[RFC][PATCH] BKL reduction in do_exit

[Dave Hansen <haveblue@us.ibm.com>]

[-- Attachment #1: Type: text/plain, Size: 1091 bytes --]

A week ago, I posted this:
http://groups.google.com/groups?selm=linux.kernel.3CA20C9B.20309%40us.ibm.com

Nobody had anything to sayabout it, so here's a patch.  It moves the 
disassociate_ctty(1) up, and releases the BKl after it gets done.  Is 
this a sane thing to do, or do some of those exit_*() functions still 
need the tty?

The patch reduces hold times of the BKL in do_exit() by a factor of 100. 
   They were on the order of 200us, now they're about 1.5us.  However, 
those numbers were on Martin Bligh's NUMA-Q box, so they represent a 
serious worst-case scenario.

The patch is stable, but I don't properly understand the consequences of 
moving the disassociate_ctty(1) up.  I guess we could do this instead:

  	lock_kernel();
  	sem_exit();
	unlock_kernel();
	
  	__exit_files(tsk);
  	__exit_fs(tsk);
  	exit_namespace(tsk);
  	exit_sighand(tsk);
  	exit_thread();
	
	lock_kernel();
	if (current->leader)
		disassociate_ctty(1);
	unlock_kernel();

But, I hesitated to do that because of the lock bouncing consequences on 
the NUMA-Q.

-- 
Dave Hansen
haveblue@us.ibm.com



[-- Attachment #2: do_exit-bkl_reduction-2.5.7.patch --]
[-- Type: text/plain, Size: 525 bytes --]

--- linux-2.5.7-clean/kernel/exit.c	Tue Apr  2 10:43:28 2002
+++ linux//kernel/exit.c	Wed Apr  3 08:56:00 2002
@@ -500,15 +500,16 @@
 
 	lock_kernel();
 	sem_exit();
+	if (current->leader)
+		disassociate_ctty(1);
+	unlock_kernel();
+	
 	__exit_files(tsk);
 	__exit_fs(tsk);
 	exit_namespace(tsk);
 	exit_sighand(tsk);
 	exit_thread();
 
-	if (current->leader)
-		disassociate_ctty(1);
-
 	put_exec_domain(tsk->thread_info->exec_domain);
 	if (tsk->binfmt && tsk->binfmt->module)
 		__MOD_DEC_USE_COUNT(tsk->binfmt->module);