acpi suspend to ram on FC5 and IBM thinkpad T22

[henri Michel]

I would like to know wether suspend to RAM using acpi works for if you're using FC5 on a laptop.
I'm particularly interested by the answer if you happen to own an IBM thinkpad T20,T21 or T22.

Here's what I'm observing on either kernel 2.6.15-1.2054_FC5
(default for FC5) or 2.6.17-1.2145_FC5 (updated)

I run

echo -n mem > /sys/power/state

The "moon crescent LED" will turn green and disk will spin down -> everything seems OK

When hitting a key the laptop APPARENTLY returns from suspend but
. the "moon crescent LED" stays blinking
. the command "echo -n mem > /sys/power/state" does not exit (the terminal either xterm.. or a virtual console) in which I typed the command is frozen.

several observations
.) the pb is present even if X11 is not started
.) I can't not run "echo -n mem > /sys/power/state" twice. A following attempt will provoke a message of the form "file locked by another process" (I don't remember exactly the phrasing and don't want to freeze my laptop again
.) sometimes the laptop is totally frozen, need to remove the power chord and the battery to restart it.
+ there seems to be some state in gdm, since a normal boot which should normally end up in the gdm greeting screen ends up in a black screen with no answer from the keyboard (even though the first "rhgb" part of boot was OK. Only remedy: start in runlevel 3 run xinit /usr/bin/xterm which
seems to clean this "gdm state"
----------------------------------
I must add that suspend to RAM works with apm
. booting with kernel arguments

acpi=off apm=on

. running

apm --suspend

--------------------------------------
Here's is the output of lspci if that gives more clues:

00:00.0 Host bridge: Intel Corporation 440BX/ZX/DX - 82443BX/ZX/DX Host bridge (rev 03)
00:01.0 PCI bridge: Intel Corporation 440BX/ZX/DX - 82443BX/ZX/DX AGP bridge (rev 03)
00:02.0 CardBus bridge: Texas Instruments PCI1450 (rev 03)
00:02.1 CardBus bridge: Texas Instruments PCI1450 (rev 03)
00:03.0 Ethernet controller: Intel Corporation 82557/8/9 [Ethernet Pro 100] (rev 0c)
00:03.1 Serial controller: Agere Systems LT WinModem (rev 01)
00:05.0 Multimedia audio controller: Cirrus Logic CS 4614/22/24 [CrystalClear SoundFusion Audio Accelerator] (rev 01)
00:07.0 Bridge: Intel Corporation 82371AB/EB/MB PIIX4 ISA (rev 02)
00:07.1 IDE interface: Intel Corporation 82371AB/EB/MB PIIX4 IDE (rev 01)
00:07.2 USB Controller: Intel Corporation 82371AB/EB/MB PIIX4 USB (rev 01)
00:07.3 Bridge: Intel Corporation 82371AB/EB/MB PIIX4 ACPI (rev 03)
01:00.0 VGA compatible controller: S3 Inc. 86C270-294 Savage/IX-MV (rev 13)

[undertoad]

try setting kernel boot param:

ec_intr=0

[henri Michel]

Many thanks.

Sorry to have bothered you all since I could have found the answer myself by close reading ot thinkwiki
in particular Problems with ACPI suspend to ram

2 remarks/questions before closing
1) this is not a 100% fix. In very rare occasions I still have complete failures to return from suspend.
Typical situation:
. LID is closed, my IBM thinkpad T22 is quietly suspended to RAM
. A beep occurs
. I open the LID
. screen is black, keyboard is dead -> no alternative other than removing the power chord + battery to restart
I am tempted to think that this beep is produced by the BIOS which tries to do some "Power saving" ... APM functions, or reporting errors
My question is thus : "Is APM (aka BIOS cleverness) disabled when acpi is on or does it still try to interact ?" .
I disabled all power managing functions from Bios configurations but may have forgotten something. I wonder in particular about what is happening when acpi is on (FC5 default) and the BIOS setting "Suspend when LID is closed" is not disabled (IBM T22 Bios default)
Pointers to info on these fascinating ACPI, power management matters welcome.
So far I just starting browsing source forge ACPI doc, and 2.6.17 acpi kernel sources, but learning curve is slow.

2) These brutal restarts (removing the battery) exposed the fact that my conf (IBM thinkpad T22, 2.6.17-1.2145_FC5, Savage/IX-MV graphics card, regular FC5 Xorg server) was affected by
a DRI module problem. DRI works in general but I get dead laptop when booting to runlevel 5 after a brutal restart.
Need to disable DRI in xorg.conf, or sometimes after several reboot (in runlevel 3 staring Xorg manually, or to FC3 installation I still have around) the card is reset correctly.
... I will continue this topic on "Hardware forum".

[cajudosu]

Could you please explain to me, what you exactly did to get the resume work?
I also found the thinkwiki site, but nothing of that worked for me. Perhaps I overread something. Would be nice, if you answered...

cajudosu

[henri Michel]

i just added ec_intr=0 to the boot command line.
My /boot/grub/grub.conf for FC5 is thus the following:
. title Fedora Core 5 (2.6.17.2145_FC5)
root (hd0,09)
kernel /boot/vmlinuz-2.6.17-1.2145_FC5 ro root=LABEL=/fc5 ec_intr=0 rhgb quiet
initrd /boot/initrd-2.6.17-1.2145_FC5.img
As I mentionned in my previous post. There are this room for further investigation on this topic, it works most of the time but not always.

[dshaw256]

I don't know if this helps, but you might have a look at the second post in this thread:

http://forums.fedoraforum.org/forum/...d.php?t=117390

I ran across a script somewhere on the 'net that specifically addresses the problem of restoring the X environment to avoid a black screen on start-up. This script works fairly well for me on my Toshiba Tecra under FC4. I do occasionally have the screen blank after a suspend, and I have to briefly press/release the lid switch to wake it back up. I'd say I'm 95% good with this script.

The third post is something you might look at. My lid switch is not read by ACPI so I have to manually suspend then use the power button to restore the system. ACPI did see the power button, and was using the example rule to go into shut down. Killing the example rule cleared this up.

Again, hope this helps.

[techmum]

Originally Posted by henri Michel

screen is black, keyboard is dead -> no alternative other than removing the power chord + battery to restart

Following won't solve your problem, but will make it less frustrating! >

There is no need to remove your battery to force a poweroff. Simply depress and hold the power button on your notebook/PC for about 5 seconds. This will force it to poweroff.