medium
* No-change backport of bugfix-only release to Noble.
cloud-initramfs-tools (0.49) oracular; urgency=medium
[ Benjamin Drung ]
* overlayroot: C
This bug was fixed in the package lvm2 - 2.03.16-3ubuntu3.1
---
lvm2 (2.03.16-3ubuntu3.1) noble; urgency=medium
* initramfs-tools hook: Combine calls to manual_add_modules (LP:
#2065180)
-- Benjamin Drung Mon, 01 Jul 2024 20:45:03 +0200
** Changed in: lvm2 (Ubuntu Noble)
This bug was fixed in the package initramfs-tools - 0.142ubuntu25.2
---
initramfs-tools (0.142ubuntu25.2) noble; urgency=medium
[ Jeffery To ]
* Fix resume failure due to resume=UUID=... in certain cases (LP: #1769297)
* hooks/resume: Fix resume device type check
[ Benjamin D
This bug was fixed in the package cryptsetup - 2:2.7.0-1ubuntu4.1
---
cryptsetup (2:2.7.0-1ubuntu4.1) noble; urgency=medium
* initramfs hook: Combine calls to manual_add_modules (LP: #2065180)
-- Benjamin Drung Mon, 01 Jul 2024 20:38:08 +0200
** Changed in: cryptsetup (Ubuntu N
Tested on amd64 laptop in a chroot. Without noble-proposed:
```
$ schroot-wrapper -c noble -u root -p
initramfs-tools,cloud-initramfs-copymods,cloud-initramfs-rescuevol,cloud-initramfs-dyn-netconf,cloud-initramfs-growroot,cloud-initramfs-rooturl,cloud-initramfs-updateroot,overlayroot,hyperfine,li
Hello Viraniac, or anyone else affected,
Accepted cloud-initramfs-tools into noble-proposed. The package will
build now and be available at
https://launchpad.net/ubuntu/+source/cloud-initramfs-tools/0.49~24.04.1
in a few hours, and then in the -proposed repository.
Please help us by testing this
This bug was fixed in the package dracut - 060+5-1ubuntu3.2
---
dracut (060+5-1ubuntu3.2) noble; urgency=medium
* Cherry-pick upstream performance fixes (LP: #2065180):
- perf(dracut-install): memoize find_kmod_module_from_sysfs_node
- perf(dracut-install): use driver/module
This bug was fixed in the package thin-provisioning-tools -
0.9.0-2ubuntu5.1
---
thin-provisioning-tools (0.9.0-2ubuntu5.1) noble; urgency=medium
* initramfs-hook: Combine calls to manual_add_modules (LP: #2065180)
-- Benjamin Drung Mon, 01 Jul 2024 20:48:29 +0200
** Changed in
This bug was fixed in the package open-iscsi - 2.1.9-3ubuntu5.1
---
open-iscsi (2.1.9-3ubuntu5.1) noble; urgency=medium
* initramfs-hook: Combine calls to manual_add_modules (LP: #2065180)
-- Benjamin Drung Mon, 01 Jul 2024 21:01:25 +0200
** Changed in: open-iscsi (Ubuntu Noble
Note that initramfs-tools is also fixing
https://bugs.launchpad.net/ubuntu/+source/initramfs-tools/+bug/1769297,
which has not been verified yet.
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
There are a bunch of autopkgtest failures. I retriggered some runs,
after asking #is to restart neutron in the s390x cloud, as per standing
workaround for RT #155441. I'll try to check back on these results still
during my shift today.
--
You received this bug notification because you are a membe
Regarding cloud-initrafms-tools (comment #51): initramfs-tools calls
/usr/lib/dracut/dracut-install with the parameter -o. This ignores
missing kernel modules. So combining the manual_add_modules calls in
cloud-initrafms-tools are okay.
