Hi!

quoted

Understand, however, there are systems which suspend/resume reliably
many times per second, making policy choice of having the kernel hard-code
a sys_sync() into the suspend path a bad idea.

My current view on that is that whether or not to do a sync() before suspending
ultimately is a policy decision and should belong to user space as such (modulo
the autosleep situation when user space may not know when the suspend is going
to happen).

Moreover, user space is free to do as many sync()s before suspending as it
wants to and the question here is whether or not the *kernel* should sync()
in the suspend code path.

sync()s from userspace do not work, as userspace is still running.

sync() from kernel happens with tasks stopped. ... so it should really
get consistent image on disk.

And there are already interfaces that can s2ram without sync, just use
uswsusp ioctls, not the sysfs writes.

If you are doing multiple suspends per second, a) you are doing
something wrong and b) you'd better use ioctl anyway.
									Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html