Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 28 Apr 2005 18:29:31 +1000
From:      Peter Jeremy <PeterJeremy@optushome.com.au>
To:        /dev/null <null@dnswatch.com>
Cc:        freebsd-current@freebsd.org
Subject:   Re: can't build kernel [with CONFIG this time][[[FIXED]]]
Message-ID:  <20050428082931.GA232@cirb503493.alcatel.com.au>
In-Reply-To: <59642.216.177.243.35.1114655240.localmail@webmail.dnswatch.com>
References:  <61364.216.177.243.42.1114593685.localmail@webmail.dnswatch.com> <84dead7205042706554f32484@mail.gmail.com> <57775.216.177.243.35.1114633009.localmail@webmail.dnswatch.com> <20050427204939.GA796@uk.tiscali.com> <53015.216.177.243.35.1114635611.localmail@webmail.dnswatch.com> <20050427212835.GA87673@uk.tiscali.com> <64018.216.177.243.35.1114641057.localmail@webmail.dnswatch.com> <42701D35.6020307@gmail.com> <42703671.8060707@attglobal.net> <59642.216.177.243.35.1114655240.localmail@webmail.dnswatch.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, 2005-Apr-27 19:27:20 -0700, /dev/null wrote:
>In my case I think it was just "sync" issues between the CPU cache and
>the RAM. Although it could have been just that this processor drove
>the RAM stick hard enough to make it fail. No matter, the stick of RAM is
>working flawlessly in a different box. :)

Modern hardware is pushed very close to the limits.  Maybe the RAM has
some pattern sensitivity that just happens to be hit by buildkernel or
buildworld.  Maybe a combination of marginal bypass capacitors,
marginal memory bus drivers means that the logic transition is a few
picoseconds longer than desirable if a large number of bits change
state simultaneously so that the transition isn't seen by the marginal
receiver.

-- 
Peter Jeremy



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050428082931.GA232>