[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <AANLkTi=vN-YOFLTcN4Lw_h9CqGLqko=yvmmsuhw0kAK7@mail.gmail.com>
Date: Tue, 21 Dec 2010 20:23:46 -0800
From: Dhiru Kholia <dhiru.kholia@...il.com>
To: john-users@...ts.openwall.com
Subject: Re: NTLM and OpenCL 1.0
On Tue, Dec 21, 2010 at 6:27 PM, Dhiru Kholia <dhiru.kholia@...il.com> wrote:
> I tried running your patch on a ATI Radeon 4870 system (with ATI
> Stream SDK 2.3, Catalyst 10.12, Linux 64-bit system, GCC 4.5.2). The
> OpenCL kernel fails to compile with "goto statement not allowed"
> errors (log attached). It seems that ATI's OpenCL implementation is
> somewhat more "strict" than NVIDIA's.
Attached patch (for OpenCL kernel) seems to work fine on both ATI 4870
and NVIDIA GT 240.
Benchmarking: NT MD4 [OpenCL 1.0]...
OpenCL Platform: <<<ATI Stream>>> and device: <<<ATI RV770>>>
DONE
Raw: 11707K c/s real, 11822K c/s virtual
CPU : AMD X3 720
Benchmarking: NT MD4 [OpenCL 1.0]...
OpenCL Platform: <<<NVIDIA CUDA>>> and device: <<<GeForce GT 240>>>
DONE
Raw: 23359K c/s real, 41391K c/s virtual
CPU : Core i5 750
The low performance numbers for 4870 don't make much sense (different
underlying CPUs?). Any guesses?
--
Cheers,
Dhiru
View attachment "nt_opencl_kernel_ati_hack.patch" of type "text/x-patch" (1154 bytes)
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.
