Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Fri, 12 Oct 2012 00:04:07 +0200
From: magnum <>
Subject: Re: Shared find_best_workgroup

On 11 Oct, 2012, at 21:30 , Claudio André <> wrote:

> Em 10-10-2012 10:25, magnum escreveu:
>> But it's not the end of the world. I'll give this up for a while and concentrate on other run-time tweaks: I'm currently experimenting with a way for the formats to pass stuff like "-DHASH_LOOP=1024" when building the kernel. This way things can be fully unrolled or otherwise optimised even when parameters are chosen at run-time.
> And the compilation time? I had problems when i tried to do "more optimizations" (e.g. unrolls).
> I saw one compilation process that took 20 minutes.

I've seen like 20 seconds, but never 20 minutes. I have also seen (fairly weak) evidence that my kernel splits are resulting in better performance instead of worse, likely because the resulting kernels are using a fixed and fairly low number of loops [so are unrolled] as opposed to a variable and fairly high number [so are not].


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.