netdev - netlink & rhashtable status

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-next>] [day] [month] [year] [list]

Message-ID: <1431495036.566.126.camel@edumazet-glaptop2.roam.corp.google.com>
Date:	Tue, 12 May 2015 22:30:36 -0700
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	David Miller <davem@...emloft.net>, Thomas Graf <tgraf@...g.ch>,
	Herbert Xu <herbert@...dor.apana.org.au>
Cc:	netdev <netdev@...r.kernel.org>
Subject: netlink & rhashtable status


linux-3.17, 3.18, 3.19 and 4.0.2 have issues with netlink/rhashtable,
making things like getaddrinfo() not working after a while :

socket(PF_NETLINK, SOCK_RAW, NETLINK_ROUTE) = 3
bind(3, {sa_family=AF_NETLINK, pid=0, groups=00000000}, 12) = 0
getsockname(3, {sa_family=AF_NETLINK, pid=24658, groups=00000000}, [12]) = 0
sendto(3, "\24\0\0\0\26\0\1\3)\320RU\0\0\0\0\0\0\0\0", 20, 0, {sa_family=AF_NETLINK, pid=0, groups=00000000}, 12) = -1 ECONNREFUSED (Connection refused)


lpaa23:~# tc qd
Cannot send dump request: Connection refused
lpaa23:~# ip ro ls
Cannot send dump request: Connection refused


Before the problem we have following visible sockets.
 
lpaa23:~# cat /proc/net/netlink 
sk       Eth Pid    Groups   Rmem     Wmem     Dump     Locks     Drops     Inode
ffff883ff23e7400 0   0      00000000 0        0        0 2        0        3       
ffff883fff0f5800 4   0      00000000 0        0        0 2        0        13314   
ffff881ff0f41800 8   0      00000000 0        0        0 2        0        11308   
ffff883ff0902400 9   0      00000000 0        0        0 2        0        10241   
ffff883ff167c800 10  0      00000000 0        0        0 2        0        9719    
ffff881ff13cfc00 11  0      00000000 0        0        0 2        0        2051    
ffff881fde863800 15  7460   00000001 0        0        0 2        0        509     
ffff883fff055000 15  4294963199 00000001 0        0        0 2        0        30754   
ffff881ff2204c00 15  0      00000000 0        0        0 2        0        5       
ffff881ff13cf400 16  0      00000000 0        0        0 2        0        2054    
ffff881ff0a0f000 20  0      00000000 0        0        0 2        0        2084    


When problem happens we have : (note the first socket disappeared)

lpaa23:~# cat /proc/net/netlink 
sk       Eth Pid    Groups   Rmem     Wmem     Dump     Locks     Drops     Inode
ffff883fff0f5800 4   0      00000000 0        0        0 2        0        13314   
ffff881ff0f41800 8   0      00000000 0        0        0 2        0        11308   
ffff883ff0902400 9   0      00000000 0        0        0 2        0        10241   
ffff883ff167c800 10  0      00000000 0        0        0 2        0        9719    
ffff881ff13cfc00 11  0      00000000 0        0        0 2        0        2051    
ffff881fde863800 15  7460   00000001 0        0        0 2        0        509     
ffff883fff055000 15  4294963199 00000001 0        0        0 2        0        30754   
ffff881ff2204c00 15  0      00000000 0        0        0 2        0        5       
ffff881ff13cf400 16  0      00000000 0        0        0 2        0        2054    
ffff881ff0a0f000 20  0      00000000 0        0        0 2        0        2084    


We probably need to find out what fixes are needed for stable kernels,
if not already done, sorry if I missed something.

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