Date: Fri, 2 Dec 2005 16:45:51 +0100 From: Csaba Henk <csaba-ml@creo.hu> To: Robert Watson <rwatson@freebsd.org> Cc: freebsd-current@freebsd.org, Kris Kennaway <kris@obsecurity.org> Subject: Re: double close strikes panic if md attaching a corrupt file Message-ID: <20051202154551.GP2911@beastie.creo.hu> In-Reply-To: <20051126124248.L81764@fledge.watson.org> References: <20051125214738.GL2911@beastie.creo.hu> <20051126025620.GA62284@xor.obsecurity.org> <20051126054748.GM2911@beastie.creo.hu> <20051126124248.L81764@fledge.watson.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Nov 26, 2005 at 12:43:36PM +0000, Robert Watson wrote: > > On Sat, 26 Nov 2005, Csaba Henk wrote: > > >On Fri, Nov 25, 2005 at 09:56:20PM -0500, Kris Kennaway wrote: > >>You probably should do so ;-) This isn't the easiest thing for someone > >>to test without such corrupted media. [...] > Well, a very easy way to test would be to add a local modification to UFS > to create a file flag that causes VOP_READ to return EIO :-). That way > you can easily mark files to return EIO. OK, I've been there, done that. Tried both with UFS + an SF_IOERROR flag and with a hacked reiserfs, where reading from regular files always fails. Original md code does choke on EIO returned. After I applied my first patch (md-doubleclose-1.diff), all worked fine, both in the case when reading returns error and when reading works normally. Csaba
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20051202154551.GP2911>