Date: Wed, 11 Sep 1996 15:50:45 -0500 (EST) From: John Dyson <dyson@dyson.iquest.net> To: bde@zeta.org.au (Bruce Evans) Cc: bde@zeta.org.au, dyson@dyson.iquest.net, current@FreeBSD.org, dg@Root.COM, kato@eclogite.eps.nagoya-u.ac.jp Subject: Re: patch for Cyrix/Ti 486SLC/DLC CPU bug Message-ID: <199609112050.PAA00852@dyson.iquest.net> In-Reply-To: <199609112023.GAA10051@godzilla.zeta.org.au> from "Bruce Evans" at Sep 12, 96 06:23:04 am
next in thread | previous in thread | raw e-mail | index | archive | help
>
> >> pmap_update_{1,2}pg(). The magic .byte's are a bad way of writing
> >> `invlpg' (even gas understands this).
> >>
> >I don't know the syntax -- change it, and I'll propagate it!!!
>
> Intel gives the syntax `invlpg m' where m is a memory address. This
> maps nicely to gcc inline asm (I think the contstraint is "m").
>
> I would put this in cpufunc.h since it is a single instruction and
> might be useful elsewhere. Is it reasonable for device drivers to
> invalidate pages directly?
>
It is unreasonable (IMO) but unavoidable if the system doesn't provide
a needed service. It is difficult to imagine a machine independent
reason to do a pmap_update (but I guess that it might be possible.)
So, to kind-of answer your comment -- I will put it in cpufunc, but
the only module that should normally need it would be pmap.c???
John
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199609112050.PAA00852>