linux-kernel - Re: [PATCH] avoid second smp_processor_id() call in __touch

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Sat, 25 Sep 2010 20:43:59 +0300
From:	Sergey Senozhatsky <sergey.senozhatsky@...il.com>
To:	Don Zickus <dzickus@...hat.com>
Cc:	Sergey Senozhatsky <sergey.senozhatsky@...il.com>,
	Andrew Morton <akpm@...ux-foundation.org>,
	Peter Zijlstra <a.p.zijlstra@...llo.nl>,
	Ingo Molnar <mingo@...e.hu>, linux-kernel@...r.kernel.org,
	Cyrill Gorcunov <gorcunov@...il.com>,
	Frederic Weisbecker <fweisbec@...il.com>
Subject: Re: [PATCH] avoid second smp_processor_id() call in
 __touch_watchdog

On (09/24/10 15:34), Don Zickus wrote:
> On Wed, Sep 22, 2010 at 12:00:12PM +0300, Sergey Senozhatsky wrote:
> > Hello,
> > 
> > Per our previous conversation:
> > 
> > Andrew Morton wrote:
> > > Fair enough, although strictly speaking this should be done in a
> > > separate and later patch.
> > > 
> > 
> > 
> > Avoid double smp_processor_id() call in __touch_watchdog (smp_processor_id() 
> > itself and later call in __get_cpu_var())
> 
> Dropping this patch then based on Peter's and Frederic's feedback that the
> compiler probably already optimizes for this, leaving readability as a good
> excuse not to change anything.
>

Hello,

gcc 4.5.1, x86_64

Dump of assembler code for function __touch_watchdog:
   <+0>:	push   %rbp
   <+1>:	mov    %rsp,%rbp
   <+4>:	push   %r12
   <+6>:	push   %rbx
   <+7>:	mov    $0x0,%rbx
   <+14>:	callq  0x43 <__touch_watchdog+19>
   <+19>:	mov    %eax,%r12d
   <+22>:	callq  0x4b <__touch_watchdog+27>
   <+27>:	mov    %eax,%eax
   <+29>:	mov    %r12d,%edi
   <+32>:	add    0x0(,%rax,8),%rbx
   <+40>:	callq  0x5d <__touch_watchdog+45>
   <+45>:	shr    $0x1e,%rax
   <+49>:	mov    %rax,(%rbx)
   <+52>:	pop    %rbx
   <+53>:	pop    %r12
   <+55>:	leaveq 
   <+56>:	retq   


patched __touch_watchdog:
Dump of assembler code for function __touch_watchdog:
   <+0>:	push   %rbp
   <+1>:	mov    %rsp,%rbp
   <+4>:	push   %rbx
   <+5>:	mov    $0x0,%rbx
   <+12>:	sub    $0x8,%rsp
   <+16>:	callq  0x45 <__touch_watchdog+21>
   <+21>:	movslq %eax,%rdx
   <+24>:	mov    %eax,%edi
   <+26>:	add    0x0(,%rdx,8),%rbx
   <+34>:	callq  0x57 <__touch_watchdog+39>
   <+39>:	shr    $0x1e,%rax
   <+43>:	mov    %rax,(%rbx)
   <+46>:	pop    %rax
   <+47>:	pop    %rbx
   <+48>:	leaveq 
   <+49>:	retq
 

	Sergey

> > 
> > Signed-off-by: Sergey Senozhatsky <sergey.senozhatsky@...il.com>
> > 
> > ---
> > 
> > diff --git a/kernel/watchdog.c b/kernel/watchdog.c
> > index 7f9c3c5..03d97c5 100644
> > --- a/kernel/watchdog.c
> > +++ b/kernel/watchdog.c
> > @@ -116,8 +116,7 @@ static unsigned long get_sample_period(void)
> >  static void __touch_watchdog(void)
> >  {
> >  	int this_cpu = smp_processor_id();
> > -
> > -	__get_cpu_var(watchdog_touch_ts) = get_timestamp(this_cpu);
> > +	per_cpu(watchdog_touch_ts, this_cpu) = get_timestamp(this_cpu);
> >  }
> >  
> >  void touch_softlockup_watchdog(void)
> > 
> 

Content of type "application/pgp-signature" skipped