[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Mon, 08 Jul 2013 21:06:17 +0530
From: Sayantan Datta <std2048@...il.com>
To: john-dev@...ts.openwall.com
Subject: Re: bug: GPU use in CPU-only formats
On Friday 05 July 2013 05:23 AM, magnum wrote:
> On 5 Jul, 2013, at 1:24 , magnum <john.magnum@...hmail.com> wrote:
>> I'm looking into some weirdness with mscash2-opencl when using multiple devices, I may have to tweak something.
> That problem was not introduced now so I'll leave it to Sayantan. It only happens with --test=0:
>
> $ ../run/john -t=0 -form:mscash2-opencl -dev=gpu
> Warning: doing quick benchmarking - the performance numbers will be inaccurate
> Device 0: GeForce GTX 570
> Device 1: Tahiti (AMD Radeon HD 7900 Series)
> Optimal Work Group Size:256
> Kernel Execution Speed (Higher is better):1.565550
> Optimal Global Work Size:253952
> WARNING: Device 0 is too slow and might cause degradation in performance.
> WARNING: Devices are NOT arranged in DECREASING order of performance and might cause degradation in performance.
> Benchmarking: mscash2-opencl, M$ Cache Hash 2 (DCC2) [PBKDF2-SHA1 OpenCL]... OpenCL error (CL_INVALID_WORK_GROUP_SIZE) in file (common_opencl_pbkdf2.c) at line (353) - (Enque Kernel Failed)
>
> $ ../run/john -t -form:mscash2-opencl -dev=gpu
> Device 0: GeForce GTX 570
> Optimal Work Group Size:128
> Kernel Execution Speed (Higher is better):0.470798
> Optimal Global Work Size:81920
> Device 1: Tahiti (AMD Radeon HD 7900 Series)
> Optimal Work Group Size:256
> Kernel Execution Speed (Higher is better):1.565677
> Optimal Global Work Size:253952
> WARNING: Devices are NOT arranged in DECREASING order of performance and might cause degradation in performance.
> Benchmarking: mscash2-opencl, M$ Cache Hash 2 (DCC2) [PBKDF2-SHA1 OpenCL]... DONE
> Raw: 64291 c/s real, 64129 c/s virtual
>
> I think that "decreasing order" requirement should be taken care of by the format, not the user. When using --device=gpu you select all GPU devices available but on Bull they happen to come in increasing order of performance.
>
> magnum
Fixed both of the issues in bleeding.
std2048@...l:~/Jtr-bleeding/run$ ./john -te -format=mscash2-opencl -dev=gpu
Device 0: GeForce GTX 570
Optimal Work Group Size:128
Kernel Execution Speed (Higher is better):0.472683
Optimal Global Work Size:81920
Device 1: Tahiti (AMD Radeon HD 7900 Series)
Optimal Work Group Size:128
Kernel Execution Speed (Higher is better):1.566039
Optimal Global Work Size:335872
Benchmarking: mscash2-opencl, M$ Cache Hash 2 (DCC2) [PBKDF2-SHA1
OpenCL]... DONE
Raw: 130689 c/s real, 130182 c/s virtual
std2048@...l:~/Jtr-bleeding/run$ ./john -te=0 -format=mscash2-opencl
-dev=gpu
Warning: doing quick benchmarking - the performance numbers will be
inaccurate
Device 0: GeForce GTX 570
Device 1: Tahiti (AMD Radeon HD 7900 Series)
Benchmarking: mscash2-opencl, M$ Cache Hash 2 (DCC2) [PBKDF2-SHA1
OpenCL]... DONE
Raw: 23076 c/s real, 23076 c/s virtual
Regards,
Sayantan
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.
