Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 16 Jun 2002 15:46:41 -0700 (PDT)
From:      John Polstra <jdp@polstra.com>
To:        Matthew Dillon <dillon@apollo.backplane.com>
Cc:        obrien@FreeBSD.ORG, freebsd-hackers@FreeBSD.ORG, freebsd-current@FreeBSD.ORG, Alfred Perlstein <bright@mu.org>
Subject:   Re: rtld-elf patches for non-i386 architectures (review / test r
Message-ID:  <XFMail.20020616154641.jdp@polstra.com>
In-Reply-To: <200206162211.g5GMBgri042006@apollo.backplane.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Matthew Dillon wrote:
>:
>:Why memset(3) the anon memory to zero, isn't that what it's supposed
>:to be initialized to anyway?
>:
>:-- 
>:-Alfred Perlstein [alfred@freebsd.org]
> 
>     I was just being ultra-conservative.  You are absolutely correct
>     in regards to the anonymous area already being zero.  I would like
>     to commit with the memset()'s in just for uniformity and then make 
>     another pass to remove them.  (there's no real issue of cpu waste
>     since we are going to take faults on the anonymous pages anyway).

Sounds OK to me, although I worry a little bit about relying on the
zeroing of the memory by mmap.  I don't see it documented in the man
page, so there's a (slight) risk that somebody could remove that
feature in the future.

John

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.20020616154641.jdp>