On Monday, 26 February 2007 10:17, Stefan Seyfried wrote: > On Sun, Feb 25, 2007 at 10:26:51PM +0100, Rafael J. Wysocki wrote: > > Hi, > > > > I've just notice that we "forgot" to implement the "platform" mode in > > resume. > > The appended patch fixes this problem. > > I did not read the ACPI spec on this, but we never did that in the in-kernel > suspend, and it seems to have worked fine until now (and actually, Patricks > version seemed to be built upon reading the spec). > > With this change, if i read it correctly, we are preparing a suspend (during > resume), without entering it. Then we finish the suspend. > I would not bet that this does not confuse the BIOSes that actually care about > this stuff. > > Do you have a machine that behaves differently with this change?
On my box the AC status is not updated properly without it, or so it seems. > I don't think it will harm anybody, however, if we do not have a case where > we need this, i am a bit hesitant to change any BIOS-touching codepath ;-) Well, during the resume we're suspending devices too, so I think we should call platform_prepare(). And we don't finish it, unless the atomic restore fails. The suspend code finishes it and that's the point. Please read kernel/power/disk.c:software_resume(), preferably in 2.6.21-rc1, and you'll see that it calls platform_prepare() (even earlier, but I don't think that is really necessary). Greetings, Rafael ------------------------------------------------------------------------- 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
