[Bug 37243] Will not resume from hibernation or sleep

Tobi Vollebregt tobivollebregt at gmail.com
Thu Mar 30 08:32:19 UTC 2006


Public bug report changed:
https://launchpad.net/malone/bugs/37243

Comment:
I'm not really sure it it's caused by the same bug, but I've been experiencing some weird behaviours here too. I'm running:

Asrock 939Dual-SATA2 w/ AMD64 3200+
1024MB DDR2 Dual Channel
nVidia Geforce 4 Ti 4200 w/ 64MB
1 ide hard drive, 1 ide dvd rw drive
ps/2 keyboard and (optical) mouse

First of all, I dualboot windows and standby works fine in it and it (win) doesn't POST (power on self test) the video on resume. Ubuntu POST the video on resume by default, which causes my video card to go crazy (hibernate and sleep didn't work with POST_VIDEO=true). It only displayed (pure, without any blue or red components) green tetris like blocks which jumped a bit when the screen content changed (by typing something for example). So I set POST_VIDEO=false in /etc/default/acpi-support.

Because Ubuntu apparently used a different sleep (S1 vs S3?) then windows (everything off in Ubuntu, fans still on on windows), I set ACPI_SLEEP_MODE=standby in /etc/default/acpi-support. ACPI_SLEEP_MODE=mem causes the video card to not wake up at all on resume (ie. my monitor stays on standby because it doesn't get a signal). One time the kernel also didn't get loaded correctly (alt+sysreq keys to sync,unmount,poweroff didn't work). Another time it did get loaded correctly.

With these settings (no POST, sleep mode standby), either with nv or nvidia driver, hibernate and sleep both work, though when hibernating/sleeping from the console I need to switch to and from X to restore the console (green blocks otherwise). Hibernating/sleeping from X (+kdm) seems to work fine.

Different vga= parameters to the kernel don't seem to make a difference (I run vga=0x314 [800x600] normally, tested that one and the default [no vga= param]).

I can't say anything about the reliability though, I only tested everything once (or twice).

Anyway, long story short, I suspect the GF4 could be a problem (in my case it seems to be _the_ problem).

Also, I set a jumper on my motherboard which causes the ps/2 and usb devices to not get standby power, which caused that I could only resume with the power button on ACPI_SLEEP_MODE=mem sleep. (might be a hint too.)




More information about the kernel-bugs mailing list