linux-kernel - Re: [PATCH V2] tipc: Use bsearch library function

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:   Sat, 16 Sep 2017 03:17:41 -0700
From:   Joe Perches <joe@...ches.com>
To:     Ying Xue <ying.xue@...driver.com>, Thomas Meyer <thomas@...3r.de>,
        jon.maloy@...csson.com, netdev@...r.kernel.org,
        tipc-discussion@...ts.sourceforge.net,
        linux-kernel@...r.kernel.org, davem@...emloft.net
Subject: Re: [PATCH V2] tipc: Use bsearch library function

On Sat, 2017-09-16 at 18:10 +0800, Ying Xue wrote:
> On 09/16/2017 05:58 PM, Joe Perches wrote:
> > On Sat, 2017-09-16 at 17:36 +0800, Ying Xue wrote:
> > > On 09/16/2017 05:26 PM, Joe Perches wrote:
> > > > On Sat, 2017-09-16 at 17:02 +0800, Ying Xue wrote:
> > > > > On 09/16/2017 03:50 PM, Thomas Meyer wrote:
> > > > > > Use common library function rather than explicitly coding
> > > > > > some variant of it yourself.
> > > > > > 
> > > > > > Signed-off-by: Thomas Meyer <thomas@...3r.de>
> > > > > 
> > > > > Acked-by: Ying Xue <ying.xue@...driver.com>
> > > > 
> > > > Are you sure you want to do this?
> > > > 
> > > > Note the comment above nameseq_find_subseq
> > > > 
> > > >  * Very time-critical, so binary searches through sub-sequence array.
> > > > 
> > > > What impact does this change have on performance?
> > > 
> > > Sorry, I couldn't see any essential difference between this new
> > > implementation and the original one except that the former tries to use
> > > the library function - bsearch() to replace the original binary search
> > > algorithm implemented in TIPC itself. Therefore, I don't think the
> > > change will have a big impact on performance.
> > > 
> > > If I miss something, please let me know.
> > 
> > Comparison via a function pointer in bsearch is slower
> > than direct code without the function call overhead.
> > 
> 
> Right, but probably we can tolerate the slight sacrifice here.

What part of "very time critical" have you verified
and benchmarked as inconsequential?

Please post your results.