On Fri, Mar 09, 2007 at 06:51:18PM -0600, neil verplank wrote: > > Hello all, > > First, I'm responding based on the page on opensuse > (http://en.opensuse.org/S2ram). And I'm responding because under > suse10.1, my laptop went to sleep and woke up just fine, but under suse > 10.2, it no longer does (which is why it seems like a bug, as opposed to > a machine that's simply not supported). I should point out that I am > loading a repaired DSDT table at boot, and that I was doing so under > 10.1 as well. > > The output of s2ram -i : > > This machine can be identified by: > sys_vendor = "Intel Corporation" > sys_product = "Calistoga & ICH7M Chipset" > sys_version = "Not Applicable" > bios_version = "Q3B21 " > > This is a core duo 2.0G machine, built by Quanta, distributed through a > reseller program sponsored by Intel. I attached the output of lsmod at > the end of the message, in the event that's helpful.
If this is an "original" intel machine, you should not load a "repaired" DSDT but instead tell the intel developers at the linux-acpi list. They will either fix the ACPI code or try to get the vendor to ship a fixed DSDT. > As I understand it, suse 10.1 uses the powersaved, and 10.2 uses > pm-utils, but they both rely on s2ram. I'm a little unclear on that, > obviously if s2ram is completely new to suse 10.2, that's a different > story than "used to work and now doesn't." No, it should still work the same as in 10.1. We are per default not unloading any drivers in 10.2 as we did with 10.1, which could effect suspend and resume, but not when trying with init=/bin/bash. > Under suse 10.1, I simply set > /etc/powersave/sleep:SUSPEND2RAM_FORCE="yes", and made sure the lid > button event caused suspend to ram, and it worked flawlessly. This is the same as plain "s2ram -f". > Under suse 10.2, when I try s2ram -f, the machine does sleep, and it > does wake up (in X mode), but it's extremely slow and unresponsive, and > quickly comes to need a hard reboot. So, I tried all the options (I > think) from the s2ram page above, both under X, and using > init=/bin/bash. Using console mode only, I found: > > -f no screen on wakeup (no) > -f -a1 abc superslow mode (slow) > -f a2 slow > -f a3 slow > -f -p -m no > -f -m no > -f -s slow > -f -p no > -f -m -v slow > -f -m -v slow If you still have the 10.1 around, you should try if it worked there from text mode (i doubt it) and if you can find options that work there. The new intel X server (i assume that that machine has an intel gfx chip) in 10.2 has gotten more sensitive and fails e.g. on many machines that worked with "-p -s" before and now only work with "-p -m". > Note that I rebooted after each attempt, to make sure the errors didn't > compound on each other. I also tried (I think) every option with vga=0, > I'd be happy to try other options, or to retry something specific (I > *tried* to cover all the possibilities, but there are many ;-), but > given that the machine seemed to sleep / resume just fine under suse > 10.1, and now it *does* wake up, but it's just realllly slow, I thought > I'd consult the developers to see if I'm missing something, have found a > new one, of if there's something else I can try. > > Thanks in advance for any assistance you can provide. Generally this looks more like a kernel bug than a s2ram bug, i mean, you can revive your graphics card, but the whole system somehow is broken in places where we do nothing ;-) I'd suggest taking this up with the intel developers over at the acpi-devel list. -- Stefan Seyfried "Any ideas, John?" "Well, surrounding them's out." ------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys-and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV _______________________________________________ Suspend-devel mailing list Suspend-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/suspend-devel
