netdev - Re: Broken IPv6 multicast binding

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:	Wed, 18 Jun 2014 00:43:59 -0700
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	James Ward <j.ward@...r.usyd.edu.au>
Cc:	"David S. Miller" <davem@...emloft.net>,
	Alexey Kuznetsov <kuznet@....inr.ac.ru>,
	James Morris <jmorris@...ei.org>,
	Hideaki YOSHIFUJI <yoshfuji@...ux-ipv6.org>,
	Patrick McHardy <kaber@...sh.net>, netdev@...r.kernel.org
Subject: Re: Broken IPv6 multicast binding

On Tue, 2014-06-17 at 14:12 +1000, James Ward wrote:
> My colleagues and I have uncovered what we believe is a bug around 
> multicast IPv6 introduced to the kernel recently.
> 
> When more than one socket is bound to different multicast group 
> addresses on the same interface and port, only the last bound socket 
> will receive packets sent to it.
> 
> I have bisected the kernel and identified that commit efe4208f47f907  - 
> https://github.com/torvalds/linux/commit/efe4208f47f907
> was the one that introduced this behaviour. Prior commits behave as 
> expected.
> 
> Attached are two demonstrations of the behaviour.
> 
> C:
> Run "mcreceive6 ff15::1 14678 0"
> Run "mcreceive6 ff15::2 14678 0"
> Run "mcsend6 ff15::1 14678"
> Run "mcsend6 ff15::2 14678"
> 
> Expected: all messages typed into both mcsend6 instances will be echoed 
> to corresponding mcreceive6 instances.
> Actual: only messages sent to ff15::2 are echoed. Messages sent to 
> ff15::1 are not echoed.
> Killing mcreceive6 on ff15::2 will mean messages on ff15::1 are received 
> again.
> 
> Python:
> Run "test_multicast.py"
> Test involving subscription to multiple groups fails.
> 
> James
> 

Hmm... next time you find a regression caused by a particular commit,
make sur to CC patch author, instead of various people found in
MAINTAINERS.

I believe this problem is fixed by commit
3bfdc59a6c24608ed23e903f670aaf5f58c7a6f3
("ipv6: Fix regression caused by efe4208 in udp_v6_mcast_next()")

Thanks



--
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