From mboxrd@z Thu Jan 1 00:00:00 1970 From: Hans Reiser Subject: Re: [PATCH] various allocator optimizations Date: Wed, 12 Mar 2003 17:00:32 +0300 Message-ID: <3E6F3D80.80306@namesys.com> References: <1047400482.8215.312.camel@tiny.suse.com> <20030311194205.A4493@namesys.com> <1047403968.8219.337.camel@tiny.suse.com> <3E6E584D.4080809@namesys.com> <1047421551.8219.448.camel@tiny.suse.com> <3E6E674E.4040305@namesys.com> <1047433739.8215.487.camel@tiny.suse.com> <20030312101222.A8452@namesys.com> <1047475876.8215.507.camel@tiny.suse.com> Mime-Version: 1.0 Content-Transfer-Encoding: 7bit Return-path: list-help: list-unsubscribe: list-post: Errors-To: flx@namesys.com In-Reply-To: <1047475876.8215.507.camel@tiny.suse.com> List-Id: Content-Type: text/plain; charset="us-ascii"; format="flowed" To: Chris Mason Cc: Oleg Drokin , reiserfs-list@namesys.com Chris Mason wrote: >On Wed, 2003-03-12 at 02:12, Oleg Drokin wrote: > > > >>>>No, it uses the left neighbor in the tree. Please correct me if I am >>>>wrong, because if I am wrong we have a bug. >>>> >>>> >>>It uses the left neighbor in the tree once the file has a starting >>>block. But the first starting block comes by searching from block #0. >>> >>> >>No, this is not correct. >>We look at the left neighbor. >>See the get_left_neighbor() in bitmap.c, >>first we make the search_start to be a block number of last path >>element (if there is path at all of course which may be not there in some >>cases), ans then check if we are allocating indirect block and if there is >>non zero element in current indirect item in path, if there is non-zero element, >>we take it as new search_start, otherwise it still contains blocknumber of >>left neighbor. >> >> > >Right, I missed the hint->search_start = bh->b_blocknr for when we don't >find an indirect item. > >This is better, so the search will start from the bh with the stat data >for a new file, but it depends on the last path element being in a good >position relative to the rest of the files in the directory. > Where other than near the stat data do you propose? > >-chris > > > > > > > > -- Hans