On Wed, May 02, 2007 at 12:45:06PM +0200, Martin Schwidefsky wrote:
> I thought about using a direct call for s390 as well. The advantage of a
> direct call is that it avoids the overhead of a notifier call even if
> kprobes is running. The disadvantage is that there cannot be a second
> consumer o
On Sat, 2007-04-28 at 09:34 +0100, Christoph Hellwig wrote:
> Well, here's my powerpc patch to do the direct call that I sent out
> a few weeks ago. It not just speed up the pagefault path a lot,
> but also is a major code cleanup. Andi and Anton didn't like it
> because they have ambition for m
Christoph Hellwig writes:
> Well, here's my powerpc patch to do the direct call that I sent out
> a few weeks ago. It not just speed up the pagefault path a lot,
> but also is a major code cleanup. Andi and Anton didn't like it
> because they have ambition for murky debugger code to possibly
>
On Sat, Apr 28, 2007 at 02:49:55PM +1000, Paul Mackerras wrote:
> Martin Schwidefsky writes:
>
> > The minor fault path has grown a lot in terms of cycles. In particular
> > the kprobes hook is very costly. Optimize the path to save a couple of
> > cycles. If kprobes is enabled more than 300 cycle
Martin Schwidefsky writes:
> The minor fault path has grown a lot in terms of cycles. In particular
> the kprobes hook is very costly. Optimize the path to save a couple of
> cycles. If kprobes is enabled more than 300 cycles can be avoided if
> kprobes_running() is false.
There's no good reason
From: Martin Schwidefsky <[EMAIL PROTECTED]>
The minor fault path has grown a lot in terms of cycles. In particular
the kprobes hook is very costly. Optimize the path to save a couple of
cycles. If kprobes is enabled more than 300 cycles can be avoided if
kprobes_running() is false.
Signed-off-b
6 matches
Mail list logo
