tech-kern archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: Heads up: moving some uvmexp stat to being per-cpu

Eduardo Horvath wrote:

> On Wed, 15 Dec 2010, Martin Husemann wrote:
> > I have one stupid question: why can't we leave the size of the counters
> > at 32bit on a per arch basis?
> Or use 64-bit counters but just update the lower 32-bits of them.  Is 
> there some danger that a 32-bit counter will overflow?

My home server has been up for 70 days, and the largest counter is

        3826338196 software interrupts

which is pretty close to 2^32 (if I counted the digits correctly).

I haven't looked at Matt's diff yet, but 64 bit uvmexp counters was
always one of those little projects I meant to get around to one day
but didn't.  Nice to see this at last.


Home | Main Index | Thread Index | Old Index