--
You received this bug notification because you are a memb
** Tags removed: verification-needed-noble
** Tags added: verification-done-noble
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
performance regression in dracut-install 060
To manage
I successfully verified the SRU on my Raspberry Pi Zero 2W:
```
bdrung@zero2w:~$ sudo cp /boot/initrd.img-6.8.0-1007-raspi
/boot/initrd.img-6.8.0-1007-raspi.current-noble
bdrung@zero2w:~$ sudo sh -c '3cpio -t
/boot/initrd.img-6.8.0-1007-raspi.current-noble >
/boot/initrd.img-6.8.0-1007-raspi.cu
> The proposal in comment #2065180-44 is interesting. That would address
user hooks. But what about hooks shipped by packages outside the Ubuntu
archive?
We don't support packages shipped outside the Ubuntu archive. If they
break, they break.
But it's not *hugegly* likely that they'll break, eith
The proposal in comment #2065180-44 is interesting. That would address
user hooks. But what about hooks shipped by packages outside the Ubuntu
archive?
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/206
Ok. Everything but the cloud-initrafms-tools looks OK (you're still
welcome to upload something along the lines of
https://bugs.launchpad.net/ubuntu/+source/dracut/+bug/2065180/comments/44).
For cloud-initramfs-tools it seems like combining the silent-failure on
lack of intel-aesni with the rest m
Hello Viraniac, or anyone else affected,
Accepted dracut into noble-proposed. The package will build now and be
available at
https://launchpad.net/ubuntu/+source/dracut/060+5-1ubuntu3.2 in a few
hours, and then in the -proposed repository.
Please help us by testing this new package. See
https://
Hello Viraniac, or anyone else affected,
Accepted cryptsetup into noble-proposed. The package will build now and
be available at
https://launchpad.net/ubuntu/+source/cryptsetup/2:2.7.0-1ubuntu4.1 in a
few hours, and then in the -proposed repository.
Please help us by testing this new package. Se
Hello Viraniac, or anyone else affected,
Accepted initramfs-tools into noble-proposed. The package will build now
and be available at https://launchpad.net/ubuntu/+source/initramfs-
tools/0.142ubuntu25.2 in a few hours, and then in the -proposed
repository.
Please help us by testing this new pack
I'm not a bash expert by any means, but something like this would appear to
work?
```
function manual_add_modules() {
... normal stuff goes here ...
if [ $IN_USER_CONFIG -gt 0 ]; then
apply_add_modules
fi
}
...
in mkinitramfs:
...
IN_USER_CONFIG=0
...
apply_add_modules
run_scrip
This bug was fixed in the package dracut - 102-3ubuntu4
---
dracut (102-3ubuntu4) oracular; urgency=medium
* Cherry-pick upstream fixes for systemd 256 (LP: #2069290):
- fix(test): use --add instead of --modules to create test-makeroot
- test: avoid writing to rootfs as it m
Resetting dracut since there are more performance fixes for dracut-
install
** Changed in: dracut (Ubuntu)
Status: Fix Released => Fix Committed
** Changed in: dracut (Ubuntu Noble)
Status: Fix Released => New
--
You received this bug notification because you are a member of Ubunt
Another dracut-install commit that we can pick:
https://github.com/dracut-ng/dracut-ng/pull/479
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
performance regression in dracut-install
** Changed in: initramfs-tools (Ubuntu Noble)
Status: Incomplete => New
** Tags removed: foundations-todo
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
performance regression
The problem is that a custom hook might rely on the behavior that all
kernel modules were copied to $DESTDIR.
For the SRU I am playing the safe card now as documented in the "Reduce
manual_add_modules calls" section of the bug description.
--
You received this bug notification because you are a
> I don't know how to differentiate between if `manual_add_modules` was
called by a script in /usr/share/initramfs-tools or from outside.
I *think* you could have `manual_add_modules` check the environment for
something like `INITRAMFS_TOOLS_IN_USER_CONFIG=1` or something, and
change behaviour bas
This bug was fixed in the package cloud-initramfs-tools - 0.49
---
cloud-initramfs-tools (0.49) oracular; urgency=medium
[ Benjamin Drung ]
* overlayroot: Combine calls to manual_add_modules (LP: #2065180)
-- Paride Legovini Tue, 02 Jul 2024 12:38:14 +0200
** Changed in: clou
The second iteration of the noble SRU: Reduce the number of dracut-
install calls from 51 down to 26. This involves adjusting six packages:
cryptsetup (2 -> 1), lvm2 (8 -> 1), thin-provisioning-tools (3 -> 1),
open-iscsi (9 -> 1), cloud-initramfs-tools (5 -> 1), and initramfs-tools
itself (8 -> 5).
