From mboxrd@z Thu Jan  1 00:00:00 1970
Return-Path: <linux-kernel-owner@vger.kernel.org>
Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand
	id S1754681AbbEUHvQ (ORCPT <rfc822;w@1wt.eu>);
	Thu, 21 May 2015 03:51:16 -0400
Received: from mx1.redhat.com ([209.132.183.28]:42939 "EHLO mx1.redhat.com"
	rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP
	id S1753636AbbEUHvN (ORCPT <rfc822;linux-kernel@vger.kernel.org>);
	Thu, 21 May 2015 03:51:13 -0400
Message-ID: <555D8E4E.9030109@redhat.com>
Date: Thu, 21 May 2015 09:50:38 +0200
From: Denys Vlasenko <dvlasenk@redhat.com>
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:24.0) Gecko/20100101 Thunderbird/24.2.0
MIME-Version: 1.0
To: Brian Norris <computersforpeace@gmail.com>
CC: David Woodhouse <David.Woodhouse@intel.com>,
        Dan Carpenter <dan.carpenter@oracle.com>,
        Jingoo Han <jg1.han@samsung.com>, Aaron Sierra <asierra@xes-inc.com>,
        Artem Bityutskiy <Artem.Bityutskiy@linux.intel.com>,
        linux-mtd@lists.infradead.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH] mtd: cfi: Deiline large functions
References: <1431946720-32281-1-git-send-email-dvlasenk@redhat.com> <20150520185641.GI11598@ld-irv-0074>
In-Reply-To: <20150520185641.GI11598@ld-irv-0074>
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Sender: linux-kernel-owner@vger.kernel.org
List-ID: <linux-kernel.vger.kernel.org>
X-Mailing-List: linux-kernel@vger.kernel.org

On 05/20/2015 08:56 PM, Brian Norris wrote:
> On Mon, May 18, 2015 at 12:58:40PM +0200, Denys Vlasenko wrote:
>> With this .config: http://busybox.net/~vda/kernel_config,
>> after uninlining these functions have sizes and callsite counts
>> as follows:
> 
> Most of this is probably good, thanks. But I'm curious about one:
> 
>> cfi_udelay(): 74 bytes, 26 callsites
> 
> ^^ This is pretty dead-simple. If it's generating bad code, we might
> look at fixing it up instead. Almost all of its call sites are with
> constant input, so it *should* just become:
> 
> 	udelay(1);
> 	cond_resched();
> 
> in most cases. For the non-constant cases, we might still do an
> out-of-line implementation. Or maybe we just say it's all not worth it,
> and we just stick with what you have. But I'd like to consider
> alternatives to out-lining this one.

You want to consider not-deinlining (IOW: speed-optimizing)
a *fixed time delay function*?

Think about what delay functions do...