Hi! > >> I own an older Sony Vaio SRX51P, European Model, P3 based. s2ram > >> identifies it with > >> > >> sys_vendor = "Sony Corporation " > >> sys_product = "PCG-SRX51P(DE) " > >> sys_version = "01 " > >> bios_version = "R0232U2" > >> > >> Suspend to RAM by using "s2ram -v -m -f" actually works and the > >> laptop comes back to life, is accessible by network, etc. kudos > >> so far. > >> The only but serious problem is, that the lcd stays off after > >> resume. No matter what kind of options for s2ram I try, if I disable > > > > Is the _lcd_ off or the _backlight_ off? Use bright flashlight to > > tell. > > The lcd. If you press the lid button you get the same effect (but you > cannot use it to turn the lcd on again :-(( ). But I'll doublecheck with > a flashlight nevertheless. If lid button breaks the lcd, even in grub boot loader... then I'd call your machine broken hardware, complain to the manufacturer. (Are you at latest BIOS?) > >> framebuffer or suspend from X, the lcd always stays off. Only > >> a reboot fixes this. Note that the X driver also cannot dis-/enable > >> the lcd (xset dpms force off). It always stays lit. I also tried > >> with i810switch, but that also does not affect the lcd. > >> spicctrl -b42 neither. > >> Latest kernel I tested is 2.6.20-git11 from today. > >> > >> So has anyone of you suspend, acpi, sonypi, fb people an idea how to fix > >> this? I suspect one has to call some magic ACPI method upon resume? What > >> other kind of information would be needed to debug this? Anything more > >> to try? Are there some sony people here listening who can fix this? > > > > sonypi people actually might know how to help... > > > >> 00:02.0 VGA compatible controller: Intel Corporation 82815 CGC > > [Chipset Graphics Controller] (rev 11) > > > > Wait a moment, did not Intel release nice, commented, GPLed sources > > for their graphics cards somewhere? That might help. > > URL? But I suspect the lcd on/off is controlled by some embedded controller > or such (reachable via acpi, at least I've an EC0 in /proc/acpi/). Google a bit, it was a _big_ announcement. Of course it is _possible_ EC is responsible, but I don't quite think so. Pavel -- (english) http://www.livejournal.com/~pavelmachek (cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
