From mboxrd@z Thu Jan 1 00:00:00 1970 From: Lukasz Majewski Date: Mon, 24 Aug 2015 09:48:41 +0200 Subject: [U-Boot] Question on fitupd In-Reply-To: References: Message-ID: <20150824094841.172ae381@amdc2363> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de Hi Simon, Brian, > +Jagan, Lukasz > > Hi Brian, > > On 19 August 2015 at 15:07, wrote: > > Dell Customer Communication > > > > > > > > > > > > PURPOSE: > > > > This email is with respect to the ?fitupd? command. > > > > In looking at the source in the current u-boot tree, fitupd > > supports a NOR flash only. > > > > > > > > > > > > OUR GOAL: > > > > We would like to expand fitupd to be able to flash a SPI NOR, and > > MMC block devices. I've already prepared some patches (which will be accepted shortly) to extend update_tftp() function, which is used directly by fitupd command. Please refer to the following patch series if they suit your needs: http://thread.gmane.org/gmane.comp.boot-loaders.u-boot/231775/focus=231772 In short - support for eMMC and other DFU supported mediums have been provided. > > > > This would require support for, file writes (ex. ext4write), raw > > block writes (ex. mmc write), etc. > > > > Our design also supports a master boot record with two bootable > > partitions. > > > > > > > > > > > > QUESTIONS: > > > > Is expanding fitupd support for block devices a desirable feature to > > contribute back? > > > > Is this something that the u-boot projects would want? > > > > > > > > Would you want this in the ../common path for everyone to have > > access to? > > > > Or, prefer that this support be added to ../board// > > for our unique support only? > > > > > > > > > > > > > > > > I am leaving out all the details for now, we are first trying to > > understand the general direction we should go. > > > > > > > > Thank you for any advice in this area. > > > > Brian Brelsford > > This sounds useful to me. I added a few people on CC you may have > more thoughts. > > It would be good if this could somehow use driver model, but it would > need some thought. > > Regards, > Simon -- Best regards, Lukasz Majewski Samsung R&D Institute Poland (SRPOL) | Linux Platform Group