Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 22 May 2012 22:29:52 +0200
From: magnum <john.magnum@...hmail.com>
To: john-dev@...ts.openwall.com
Subject: Re: Fwd: bash auto-completion for john

On 05/22/2012 07:39 PM, Frank Dittrich wrote:
> I just realized it might be more important to add support for GPU
> specific options.
> 
> (E.g., I saw -pla=1 mentioned in some john-dev threads.)
> 
> Can someone post john's usage output for a CUDA/OpenCL build

Ah, yes. For CUDA builds there is -device only. For OpenCL builds, there
are -platform and -device. A GPU build has both CUDA and OpenCL (and
then the -device option will mean CUDA device or OpenCL device depending
on what format you chose). They will show up like this in the usage blob:

CUDA build:
--device=N                set CUDA device, default 0

OpenCL build:
--platform=N (or =LIST)   set OpenCL platform, default 0
--device=N                set OpenCL device, default 0

GPU build:
--platform=N (or =LIST)   set OpenCL platform, default 0
--device=N                set OpenCL or CUDA device, default 0


> I probably need to install a proprietary driver, and the GF108 isn't the
> fastest nVidia card, but this shouldn't prevent me from using the GPU
> for john, even if only to find out which additional options need to be
> supported by the bash completion script.

The question is mostly about what distribution you use. Latest Ubuntu
versions of "nvidia-current" are OK for anything but GTX6xx as far as I
know. If you are to install the nvidia drivers off their web you may
need to blacklist nouveau first.

> (Do we even need additional --list= values, to list the available GPUs...)

--device=LIST (or --platform=LIST which is currently exactly the same)
works for OpenCL while CUDA values are in the blind for now. But I think
it's enough to just expand the options, not the arguments. Anyway here's
an example output (CPU only):

$ ../run/john --plat=list
Platform #0 name: Intel(R) OpenCL
Platform version: OpenCL 1.1 LINUX
	Device #0 name:		Intel(R) Core(TM)2 Duo CPU     P8600  @ 2.40GHz
	Device vendor:		Intel(R) Corporation
	Device type:		CPU (LE)
	Device version:		OpenCL 1.1 (Build 15293.6649)
	Driver version:		1.1
	Global Memory:		3855 MB
	Global Memory Cache:	3072 KB
	Local Memory:		32 KB (Global)
	Max clock (MHz) :	2400
	Max Work Group Size:	1024
	Parallel compute cores:	2

Platform #1 name: AMD Accelerated Parallel Processing
Platform version: OpenCL 1.1 AMD-APP-SDK-v2.5 (684.213)
	Device #0 name:		Intel(R) Core(TM)2 Duo CPU     P8600  @ 2.40GHz
	Device vendor:		GenuineIntel
	Device type:		CPU (LE)
	Device version:		OpenCL 1.1 AMD-APP-SDK-v2.5 (684.213)
	Driver version:		2.0
	Global Memory:		3855 MB
	Global Memory Cache:	32 KB
	Local Memory:		32 KB (Global)
	Max clock (MHz) :	800
	Max Work Group Size:	1024
	Parallel compute cores:	2


I suppose you could grep for /^Platform #([0-9]+)/ and /^\tDevice
#([0-9]+)/ if you like.

magnum

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.