On Tue, 30 Sep 2008 00:10:37 +0200 "Rafael J. Wysocki" <rjw@sisk.pl> wrote:
Violent objections.
We just don't know what this change will do. It potentially affects
every code site in the kernel which looks at system_state. We've had
problems in the past with this thing and the more complex we make it,
the worse any future problems will be. It's Just Bad.
Can we just create a new global?
extern bool system_entering_hibernation_or_whatever;
?
It'll add four bytes of bss, it'll save *more* than four bytes of
text and it is plainly obviously non-injurious to present and future
code.
hmm?
--