From mboxrd@z Thu Jan 1 00:00:00 1970 From: Yury Umanets Subject: Re: Reiser4 support on parted program ... Date: Wed, 15 Sep 2004 16:31:53 +0300 Message-ID: <41484449.8060301@clusterfs.com> References: <41472F46.4000109@futuretg.com> <41477625.304@slaphack.com> <4147E95B.8090108@futuretg.com> <20040915094952.GA5137@backtop.namesys.com> <414840FC.4090505@clusterfs.com> <20040915132013.GG26192@nysv.org> Mime-Version: 1.0 Content-Transfer-Encoding: quoted-printable Return-path: list-help: list-unsubscribe: list-post: Errors-To: flx@namesys.com In-Reply-To: <20040915132013.GG26192@nysv.org> List-Id: Content-Type: text/plain; charset="iso-8859-9"; format="flowed" To: =?ISO-8859-1?Q?Markus_T=F6rnqvist?= Cc: Alex Zarochentsev , "Dr. Giovanni A. Orlando" , David Masover , reiserfs-list@namesys.com Markus T=F6rnqvist wrote: >On Wed, Sep 15, 2004 at 04:17:48PM +0300, Yury Umanets wrote: > =20 > >>Alex is right in both. To add such a support one should write resizing=20 >>stuff in libreiser4 first. Then write glue code in parted. IMHO both are = >>not so difficult. Probably Vitaly or Alex self could do that in week=20 >>(may be few days). >> =20 >> > >How long for Dr. Orlando and his crew in cooperation with Namesys?-) > =20 > Have no idea :) I do not know anything about Dr. Orlando and his crew. Suff itself is not difficult, as there is not tree modifications, etc. Rough sequence is like this: (1) make sure, that there is no not replied transactions. (2) tree traverse and move all nodes and unformatted blocks which do not=20 fit to new range (new_start - new_end) one by one to new area. Block=20 allocator info should be updated on this stage too. (3) update some fields in super block. Probably one more update of block=20 allocator. The framework for all these actions exists in libreiser4. Parted's part is trivial -- just follow the API. > =20 > >>The only problem Hans seems does not want such a support or does not=20 >>count it as somethinf serious attention should be paid to :( >> =20 >> > >I think his followers would disagree ;) > > =20 > --=20 umka