netdev - Re: TCP-MD5 checksum failure on x86

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]

Date:	Tue, 11 May 2010 08:27:47 +0200
From:	Eric Dumazet <eric.dumazet@...il.com>
To:	Bijay Singh <Bijay.Singh@...vus.com>
Cc:	Stephen Hemminger <shemminger@...tta.com>,
	David Miller <davem@...emloft.net>,
	"<bhaskie@...il.com>" <bhaskie@...il.com>,
	"<bhutchings@...arflare.com>" <bhutchings@...arflare.com>,
	"<netdev@...r.kernel.org>" <netdev@...r.kernel.org>
Subject: Re: TCP-MD5 checksum failure on x86_64 SMP

Le mardi 11 mai 2010 à 04:08 +0000, Bijay Singh a écrit :
> Hi Eric,
> 
> I guess that makes me the enviable one. So I am keen to test out this feature

>  completely, as long as I know what to do as a next step, directions, patches.
> 

MTU > 4000 is not reliable because of high order allocations on typical
NICS. I am afraid you need NIC able to deliver page fragments.

Its working here (32bit kernel) with a tg3 NIC, but I got following
message :

ifconfig eth3 mtu 9000
...
[51492.936500] 167731 total pagecache pages
[51492.936500] 0 pages in swap cache
[51492.936500] Swap cache stats: add 0, delete 0, find 0/0
[51492.936500] Free swap  = 4192928kB
[51492.936500] Total swap = 4192928kB
[51492.936500] 1114110 pages RAM
[51492.936500] 885761 pages HighMem
[51492.936500] 77073 pages reserved
[51492.936500] 134483 pages shared
[51492.936500] 159131 pages non-shared
[51492.953027] tg3 0000:14:04.1: eth3: Using a smaller RX standard ring.
Only 183 out of 511 buffers were allocated successfully

$ ethtool -g eth3
Ring parameters for eth3:
Pre-set maximums:
RX:		511
RX Mini:	0
RX Jumbo:	0
TX:		511
Current hardware settings:
RX:		183
RX Mini:	0
RX Jumbo:	0
TX:		511

$ cat /proc/buddyinfo
Node 0, zone      DMA      5      2      1      1      2      2      2      1      0      1      0 
Node 0, zone   Normal   4285   1823    248     73      9      5      0      0      0      0      0 
Node 0, zone  HighMem     97    199    921    583    383    261    155    117     69     41    649 

I know that if I try to stress RX path, I'll get failures.

Could you explain me why you need both big MTUS and TCP-MD5 ?



--
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