From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from fed1mtao01.cox.net ([68.6.19.244]) by pentafluge.infradead.org with esmtp (Exim 3.22 #1 (Red Hat Linux)) id 17uF6X-0003no-00 for ; Wed, 25 Sep 2002 17:33:38 +0100 Subject: RE: Interest in DOC and YAFFS? --> YAFFS bootloading From: Russ Dill To: Henrik =?ISO-8859-1?Q?Nordstr=F6m?= Cc: Srinivasan.Ramasubramaniam@nokia.com, elf@buici.com, kenneth.johansson@etx.ericsson.se, linux-mtd@lists.infradead.org In-Reply-To: References: Content-Type: text/plain; charset=ISO-8859-1 Content-Transfer-Encoding: quoted-printable Date: 25 Sep 2002 09:34:13 -0700 Message-Id: <1032971654.13282.18.camel@russ> Mime-Version: 1.0 Sender: linux-mtd-admin@lists.infradead.org Errors-To: linux-mtd-admin@lists.infradead.org List-Help: List-Post: List-Subscribe: , List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: On Wed, 2002-09-25 at 06:34, Henrik Nordstr=F6m wrote: > On 25 Sep 2002, Russ Dill wrote: >=20 > > the compressed cramfs image will be loaded into ram, and then, from > > there, as pages are needed, they will be decompressed again and put int= o > > ram (but again, if memory is in a pinch, these pages can be dropped). I > > don't know if an ext2 initrd will do XIP (execute in place), but I > > somehow doubt that anyone has worked on that, meaning that the pages > > would need to be copied again, in which case, a cramfs would use less > > ram. >=20 > I have never succeded in getting cramfs to work from a ramdisk.. it does=20 > something magic with blocksizes that seems to srew up the ramdisk data=20 > (first blocks get "forgotten"). the block size needs to be changed to 4k, there are patches floating around =20 > As the ramdisk is living within the buffer cache a ext2 image most likely= =20 > will do XIP. I don't know about that, the data would need to be aligned to 4k, I would think that the first part of most 4k pages would have fs information or some such, someone can call me on this though