[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20120203224240.GA18521@openwall.com>
Date: Sat, 4 Feb 2012 02:42:40 +0400
From: Solar Designer <solar@...nwall.com>
To: john-users@...ts.openwall.com
Subject: Re: DES with OpenMP
On Fri, Feb 03, 2012 at 09:57:34PM +0000, Alex Sicamiotis wrote:
> Btw #2, I've also run into plenty of trouble with the Open64 (5.0) and AMD Open 64 (4.5.1) which I've been experimenting with (it's generally quite slower than gcc and icc). Too many compilation and runtime issues which is almost certainly due to compiler maturity or inherent limitations. If it was producing faster binaries it might even worth a look into making the code more open64-compatible.
I don't mind making the source code more compatible with another
compiler even if the resulting binaries are slower right now. This may
help improve code portability in general, and it may assist in more
extensive testing of further changes to John. So you may report
specific compile errors you're getting - perhaps to john-dev since
builds with Open64 are in fact of little relevance to john-users now and
since the postings will be expected to result in discussions of the
relevant places in the source code.
> I am not sure if this applies to bulldozer or xop-specific customisations. There was a benchmark in phoronix that was quite impressive (c-ray):
>
> http://www.phoronix.com/scan.php?page=article&item=amd_bulldozer_compilers&num=2
>
> 3d rendering (which is pretty intensive in calculations) got very fast on the bulldozer with the Open64, "bulldozing" even the overclocked i5's and 6-core i7's: http://www.phoronix.com/scan.php?page=article&item=intel_corei7_3960x&num=7 ...this MAY imply cracking potential that is simply unavailable in GCC (?).
Hardly. This is more likely either a special case that would not apply
here or even just an artifact of Phoronix's benchmarking. Speaking of
the latter, Phoronix is notorious for not mentioning the details of how
the software was built. For example, for JtR - while I am grateful that
it's often included in their benchmarks - it is almost never known what
make target was used (well, maybe it may be found somewhere in "pts" -
but I am not sure if that would reliably reflect benchmarks included in
specific Phoronix news stories). Phoronix posted benchmarks for JtR on
Bulldozer fairly early:
http://www.phoronix.com/scan.php?page=article&item=amd_bulldozer_compilers&num=5
but I could not make any use of those since it was not known if they
were SSE2, AVX, or XOP. It's only much later that I finally got
reliable XOP benchmarks (vs. AVX on the same CPU) on this forum:
http://www.linux.org.ru/news/hardware/7282032/page1#comment-7284811
Note the level of detail - just what is needed and just what we don't
get from Phoronix. So I was able to add these newer results to the wiki:
http://openwall.info/wiki/john/benchmarks
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.
