[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20140508035059.GA23084@openwall.com>
Date: Thu, 8 May 2014 07:50:59 +0400
From: Solar Designer <solar@...nwall.com>
To: john-dev@...ts.openwall.com
Subject: Re: ARCH_INT_GT_32
On Wed, May 07, 2014 at 03:09:28PM -0500, jfoug wrote:
> I thought this was supposed to be set to 1 for all CPU's larger than 32 bit
> int's. But looking at all headers (and source), it is always set to 0. Did
> this used to be set to 1, or is something not right?
Neither. We just don't have any explicit make targets for platforms
weird enough that they'd have larger than 32-bit int. That would be
e.g. Cray T3E:
http://gcc.gnu.org/ml/gcc-patches/2001-06/msg01752.html
"on the Cray T3E, short is 32 bits and int is 64 bits."
And it's not about the CPUs - rather, it's about the integer types as
seen from C. Other machines that used the same Alpha CPUs as Cray T3E
did commonly had 32-bit int in their native C compilers.
Our detect.c should correctly set ARCH_INT_GT_32 to 1 when compiling on
one of those rare platforms with "make generic". The logic is in there.
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.
