On Wednesday 31 August 2005 01:54, Christoph Lameter wrote:
> Certainly not a big effect (if we make sure the compiler knows that
> this test mostly fails and insure that the variable is in
> __mostly_read)
Currently neither, but that could be easily fixed.
> but this is a frequently executed
From: Andi Kleen <[EMAIL PROTECTED]>
Subject: Re: [PATCH] Only process_die notifier in ia64_do_page_fault if KPROBES
is configured.
Date: Wed, 31 Aug 2005 01:38:08 +0200
> On Wednesday 31 August 2005 01:05, Luck, Tony wrote:
> > >Please do not generate any code if the feature cannot ever be
> >
On Wed, 31 Aug 2005, Andi Kleen wrote:
> Also with the inline the test should be essentially a single test of
> a global variable and jump. Hardly a big performance issue, no?
There are multiple effects of this code.
- Additional cacheline in use in the page fault handler
increasing the
On Wednesday 31 August 2005 01:05, Luck, Tony wrote:
> >Please do not generate any code if the feature cannot ever be
> >used (CONFIG_KPROBES off). With this patch we still have lots of
> >unnecessary code being executed on each page fault.
>
> I can (eventually) wrap this call inside the #ifdef
>Please do not generate any code if the feature cannot ever be
>used (CONFIG_KPROBES off). With this patch we still have lots of
>unnecessary code being executed on each page fault.
I can (eventually) wrap this call inside the #ifdef CONFIG_KPROBES.
But I'd like to keep following leads on
Please do not generate any code if the feature cannot ever be
used (CONFIG_KPROBES off). With this patch we still have lots of
unnecessary code being executed on each page fault.
I can (eventually) wrap this call inside the #ifdef CONFIG_KPROBES.
But I'd like to keep following leads on making
On Wednesday 31 August 2005 01:05, Luck, Tony wrote:
Please do not generate any code if the feature cannot ever be
used (CONFIG_KPROBES off). With this patch we still have lots of
unnecessary code being executed on each page fault.
I can (eventually) wrap this call inside the #ifdef
On Wed, 31 Aug 2005, Andi Kleen wrote:
Also with the inline the test should be essentially a single test of
a global variable and jump. Hardly a big performance issue, no?
There are multiple effects of this code.
- Additional cacheline in use in the page fault handler
increasing the
From: Andi Kleen [EMAIL PROTECTED]
Subject: Re: [PATCH] Only process_die notifier in ia64_do_page_fault if KPROBES
is configured.
Date: Wed, 31 Aug 2005 01:38:08 +0200
On Wednesday 31 August 2005 01:05, Luck, Tony wrote:
Please do not generate any code if the feature cannot ever be
used
On Wednesday 31 August 2005 01:54, Christoph Lameter wrote:
Certainly not a big effect (if we make sure the compiler knows that
this test mostly fails and insure that the variable is in
__mostly_read)
Currently neither, but that could be easily fixed.
but this is a frequently executed code
On Mon, 29 Aug 2005, Rusty Lynch wrote:
> So, assuming inlining the notifier_call_chain would address Christoph's
> conserns, is the following patch something like what you are sugesting?
> This would make all the kdebug.h::notify_die() calls use the inline version.
Please do not generate any
On Tuesday 30 August 2005 02:19, Rusty Lynch wrote:
>
> So, assuming inlining the notifier_call_chain would address Christoph's
> conserns, is the following patch something like what you are sugesting?
Yes.
Well in theory you could make fast and slow notify_die too, but that's
probably not
On Sat, Aug 27, 2005 at 02:24:25AM +0200, Andi Kleen wrote:
> On Saturday 27 August 2005 01:05, Christoph Lameter wrote:
> > On Fri, 26 Aug 2005, Rusty Lynch wrote:
> > > Just to be sure everyone understands the overhead involved, kprobes only
> > > registers a single notifier. If kprobes is
On Tuesday 30 August 2005 02:19, Rusty Lynch wrote:
So, assuming inlining the notifier_call_chain would address Christoph's
conserns, is the following patch something like what you are sugesting?
Yes.
Well in theory you could make fast and slow notify_die too, but that's
probably not worth
On Mon, 29 Aug 2005, Rusty Lynch wrote:
So, assuming inlining the notifier_call_chain would address Christoph's
conserns, is the following patch something like what you are sugesting?
This would make all the kdebug.h::notify_die() calls use the inline version.
Please do not generate any
On Sat, Aug 27, 2005 at 02:24:25AM +0200, Andi Kleen wrote:
On Saturday 27 August 2005 01:05, Christoph Lameter wrote:
On Fri, 26 Aug 2005, Rusty Lynch wrote:
Just to be sure everyone understands the overhead involved, kprobes only
registers a single notifier. If kprobes is disabled
On Saturday 27 August 2005 01:05, Christoph Lameter wrote:
> On Fri, 26 Aug 2005, Rusty Lynch wrote:
> > Just to be sure everyone understands the overhead involved, kprobes only
> > registers a single notifier. If kprobes is disabled (CONFIG_KPROBES is
> > off) then the overhead on a page fault
On Fri, 26 Aug 2005, Rusty Lynch wrote:
> Just to be sure everyone understands the overhead involved, kprobes only
> registers a single notifier. If kprobes is disabled (CONFIG_KPROBES is
> off) then the overhead on a page fault is the overhead to execute an empty
> notifier chain.
Its the
On Fri, 26 Aug 2005, Rusty Lynch wrote:
Just to be sure everyone understands the overhead involved, kprobes only
registers a single notifier. If kprobes is disabled (CONFIG_KPROBES is
off) then the overhead on a page fault is the overhead to execute an empty
notifier chain.
Its the
On Saturday 27 August 2005 01:05, Christoph Lameter wrote:
On Fri, 26 Aug 2005, Rusty Lynch wrote:
Just to be sure everyone understands the overhead involved, kprobes only
registers a single notifier. If kprobes is disabled (CONFIG_KPROBES is
off) then the overhead on a page fault is the
20 matches
Mail list logo