Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 13 Dec 1998 00:06:59 +0100
From:      Eivind Eklund <eivind@yes.no>
To:        Dag-Erling Smorgrav <des@flood.ping.uio.no>, Matt Dillon <dillon@FreeBSD.ORG>
Cc:        hackers@FreeBSD.ORG
Subject:   Re: cvs commit: src/sys/kern vfs_syscalls.c
Message-ID:  <19981213000659.A5444@follo.net>
In-Reply-To: <xzppv9pgemn.fsf@flood.ping.uio.no>; from Dag-Erling Smorgrav on Sat, Dec 12, 1998 at 11:49:20PM %2B0100
References:  <199812122107.NAA12093@freefall.freebsd.org> <xzppv9pgemn.fsf@flood.ping.uio.no>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Dec 12, 1998 at 11:49:20PM +0100, Dag-Erling Smorgrav wrote:
> Speaking of kernel dumps, what happens to physical memory pages when
> they're not in use? Are they zeroed? If not, is it possible to add a
> kernel option that zeroes out unused pages, or fills them with a fixed
> pattern (e.g. f001f001 og deadbeef)? That would make kernel dumps more
> compressible...

They are zeroed so approx 1/2 the free pages are zeroed.  The
percentage was based on John's intuition.  The reason to not zero all
is that zeroing still cost CPU cache lines, I think.

The zeroing is done in the idle loop, or when there are "way too few
zeroed pages available".  All of this is from my memory of the code,
so you should check with reality (the code itself).

Eivind.

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



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?19981213000659.A5444>