From mboxrd@z Thu Jan 1 00:00:00 1970 From: Neil Brown Date: Tue, 14 Apr 2009 01:33:34 +0000 Subject: Re: [PATCH] 64-md-raid.rules fixes v2 Message-Id: <18915.59374.663499.384069@notabene.brown> List-Id: References: <7fe3a91a4b0971f1be490bdaccf37479047ac0a0.1237151296.git.soltys@ziu.info> In-Reply-To: <7fe3a91a4b0971f1be490bdaccf37479047ac0a0.1237151296.git.soltys@ziu.info> MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable To: linux-hotplug@vger.kernel.org On Sunday March 15, soltys@ziu.info wrote: (sorry for taking nearly a month to reply...) > 1) introduced in commit b822542608326092e177fd1707ca7fb53b2846c4 >=20 > md/array_state in case of partition doesn't exists, so all uevents > for partitions didn't execute any SYMLINK rules Thanks. I've applies this (with a slightly more verbose comment) in the version of the rules file to be released with mdadm-3.0 >=20 > 2) add compat rule, so both MD_DEVNAME (mdadm 3.0) and MD_NAME > (mdadm 2.9.x) can be used I probably don't agree with this. mdadm 2.x created dev files in /dev directly. So it does not expect udev to do anything, particularly with names. mdadm 3.x by contrast doesn't create anything in /dev (if udev seems to be installed) so the rules file must create everything. So I think that 'compat' rule is wrong. I'm happy to listen to more details arguments though. http://neil.brown.name/git?p=3Dmdadm;a=3Dcommitdiff;h=FA5090285af9401f950aa= 5c1a3a82708c69c4906 Thanks, NeilBrown