linux-kernel - Re: [PATCH] driver core: auxiliary bus: Optimize auxiliary_match

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 linux-cve-announce 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]

Message-ID: <2025090209-unlinked-remedial-555e@gregkh>
Date: Tue, 2 Sep 2025 14:51:53 +0200
From: Greg Kroah-Hartman <gregkh@...uxfoundation.org>
To: Zijun Hu <zijun_hu@...oud.com>
Cc: Dave Ertman <david.m.ertman@...el.com>, Ira Weiny <ira.weiny@...el.com>,
	Leon Romanovsky <leon@...nel.org>,
	"Rafael J. Wysocki" <rafael@...nel.org>,
	Danilo Krummrich <dakr@...nel.org>, linux-kernel@...r.kernel.org,
	Zijun Hu <zijun.hu@....qualcomm.com>
Subject: Re: [PATCH] driver core: auxiliary bus: Optimize auxiliary_match_id()

On Tue, Sep 02, 2025 at 08:42:24PM +0800, Zijun Hu wrote:
> On 2025/9/2 20:20, Greg Kroah-Hartman wrote:
> > On Tue, Sep 02, 2025 at 08:05:32PM +0800, Zijun Hu wrote:
> >> From: Zijun Hu <zijun.hu@....qualcomm.com>
> >>
> >> Variable @match_size is fixed in auxiliary_match_id().
> >>
> >> Optimize the function by moving the logic calculating the variable
> >> out of the for loop.
> > Optimize it how?  Does this actually result in a difference somehow, if
> > so, what?
> > 
> 
> auxiliary_match_id() repeatedly calculates variable @match_size in the
> for loop. however, the variable is fixed actually. so it is enough to
> calculate the variable once.
> 
> Optimize it by moving the logic calculating the variable out of the for
> loop, and it result in:
> 
> 1) calculate the variable once instead of repeatedly.
> 2) it will return as early as possible if device name is unexpected,
> namely, (@p == NULL).
> 
> so this fix will improve performance.

This isn't a "performant" path at all, is it?

I'm not disagreeing that calculating this all the time is a bad idea,
but please be reasonable.  If it can't be measured, you can't really say
it will "improve performance" :)

thanks,

greg k-h