Question about pm-suspend backend...

pogeymanz · 2011-01-10 23:18:47

This is an excerpt from the default config for pm-utils

# The default sleep/wake system to use.  Valid values are:
#   kernel    The built-in kernel suspend/resume support.
#             Use this if nothing else is supported on your system.
#   uswsusp   If your system has support for the userspace
#             suspend programs (s2ram/s2disk/s2both), then use this.
#   tuxonice  If your system has support for tuxonice, use this.
#
# The system defaults to "kernel" if this is commented out.

It says to use kernel if nothing else is supported. Does this imply a shortcoming in the kernel suspend routine? Is using uswsusp somehow better/safer/faster than the kernel one?

I've switched to uswsusp and I'm not experiencing the intermittent resume failures (YET!) that I was with the default kernel method. Although, it's uglier because it switches to VT1 before suspending.

tomk · 2011-01-10 23:25:02

No shortcoming implied - it just means what it says. "if nothing else is supported" could be more accurately expressed as "if you have not installed any other backend".

uswsusp does provide additional functionality compared with the native kernel system e.g. hibernate to swapfile. It's your call which of them is better/safer/faster.

Ramses de Norre · 2011-01-10 23:30:47

I used to use uswsusp because, due to its option to compress the image, it was considerably faster than the kernel method, since linux 2.6.37, however, (which is in testing) the kernel hibernation now does this too and I've switched to the kernel method.

Uswsusp has more options and I particularly liked the progress that is shown while hibernating, but it has been unmaintained for a while and that's starting to show off (e.g. 32bit userland on a 64bit kernel is not supported).

I've heard very good things about TuxOnIce, but currently for me the benefits don't outweigh the disadvantage of regularly recompiling the kernel.

Last edited by Ramses de Norre (2011-01-10 23:32:03)

pogeymanz · 2011-01-11 02:30:30

Thanks for the replies. That's basically what I thought it meant: "Just use this if you aren't using something else."

My thought process is: The kernel eventually has to suspend/hibernate, so why add more moving parts to it if I don't need to. Just seems like one more thing to debug if/when it fails.

ngoonee · 2011-01-11 06:22:02

Ramses de Norre wrote:

I've heard very good things about TuxOnIce, but currently for me the benefits don't outweigh the disadvantage of regularly recompiling the kernel.

Its normally okay to just keep the kernel and not update (its JUST security fixes). Besides, with a streamlined config my kernel26-ice takes 10 minutes to compile (add another 2 minutes for vbox and nvidia drivers recompiling and you're done).

Arch Linux

#1 2011-01-10 23:18:47

Question about pm-suspend backend...

#2 2011-01-10 23:25:02

Re: Question about pm-suspend backend...

#3 2011-01-10 23:30:47

Re: Question about pm-suspend backend...

#4 2011-01-11 02:30:30

Re: Question about pm-suspend backend...

#5 2011-01-11 06:22:02

Re: Question about pm-suspend backend...

Board footer