Re: Some success with laptop sleep (ACPI S3)

[Roy Bixler <rcbixler%nyx.net@localhost>]

On Sun, Feb 07, 2016 at 12:31:33PM +0000, coypu%SDF.ORG@localhost wrote:
> Hello, I know a lot of people have been unable to put their laptops to
> sleep and wake them back up.
> I've had marginally more success with it and can do a sleep-wake cycle
> once with my machine with the following changes:
> sysctl -w hw.acpi.sleep.vbios=2 (Needed for when you have DRMKMS)
> sysctl -w hw.acpi.sleep.state=3 (Put it to sleep)
> 
> Some trouble I've had:
> Sometimes it doesn't fully wake up.
> This is less likely with Xorg running, and once I've woken it up with
> ssh from another machine.
> 
> Sleep a second time:
> This either doesn't work (it will revert the changes), or it will go
> into an intermediate mode. 
> In this intermediate mode, waking the laptop up with power plugged in
> will do nothing. with no power plugged in, it will reboot.
> 
> It's progress!
> Please attempt the same on your laptops if possible.
> 
> I've filed this as port-amd64/50733.
> Link for the lazy: http://gnats.netbsd.org/50733

I've never had sleep-wake work for me under NetBSD.  For comparison,
it's also pretty spotty under Linux.

Since you ask, I decided to try on a couple of Lenovo Thinkpad laptops
I have available.  I had slightly better luck on the first one, which
is an older model with all Intel hardware.  It uses the i915 DRMKMS
driver under NetBSD 7 and generally works perfectly.  One problem is
that I use a USB Ethernet connector for wired connectivity and I had
to pull it first before trying suspend.  Otherwise, the OS complains
about "devices without power management support."  After yanking the
USB connection, the system appears to suspend successfully.  However,
it never wakes up and instead crashes and reboots.

I also tried this on a newer Lenovo Thinkpad under NetBSD-current.
It's a bit different in that it has AMD hardware and Radeon graphics.
I wasn't quite able to install NetBSD 7 on this one.  It generally
works well with the exception of missing power management
functionality (i.e. no shutdown with the power button) and no wireless
driver.  Since power management doesn't work well, it's no surprise
that it doesn't suspend at all.  It starts well enough and syncs the
disk, but the screen then goes dark and the system remains fully on.

-- 
Roy Bixler <rcbixler%nyx.net@localhost>
"The fundamental principle of science, the definition almost, is this: the
sole test of the validity of any idea is experiment."
-- Richard P. Feynman