Re: [patch][rfc] enable uptime display > 497 days on 32 bit

[Tim Schmielau]

On Tue, 26 Feb 2002, Andreas Dilger wrote:
> I'm
> not quite sure why we need to use the irq spinlock, since we already
> make a local copy of jiffies so another timer IRQ changing the jiffies
> value shouldn't affect the return value of get_jiffies64().  Then again,
> that isn't exactly stuff I'm familiar with, so I could be totally
> off-base here.

Indeed, the outcome of get_jiffies64() cannot be affected. The lock is 
just to prevent the tiny chance of jiffies_hi getting incremented twice.

Tim