From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 13083 invoked by alias); 6 Aug 2004 17:19:23 -0000 Mailing-List: contact ecos-discuss-help@ecos.sourceware.org; run by ezmlm Precedence: bulk List-Subscribe: List-Archive: List-Post: List-Help: , Sender: ecos-discuss-owner@ecos.sourceware.org Received: (qmail 13076 invoked from network); 6 Aug 2004 17:19:22 -0000 Received: from unknown (HELO willers.employees.org) (192.83.249.36) by sourceware.org with SMTP; 6 Aug 2004 17:19:22 -0000 Received: by willers.employees.org (Postfix, from userid 10726) id A3C285C79B; Fri, 6 Aug 2004 10:19:21 -0700 (PDT) To: DAM@tt.dk (David Marqvar), andrew@lunn.ch (Andrew Lunn) Date: Fri, 06 Aug 2004 17:19:00 -0000 Cc: ecos-discuss@ecos.sourceware.org, Vivek.Kumar@zeevo.com In-Reply-To: <0175EEA85416D811893E009027D0F5D79389D4@MAILHOST> from "Kumar, Vivek" at Aug 06, 2004 10:12:55 AM MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Message-Id: <20040806171921.A3C285C79B@willers.employees.org> From: vivek@employees.org (Vivek Kumar) Subject: Re: [ECOS] Flash infrastructure rework X-SW-Source: 2004-08/txt/msg00124.txt.bz2 Hi, The exact cfi specs can be downloaded from www.jedec.com (registration is probably required). AMD and Intel would cover all bases. SST has its own extensions (so does Mitsubishi) but they are almost similar to the AMD type. The implementation is really simple, but the the real pain was to take care of is the 'diversions' which the vendors take and not tell you. The things I remember offhand are - SST flash lists its sectors twice, so if you add all of them up you get double the flash size. - SST also needs some extended commands to enter the CFI mode. - AMD extensions V1.1 and below do not tell whether its a top boot flash or a bottom boot and the sectors are sometime stored as smallest first or the largest first. We had to read the Flash ID to determine what to do with the sectors. and on and on and on At one point we were thinking why the hell did we try cfi? But having one that I can say for sure its really good we did that. Marketing now does not now come to me every other day asking if I can add driver support for xxx flash, which most of the time meant reading the spec adding another table and testing it. Whew.. -Vivek http://www.vivekkumar.net > -----Original Message----- > From: David Marqvar (DAM) [mailto:DAM@tt.dk] > Sent: Friday, August 06, 2004 6:45 AM > To: Andrew Lunn > Cc: ecos-discuss@ecos.sourceware.org > Subject: RE: [ECOS] Flash infrastructure rework > > > Sounds great. > > I've been wanting to make a general flash driver based on CFI which > would support the two main programming algorithm's around: Intel > (Strata) and AMD (don't know if it's really these companies that > invented the algorithm, anyway I think you know what I mean). Is SST yet > another programming algorithm or == ADM? > > Though CFI I can read the physical layout of the flash and the > programming algorithm to use. > This way the flash-driver could be generel, not limited to one or more > certian devices. > > Do you see any show-stoppers for creating such driver? > > /David > > -----Original Message----- > From: ecos-discuss-owner@ecos.sourceware.org > [mailto:ecos-discuss-owner@ecos.sourceware.org] On Behalf Of Andrew Lunn > Sent: 6. august 2004 15:23 > To: eCos Disuss > Subject: Re: [ECOS] Flash infrastructure rework > > > I put the code on a branch so that a few people can test it and so we > > get a better idea how stable the code is with hardware i don't have > > access to. > > This is what i decided to do. In cvs there is now a branch called > flash_v2. If you checkout/update to that branch you will get my new > flash code. There is also new generic drivers for SST and Strata. Since > so many targets use strata i did not want to modify them all to use the > new driver. So the strata driver is a new packets along side the old > one. For the SST driver i have modified all targets that use it. That > was easier since only the e7t and the aim711 use this driver. > > I've attached a hardware dependent driver for our platform which makes > use of the sst and strata device. This could be used as an example as to > how to modify the hardware dependent part of a flash driver to use the > new code. > > One thing to watch out for is that the hardware dependent driver has to > go into libextras.a. If you forget this the driver will be thrown away > at link time... > > And lastly a warning to testers.... There could be bugs which destroy > your boot loader etc. Make sure you can restore the device with jtag > before playing with this code. > > Thanks > Andrew > > -- > Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos > and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss > > > _____________________ > Confidentiality note: This email and any attachments may contain private, confidential, and privileged material for the sole use of the intended recipient. Any unauthorized review, use, disclosure or distribution is strictly prohibited. If you are not the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message. > -- Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss