Suspend to RAM (FC4)

[oleksandr korneta]

> On 1/11/06, Mauriat Miranda <mirandam gmail com> wrote:
> 
> Are you doing the tests with or with the X-server running?
> If you haven't already tried, test in init-3 without X.
> Additionally, sometimes video card drivers can cause problems (I've
> had this in the past with NV binary drivers).
just tried runlevel 3. Result is almost the same. The only difference is 
additional message below the old one.

Back to C!
ACPI: PCI Interrupt 0000:00:13.0[A] ->GSI19 (level, low) -> IRQ209


this is USB controller
/sbin/lspci | grep 00:13.0
00:13.0 USB Controller: ATI Technologies Inc IXP SB400 USB Host Controller

> 
>> I tried to sent the system to sleep mode both manually and by using
>> special script with all this fancy kernel modules reloading and stuff,
>> but the result is the same...
> 
> Did you look at your logs when you reboot to see if something could
> not wake properly or where the fault is? Are you sure it fails when
> going to sleep or when waking up?
> /var/log/messages ?

I have no idea. When I watch /var/log/messages from the remote machine 
it says:

Jan 12 10:47:47 localhost kernel: ehci_hcd 0000:00:13.2: remove, state 1
Jan 12 10:47:47 localhost kernel: usb usb1: USB disconnect, address 1
Jan 12 10:47:47 localhost kernel: ehci_hcd 0000:00:13.2: USB bus 1 
deregistered
Jan 12 10:47:47 localhost kernel: ACPI: PCI interrupt for device 
0000:00:13.2 disabled

and after that laptop goes to sleep. When I try to wake it up there is 
no messages at all, therefore presumably the problem is with waking up

  > - disabling framebuffer consoles sometimes helped

that is the first thing I've done. I boot with vga=normal

> - I've had better success with 2.6.9-12 kernels, 13-15 have been buggy with S3
hm... mine is 2.6.14-1.1653_FC4, shall I switch?

> - using USB cardreaders or hubs seems to give me problems --- try
> without any extra accessories plugged in
nothing is plugged. I tried with usb mouse plugged, there is really no 
difference.

> 
> In the end, I had to try multiple different kernels (often
> recompiling) till I got it right. In one case FC4 2.6.11 worked out of
> the box, however FC4 2.6.14 blows up. ... Also sometimes, those
> "special scripts" with fancy load/unload may be the only way to get
> things to work.

here is what I'm using

#!/bin/sh
# Script to drive the system to S3
# suspend-to-ram
#
# JRA Jan 2005

# Step 1: preparing sleep

if (-e /tmp/suspending) then
  echo "already in the process of suspending. Please be patient."
else
/bin/touch /tmp/suspending
/usr/bin/killall -s SIGUSR1 cpuspeed
/usr/bin/chvt 1 # necessary to make DRI work
/sbin/service anacron stop
# USB doesn't suspend without unloading first
/sbin/rmmod ehci_hcd
/sbin/rmmod uhci_hcd
#save the system time
/sbin/hwclock --adjust
#/sbin/rmmod wacom # only necessary of you have wacom stylus
/sbin/rmmod -as
/bin/sync
# Step 2: send sleep command via ACPI
echo mem > /sys/power/state

# Step 3: wake-up and reload
/sbin/hwclock --hctosys
/sbin/modprobe ehci_hcd
/sbin/modprobe uhci_hcd
# initialise X
xinit /bin/false -- :1
/usr/bin/chvt 7
/sbin/service anacron start
/sbin/service cpuspeed restart
rm /tmp/suspending
endif

> ps. The newer versions of SUSE come with a *very very* nice
> configuration (scripts+tools built in) to go to S3 (S4, etc). I have
> this working perfectly on an x86_64 desktop. Not sure when things like
> this will make it into Fedora.

I was told about that before, just wanted to have fedora on all the 
machines. Probably that was a mistake.