** Description changed:
[ Impact ]
When compared to Ubuntu 23.10, creating intramfs files with update-
initramfs takes 2 to 5 times more time on ARM devices.
IIUC, dracut-install usage was added to initramfs-tools to speed up the
process. But now its way slower. Even running update
Hi Benjamin,
I am also not seeing much difference on VIM4. The dracut-install version
060+5-1ubuntu3.2~ppa1 is providing only 200-300 ms improvement over
060+5-1ubuntu3.1 version.
On Tue, Jul 2, 2024 at 12:10 AM Benjamin Drung <2065...@bugs.launchpad.net>
wrote:
> There is another performance i
** Merge proposal unlinked:
https://code.launchpad.net/~bdrung/ubuntu/+source/initramfs-tools/+git/initramfs-tools/+merge/468583
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
perf
** Merge proposal linked:
https://code.launchpad.net/~bdrung/ubuntu/+source/initramfs-tools/+git/initramfs-tools/+merge/468583
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
perfor
** Changed in: cloud-initramfs-tools (Ubuntu)
Status: New => Fix Committed
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
performance regression in dracut-install 060
To manage
** Also affects: cloud-initramfs-tools (Ubuntu)
Importance: Undecided
Status: New
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
performance regression in dracut-install 060
** Merge proposal linked:
https://code.launchpad.net/~bdrung/cloud-initramfs-tools/+git/cloud-initramfs-tools/+merge/468579
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
performan
** Also affects: open-iscsi (Ubuntu)
Importance: Undecided
Status: New
** Changed in: open-iscsi (Ubuntu)
Status: New => Invalid
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
There is another performance improvement upstream:
https://github.com/dracut-ng/dracut-ng/pull/408
I tested this change a Raspberry Pi Zero 2W, but it had no measurable
performance improvement:
```
$ sudo hyperfine --warmup 1 -r 10 "update-initramfs -u"
Benchmark 1: update-initramfs -u
Time (me
There are 51 dracut-install calls on my Raspberry Pi Zero 2W on Ubuntu
24.04:
```
dracut-install -m -P
/hid-(a4tech|cypress|dr|elecom|gyration|icade|kensington|kye|lcpower|magicmouse|ntrig|petalynx|picolcd|pl|ps3remote|quanta|roccat-ko.*|roccat-pyra|saitek|sensor-hub|sony|speedlink|tivo|twinhan|u
** Description changed:
[ Impact ]
When compared to Ubuntu 23.10, creating intramfs files with update-
initramfs takes 2 to 5 times more time on ARM devices.
IIUC, dracut-install usage was added to initramfs-tools to speed up the
process. But now its way slower. Even running update
** Description changed:
[ Impact ]
When compared to Ubuntu 23.10, creating intramfs files with update-
initramfs takes 2 to 5 times more time on ARM devices.
IIUC, dracut-install usage was added to initramfs-tools to speed up the
process. But now its way slower. Even running update
** Tags removed: fountations-todo
** Tags added: foundations-todo
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
performance regression in dracut-install 060
To manage notifications a
** Tags added: fountations-todo
** Changed in: initramfs-tools (Ubuntu Noble)
Assignee: (unassigned) => Benjamin Drung (bdrung)
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
per
This bug was fixed in the package dracut - 060+5-1ubuntu3.1
---
dracut (060+5-1ubuntu3.1) noble; urgency=medium
* perf(dracut-install): preload kmod resources for quicker module lookup
(LP: #2065180)
-- Benjamin Drung Tue, 04 Jun 2024 17:21:56 +0200
** Changed in: dracut (U
I don't know how to differentiate between if `manual_add_modules` was
called by a script in /usr/share/initramfs-tools or from outside.
The only safer solution that I can come up with: Keep
`manual_add_modules` as it is and introduce a new function (e.g.
`manual_stage_modules`) that introduces the
Ooooh, right. The hook might want to call something like `depmod`
against the modules.
