** Changed in: linux (Ubuntu)
Status: Incomplete => Won't Fix
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has become very slow
Status in l
** Changed in: linux (Ubuntu Yakkety)
Status: In Progress => Incomplete
** Changed in: linux (Ubuntu)
Status: In Progress => Incomplete
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.
Still getting slow boot and hangs with latest 4.8.0-37.39.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has become very slow
Status in linux packa
The fix for bug 1649905 should have helped impove this issue.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has become very slow
Status in linux pa
I booted to 4.8 again and saw the dmesg.
This is the error:
[3.375735] ...
[ 12.768084] [drm:drm_atomic_helper_commit_cleanup_done [drm_kms_helper]]
*ERROR* [CRTC:26:pipe A] flip_done timed out
Apparently it is a known problem
https://bbs.archlinux.org/viewtopic.php?id=218581 related to th
So I did a little bisecting today on mainline x86_64 kernels with
CONFIG_SLUB:
v4.4 Startup finished in 2.718s (kernel) + 11.831s
(userspace) = 14.549s
v4.5 Startup finished in 2.707s (kernel) + 12.777s
(userspace) = 15.484s
v4.6-rc4-1018-g5a5f0792
For completeness, the latest mainline.
4.10.0rc6
Startup finished in 15.387s (kernel) + 25.344s (userspace) = 40.731s
We can definitely conclude something happens between 4.7.10 and
4.8.0rc1. Detailed bisection should happen there.
--
You received this bug notification because you are a member
I decided to test with some of the mainline kernel images posted at
http://kernel.ubuntu.com/~kernel-ppa/mainline/
4.6.4
Startup finished in 4.991s (kernel) + 23.504s (userspace) = 28.496s
4.6.6
Startup finished in 5.019s (kernel) + 24.044s (userspace) = 29.064s
4.6.7
Startup finished in 5.063s
I had slow boot with 16.10 few months ago. I moved back to 16.04. Today
I had time to test some of the kernels above from the bisection process.
These are the results
4.4.0-59-generic# Ubuntu 16.04 official January 2017
Startup finished in 4.889s (kernel) + 28.001s (userspace) = 32.890s
4
Are you guys aware of any other patches that went to mainline and may
affect this? I've been hit by this issue since 4.6/4.7-ish days and can
still reproduce it with 4.9.7. The two patches do help, but it's far
from perfect:
4.9.7:
Startup finished in 3.498s (kernel) + 20.462s (userspace) = 23.961
The patches we testing have now landed in mainline. Does this bug still
happen on Yakkety with the latest updates?
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 reg
We did in fact test the yakkety kernel with those two patched on Oct 21,
per comments 23 and 24.
However, the test kernel I built with those two patches was 4.8.0-26.
The last 4.8 based test kernel I posted was built with the two patches
against Yakkety -proposed. That kernel is now at 4.8.0-29.
> http://kernel.ubuntu.com/~jsalisbury/lp1626436/yakkety-proposed-
patched
This is a GOOD kernel as well -- booting is almost as fast as with 4.4,
and the load doesn't explode under sbuild either.
I'm a bit confused now, I thought we already tried the two patches in
isolation -- but I think *only
Yes, that result is unexpected. So it sounds like v4.9-rc5 without the
patches is still bad, but adding the two patches fixes the bug.
I built another -proposed Yakkety test kernel with those two patches. It can be
downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/yakkety-proposed
> I noticed the comment does not have the download link. It is:
> http://kernel.ubuntu.com/~jsalisbury/lp1626436/mailine-with-two-commits/
Bazinga! Perhaps unexpectedly, this is a GOOD kernel -- boot speed is
comparable with 4.4, there is no high load after booting any more, and
shutdown is fast a
The kernel in #57 did not include the patches, but the kernel in comment
#59 is current mainline and it DOES have the patches.
I noticed the comment does not have the download link. It is:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/mailine-with-two-commits/
I just wanted to confirm the mainl
ah, just 4 sounds fine indeed. I did test 4.9-rc5 yesterday already (see
comment #57), just not sure if that includes "the" the two commits.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/
We only have about 4 more kernels to test. We could try making changes
to the changes with cgroups. We could even revert some of them.
However, it could be hit or miss. A bisect will tell us exactly which
commit started this, and then we can examine that commit to know what is
going on.
Just as
> http://kernel.ubuntu.com/~kernel-ppa/mainline/v4.9-rc5/
This still shows unchanged behaviour: slow boot and shutdown, and high
load after boot. So unfortunately it didn't magically fix itself :(
--
You received this bug notification because you are a member of Kernel
Packages, which is subscri
Is there something which could help to cut down the bisect? I had a
theory in comment 12 about the plethora of cgroup related uevents.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Since the mainline kernel is now at v4.9-rc5, it might be worth giving
it a test. If it fixes the boot slowness, time would be better spent
looking for the fix, versus continuing a search on what broke it.
