Comment 61 for bug 1328727

> So you got hibernate working now with pm-utils*and* the prop. Nvidia
> drivers. That's good - although a bit contrary to what you said in
> Comment 29:
>
I was told so, long time ago struggling to get nvidia prop to resume
from hibernation, I found out that uswsusp was better for it (googling
or on irc), and it indeed worked better, or with less effort at least,
back  then.  I made that comment thinking this was true but I just
proved myself wrong...

> till puzzles me is that while others are having problems,
> suspend-utils/uswsusp work for me almost 100 % of the time, except for a
> few extreme test-cases in the past. You also said that it worked
> "flawlessly" for yo

Yes! It worked pretty good on version 18.04.1 of ubuntu with kernel
4.15.0-42 and 41 using uswsusp. There was a long problem with nvidia
props that wouldn't let the system resume, but this was fixed when I
upgraded to the latest version of the 415 nvidia driver. I kept like one
month just hibernating to switch to windows and coming back to the
restored snapshot of linux.  You can check my apt history here:
https://launchpadlibrarian.net/415602746/aptHistory.log. At the
Start-Date: 2019-02-02 15:40:45, I'm 100% sure it was perfect. I am 100%
sure that it wasn't already working anymore having the s2disk freeze
issue at Start-Date: 2019-03-05 10:38:4.

uswsusp also worked fine on ubuntu 16.04, but I dont remember the kernel
versions. Now I'm currently with the nvidia 418.56, ubuntu 18.04.2,
kernel 4.18.0-17-generic and hibernation with pm-utils works. I haven't
found any major problem with it besides failing to suspend to ram
yesterday, which  I don't know if is related to it or not, but today I
tested it after and before hibernation and seems to be ok.

> So I'm wondering whether used-up swap space might play a role in this
> matter, too. At least for the cases that I've seen on my system, I can't
> rule this out. And when I look at the screenshot you provided in Comment
> 27 (https://launchpadlibrarian.net/417327528/i915.jpg), sparse
> swap-space could have been a factor in that case as well. Because
> roughly 3.5 GB free swap-space doesn't seem much for a 16-GB-RAM box.

On my many tests with uswsusp and a 16gb swap partition and 16 gb of
ram, I noticed that it would be less likely to fail when less than
something about 2 gb of ram, like just after boot up, it would though
after the 3rd or 4th followed hibernation cycle. If after the boot up I
allocate more than that value if would be much more likely to happen
like always on the 2nd attempt, and if more than around 6gb would fail
on the first attempt.

Those aren't sure values, sometimes it failed regardless of ram usage,
specially on my latest tests. also once it hibernated with more than
11gb ram usage and failed on the second attempt. So this is all
happening pretty randomly. What I described above is just most of the
cases and maybe this is just random anyway.