Re: Testers wanted! (was Re: "movb" for spin-unlock)

From: Oliver Xymoron (
Date: Mon Apr 24 2000 - 10:53:08 EST

On Mon, 24 Apr 2000, Alex Buell wrote:

> On Sun, 23 Apr 2000, Oliver Xymoron wrote:
> > Anyone with a dual x86, please give this a quick run. I want to hear
> > about any failures (it'll abort with "state was 1!") and if you have a
> > PPro with a stepping of less than 9 (which have been rumored to have a
> > problem with this), tell me about your successes as well. Send me a
> > copy of your /proc/cpuinfo too. If we can get a solid set of success
> > reports for early P6s and no failures elsewhere, perhaps we can get
> > this faster locking method into the kernel.
> With the new program, with MOVB defined, completes successfully, with MOVB
> undefined, completes successfully. Both works. Hmm! Using gcc 2.95.2,
> glibc 2.1.3 w/linuxthreads.

No surprise - without MOVB defined, we use the same unlock that the
current kernel is using.

 "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