So, that's a reasonable thing for a user-configured hook to do, which
means we can't break it in an SRU. It *would* still be nice to get the
significant speedup you've got here, though. Could this change to
`man
`manual_add_modules` is exposed to user-configuration. We run
`apply_add_modules` after the hooks has been run (CONFDIR is the user
configuration directory):
```
run_scripts_optional "${CONFDIR}"/hooks
# cache boot run order
for b in $(cd "${DESTDIR}/scripts" && find . -mindepth 1 -type d); do
Correct me if I'm wrong, but isn't `manual_add_modules` exposed to user-
configuration? User configuration in /etc/initramfs-tools/hooks is
expected to include /usr/share/initramfs-tools/hook-functions, and so
making this API break has the possibility of causing currently working
user configuration
Thanks. So marking it as verification-done-noble for dracut.
** Tags removed: verification-needed verification-needed-noble
** Tags added: verification-done verification-done-noble
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https
Sorry forgot to mention about the contents before. I did checked the
contents. They were exactly the same.
On Mon, Jun 17, 2024 at 6:01 PM Benjamin Drung <2065...@bugs.launchpad.net>
wrote:
> Thanks for the test on the VIM4. Can you also verify that the content of
> the initrd hasn't change (see
Thanks for the test on the VIM4. Can you also verify that the content of
the initrd hasn't change (see "Test Plan" in the bug description)?
Further speedup will be achieved by the initramfs-tools change (that is
waiting in the SRU upload queue).
--
You received this bug notification because you
Tried on VIM4. It took 30s to build the initrd which is definitely way
better than 1m30s it was taking before. It still doesn't beat the 20s time
that is achievable just by reverting 131822e and 3de4c73 commits.
On Fri, Jun 14, 2024 at 8:01 PM Timo Aaltonen <2065...@bugs.launchpad.net>
wrote:
>
Hello Viraniac, or anyone else affected,
Accepted dracut into noble-proposed. The package will build now and be
available at
https://launchpad.net/ubuntu/+source/dracut/060+5-1ubuntu3.1 in a few
hours, and then in the -proposed repository.
Please help us by testing this new package. See
https://
As requested, results from running stock noble on the same Pi 4B with
the same SD card as before. First, stock noble (with all available
upgrades):
$ sudo hyperfine --warmup 1 -r 5 "update-initramfs -u"
Benchmark 1: update-initramfs -u
Time (mean ± σ): 193.558 s ± 2.334 s[User: 77.577 s
** Merge proposal unlinked:
https://code.launchpad.net/~jefferyto/ubuntu/+source/initramfs-tools/+git/initramfs-tools/+merge/467015
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
p
** Merge proposal linked:
https://code.launchpad.net/~jefferyto/ubuntu/+source/initramfs-tools/+git/initramfs-tools/+merge/467015
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
per
This bug was fixed in the package dracut - 060+5-8ubuntu2
---
dracut (060+5-8ubuntu2) oracular; urgency=medium
* perf(dracut-install): preload kmod resources for quicker module lookup
(LP: #2065180)
-- Benjamin Drung Tue, 04 Jun 2024 16:33:13 +0200
** Changed in: dracut (Ub
This bug was fixed in the package initramfs-tools - 0.142ubuntu28
---
initramfs-tools (0.142ubuntu28) oracular; urgency=medium
* hook-functions: Use firmware search order from kernel
* mkinitramfs: Resolve hidden dependencies after all modules were copied
* reduce number of drac
This bug was fixed in the package miniramfs - 1.0.2ubuntu1
---
miniramfs (1.0.2ubuntu1) oracular; urgency=medium
* Call apply_add_modules after manual_add_modules (LP: #2065180)
-- Benjamin Drung Tue, 04 Jun 2024 16:58:24 +0200
** Changed in: miniramfs (Ubuntu)
Status: F
I uploaded dracut, initramfs-tools, and miniramfs to oracular and the
SRUs for noble. For easier testing, I also uploaded the noble SRU
packages to https://launchpad.net/~bdrung/+archive/ubuntu/ppa
With the dracut and initramfs-tools SRUs the execution time on the
Raspberry Pi Zero 2W reduces to:
** Changed in: dracut (Ubuntu)
Importance: Undecided => High
** Changed in: dracut (Ubuntu)
Status: New => Fix Committed
** Changed in: miniramfs (Ubuntu)
Importance: Undecided => High
** Changed in: miniramfs (Ubuntu)
Status: New => Fix Committed
** Also affects: initramfs-
** Description changed:
+ [ Impact ]
+
When compared to Ubuntu 23.10, creating intramfs files with update-
initramfs takes 2 to 5 times more time on ARM devices.
