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: <20241002073156.447d06c4@kernel.org>
Date: Wed, 2 Oct 2024 07:31:56 -0700
From: Jakub Kicinski <kuba@...nel.org>
To: Kory Maincent <kory.maincent@...tlin.com>
Cc: netdev@...r.kernel.org, linux-kernel@...r.kernel.org, Kyle Swenson
 <kyle.swenson@....tech>, Simon Horman <horms@...nel.org>, Oleksij Rempel
 <o.rempel@...gutronix.de>, thomas.petazzoni@...tlin.com, "David S. Miller"
 <davem@...emloft.net>, Eric Dumazet <edumazet@...gle.com>, Paolo Abeni
 <pabeni@...hat.com>
Subject: Re: [PATCH net v2] net: pse-pd: tps23881: Fix boolean evaluation
 for bitmask checks

On Wed, 2 Oct 2024 14:53:02 +0200 Kory Maincent wrote:
> On Wed, 2 Oct 2024 05:27:32 -0700
> Jakub Kicinski <kuba@...nel.org> wrote:
> 
> > On Wed, 2 Oct 2024 05:24:31 -0700 Jakub Kicinski wrote:  
> > > On Wed,  2 Oct 2024 12:23:40 +0200 Kory Maincent wrote:    
> > > > In the case of 4-pair PoE, this led to incorrect enabled and
> > > > delivering status values.      
> > > 
> > > Could you elaborate? The patch looks like a noop I must be missing some
> > > key aspect..    
> > 
> > Reading the discussion on v1 it seems you're doing this to be safe,
> > because there was a problem with x &= val & MASK; elsewhere.
> > If that's the case, please resend to net-next and make it clear it's
> > not a fix.  
> 
> Indeed it fixes this issue.

Is "this" here the &= issue or the sentence from the commit message?

> Why do you prefer to have it on net-next instead of a net? We agreed with
> Oleksij that it's where it should land. Do we have missed something?

The patch is a noop, AFAICT. Are you saying it changes how the code
behaves? 

The patch only coverts cases which are 

	ena = val & MASK;

the automatic type conversion will turn this into:

	ena = bool(val & MASK);
which is the same as:
	ena = !!(val & MASK);

The problem you were seeing earlier was that:

	ena &= val & MASK;

will be converted to:

	ena = ena & (val & MASK);

and that is:

	ena = bool(int(ena) & (val & MASK));
                   ^^^

IOW ena gets promoted to int for the & operation.
This problem does not occur with simple assignment.

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