On Mon, Dec 10, 2007 at 10:26:46PM -0500, Daniel Gryniewicz wrote:
> Okay, attached is a patch against 2.6.23.9 + skas3 patch to make it
> build (and run) as a UML guest. I've only tested it on x86_64 (but I
> did use SUBARCH=i386). I went the route of minimal changes to get it to
> build, as jus
On Mon, 2007-12-10 at 10:45 -0500, Daniel Gryniewicz wrote:
> On Sat, 2007-12-08 at 09:17 -0500, Jeff Dike wrote:
> > A skas3 patch which works on 2.6.23 is below.
>
> BTW, it would appear that applying this patch breaks building the
> resulting kernel as a guest kernel, at least on x86_64.
> as
On Sat, 2007-12-08 at 09:17 -0500, Jeff Dike wrote:
> A skas3 patch which works on 2.6.23 is below.
BTW, it would appear that applying this patch breaks building the
resulting kernel as a guest kernel, at least on x86_64.
asm-x86_64/proc_mm.h doesn't end up in asm/, and the #includes don't use
as
On 8 Dec 2007, Karol Swietlicki outgrape:
> I can't wait to test this.
WORKSFORME, nice and fast again just in time for weeks of unattended
operation as I head out for a nice long Newtonmass break. :)
(so it had better not crash! ;} )
--
`The rest is a tale of post and counter-post.' --- Ian Ra
On 08/12/2007, Jeff Dike <[EMAIL PROTECTED]> wrote:
> A skas3 patch which works on 2.6.23 is below. There were a couple of
> problems that I fixed which are described in the changelog. With
> this on the host and a UML close to what's currently in -mm, I get ~85%
> of native performance on a kern
A skas3 patch which works on 2.6.23 is below. There were a couple of
problems that I fixed which are described in the changelog. With
this on the host and a UML close to what's currently in -mm, I get ~85%
of native performance on a kernel build. With skas0, the best I've
seen is ~75%.
Thanks t
