On 10/12/12 00:08, Alexander Hall wrote:
On 09/23/12 22:24, Ted Unangst wrote:
On Sat, Sep 22, 2012 at 14:01, Theo de Raadt wrote:
Only concern I have is whether #include <machine/cpu.h> could have
side-effects on architectures I don't know of (which, admittedly, is
quite a few).
it is a bad idea to pull machine/ header files.
I was going to agree, but that's how sbin/sysctl gets at these
defines. So it's that or sys/sysctl.h needs to pull it in itself.
So, if the consensus indeed is that this is the right way to do this,
I'd like to see this tested on various archs to make sure there is
nothing funky going on.
So please test this whether you have a lid to suspend with or not.
OK?
It's now committed, so the plea now changes to "please report any
regression".
/Alexander