V4.9-rc5 can be downloaded from:
http://kernel.ubuntu.com/~kernel-ppa/mainline/v4.9-rc5/
-
Ignore my last test kernel, I forgot to add the two patches. I'll build
a new one.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has become very sl
I built the next test kernel, up to the commit:
caee57ec71e1a0cb50b6028c706cfe541dcb080a
This kernel also has the two patches.
The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/caee57ec71e1a0cb50b6028c706cfe541dcb080a
Can you test that kernel and report back
>
http://kernel.ubuntu.com/~jsalisbury/lp1626436/e257ef55ce51d7ec399193ee85acda8b8759d930
Still GOOD.
Thanks!
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regress
I built the next test kernel, up to the commit:
e257ef55ce51d7ec399193ee85acda8b8759d930
This kernel also has the two patches.
The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/e257ef55ce51d7ec399193ee85acda8b8759d930
Can you test that kernel and report back
>
http://kernel.ubuntu.com/~jsalisbury/lp1626436/58f8b094e96f12e899bf767fc658c165908065d4
GOOD
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has be
I built the next test kernel, up to the commit:
58f8b094e96f12e899bf767fc658c165908065d4
This kernel also has the two patches.
The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/58f8b094e96f12e899bf767fc658c165908065d4
Can you test that kernel and report back
>
http://kernel.ubuntu.com/~jsalisbury/lp1626436/1acd010152138644f63d743e165161edc780fc32
GOOD
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has be
I built the next test kernel, up to the commit:
1acd010152138644f63d743e165161edc780fc32
This kernel also has the two patches.
The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/1acd010152138644f63d743e165161edc780fc32
Can you test that kernel and report back
>
http://kernel.ubuntu.com/~jsalisbury/lp1626436/a0d3c7c5c07cfbe00ab89438ddf82482f5a99422
Still GOOD.
I'm becoming a bit nervous, but I just re-tested 4.7rc1 again and it's
still bad.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux
I built the next test kernel, up to the commit:
a0d3c7c5c07cfbe00ab89438ddf82482f5a99422
This kernel also has the two patches.
The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/a0d3c7c5c07cfbe00ab89438ddf82482f5a99422
Can you test that kernel and report back
>
http://kernel.ubuntu.com/~jsalisbury/lp1626436/07b75260ebc2c789724c594d7eaf0194fa47b3be
GOOD
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has be
I built the next test kernel, up to the commit:
07b75260ebc2c789724c594d7eaf0194fa47b3be
This kernel also has the two patches.
The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/07b75260ebc2c789724c594d7eaf0194fa47b3be
Can you test that kernel and report back
> The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/0efacbbaee1e94e9942da0912f5b46ffd45a74bd
This is a GOOD kernel.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchp
I built the next test kernel, up to the same commit:
0efacbbaee1e94e9942da0912f5b46ffd45a74bd
This kernel also has the two patches.
The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/0efacbbaee1e94e9942da0912f5b46ffd45a74bd
Can you test that kernel and report
>
http://kernel.ubuntu.com/~jsalisbury/lp1626436/4a5219edcdae52bfb5eea0dfc2a7bd575961dad7
I do see the files now. Sorry for the delay, travel/sprint and all.
This is a GOOD kernel.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in
Can you see the files now?
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has become very slow
Status in linux package in Ubuntu:
In Progress
Stat
>
http://kernel.ubuntu.com/~jsalisbury/lp1626436/4a5219edcdae52bfb5eea0dfc2a7bd575961dad7
this is empty
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] bo
I built the another test kernel, up to the same commit:
4a5219edcdae52bfb5eea0dfc2a7bd575961dad7
This kernel also has the two patches.
The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/4a5219edcdae52bfb5eea0dfc2a7bd575961dad7
Can you test that kernel and repo
> I built the another test kernel, up to the same commit:
a7fd20d1c476af4563e66865213474a2f9f473a4
This is GOOD.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regre
I built the another test kernel, up to the same commit:
a7fd20d1c476af4563e66865213474a2f9f473a4
This kernel also has the two patches.
The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/a7fd20d1c476af4563e66865213474a2f9f473a4
Can you test that kernel and repo
> In the meantime, it might be worthwhile to test v4.9-rc2. It's available from:
> http://kernel.ubuntu.com/~kernel-ppa/mainline/v4.9-rc2/
Still slow boot/high load, i. e. BAD.
