On Mon, 22 Oct 2007, Matthew Wilcox wrote:

We certainly don't want to encourage people to blindly make those
conversions ... and I've seen the results of encouraging kernel janitors
to do things a certain way.

There's another issue: the "irqsave/irqrestore" versions are much safer 
than the plain "irq" versions, in case the caller already has interrupts 
disabled. 

So anybody making the change not only would need to make the performance 
argument, he'd better not be a janitor that blindly does the change 
without thinking about all call-sites etc..

		Linus