linux-kernel - Re: [PATCH 1/2] lib/string: Improve strstarts() performance

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: <Z_PjksaIc0B2AQhs@smile.fi.intel.com>
Date: Mon, 7 Apr 2025 17:39:14 +0300
From: Andy Shevchenko <andy@...nel.org>
To: Zijun Hu <zijun_hu@...oud.com>
Cc: Kees Cook <kees@...nel.org>, Andrew Morton <akpm@...ux-foundation.org>,
	linux-hardening@...r.kernel.org, linux-kernel@...r.kernel.org,
	Zijun Hu <quic_zijuhu@...cinc.com>,
	"Rob Herring (Arm)" <robh@...nel.org>
Subject: Re: [PATCH 1/2] lib/string: Improve strstarts() performance

On Mon, Apr 07, 2025 at 10:33:34PM +0800, Zijun Hu wrote:
> On 2025/4/7 21:51, Andy Shevchenko wrote:
> > First of all, this function is supposed to be run against constant string literals.
> 
> for strstarts(s, "prefix"), strlen("prefix") should *NOT* be compile
> time constant. it is a loop and unavoidable to have strlen("prefix")
> iterations.

What do you mean by that? Compiler uses __builtin_strlen() and it *IS*
a compile-time constant. Just check it.

> > Second, this commit message has zero proofs to tell if there is actual performance
> > downgrage even in the case when prefix is not a constant string literal.
> 
> for either constant string or non constant string. this patch
> eliminating a loop which have strlen()iterations. i feel it is obvious
> it can improve performance.

No, it's not.

-- 
With Best Regards,
Andy Shevchenko