Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1503250

Re: [PATCH 3/3] early_printk: Add simple serialization to early_vprintk()

From Peter Zijlstra <peterz@infradead.org>
Newsgroups linux.kernel
Subject Re: [PATCH 3/3] early_printk: Add simple serialization to early_vprintk()
Date 2016-10-18 19:40 +0200
Message-ID <stGwN-4WW-23@gated-at.bofh.it> (permalink)
References <stGdr-4O9-19@gated-at.bofh.it> <stGds-4O9-33@gated-at.bofh.it> <stGn7-4Sx-5@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw


On Tue, Oct 18, 2016 at 01:19:51PM -0400, Steven Rostedt wrote:
> 
> hehe, I have the exact same patch for my -rt work.

> > +static int early_printk_cpu = -1;
> > +
> >  static int early_vprintk(const char *fmt, va_list args)
> >  {
> > +	int n, cpu, old;
> >  	char buf[512];
> > +
> > +	cpu = get_cpu();
> > +	for (;;) {
> > +		old = cmpxchg(&early_printk_cpu, -1, cpu);
> > +		if (old == -1 || old == cpu)
> > +			break;

Looking at this again, we really should not spin using cmpxchg(), that
thrashes the cacheline.

The below is slightly better spinning... then again, this isn't
performance code.

--- a/kernel/printk/printk.c
+++ b/kernel/printk/printk.c
@@ -376,10 +376,16 @@ static int early_vprintk(const char *fmt
 
 	cpu = get_cpu();
 	for (;;) {
-		old = cmpxchg(&early_printk_cpu, -1, cpu);
-		if (old == -1 || old == cpu)
+		old = READ_ONCE(early_printk_cpu);
+		if (old == cpu)
 			break;
 
+		if (old == -1) {
+			old = cmpxchg(&early_printk_cpu, -1, cpu);
+			if (old == -1 || old == cpu)
+				break;
+		}
+
 		cpu_relax();
 	}
 

Back to linux.kernel | Previous | NextPrevious in thread | Next in thread | Find similar | Unroll thread


Thread

[PATCH 0/3] make printk work again Peter Zijlstra <peterz@infradead.org> - 2016-10-18 19:20 +0200
  [PATCH 1/3] printk: Fix kdb_trap_printk placement Peter Zijlstra <peterz@infradead.org> - 2016-10-18 19:20 +0200
    Re: [PATCH 1/3] printk: Fix kdb_trap_printk placement Petr Mladek <pmladek@suse.com> - 2016-10-19 16:50 +0200
      Re: [PATCH 1/3] printk: Fix kdb_trap_printk placement Peter Zijlstra <peterz@infradead.org> - 2016-10-19 17:20 +0200
    Re: [PATCH 1/3] printk: Fix kdb_trap_printk placement Sergey Senozhatsky <sergey.senozhatsky@gmail.com> - 2016-10-20 15:10 +0200
  [PATCH 3/3] early_printk: Add simple serialization to early_vprintk() Peter Zijlstra <peterz@infradead.org> - 2016-10-18 19:20 +0200
    Re: [PATCH 3/3] early_printk: Add simple serialization to  early_vprintk() Steven Rostedt <rostedt@goodmis.org> - 2016-10-18 19:30 +0200
      Re: [PATCH 3/3] early_printk: Add simple serialization to  early_vprintk() Peter Zijlstra <peterz@infradead.org> - 2016-10-18 19:40 +0200
        Re: [PATCH 3/3] early_printk: Add simple serialization to  early_vprintk() Steven Rostedt <rostedt@goodmis.org> - 2016-10-18 20:00 +0200
  [PATCH 2/3] early_printk: Add force_early_printk kernel parameter Peter Zijlstra <peterz@infradead.org> - 2016-10-18 19:20 +0200
  Re: [PATCH 0/3] make printk work again Peter Zijlstra <peterz@infradead.org> - 2016-10-19 16:30 +0200
  Re: [PATCH 0/3] make printk work again Sergey Senozhatsky <sergey.senozhatsky@gmail.com> - 2016-10-19 16:30 +0200
  Re: [PATCH 0/3] make printk work again Jan Kara <jack@suse.cz> - 2016-10-19 17:00 +0200
    Re: [PATCH 0/3] make printk work again Peter Zijlstra <peterz@infradead.org> - 2016-10-19 18:20 +0200

csiph-web