Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 10 Oct 1995 14:56:53 -0600
From:      Warner Losh <imp@village.org>
To:        Paul Traina <pst@shockwave.com>
Cc:        current@FreeBSD.ORG
Subject:   Re: tail dumps core 
Message-ID:  <199510102056.OAA18756@rover.village.org>
In-Reply-To: Your message of Tue, 10 Oct 1995 12:45:13 PDT

next in thread | raw e-mail | index | archive | help
: [code deleted that does a memset() on an integer]
:
: How could this possibly be allowed by the C standard?  I'm utterly
: confused.

I think that on one's complement machines you can have a zero that is
"+0" and one that is "-0", one of which may or may not be the
"standard" zero on that platform.  +0 is typically all zeros on this
machine but -0 typically has the sign bit set, and all the rest of the
bits clear.  However, on a machine like this one would expect the zero
to either be normalized, or the sign bit to be masked out.

It is a longshot, but I think that it is allowed.  It is certainly
allowed for pointers, but this isn't a pointer....

Warner





Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199510102056.OAA18756>