Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 11 Mar 2004 14:08:42 +0100
From:      "Poul-Henning Kamp" <phk@phk.freebsd.dk>
To:        Bruce Evans <bde@zeta.org.au>
Cc:        Colin Percival <colin.percival@wadham.ox.ac.uk>
Subject:   Re: cvs commit: src/sys/sys mdioctl.h src/sys/dev/md md.c src/sbin/mdconfig mdconfig.8 mdconfig.c 
Message-ID:  <50520.1079010522@critter.freebsd.dk>
In-Reply-To: Your message of "Fri, 12 Mar 2004 00:04:47 %2B1100." <20040311230444.G6384@gamplex.bde.org> 

next in thread | previous in thread | raw e-mail | index | archive | help
In message <20040311230444.G6384@gamplex.bde.org>, Bruce Evans writes:

>It might also have fewer deadlock possibilities.  VOP_READ/VOP_WRITE are
>inherently more blocking than VOP_STRATEGY.

While true, this is not really relevant any more because the actual
VOP_READ/VOP_WRITE calls happen from a kthread dedicated to that
purpose.

The deadlock I fixed (or at least made optional) was caused by the
layering violation which stems from the fact that our buffer cache
is not sufficiently general and our syncer suffers from hurry-up-and-wait
ideology.

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?50520.1079010522>