Please test and stabilize =sys-kernel/tuxonice-sources-2.6.38-r1. Testing involves running the kernel for some time and using suspend to ram/disk a few times. System should come back without any issues.
amd64: I use hibernate as the norm, though I cannot get the hibernate cmd to work. archtester testuser # hibernate /usr/share/hibernate/scriptlets.d/tuxonice: line 525: echo: write error: Device or resource busy. However, s2ram -f and the other CL options all work AOK, and system returns perfectly.
"suspend to ram/disk a few times". Similarly, archtester testuser # pm-hibernate archtester testuser # pm-hibernate --suspend both achieve nothing. # echo mem > /sys/power/state invokes suspend to ram. # echo disk > /sys/power/state results in nothing. I can't find a solution to get supsend to disk in any form to work, but I suspect it's not the tux kernel.
the tuxonice patches work perfectly for me. amd64 done
I know you are probably aware but there are power consumption issues in 2.6.38, have they been addressed in tuxonice as tuxonice users are probably concerned about battery life? If not we may want to remask it so that people know it'll eat up their battery.
Don't know if the issue still remains, but just following Pawel's call at gentoo.org. I have two machines (x86 and amd64) both running tuxonice 2.6.38-r2 for some time now without any problems. I updated the amd64 machine recently to 3.0.6 and I did not see any problems either, besides the laptop-mode issue. cheers
I recall 2.6.38 being reasonably stable, though I've moved on to 3.0.7 on the laptop and I'm using zen-sources for reiser4 on the desktop. Unlike Ian, I don't use suspend to ram and always use hibernate, to a file on the root partition which I identify by UUID in the kernel command line. I didn't have much trouble with that configuration. 3.0+ kernels have issues with nuking USB URBs (or something like that) which disables all USB devices until next reboot, but <3.0 was fine with that. I'm on ~x86, with a core i5 on the laptop. I've never noticed power consumption issues because I keep it plugged in all the time, though I couldn't get the KMS intel drivers working until 3.0+ (which may have been because I didn't enable the backlight drivers).
x86 stable, closing, thanks for testing
