From mboxrd@z Thu Jan 1 00:00:00 1970 From: Jean Delvare Subject: Re: [PATCH] video: limit stack usage of ir-kbd-i2c.c Date: Thu, 28 Feb 2008 19:29:24 +0100 Message-ID: <20080228192924.1ab1dc66@hyperion.delvare> References: <20080225205055.GA27455@joi> <20080226133222.7af260b2@hyperion.delvare> <20080226210307.GA6085@joi> <20080226232320.2df756d6@hyperion.delvare> <20080227102309.GA6698@joi> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20080227102309.GA6698@joi> Sender: linux-kernel-owner@vger.kernel.org To: Marcin Slusarz Cc: LKML , Mauro Carvalho Chehab , i2c@lm-sensors.org, video4linux-list@redhat.com List-Id: linux-i2c@vger.kernel.org On Wed, 27 Feb 2008 11:23:26 +0100, Marcin Slusarz wrote: > On Tue, Feb 26, 2008 at 11:23:20PM +0100, Jean Delvare wrote: > > * You can move the handling of some ioctls to dedicated functions, just > > like I did in i2c-dev: > > http://lists.lm-sensors.org/pipermail/i2c/2008-February/003010.html > > However there is a risk that gcc will inline these functions (that's > > what happened to me...) Not sure how to prevent gcc from inlining. > > There's "noinline" attribute in linux/compiler.h (compiler-gcc.h actually) > for these situations. I didn't know about noinline, thanks for the tip. It works as expected, however in the case of i2cdev_ioctl I'm not sure if it's worth it. Without it, I have: 0x02ae i2cdev_ioctl [i2c-dev]: 192 With noinline, I have: 0x02ac i2cdev_ioctl_smbus [i2c-dev]: 112 0x05bc i2cdev_ioctl_rdrw [i2c-dev]: 112 0x087e i2cdev_ioctl [i2c-dev]: 32 32 + 112 = 144, 192 - 144 = only 48 bytes saved on the stack, at the price of some performance loss. OTOH this makes the binary slightly smaller (not sure why), which is always nice: add/remove: 2/0 grow/shrink: 0/1 up/down: 1121/-1182 (-61) function old new delta i2cdev_ioctl_rdrw - 696 +696 i2cdev_ioctl_smbus - 425 +425 i2cdev_ioctl 1482 300 -1182 -- Jean Delvare