> I built the another test kernel, up to the same commit:
> 7afd16f882887c9adc69cd1794f5e5723217
That's GOOD.
Than
I built the another test kernel, up to the same commit:
7afd16f882887c9adc69cd1794f5e5723217
This kernel also has the two patches.
The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436
Can you test that kernel and report back if it has the bug or not? I
will
Yes, I'll rebuild that test kernel with the two patches. The second
patched does not apply cleanly due to where the bisect is in time
currently.
In the meantime, it might be worthwhile to test v4.9-rc2. It's available from:
http://kernel.ubuntu.com/~kernel-ppa/mainline/v4.9-rc2/
--
You receive
> I built the first test kernel, up to the following commit:
> 7afd16f882887c9adc69cd1794f5e5723217
Like Florian, the result isn't unambiguous. It for sure boots much
faster than 4.7/4.8 (~ 3s kernel + 5s userspace), but still much slower
than 4.4. But this could be attributed to the #1 issue
Thanks for the update, Florian. Can you also test the kernel posted in
comment #29, Martin? If you also see good boot speed, I'll mark that
kernel as good and build the next one.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in U
Thanks Joseph!
I now believe I'm seeing two different issues in 4.8 that are not in 4.4 and
slow down my boot, which is very confusing.
- #1: High number of kworkers (fixed in 4.8 using the two patches and bisected
by Doug Smythies already)
- #2: Unknown other issue. I've noticed that when compi
I started a kernel bisect between v4.6 final and v4.7-rc1. The kernel
bisect will require testing of about 7-10 test kernels.
I built the first test kernel, up to the following commit:
7afd16f882887c9adc69cd1794f5e5723217
The test kernel can be downloaded from:
http://kernel.ubuntu.com/~jsali
** Changed in: linux (Ubuntu)
Importance: Medium => High
** Changed in: linux (Ubuntu Yakkety)
Importance: Medium => High
** Changed in: linux (Ubuntu)
Status: Triaged => In Progress
** Changed in: linux (Ubuntu Yakkety)
Status: Triaged => In Progress
** Changed in: linux (U
@Florian, I can built the test kernels for the bisect. Per comments
#17 and #19, we know that 4.6 is good and 4.7-rc1 is bad, so I'll bisect
between those two versions. I'll build the first test kernel and post
it shortly.
--
You received this bug notification because you are a member of Kerne
So, as another datapoint, I have built the same package with sbuild on
4.4 and 4.8.
I grabbed 4.4 from xenial, rebooted into it. built the same trival package that
i built yesterday on 4.8
$ ( cd /home/smoser/ubuntu/logs/ && grep "^Build needed"
smello_0.4~ppa1_amd64-2016-10-2*)
smello_0.4~ppa1_
My boot time comparison: (Wall clock time from GRUB to lightdm ready for login)
4.4: 12.28s
4.8.0-26 + Patches (Joseph's kernel): 21.78s
4.8.0-22 (Yakkety): 25.06s
This is with NetworkManager-wait-online.service enabled, which seems to
take ~8s in my case.
The attached file has systemd-analyze (b
Thanks Joseph! I'm seeing the same effect described by Martin.
Boot+lightdm+loading Unity 7 is not as fast as it used to be in 16.04.
I need to compare timings (4.8 patched vs 4.8 vs 4.4) tomorrow to make
sure, but I also feel like it's a little faster now with the patches
applied.
The number of
Thanks Florian and Joseph! With the two patches it's definitively a lot
better! Boot times without NM and lightdm:
4.4: 1.5s
4.8.0-22 (zesty): 16.8s
4.8.0-26 (Joseph's kernel): 8.4s
So still quite far from what we used to have, but already twice as fast
as the current y/z kernel.
--
You receive
I built a Yakkety test kernel with those two patches. It can be downloaded
from:
http://kernel.ubuntu.com/~jsalisbury/lp1626436/
Can you test this kernel and see if it resolves this bug?
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to l
I'll build a test kernel with those two patches and post a link to it
shortly.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has become very slow
S
Well, I guess the other patch is required as well for SLUB (not only SLAB as I
assumed at first):
https://patchwork.kernel.org/patch/9361853/
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs
My problem is probably the same that Doug Smythies described in bug
1626564 . I used the "echo workqueue:workqueue_queue_work >
/sys/kernel/debug/tracing/set_event" method and had about 1000 work
items enqueued by "memcg_kmem_cache_create_func" all at pretty much the
same time, which correlates nic
> 4.7-rc1: http://kernel.ubuntu.com/~kernel-
ppa/mainline/v4.7-rc1-yakkety/
bad
Thanks Joseph for taking this!
