lkml.org 
[lkml]   [2002]   [Dec]   [10]   [last100]   RSS-feed
Views: [more markup]  [less markup]  [headers]  [forward] 
 
Messages in this thread
Complete thread (EXPERIMENTAL)
/
DateMon, 09 Dec 2002 22:55:45 -0800
Fromgeorge anzinger <>
SubjectRe: [PATCH 1/3] High-res-timers part 1 (core) take 20
Digg This
Arjan van de Ven wrote:

On Mon, Dec 09, 2002 at 04:27:56AM -0800, george anzinger wrote:

that's why spinlocks are effectively nops on UP.
What you say is true of just about every spinlock user, and no they shouldn't all do some IF_SMP() thing; the spinlock itself should be (and is) zero on UP

But with preemption, they really are not nops on UP...

that doesn't justify fuglyfying the kernel code. If you can't live with the overhead of preemption, disable preemption. Simple. We DON'T want
spin_lock_nop_on_preempt()
...

spin_unlock_nop_on_preempt()

really, I don't, and I can't see anyone else wanting that either

Well, I just thought it was an optimization. I will leave it the way it is.

-- 
George Anzinger   george@mvista.com
High-res-timers: 
http://sourceforge.net/projects/high-res-timers/
Preemption patch:
http://www.kernel.org/pub/linux/kernel/people/rml
\
ISP Services
Valid XHTML 1.0!\ /
Valid CSS! Last update: 2005-03-22 12:31    [W:0.115 / U:0.000 seconds]
©2003-2005 Jasper Spaans