linux-hardening - Re: [PATCH v3] iio: sx9324: avoid copying property strings

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] [day] [month] [year] [list]

Message-ID: <9a9e84bf64fc95840ff852aeb276ef8f724b60be.camel@perches.com>
Date: Sun, 17 Dec 2023 05:41:12 -0800
From: Joe Perches <joe@...ches.com>
To: Jonathan Cameron <jic23@...nel.org>
Cc: Justin Stitt <justinstitt@...gle.com>, Lars-Peter Clausen
 <lars@...afoo.de>,  Stephen Boyd <swboyd@...omium.org>,
 linux-iio@...r.kernel.org, linux-kernel@...r.kernel.org, 
 linux-hardening@...r.kernel.org
Subject: Re: [PATCH v3] iio: sx9324: avoid copying property strings

On Sun, 2023-12-17 at 13:24 +0000, Jonathan Cameron wrote:
> On Mon, 11 Dec 2023 22:30:12 -0800
> Joe Perches <joe@...ches.com> wrote:
> 
> > On Tue, 2023-12-12 at 00:42 +0000, Justin Stitt wrote:
> > > We're doing some needless string copies when trying to assign the proper
> > > `prop` string. We can make `prop` a const char* and simply assign to
> > > string literals.  
> > 
> > trivia:
> > 
> > I would have updated it like this moving the
> > various declarations into the case blocks
> > where they are used and removing a few unused
> > #defines
> 
> I'd definitely like to see those defines gone.
> Arguably an unrelated change as I guess they are left from a previous refactor
> of this code.
> 
> Why prop to type renaming?

random, no specific need, though I prefer not reusing
identifiers with different types in separate local scopes.

>   It's getting passed into calls where the parameter
> is propname so I'd understand renaming to that, but type just seems a bit random
> to me.  I do wonder if we are better off having some long lines and getting rid
> of the property naming local variables completely by just duplicating
> the device_property_read_u32() call and passing them in directly.

maybe, give it a try and see what you think.