Follow @Openwall on Twitter for new release announcements and other news
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Sun, 12 Jul 2015 19:52:18 +0300
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: extend SIMD intrinsics

Alain,

On Sun, Jul 12, 2015 at 12:27:40PM -0400, Alain Espinosa wrote:
> I use __m128i pointers and static stack allocated __m128i arrays. Both generate unaligned loads. For example this generate an unaligned load and the compiler had all information to use aligned:
> 
> void sample_func()
> {
>   __m128i test_array[16];
> 
>   ...
>   __m128i test_var = test_array[6];
>   ...
> }

Right.  However, what if you move test_array to global scope, or declare
it "static" in the function?  I am wondering if the compiler possibly
doesn't want to rely on the stack being 16-byte aligned (as it normally
is per x86_64 ABI).

Alexander

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.