Kernel upgrade has broken suspend

romstor · 2019-12-09 14:21:35

Flapper wrote:

Have we solved this (it's not solved for everyone by the looks of it)?

It solves the original issue you've posted - so I'd vote yes. I'll need to do my own git bisect process -- but it will take many days on that old laptop. Will see - may be over coming holidays.

loqs · 2019-12-09 21:26:28

@romstor have you checked the contents of /proc/cmdline for init_on_alloc=0

bambooCZ · 2019-12-10 05:36:52

Same here. LTS kernel sleeps like a baby, normal doesn't sleep at all and refuses to wake. init_on_alloc=0 makes no difference. HP Elitebook 840 G6

romstor · 2019-12-10 12:58:12

loqs wrote:

@romstor have you checked the contents of /proc/cmdline for init_on_alloc=0

Yes, it was properly picked up. I even tried to add both init_on_alloc=0 and init_on_free=0 (and combinations thereof) - no luck. May be my issue is different.

bambooCZ · 2019-12-10 20:39:18

I forgot to turn off automatic suspend after a time and strangely enough the PC was successfully suspended when I came to it. I have troubleshooted audio problems before that. So maybe snd_hda_intel might interfere here? Possibly snd_hda_intel.dmic_detect=0 might have made it work?

Maybe the PC have enough time to finish the suspend procedure? I have 32GB of RAM. So if it should be dumped to disk it could take a while.

I have no time to debug it tho. I am just sharing my experience. Suspend now works as expected for me.

loqs · 2019-12-10 21:01:33

@bambooCZ was your issue triggered by the 5.3.12 kernel update?

bambooCZ · 2019-12-10 22:45:47

@loqs Not exactly. My issue was on new installation. I have tried LTS kernel when I was troubleshooting it. The system was about 5 hours old.

Sound was also working with the LTS kernel, but not with the normal one.

Last edited by bambooCZ (2019-12-10 22:47:35)

ugjka · 2019-12-11 04:52:25

https://01.org/blogs/rzhang/2015/best-p … ate-issues

throbscottle · 2019-12-15 22:06:28

I had a version of same problem, see separate thread if interested: https://bbs.archlinux.org/viewtopic.php … 3#p1877773
Thanks to everyone who worked towards this fix

chrisdb · 2019-12-16 07:57:00

loqs wrote:

lschu wrote:
I plan to give those a go this weekend if I have time to get around to it.
I would suggest running the first two builds overnight as they can take a long time.

Just curious to know, but the OP stated the build took longer than 6 hours, how much would you benefit if you used parallel builds ? I know OP does not have this option, just asking for myself because in my case I'm not always in the possibility to run things overnight....
Would it also be required to run overnight?

Wild Penguin · 2020-01-20 09:27:46

Getting OT (and old-ish thread), sorry about this...

chrisdb wrote:

Just curious to know, but the OP stated the build took longer than 6 hours, how much would you benefit if you used parallel builds ? I know OP does not have this option, just asking for myself because in my case I'm not always in the possibility to run things overnight....
Would it also be required to run overnight?

... but as this is the newbie corner I believe a few pieces of information is in order

Parallel builds means to use all / most cores of a CPU (depends on configuration). It will make a massive difference on modern systems, of course depending on the number of cores the CPU/system has. Even the least efficient CPUs have at least two cores -> compilation time will be at least halved on almost any computer these days! It is not uncommon to have 4, 6 or even 8 cores, however for a single compilation it might not scale that well to N cores (there's a limit to how well the compilation can parallelize, as some parts depend on others). By my gut feeling (*rumble*) I say it is possible to get a reduction of 80% (with >4 cores) for large compile tasks like the Kernel, but probably not much more than that.

Of course running overnight is just an suggestion (instead of overnight you might run it during a lunch break, while going to the GYM or, while browsing Reddit or Arch Forums .... YMMV!) :-D

(also having hibernate / suspend issues, but not at all sure if this thread is relevant to my problem; will post elsewhere once I'm sure where to post / if I have the time...)

Last edited by Wild Penguin (2020-01-20 16:08:32)

auxym · 2020-01-23 12:52:12

I'm also experiencing the same issue, and init_on_alloc=0 does not fix it. I did confirm it is present in /proc/cmdline.

PC is an old HP Elitebook 6930p from 2010ish. Intel graphics. Journalctl has nothing in it between a successful suspend, and then the hard reboot.

I will try to get my hands on an older kernel. Is this bug being tracked somewhere?

Edit: words are hard

Last edited by auxym (2020-01-23 23:36:45)

loqs · 2020-01-23 23:29:53

@auxym old packages are available in the ALA. No one has mentioned reporting the issue upstream.

auxym · 2020-01-25 00:22:43

Alright, I installed kernel v5.3.18-1 and suspend works fine. It does appear to be an issue with v5.4, and init_on_alloc=0 does not fix it for me.

loqs · 2020-01-25 00:37:24

@auxym I would start a new thread as your issue is not triggered by 5.3.12 or fixed by init_on_alloc=0
https://bbs.archlinux.org/viewtopic.php … 6#p1877006 contains step by step instructions to bisect between 5.3 and 5.4 to find the causal commit.

Arch Linux

#26 2019-12-09 14:21:35

Re: Kernel upgrade has broken suspend

#27 2019-12-09 21:26:28

Re: Kernel upgrade has broken suspend

#28 2019-12-10 05:36:52

Re: Kernel upgrade has broken suspend

#29 2019-12-10 12:58:12

Re: Kernel upgrade has broken suspend

#30 2019-12-10 20:39:18

Re: Kernel upgrade has broken suspend

#31 2019-12-10 21:01:33

Re: Kernel upgrade has broken suspend

#32 2019-12-10 22:45:47

Re: Kernel upgrade has broken suspend

#33 2019-12-11 04:52:25

Re: Kernel upgrade has broken suspend

#34 2019-12-15 22:06:28

Re: Kernel upgrade has broken suspend

#35 2019-12-16 07:57:00

Re: Kernel upgrade has broken suspend

#36 2020-01-20 09:27:46

Re: Kernel upgrade has broken suspend

#37 2020-01-23 12:52:12

Re: Kernel upgrade has broken suspend

#38 2020-01-23 23:29:53

Re: Kernel upgrade has broken suspend

#39 2020-01-25 00:22:43

Re: Kernel upgrade has broken suspend

#40 2020-01-25 00:37:24

Re: Kernel upgrade has broken suspend

Board footer