From: Andi Kleen <ak@suse.de>
To: Eric Dumazet <dada1@cosmosbay.com>
Subject: Re: [PATCH] account_system_vtime() should be a macro, not a function
Date: Thu, 19 Oct 2006 15:46:02 +0200
User-Agent: KMail/1.9.3
Cc: Andrew Morton <akpm@osdl.org>, linux-kernel@vger.kernel.org
References: <200610171747.34177.cijoml@volny.cz> <20061019014446.36410c81.akpm@osdl.org> <200610191543.18951.dada1@cosmosbay.com>
In-Reply-To: <200610191543.18951.dada1@cosmosbay.com>
MIME-Version: 1.0
Content-Disposition: inline
Message-Id: <200610191546.02566.ak@suse.de>
Content-Type: text/plain;
  charset="iso-8859-15"
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 904
Lines: 21

On Thursday 19 October 2006 15:43, Eric Dumazet wrote:
> [PATCH] account_system_vtime() should be a macro, not a function
> 
> Because the way 'current' is implemented on some archs, it's better to use a 
> null macro for account_system_vtime(current) 
> 
> I discovered that gcc was (correctly) issuing one useless instruction (to 
> load %rax with current from pda) on x86_64 on irq_enter() and __irq_exit()
> 
> This saves few bytes in kernel size, on archs where current is 'asm volatile'

I dropped the asm volatile on x86-64 some time ago. Perhaps it should
be dropped on the other architectures too.  Then such hacks wouldn't
be needed anywhere.

-Andi
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/