Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Date: Tue, 2 Jun 2015 05:16:38 +0300
From: Solar Designer <>
Subject: Re: Interleaving of intrinsics

On Mon, Jun 01, 2015 at 11:03:08PM +0800, Lei Zhang wrote:
> I tried to reduce max_keys_per_crypt as Solar suggested, but simply setting MAX_KEYS_PER_CRYPT to 1 would break the code and cause segfault.

Of course, MAX_KEYS_PER_CRYPT must not be lower than the SIMD width
times the interleave factor times the thread count.  You're probably
breaking some assumptions internal to the format when you set it lower
than that (which is why the segfault), but even if you didn't this
would run slow because most of the computed hashes would be of dummy
inputs (of keys that were not set, since at indices higher than 1).


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.