[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CA+TsHUA=FZmRvOR=ZZ4SyMxPx5pSD+d93poFpxuRfDpuBKLOtw@mail.gmail.com>
Date: Sat, 21 Apr 2012 01:49:49 +0530
From: SAYANTAN DATTA <std2048@...il.com>
To: john-dev@...ts.openwall.com
Subject: Re: cl_khr_byte_addressable_store
On Sat, Apr 21, 2012 at 1:29 AM, Milen Rangelov <gat3way@...il.com> wrote:
> Well especially for RAR on AMD, I had several attempts around that idea
> and they ended much slower than the vectorized, bitwise magic version. But
> you should leave it just because 4xxx is not supported.
>
I think none of the current opencl format except mine( opencl-mscah2) runs
on 4000 series hardware. We could have two versions of kernels one for 4000
series and another for 5000 series or above.
I know sometimes it's hard and it could get VERY UGLY (my rar kernel is
> frightening).
>
Yeah I faced the same problem while porting the cpu version of mscash2 to
gpu (ATi 4890). Lot of changes had to be made in each line and debugging
becomes a true nightmare.
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.
