From mboxrd@z Thu Jan 1 00:00:00 1970 From: Ingo Molnar Subject: Re: [RFC] speeding up the stat() family of system calls... Date: Sun, 12 Jan 2014 18:46:07 +0100 Message-ID: <20140112174607.GA26698@gmail.com> References: <20131224204625.GB20471@gmail.com> <52BD199A.8070607@zytor.com> <52BE0D89.3080405@zytor.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: Linus Torvalds , Ingo Molnar , Thomas Gleixner , Al Viro , the arch/x86 maintainers , linux-fsdevel , Linux Kernel Mailing List To: "H. Peter Anvin" Return-path: Content-Disposition: inline In-Reply-To: <52BE0D89.3080405@zytor.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-fsdevel.vger.kernel.org * H. Peter Anvin wrote: > On 12/26/2013 10:09 PM, H. Peter Anvin wrote: > > On 12/26/2013 11:00 AM, Linus Torvalds wrote: > >> > >> Interestingly, looking at the cp_new_stat() profiles, the games we > >> play to get efficient range checking seem to actually hurt us. Maybe > >> it's the "sbb" that is just expensive, or maybe it's turning a (very > >> predictable) conditional branch into a data dependency chain instead. > >> Or maybe it's just random noise in my profiles that happened to make > >> those sbb's look bad. > >> > > > > Much to my surprise, this patch adds almost 10K of text to an > > "allyesconfig" build. I wouldn't have expected it. I'll look at > > it some more tomorrow. > > Mystery solved... it is all code added by gcov &c because a new > (inline) function is added to the code base. So it is fluff, not > real. Yeah, defconfig builds are better for size comparisons, at least on x86 they are distro-config derived so a lot more relevant to real life than allyesconfig or allmodconfig. Thanks, Ingo