From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from nz-out-0102.google.com ([64.233.162.203]) by canuck.infradead.org with esmtp (Exim 4.62 #1 (Red Hat Linux)) id 1FjGBZ-0007UR-7u for linux-mtd@lists.infradead.org; Thu, 25 May 2006 09:47:40 -0400 Received: by nz-out-0102.google.com with SMTP id q3so1664374nzb for ; Thu, 25 May 2006 06:47:31 -0700 (PDT) Message-ID: <4475B575.1070903@gmail.com> Date: Thu, 25 May 2006 08:47:33 -0500 From: Nishanth Menon MIME-Version: 1.0 To: Davide CASCONE References: <001501c67fd4$b50e37c0$253682a4@nap.st.com> In-Reply-To: <001501c67fd4$b50e37c0$253682a4@nap.st.com> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Cc: linux-mtd@lists.infradead.org Subject: Re: JFFS2 Control Mode Program on Sibley List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Davide CASCONE wrote: > Does JFFS2 use the Control Mode Program on Sibley device? After a my preliminary analysis, it seems that only the Object Mode Program is used. Is it right? > That is correct Davide. the reason for that is that in Control mode, one of the A/B halfs cannot be written, in short we only get 16 bits of contiguous data in memory. The neagitve part of using only Object mode is that Of the programming regions/sector, we loose one complete region for the cleanmarker.. something like 128K(? dont remember off hand) for 8/12 bytes of cleanmarker/EBH and hence the implementation using write buffering (something we do on nand-like flash devices) on M18 flashes. Regards, Nishanth Menon