Re: ACPI failures on non-Intel hardware

[Nate Lawson <nate-Y6VGUYTwhu0@public.gmane.org>]

Matthew Garrett wrote:
> http://bugzilla.kernel.org/show_bug.cgi?id=3691 describes a fairly
> serious ACPI failure. Attempting to wake up from S3 suspend seems to
> result in the system rebooting without any wakeup code being run. I've
> seen this behaviour on two machines, and had it described to me on a
> third. Interestingly, all the machines in question were non-Intel
> hardware - one is a VIA C3, and the other two are Athlon XPs (not
> Athlon64s). On the test machine I have here, FreeBSD behaves identically
> to Linux (seemingly successful suspend, reboot on wake) and Windows
> suspends and resumes correctly.
> 
> The fact that I've only seen this on systems that didn't have Intel
> motherboard chipsets makes me wonder whether there's something going
> wrong during the suspend code that happens to work by chance on Intel
> (and a few other) chipsets, or alternatively whether these chipsets
> manage to fall outside the specs slightly. Any ideas?

Very interesting!  (Thanks for testing FreeBSD also.)  I hadn't noticed 
the common thread in S3 error complaints except that the three classes 
are 1. goes to sleep and immediately wakes, 2. resets while going to 
sleep or waking up, and 3. goes to sleep and never wakes.

One thing you should try is disabling various methods (like _SST) and 
see what happens.  I think _BFS should be run before _SST since the 
7.3.5 says no other method should be run before _BFS.  Also, enabling BM 
arb after sleep may affect things.

I'll test suspend/resume on my new EPIA-M at some point soon.

-- 
Nate


-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now. 
http://productguide.itmanagersjournal.com/