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  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, 26 Apr 2016 01:41:56 -0400 (EDT)
From:	David Miller <davem@...emloft.net>
To:	tom@...bertland.com
Cc:	netdev@...r.kernel.org, kernel-team@...com
Subject: Re: [PATCH net-next 0/3] ila: Support for checksum neutral
 translations

From: Tom Herbert <tom@...bertland.com>
Date: Sat, 23 Apr 2016 11:46:54 -0700

> This patch set updates ILA to support draft-herbert-nvo3-ila-02. The
> primary addition is support checksum neutral ILA translation.
> This allows address to be performed and still keep any transport
> layer checksums that include the addresses in their pseudo header to
> still be correct without the translator needing to parse L4.
> 
> Other items are:
>    - Structures for ILA addresses, identifiers, locators
>    - Disallow translation on non-ILA addresses (check by
>      type in identifier).
>    - Change xlat (nf_input) to translates solely based
>      on matching locators not identifiers (since identifiers
>      are not obfuscated by checksum neutral).
>    - Side effect if above is that multiple ILA domains are
>      supported. Each local locator can map to a different
>      SIR address (ILA domain), and each domain defines its
>      own identifier space.
> 
> Tested: Ran TCP_RR with 200 cnxs. ILA performance is slightly better
> than previously since we are not longer parsing L4 for checksum
> handling. I amd seeing about 1% performance overhead. Also ran
> TCP_STREAM and tested non-ILA address (type=0) are not translated.
> 
> v2: Fix complilation errors

Series applied.

There were some conflicts I had to resolve due to the recent 64-bit
padding fixes, but those weren't too difficult to deal with.

Thanks.

Powered by blists - more mailing lists