linux-kernel - Re: [PATCH net] drivers/net/wan: lapb: Replace the skb->len checks with pskb_may

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]

Date:   Sun, 04 Oct 2020 15:15:35 -0700 (PDT)
From:   David Miller <davem@...emloft.net>
To:     xie.he.0141@...il.com
Cc:     kuba@...nel.org, netdev@...r.kernel.org,
        linux-kernel@...r.kernel.org, willemdebruijn.kernel@...il.com,
        ms@....tdt.de
Subject: Re: [PATCH net] drivers/net/wan: lapb: Replace the skb->len checks
 with pskb_may_pull

From: Xie He <xie.he.0141@...il.com>
Date: Sat,  3 Oct 2020 17:46:19 -0700

> The purpose of these skb->len checks in these drivers is to ensure that
> there is a header in the skb available to be read and pulled.
> 
> However, we already have the pskb_may_pull function for this purpose.
> 
> The pskb_may_pull function also correctly handles non-linear skbs.
> 
> (Also delete the word "check" in the comments because pskb_may_pull may
> do things other than simple checks in the case of non-linear skbs.)
> 
> Cc: Willem de Bruijn <willemdebruijn.kernel@...il.com>
> Cc: Martin Schiller <ms@....tdt.de>
> Signed-off-by: Xie He <xie.he.0141@...il.com>

This is excessive.

On transmit the header will be in the linear area, especially
if it is only one byte in size.

You're adding a lot of extra checks, function calls, etc. which are
frankly unnecessary.

I'm not applying this unless you can prove that a non-linear single
header byte is possible in these code paths.  And you'll need to add
such proof to your commit message.

Thank you.