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regres
Thanks for testing. We need to narrow it down a bit more by testing
some of the v4.7 release candidates. Can you test v4.7-rc1? It can be
downloaded from:
4.7-rc1: http://kernel.ubuntu.com/~kernel-ppa/mainline/v4.7-rc1-yakkety/
--
You received this bug notification because you are a member of
> 4.5 final: http://kernel.ubuntu.com/~kernel-ppa/mainline/v4.5-wily/
good
> 4.6 final: http://kernel.ubuntu.com/~kernel-ppa/mainline/v4.6-yakkety/
good
> 4.7 final: http://kernel.ubuntu.com/~kernel-ppa/mainline/v4.7/
bad
--
You received this bug notification because you are a member of Kern
I'd like to perform a bisect to figure out what commit caused this
regression. We need to identify the earliest kernel where the issue
started happening as well as the latest kernel that did not have this
issue.
Can you test the following kernels and report back? We are looking for
the first kerne
** Description changed:
With yakkety's recent update from linux 4.4 to 4.8 booting has become a
lot slower. It's not one service in particular, but without "quiet" and
"splash" you can now easily read every single line instead of that whole
wall of text zipping by. It now takes over 20s in
This is what I get right after boot and login into Unity 7 (executed in Gnome
Terminal):
$ ps aux | grep kworker | wc -l
1004
Kernel is 4.8.0-22-generic. After a few minutes, only 21 kworkers remain. Load
after 4 minutes is not excessive (1244 sleeping processes), forgot to check
load right aft
Confirmed with 4.8.0-040800.201610022031 from http://kernel.ubuntu.com
/~kernel-ppa/mainline/v4.8/
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has
Same here, downgrading to 4.4 from Xenial improved the situation (boot
still feels a little slower than Xenial though, maybe because of the
NetworkManager bug). I use a 7 years old desktop computer based on a
Gigabyte motherboard and AMD Phenom II X4 955 CPU, So it's not Thinkpad-
related.
(From G
This could potentially be related to changes with cgroups -- creating
them did not cause any uevent in 4.4, but with 4.8 they do:
sudo systemctl stop systemd-timesyncd
udevadm monitor -k # in another terminal
sudo systemctl start systemd-timesyncd
KERNEL[393.260769] add
/kernel/slab
For the record, I tried Joseph's test bisect kernel in bug 1627108, and
this does not fix it, so this is not a duplicate of bug 1627108:
4.4.0-9136 (previous yakkety): 2.849s (kernel) + 4.639s (userspace)
4.8.0-19 (current yakkety): 5.145s (kernel) + 11.825s (userspace)
4.7.0-040700rc3 (from Josep
Still confirmed with yesterday's 4.8.0-19.21.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has become very slow
Status in linux package in Ubuntu:
I marked the "affects me too" button at the top, but just wanted to add
that we're seeing slow boots on EC2 also. With the latest yakkety AMI on
ec2 (ami-1bc5820c, us-east-1, m4.large) first and subsequent boots are
are much slower than expected.
Top 5 in the blame:
$ systemd-analyze blame
2m
Unduplicating. This still happens with 4.8.0-17 from the PPA, both the
slow boots and the high load times. It got a fair bit better, though
(load of ~ 35 instead of ~ 250), but still a huge regression compared to
4.4.
** This bug is no longer a duplicate of bug 1626564
4.8 regression: SLAB is b
*** This bug is a duplicate of bug 1626564 ***
https://bugs.launchpad.net/bugs/1626564
As far as I understand Andy, bug 1626564 is the root cause, thus duping.
** This bug has been marked a duplicate of bug 1626564
4.8 regression: SLAB is being used instead of SLUB
--
You received this b
For the record, Andy already has/knows the fix for this.
** Changed in: linux (Ubuntu)
Status: Confirmed => In Progress
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title
happens to me as well, on XPS 13 - 2013. Ubuntu always loaded before the
"second dot" became orange. Now it fills up 15 dots before loading the login
screen, and happened after upgrading 4.4 -> 4.8.
Everything else works ok.
--
You received this bug notification because you are a member of Ker
** Changed in: linux (Ubuntu)
Importance: Undecided => Medium
** Tags added: kernel-da-key
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has bec
I do not see this in QEMU (even with -smp 4), only on my laptop. At this
point I'm not sure if it is specific to my ThinkPad X230 or happens for
others too.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.laun
Confirmed with 4.8.0-14.15-generic from yakkety-proposed.
Immediately after boot the system load is also reported as ~ 250, and
then slowly goes down towards zero.
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://b
** Tags added: kernel-4.8
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1626436
Title:
[4.8 regression] boot has become very slow
Status in linux package in Ubuntu:
New
Bug descript
78 matches
Mail list logo
