[RESPONDED] Linux deep sleep

[Richard_Tango-Lowy]

@Nils

Aren’t those lockups more associated with S3 sleep than S2idle?

It’s locked up once since I’ve made the change from deep to s2idle. It goes into hybrid sleep at least three or four times each day, and it was previously locking up every two or three days.

It appears that the change has helped with the lockups.

[Yibo_Wei]

On 11th gen laptop, resuming from deep sleep is not faster than resuming from hibernation. I prefer s2idle and sleep-then-hibernate for the best experience. I set the system to sleep of 60mins then switch to hibernation, which costs only 2% battery drain at most.

[Richard_Tango-Lowy]

@Yibo_Wei

I prefer s2idle and sleep-then-hibernate for the best experience. I set the system to sleep of 60mins then switch to hibernation, which costs only 2% battery drain at most.

I’m set up the same way and can verify the same behavior. I see about 2% drain under hibernation.

And haven’t had a mouse lockup during resume in in a few months, so I expect something fixed that one.

[tokanada]

@tokanada

For some reason, when I try to wake from deep sleep, my entire filesystem gets unmounted. Not sure if its because I have the same Sabrent Rocket 4.0 that someone else had problems with, or something with my kernel.

Luckily, however, suspend then hibernate works like a dream, even with s2idle enabled.

I decided to clone my ssd to an Intel 660p and now deep sleep works properly. There is definitely some kind of problem with Sabrent Rocket ssds and Linux.

[Bacter]

Fedora 37, kernel 6.1.5-200, battery drained from 98% to 75% (23% diff) for 12h of sleep (deep sleep enabled).
No SSD or other hardware issues

[Matt_Hartley]

@Bacter

Fedora 37, kernel 6.1.5-200, battery drained from 98% to 75% (23% diff) for 12h of sleep (deep sleep enabled).

How long was this in suspend for by chance?

[Bacter]

@Matt_Hartley

How long was this in suspend for by chance?

For 12 hours, sharp

[Matt_Hartley]

@Bacter

For 12 hours, sharp

So 12 hours suspended (deep sleep, not hibernate), that sounds about right. We’re measuring power drain to improve upon it (expansion cards, etc). However, there will be some power drain as it’s not in a proper hibernation state.

[mcz]

@Bacter

Fedora 37, kernel 6.1.5-200, battery drained from 98% to 75% (23% diff) for 12h of sleep (deep sleep enabled).
No SSD or other hardware issues

If you want to join the fun here:

https://community.frame.work/t/test-results-for-standby-battery-use-of-expansion-cards

[ConfuSomu]

@Bacter

Fedora 37, kernel 6.1.5-200, battery drained from 98% to 75% (23% diff) for 12h of sleep (deep sleep enabled).

I wonder if you would have less battery drain with s2idle sleep.

I’ve used deep sleep in the past, but since a few months ago, with the 6.x kernel series, I have less battery drain, and faster resume latency, when using s2idle sleep instead of deep sleep.

To use s2idle until the next reboot (as this doesn’t change the kernel command line on disk), write s2idle to /sys/power/mem_sleep. This should work on most distros:

sudo su -c "echo s2idle > /sys/power/mem_sleep"

I should mention that I have a 12th gen Intel Core i7-1260P, if there any differences in terms of sleep with the 11th generation. Confirmed on Linux kernel versions 6.0.6-arch1-1 and 6.1.5-arch2-1.

[Matt_Hartley]

@mcz

https://community.frame.work/t/test-results-for-standby-battery-use-of-expansion-cards

Let’s keep this on one thread for tracking please.

https://community.frame.work/t/test-results-for-standby-battery-use-of-expansion-cards