Re: [parisc-linux] [RFC] rewrite kernel spinlock code to work better with gcc

["Joel Soete" <soete.joel@tiscali.be>]

Hi Dave,

I trust that I have to restart my test (32b smp, 64b smp; all seems to works
fine in 32b but some worries in 64b). With the new 2.4.23-pa1, I discover
that I let badly commented a small chunk a code (just involved toc) and I
wrongly merge Randolph's patch (just about a small #ifdef CONFIG_PA20 in
2.4: thanks for your comments :) ).

I will advise.

Thanks again,
    Joel

>-- Original Message --
>To: soete.joel@tiscali.be (Joel Soete)
>Date: Mon, 1 Dec 2003 13:30:07 -0500 (EST)
>From: "John David Anglin" <dave@hiauly1.hia.nrc.ca>
>Cc: randolph@tausq.org, parisc-linux@lists.parisc-linux.org
>Subject: Re: [parisc-linux] [RFC] rewrite kernel spinlock code to work better
>with gcc
>
>
>> (Just in case it would help) I backport your patch in 2.4 (just because
the
> only system on which I can test it is a b2k on which 2.6 doesn't boot yet)
> and build a smp 64bit kernel which boot and run fine (rebuild another kernel)
> on this b2k:
>

I'd forgotten about the "ldcw,co" errata for PA 2.0.  You only want/need
self-aligning locks for PA 1.x.  You should be able to test this on a
PA 2.0 machine if you don't define CONFIG_PA20.

Dave
-- 
J. David Anglin                          
>       dave.anglin@nrc-cnrc.gc.ca
National Research Council of Canada              (613) 990-0752 (FAX: 952-6602)



-------------------------------------------------------------------------
Tiscali ADSL: 12 mois à 29,50 €/mois! L'Internet rapide, c'est pour tout
le monde.
http://reg.tiscali.be/default.asp?lg=fr