Re: [PATCH] Futexes IV (Fast Lightweight Userspace Semaphores)

[Alan Cox]

> > It's not just 386 vs later due to cmpxchg. It's also the simple issue of
> > UP vs SMP - a UP system still wants to do locking, but it doesn't need the
> > lock prefix. And that lock prefix makes a _huge_ difference
> > performance-wise.
> 
> Fail to see why that matters. User level locking is mostly beneficial on SMPs.
> So, you lock the bus for the atomic update. This is UP, nothing's going on 
> on the bus anyway.

Lots of older x86 is too stupid to optimise exclusive cache line locked
operations. After all the bus is still shared - PCI bus masters for one

Alan