From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1758735Ab2JZNOJ (ORCPT ); Fri, 26 Oct 2012 09:14:09 -0400 Received: from mga01.intel.com ([192.55.52.88]:63648 "EHLO mga01.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1758660Ab2JZNOH (ORCPT ); Fri, 26 Oct 2012 09:14:07 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="4.80,654,1344236400"; d="scan'208";a="240543636" Date: Fri, 26 Oct 2012 21:04:08 +0800 From: Yuanhan Liu To: Stefani Seibold Cc: Yuanhan Liu , linux-kernel@vger.kernel.org, Andrew Morton , Wei Yang , Fengguang Wu , Stephen Rothwell Subject: Re: [PATCH] kfifo: remove unnecessary type check Message-ID: <20121026130408.GH2778@yliu-dev.sh.intel.com> References: <1351215971-11639-1-git-send-email-yliu.null@gmail.com> <1351229911.12511.1.camel@wall-e> <20121026061145.GA2778@yliu-dev.sh.intel.com> <1351234266.12511.23.camel@wall-e> <20121026071757.GB2778@yliu-dev.sh.intel.com> <1351243591.8719.2.camel@wall-e> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1351243591.8719.2.camel@wall-e> User-Agent: Mutt/1.5.21 (2010-09-15) Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Fri, Oct 26, 2012 at 11:26:31AM +0200, Stefani Seibold wrote: > Am Freitag, den 26.10.2012, 15:17 +0800 schrieb Yuanhan Liu: > > On Fri, Oct 26, 2012 at 08:51:06AM +0200, Stefani Seibold wrote: > > > Am Freitag, den 26.10.2012, 14:11 +0800 schrieb Yuanhan Liu: > > > > On Fri, Oct 26, 2012 at 07:38:31AM +0200, Stefani Seibold wrote: > > > > > Am Freitag, den 26.10.2012, 09:46 +0800 schrieb Yuanhan Liu: > > > > > > From: Yuanhan Liu > > > > > > > > > > > > Firstly, this kind of type check doesn't work. It does something similay > > > > > > like following: > > > > > > void * __dummy = NULL; > > > > > > __buf = __dummy; > > > > > > > > > > > > __dummy is defined as void *. Thus it will not trigger warnings as > > > > > > expected. > > > > > > > > > > > > Second, we don't need that kind of check. Since the prototype > > > > > > of __kfifo_out is: > > > > > > unsigned int __kfifo_out(struct __kfifo *fifo, void *buf, unsigned int len) > > > > > > > > > > > > buf is defined as void *, so we don't need do the type check. Remove it. > > > > > > > > > > > > LINK: https://lkml.org/lkml/2012/10/25/386 > > > > > > LINK: https://lkml.org/lkml/2012/10/25/584 > > > > > > > > > > > > Cc: Andrew Morton > > > > > > Cc: Wei Yang > > > > > > Cc: Stefani Seibold > > > > > > Cc: Fengguang Wu > > > > > > Cc: Stephen Rothwell > > > > > > Signed-off-by: Yuanhan Liu > > > > > > --- > > > [snip]... > > > > > > Also you have to build the kfifo samples, since this example code use > > > all features of the kfifo API. > > > > > > And again: The kfifo is designed to do the many things at compile time, > > > not at runtime. If you modify the code, you have to check the compiler > > > assembler output for no degradation, especially in kfifo_put, kfifo_get, > > > kfifo_in, kfifo_out, __kfifo_in and __kfifo_out. Prevent runtime checks > > > if you can do it at compile time. This is the basic reasons to do it in > > > macros. > > > > Is it enought to check kernel/kfifo.o only? I build that file with > > and without this patch. And then dump it by objdump -D kernel/fifo.o to > > /tmp/kfifo.dump.with and /tmp/kfifo.dump.without, respectively. And the > > two dump file are exactly same. > > > > No, since most of the code is inlined due performace reasons, you have > to hack the kfifo examples output code for regressions and code > increase. In my test, this patch doesn't change anything. Here are some data to prove that: $ make samples/kfifo/ $ cp samples/kfifo/*.o /tmp/before/ $ git am this-patch $ make samples/kfifo/ $ cp samples/kfifo/*.o /tmp/after/ $ for i in /tmp/before/*.o; do size $i /tmp/after/`basename $i`; done text data bss dec hex filename 1939 464 456 2859 b2b /tmp/before/bytestream-example.o 1939 464 456 2859 b2b /tmp/after/bytestream-example.o text data bss dec hex filename 1423 112 296 1831 727 /tmp/before/dma-example.o 1423 112 296 1831 727 /tmp/after/dma-example.o text data bss dec hex filename 1864 624 376 2864 b30 /tmp/before/inttype-example.o 1864 624 376 2864 b30 /tmp/after/inttype-example.o text data bss dec hex filename 1916 464 472 2852 b24 /tmp/before/record-example.o 1916 464 472 2852 b24 /tmp/after/record-example.o # You will see that it changed nothing. $ objdump -d /tmp/before/bytestream-example.o >/tmp/bytestream-example.before $ objdump -d /tmp/after/bytestream-example.o >/tmp/bytestream-example.after $ diff /tmp/bytestream.before /tmp/bytestream.after -urN --- bytestream.before 2012-10-26 20:55:33.645578668 +0800 +++ bytestream.after 2012-10-26 20:55:26.520578669 +0800 @@ -1,5 +1,5 @@ -/tmp/bytestream-example.o: file format elf64-x86-64 +/tmp/bytestream-example.o: file format elf64-x86-64 # So, as you can see, expect the filename, they are same. So, Stefani, is it what you want? Does this looks OK to you? Thanks, Yuanhan Liu