netdev - Re: [PATCH v3 2/3] packet: fill the gap of TPACKET

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:	Tue, 17 Dec 2013 01:48:45 +0900
From:	Atzm Watanabe <atzm@...atosphere.co.jp>
To:	Daniel Borkmann <dborkman@...hat.com>
Cc:	David Laight <David.Laight@...LAB.COM>, netdev@...r.kernel.org,
	Stephen Hemminger <stephen@...workplumber.org>,
	Ben Hutchings <bhutchings@...arflare.com>,
	David Miller <davem@...emloft.net>
Subject: Re: [PATCH v3 2/3] packet: fill the gap of TPACKET_ALIGNMENT with zeros

At Mon, 16 Dec 2013 13:55:43 +0100,
Daniel Borkmann wrote:
> 
> On 12/16/2013 01:41 PM, Atzm Watanabe wrote:
> > At Mon, 16 Dec 2013 11:22:35 +0100,
> > Daniel Borkmann wrote:
> >>
> >> On 12/16/2013 11:16 AM, David Laight wrote:
> >>>>> +		memset(h.h2->tp_padding, 0,
> >>>>> +		       hdrlen - offsetof(struct tpacket2_hdr, tp_padding));
> >>>
> >>> What is wrong with 'sizeof h.h2->tp_padding' ?
> >>> The compiler will probably inline the memset into a couple of word
> >>> sized writes of zero - probably not measurable.
> >>> Not zeroing them might be leaking kernel memory contents (depends
> >>> where the memory came from - might just be stale packet data).
> >>
> >> The ring buffer memory we're operating on comes from mmap(2) btw.
> >
> > Thank you for comments.
> > In struct tpacket2_hdr, it seems that a padding member was really
> > zeroing to fix information leak on commit
> > 13fcb7bd322164c67926ffe272846d4860196dc6 ("af_packet: prevent information leak").
> > So next time I'll try to zero tp_padding using sizof(h.h2->tp_padding)
> > David proposed.  If you have any thoughts on this please share it with me.
> 
> Yep, 13fcb7bd was for struct tpacket_auxdata structure in packet_recvmsg()
> that sits on the stack and copied uninitialized data to user space. But,
> okay, lets go with the memset().

Thank you, I understood.
I'll send the v4 as early as tomorrow.
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html