Re: MTRR fix

Linus Torvalds (torvalds@transmeta.com)
Tue, 1 Sep 1998 10:41:34 -0700 (PDT)

On Tue, 1 Sep 1998, Richard Gooch wrote:
>
> Hi, Linus. I've appended a patch for 2.1.119 which fixes a problem
> reported by Tom Eastep (hard lockups when mtrr_add() is called in the
> console patch with SMP kernel). The problem was that IPIs were being
> used to distribute the MTRR changes, but smp_begin() had not yet been
> called. So now I queue the calls to mtrr_add() until smp_begin() has
> been called.

I really think that it's the console patch that is broken. If it plays
with the MTRR's it should be part of the later "real" device
initialization (ie chrdev_init()) rather than called from console_init()
which happens early.

There is no excuse for playing with the MTRR's at console_init() time:
console_init() is supposed to only get the console to a usable enough
state that you can see it, not to optimize it for speed.

Linus

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.altern.org/andrebalsa/doc/lkml-faq.html