Re: "movb" for spin-unlock (was Re: namei() query)

From: Oliver Xymoron (
Date: Mon Apr 24 2000 - 11:48:07 EST

On Mon, 24 Apr 2000, Linus Torvalds wrote:

> On Mon, 24 Apr 2000, Rogier Wolff wrote:
> >
> > And as, nowadays it is possible to upgrade the CPU in-circuit, even
> > under Linux, wouldn't it be an idea to just require the microcode
> > upgrade? This is EXACTLY what the microcode upgrade is for.
> This is not going to be microcode-upgradeable, I bet. Sure, certain
> companies I could mention can change everything they do with a microcode
> update, but on normal CPU's the kind of ordering issue between simple
> operations is probably not fixable with a software upgrade (well, maybe
> the cache replacement policy might be updated or something, but we're
> talking simple operations like "mov to memory" here which are definitely
> not handled by microcode.)

The mov itself wouldn't be in microcode, as it surely maps to one uop, but
most the instruction scheduler probably is (since it's probably
table-driven). And its either the scheduling or the cache management
protocol that'd have to be iffy to result in a problem.

 "Love the dolphins," she advised him. "Write by W.A.S.T.E.." 

- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to Please read the FAQ at

This archive was generated by hypermail 2b29 : Sun Apr 30 2000 - 21:00:07 EST