Re: [patch 2/6] Notify page fault call chain for x86_64

[Keshavamurthy Anil S <anil.s.keshavamurthy@intel.com>]

On Wed, Apr 19, 2006 at 11:07:03PM +0200, Andi Kleen wrote:
> On Wednesday 19 April 2006 21:01, Anil S Keshavamurthy wrote:
> 
> You seem to be missing a description/rationale here.
Sorry about that, I had the description in [patch 0/6] and looks like 
I missed you cc'ing in that patch. Any way I am copying it again here.

   "Currently    in    the    do_page_fault()    code    path,   we   call
   notify_die(DIE_PAGE_FAULT, ...) to notify the page fault. The only 
   interested components for this page fault notifications  are  Kprobes  
   and/or  kdb. Since notify_die() is highly overloaded, this  page  fault  
   notification  is  currently  being  sent  to  other components
   registered   with  register_die_notification()  which  uses  the  same
   die_chain to loop for all the registered components.

   In order to optimize the do_page_fault() code path, this critical page
   fault notification is now moved to different call chain and the 
   test results showed great improvements.

   Patches for i386, x86_64, ia64, powerpc and sparc64 follows this mail."

This patch introduces the exclusive {register/unregister}_page_fault_notifier()
for use by Kprobes and kdb.


-Anil
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/