[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-Id: <E8965A8B-FAFD-4C9A-A1C2-B76F9F349267@gmail.com>
Date: Mon, 25 May 2015 11:39:13 +0800
From: Lei Zhang <zhanglei.april@...il.com>
To: john-dev@...ts.openwall.com
Subject: Re: interleaving in SHA256 & SHA512
> On May 24, 2015, at 11:14 AM, Solar Designer <solar@...nwall.com> wrote:
>
> Yes, but please don't give up so soon. Here's a test to perform:
> compare the speeds with and without interleaving (and with different
> factors) when running just one thread (OMP_NUM_THREADS=1 or/and a
> non-OpenMP build). Obviously, all of them will be low in absolute
> terms, and all will use this one core suboptimally (since running fewer
> than max threads on it), but we need to know whether interleaving
> achieves relative speedup in that test or not. It should, even if it
> does not achieve a speedup when running multiple threads per core.
> Whether a relative speedup is seen in this test or not may affect our
> next steps.
Here's a re-test with a non-OpenMP build:
On well:
SHA256, pwsafe
[x1]
Raw: 11232 c/s real, 11232 c/s virtual
[x2]
Raw: 9552 c/s real, 9552 c/s virtual
[x4]
Raw: 9024 c/s real, 9024 c/s virtual
[x8]
Raw: 10392 c/s real, 10496 c/s virtual
SHA512, saph
[x1]
Many salts: 1108 c/s real, 1120 c/s virtual
Only one salt: 1108 c/s real, 1098 c/s virtual
[x2]
Many salts: 1008 c/s real, 1008 c/s virtual
Only one salt: 1008 c/s real, 1008 c/s virtual
[x4]
Many salts: 847 c/s real, 847 c/s virtual
Only one salt: 855 c/s real, 855 c/s virtual
[x8]
(core dump)
On MIC:
SHA256, pwsafe
[x1]
Raw: 1129 c/s real, 1129 c/s virtual
[x2]
Raw: 878 c/s real, 878 c/s virtual
[x4]
Raw: 932 c/s real, 932 c/s virtual
[x8]
Raw: 878 c/s real, 878 c/s virtual
SHA512, saph
[x1]
Many salts: 12.6 c/s real, 25.3 c/s virtual
Only one salt: 12.6 c/s real, 25.3 c/s virtual
[x2]
Many salts: 33.5 c/s real, 33.5 c/s virtual
Only one salt: 33.3 c/s real, 33.3 c/s virtual
[x4]
Many salts: 19.7 c/s real, 19.7 c/s virtual
Only one salt: 19.6 c/s real, 19.6 c/s virtual
*** glibc detected *** run/john: free(): invalid next size (normal): 0x0000000001d13630 ***
*** glibc detected *** run/john: malloc(): memory corruption (fast): 0x0000000001d13740 ***
[x8]
Many salts: 9.4 c/s real, 9.4 c/s virtual
Only one salt: 9.4 c/s real, 9.4 c/s virtual
*** glibc detected *** run/john: free(): invalid next size (normal): 0x0000000003231630 ***
*** glibc detected *** run/john: malloc(): memory corruption (fast): 0x0000000003231740 ***
I think there's some memory bug in saph or somewhere else. I'm still investigating it.
Lei
Content of type "text/html" skipped
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.
