netdev - Re: [RFC 1/3] networking: make skb_put & friends return void pointers

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: <1497602577.2622.5.camel@sipsolutions.net>
Date:   Fri, 16 Jun 2017 10:42:57 +0200
From:   Johannes Berg <johannes@...solutions.net>
To:     Bjørn Mork <bjorn@...k.no>
Cc:     netdev@...r.kernel.org
Subject: Re: [RFC 1/3] networking: make skb_put & friends return void
 pointers

On Fri, 2017-06-16 at 09:53 +0200, Bjørn Mork wrote:

> There seem to be a large number of places where the char pointer was
> used directly. Not that I have any strong opinion either way, but
> adding lots of ugly casts like this seems to contradict the whole
> purpose of this change?:

> > -			*skb_put(info->rx_skb, 1) = x;
> > +			*(u8 *)skb_put(info->rx_skb, 1) = x;
> >  			inb(iobase + DATA_H);
> >  			info->rx_count--;
> >  
> 
> That does not look any better in my eyes, and there are ... what?
> ... hundreds of them?

Yeah it's a bit of a trade-off. There are ~140 of these, but >200 of
the removed casts. Joe also suggested we could replace these by
skb_put_u8() or so, which would be indeed be nicer. I can send a patch,
it's pretty simple.

johannes