IIUC, dracut-install usage was added to initramfs-tools to speed up the
process. But now its way slower. Even running update
** Also affects: lvm2 (Ubuntu Noble)
Importance: Undecided
Status: New
** Also affects: cryptsetup (Ubuntu Noble)
Importance: Undecided
Status: New
** Also affects: dracut (Ubuntu Noble)
Importance: Undecided
Status: New
** Also affects: thin-provisioning-tools (Ubu
While checking those packages, I found hooks that do not support zstd-
compressed kernel modules: bug #2068026
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
performance regression in
https://salsa.debian.org/kernel-team/initramfs-
tools/-/merge_requests/114 got merged after some iterations. This can be
a breaking change. To restore the previous behavior, call
`apply_add_modules` without arguments after a `manual_add_modules` call.
I checked all Ubuntu source packages that call
This bug was fixed in the package cryptsetup - 2:2.7.0-1ubuntu5
---
cryptsetup (2:2.7.0-1ubuntu5) oracular; urgency=medium
* initramfs hook: Combine calls to manual_add_modules (LP: #2065180)
-- Benjamin Drung Fri, 24 May 2024 09:48:09 +0200
** Changed in: cryptsetup (Ubuntu)
Submitted to Debian: https://salsa.debian.org/kernel-team/initramfs-
tools/-/merge_requests/114
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
performance regression in dracut-install
Results on a Pi 4B booting from SD card. Stock noble:
$ sudo hyperfine -r 5 "update-initramfs -u"
Benchmark 1: update-initramfs -u
Time (mean ± σ): 189.984 s ± 1.618 s[User: 75.720 s,
Stock noble on a Raspberry Pi Zero 2W:
```
bdrung@zero3:~$ sudo hyperfine -r 5 "update-initramfs -u"
Benchmark 1: update-initramfs -u
Time (mean ± σ): 415.664 s ± 6.015 s[User: 166.728 s, System:
232.523 s]
Range (min … max): 409.139 s … 422.632 s5 runs
```
Replace duplicate c
This bug was fixed in the package lvm2 - 2.03.16-3ubuntu4
---
lvm2 (2.03.16-3ubuntu4) oracular; urgency=medium
* initramfs-tools hook: Combine calls to manual_add_modules (LP:
#2065180)
-- Benjamin Drung Fri, 24 May 2024 09:42:08 +0200
** Changed in: lvm2 (Ubuntu)
Status
This bug was fixed in the package thin-provisioning-tools -
0.9.0-2ubuntu6
---
thin-provisioning-tools (0.9.0-2ubuntu6) oracular; urgency=medium
* initramfs-hook: Combine calls to manual_add_modules (LP: #2065180)
-- Benjamin Drung Fri, 24 May 2024 09:08:36 +0200
** Changed in:
I marked thin-provisioning-tools, lvm2, and cryptsetup as affected to
reduce the number of manual_add_modules calls in the initramfs-tools
hooks in those packages. This will help a bit, but will probably not be
enough to make it fast again.
** Also affects: thin-provisioning-tools (Ubuntu)
Impo
** Changed in: dracut
Status: Unknown => New
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
performance regression in dracut-install 060
To manage notifications about this bug
** Bug watch added: github.com/dracut-ng/dracut-ng/issues #316
https://github.com/dracut-ng/dracut-ng/issues/316
** Also affects: dracut via
https://github.com/dracut-ng/dracut-ng/issues/316
Importance: Unknown
Status: Unknown
--
You received this bug notification because you are
** Summary changed:
- performance regression in dracut-install
+ performance regression in dracut-install 060
--
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/2065180
Title:
performance regression in
78 matches
Mail list logo
