Hi, As indicated in a recent thread on Linux-PM, it's necessary to call pm_ops->finish() before devce_resume(), but enable_nonboot_cpus() has to be called before pm_ops->finish() (cf. http://lists.osdl.org/pipermail/linux-pm/2006-November/004164.html). For consistency, it seems reasonable to call disable_nonboot_cpus() after device_suspend(). This way the suspend code will remain symmetrical with respect to the resume code and it may allow us to speed up things in the future by suspending and resuming devices and/or saving the suspend image in many threads. The following series of patches reorders the suspend and resume code so that nonboot CPUs are disabled after devices have been suspended and enabled before the devices are resumed. It also causes pm_ops->finish() to be called after enable_nonboot_cpus() wherever necessary. The first patch changes the ordering of the suspend-to-RAM code and is untested, because my boxes continue refusing to resume from RAM for other reasons. If anyone can, please do me a favour and test it. The second patch changes the ordering of the built-in suspend-to-disk code, and the last two patches modify the swsusp userland interface code accordingly. As far as the last patch is concerned, I've decided to change the existing ioctls, so that the patched kernel works with the "old" userland. I don't think this change breaks anything of importance, but if it does, we can just add (yet) another ioctl to set the platform mode, although in that case the userland would have to be changed as well. All of the patches are against 2.6.20-rc1 and should not be used with any earlier kernel. Greetings, Rafael -- If you don't have the time to read, you don't have the time or the tools to write. - Stephen King
