|
Message-ID: <20111218144833.GA32628@openwall.com> Date: Sun, 18 Dec 2011 18:48:33 +0400 From: Solar Designer <solar@...nwall.com> To: john-dev@...ts.openwall.com Subject: Re: 1.7.9-jumbo On Fri, Dec 16, 2011 at 08:38:26PM +0100, magnum wrote: > I think the main theme is "Added Intrinsics". They were added to a lot > of formats that were formerly running generic code on 64-bit builds. In > many cases, a 32-bit build was faster than a 64-bit one, I believe that > is no longer the case for any format. > > Other than that, there is almost nothing apart from performance that a > user will notice. A couple new formats perhaps. No new functionality > that I can recall. Tons of tiny tweaks, little performance boosts here > and there and bug fixes. Dynamic had a makeover too. > > I'm not sure you need to include more than that in a summary. Re-phrase > as you like :) I am still trying to come up with a description of the changes in 1.7.9-jumbo-5 vs. 1.7.8-jumbo-8. Here's what relbench sees (non-OpenMP, linux-x86-64i, the older version is "file 1"): Only in file 1: Netscape LDAP SSHA:Only one salt Only in file 1: PIX MD5:Raw Only in file 1: dynamic_1008 md5($p.$s) [joomla]:Only one salt Only in file 1: dynamic_1008 md5($p.$s) [joomla]:Many salts Only in file 1: OpenLDAP SSHA:Many salts Only in file 1: Netscape LDAP SSHA:Many salts Only in file 1: OpenLDAP SSHA:Only one salt Only in file 2: Raw SHA:Raw Only in file 2: PIX MD5:Many salts Only in file 2: dynamic_1008 md5($p.$s) [RADIUS User-Password]:Many salts Only in file 2: dynamic_1009 md5($s.$p) [RADIUS Responses]:Only one salt Only in file 2: PIX MD5:Only one salt Only in file 2: dynamic_1008 md5($p.$s) [RADIUS User-Password]:Only one salt Only in file 2: NT v2:Raw Only in file 2: dynamic_1009 md5($s.$p) [RADIUS Responses]:Many salts Number of benchmarks: 153 Minimum: 0.81602 real, 0.81602 virtual Maximum: 6.70472 real, 6.63777 virtual Median: 1.00484 real, 1.00384 virtual Median absolute deviation: 0.01990 real, 0.02303 virtual Geometric mean: 1.10899 real, 1.10841 virtual Geometric standard deviation: 1.33332 real, 1.33278 virtual I need some explanation on what has replaced the old dynamic_1008 (Joomla) and the SSHA formats. Also, someone could want to identify the format that became 18% slower and see if this is reproducible and if it can be avoided (in a future version). Thanks, Alexander
Powered by blists - more mailing lists
Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.