lkml.org 
[lkml]   [2014]   [Jul]   [22]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
SubjectRe: [RFC PATCH 0/5] futex: introduce an optimistic spinning futex
On Mon, Jul 21, 2014 at 05:32:55PM -0700, Davidlohr Bueso wrote:
> On Mon, 2014-07-21 at 14:31 -0700, Andy Lutomirski wrote:
> > On Mon, Jul 21, 2014 at 2:27 PM, Peter Zijlstra <peterz@infradead.org> wrote:
> > > All this is predicated on the fact that syscalls are 'expensive'.
> > > Weren't syscalls only 100s of cycles? All this bitmap mucking is far
> > > more expensive due to cacheline misses, which due to the size of the
> > > things is almost guaranteed.
> >
> > 120 - 300 cycles for me, unless tracing happens, and I'm working on
> > reducing the incidence of tracing.
>
> fwiw here's what lmbench's lat_ctx says on my system . For 'accuracy', I
> kept the runs short.
>
> http://www.stgolabs.net/lat_ctx.png

Create a syscall without 'work body', so say something like the below.
Context switches do quite a lot of work (unfortunately).

diff --git a/arch/x86/syscalls/syscall_64.tbl b/arch/x86/syscalls/syscall_64.tbl
index ec255a1646d2..daf27ec10e29 100644
--- a/arch/x86/syscalls/syscall_64.tbl
+++ b/arch/x86/syscalls/syscall_64.tbl
@@ -323,6 +323,7 @@
314 common sched_setattr sys_sched_setattr
315 common sched_getattr sys_sched_getattr
316 common renameat2 sys_renameat2
+317 common nop sys_nop

#
# x32-specific system call numbers start at 512 to avoid cache impact
diff --git a/kernel/sys.c b/kernel/sys.c
index 66a751ebf9d9..1b86614bb551 100644
--- a/kernel/sys.c
+++ b/kernel/sys.c
@@ -2108,6 +2108,11 @@ SYSCALL_DEFINE1(sysinfo, struct sysinfo __user *, info)
return 0;
}

+SYSCALL_DEFINE0(nop)
+{
+ return 0;
+}
+
#ifdef CONFIG_COMPAT
struct compat_sysinfo {
s32 uptime;

\
 
 \ /
  Last update: 2014-07-22 10:21    [W:0.656 / U:0.056 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site