Hibernation

[michael-hiegl]

How about making hibernation to disc a default feature?

[claudiux]

https://askubuntu.com/questions/1258340/why-isnt-hibernation-available-by-default-in-ubuntu

[carsten88]

@claudiux
Thank you, so it IS possible, just takes some diskspace.

[carsten88]

Think about how much energy would be saved worldwide if this option was included and shipped as the default set option. This would be good news for the planet and all living being. Everyone I know just shuts the lid and leaves it on all night.

[cplittlfield]

I greatly prefer Hibernate to Suspend, given the price of electricity in my area. Hibernating worked just fine for me on LM 21.x, but I can't get it to work on LM 22.1. Apparently a particular parameter was removed by Ubuntu, and I haven't found where to restore it yet, even working from a solution from the Mint forums.

If you ask about hibernation on the forums, you get a large number of discouraging responses from the users who provide the best advice, on the grounds that it causes problems on many systems, which is why Ubuntu apparently stopped enabling hibernation by default. However, for users like me, who can use the feature successfully when it is enabled, it would be a boon.

[carsten88]

O well thats a shame. Maybe LMDE has this feature?

[Heavygrass]

I use hibernation every single day w/o any issue.
I can surely believe it can cause an issue on some systems, but that's not a reason to disable it for everybody.
One should device for him/herself whether its worth to use it or not.
Its especially useful for a laptop.

[alosslessdev]

It's possible but doing so would require diverting somewhat from ubuntu which has disabled it

[claudiux]

I often use 1.5 times the amount of RAM for swap, when the RAM is greater than or equal to 8 GiB (2 times the amount of RAM when the RAM is less than 8 GiB). No problems with hibernation.

[alosslessdev]

Part of the problem is that former Windows users are used to seeing a dynamic swap file, which enlarges and shrinks, depending on need. In Linux, swap space is fixed.

A package such as https://github.com/Tookmund/Swapspace could be used but then the kernel would need a patch to hibernate across several swapfiles. Maybe that swapspace code could be incorporated into the kernel too, but it would add technical debt. The right way to do things would be to remake the way the Linux kernel handles swap and implement dynamic swap in the kernel which no doubt would create pushback and would also need a lot of testing. It could be a kernel compile time option. Probably also add a limit in case a process is leaking memory.

[cplittlfield]

I'm not sure patching the kernel is necessary. I have read that Linux can use any and all swap space on the system, whether partitions or files. I can't find the passage I have in mind, but the swapon man page says

   Calls to swapon normally occur in the system boot scripts making all swap devices
   available, so that the paging and swapping activity is interleaved across several
   devices and files.

[Heavygrass]

At the very least, for hibernation to work, it is true that the size of the available swap space must exceed the amount of RAM on the system.

Are you willing to make a cash bet on your statement?
I'm sure we could find a site to do that. How about 500$+ bet?

Are you claiming that if swapfile is less than the amount of RAM the hibernation will NEVER EVER work?

So if they react this badly to 2 GB of swap, what will their reaction be when the installer creates a 38 GB swap file?

Why does installer has to create any swap file at all?
What does it have to do with Cinnamon specifically?

Just how much extra swap is required might be debatable

So how much exactly you're claiming is the minimum extra swap?

[alosslessdev]

what about crating the swap file at hibernate time, only if there's enough space? Or, if the laptop is about to shut down, do it after the battery reaches a certain threshold? maybe even based on battery wear levels? The linux kernel should have a way to calculate how much swap is needed for hibernation, maybe it is current swap use + used ram that isn't available?

[cplittlfield]

I'm going only by what I've read on the Linux Mint and AskUbuntu forums. As recently as LM 21.3, a swap file was required for writing the RAM image and restoring from it, and I am quoting the recommendations I've read. The kernel documentation at www.kernel.org says the following:

Hibernation

This state (also referred to as Suspend-to-Disk or STD) offers the greatest energy savings and can be used even in the absence of low-level platform support for system suspend. However, it requires some low-level code for resuming the system to be present for the underlying CPU architecture.

Hibernation is significantly different from any of the system suspend variants. It takes three system state changes to put it into hibernation and two system state changes to resume it.

First, when hibernation is triggered, the kernel stops all system activity and creates a snapshot image of memory to be written into persistent storage. Next, the system goes into a state in which the snapshot image can be saved, the image is written out and finally the system goes into the target low-power state in which power is cut from almost all of its hardware components, including memory, except for a limited set of wakeup devices.

Once the snapshot image has been written out, the system may either enter a special low-power state (like ACPI S4), or it may simply power down itself. Powering down means minimum power draw and it allows this mechanism to work on any system. However, entering a special low-power state may allow additional means of system wakeup to be used (e.g. pressing a key on the keyboard or opening a laptop lid).

If anyone is saying that the contents of RAM do not need to be written to disk for hibernating, I'd like to see a reference for that. If anyone is saying that the contents of RAM no longer need to be written to swap, then I'd like to see a reference for that, as well. I'd like to get hibernation re-enabled on my LM 22.1 setup (it worked just fine on LM 21.x), so if you were able to do that, please message me with instructions.

[Heavygrass]

I never asked you for any sources.
You made a generic statement, I asked you for your personal clarifications of your own statement. Period.
My questions to you are listed above (unchanged btw) and are very clear and specific and you keep avoiding every single one of my very specific questions again and again in various ways (for weeks already).
As of today you still have NOT answered a single of my questions above. Why?

[Heavygrass]

But don't try to change the goal posts, it won't work.

What of my "goal posts" you are talking about?
Please provide a citation of my "goal posts".

[cplittlfield]

I refer you to your own words:

In my world the one who makes a general statement MUST provide the proof for the statement. You made general statements about hibernation, not me.
Given that you ignored every single of my questions I am taking it as you are not standing behind any of your generic claims.
You refused to even answer what exactly is your statement about swap file size is.

These are the sources for my statement that swapfile size must equal or exceed RAM size for hibernation to work. You questioned it, I provided proof. Now, stop bothering me. I'm done here.

[Heavygrass]

You questioned it

So where have you found a request for proof?
You still refuse to even answer what exactly your statement about minimum swap size is, which I asked for weeks ago.
There's zero sense to talk about a proof until there's a clear statement.

. I'm done here.

Good finally, as I thought you don't stand behind yours statements. So no generic claims about the minimum swap size.

[Heavygrass]

These are the sources for my statement that swapfile size must equal or exceed RAM size for hibernation to work.

Let me repeat my weeks old question for you again:
"Are you claiming that if swapfile is less than the amount of RAM the hibernation will NEVER EVER work?".
FYI this is a YES/NO question.
Just in case: its NOT asking for any source/proof etc.

[alosslessdev]

Today I tested my machine a Lenovo 15ach6h on debian 13 with KDE plasma and hibernation works out of the box. No configuration required. Granted I haven't even tried installing Nvidia drivers. So I'm convinced that a big barrier is Ubuntu and undoing their changes. That maybe explains why some people are able to hibernate and others aren't. However hibernation takes much longer than windows and it only works once per boot. After you try it once subsequent attempts fail to shut down the machine I guess due to nouveau driver problems.

[cplittlfield]

That's news to me. On my HP with Intel graphics running the i915 driver, you could hibernate as many times as desired under 21.x. So you are probably right about it being something to do with the nouveau driver.