Re: [patch] preemptive kernel, preemptive-2.3.52-A7

Next message: Richard Henderson: "Re: madvice and OSF/1"
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

From: Jamie Lokier (lk@tantalophile.demon.co.uk)
Date: Tue Mar 14 2000 - 13:00:47 EST

Andrea Arcangeli wrote:
> The compiler barrier() are necessary indeed.

And can give better code on other CPUs -- register operations can be
interleaved into the multi-instruction inc/dec ops.

-- Jamie

This archive was generated by hypermail 2b29 : Thu Mar 23 2000 - 21:00:42 EST