From mboxrd@z Thu Jan 1 00:00:00 1970 Date: Fri, 6 Dec 2013 15:45:45 -0300 From: Ezequiel Garcia To: linux-mtd@lists.infradead.org, linux-arm-kernel@lists.infradead.org, Jason Cooper , Brian Norris , Andrew Lunn , Gregory Clement Subject: Merge order in Armada 370/XP (mvebu) NAND support Message-ID: <20131206184544.GA14986@localhost> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Disposition: inline Content-Transfer-Encoding: 8bit Cc: Thomas Petazzoni , Olof Johansson , Lior Amsalem List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Hello, Now that all the pxa3xx-nand code has been merged by proper maintainers and is available in each subsystem's trees I'd like to ask for suggestions regarding a potential problem. Since v3.12, the pxa3xx-nand driver supports the "armada370-nand" compatible. This was probably a mistake as little to no NAND support is present in that release. The support to handle 4K pages devices (for instance, supporting the Mirabox board) was submitted later, and is currently in MTD tree, waiting to be pulled in v3.14. On the other side, the devicetree files (dts and dtsi) needed to enable such support was taken through the mvebu tree. Currently, a board booted using mvebu's for-next branch will try to probe the NAND device (because the compatible was merged in v3.12) and there's a chance it hits on of NAND core's BUG(). The patchset containing the devicetree changes was prepared under the assumption that the driver would be merged first, so no risk existed. I admit this was a crappy assumption on my side to make (and worse: it's not the first time I assume this). Jason/Brian: What do you think we can do? -- Ezequiel García, Free Electrons Embedded Linux, Kernel and Android Engineering http://free-electrons.com