Hi Marty The only problem with this marked below: need CONFIG_MTD_PARTITIONS otherwise you don't get offered NANDSIM I am currently using this to find the problem with the mv. I think it is most likely due to the Rename changes I did not being done properly which causes a dcache inconsistency. On Friday 29 July 2005 16:13, Martin Fouts wrote: > If you wish to use the mtd nandsim device as a simulator for testing > yaffs, you can follow these steps: > > 1) Make sure you have a recent 2.6 kernel with a recent MTD tree. (If > you ask the MTD folk, they'll tell you to run the mtd patchin script > from the latest cvs version of MTD.) > > 2) Configure your kernel to enable the mtd nand support and nandsim > driver. Make sure that the following are set in .config: > > CONFIG_MTD=m > CONFIG_MTD_CHAR=m > CONFIG_MTD_BLOCK=m CONFIG_MTD_PARTITIONS=m > CONFIG_MTD_NAND=m > CONFIG_MTD_NAND_NANDSIM=m > > You can set them to 'y' rather than 'm' if you don't want the modules > form. > > 3) make and install your kernel and modules in the usual way, then boot. > > 4) To use the nandsim, you need to load some MTD modules, unless you've > configured MTD into the kernel: > > modprobe mtdblock > modprobe mtdchar > modprobe nandsim > > 5) Verify that the nandsim device has loaded properly: > > cat /proc/mtd > > You will see something like > > dev: size erasesize name > mtd0: 00800000 00002000 "NAND simulator partition" > > If everything went well. > > 6) what the nand device is called will depend on your Linux distro, and > whether you are using devfs or udev, but the '0', or whatever mtd > reports as the device associated with the simulator, in the above will > show up in the name. On my ubuntu system, if /proc/mtd reports 'mtd0' > then the block device is /dev/mtdblock0 and the char device is > /dev/mtd0. > > 7) You can mount yaffs on the nandsim in the usual way. I have > /mnt/nand as the directory I mount on, so I would mount using > > mount -t yaffs /dev/mtdblock0 /mnt/nand > > > If you find any errors in this, please let me know. It wll eventually > find its way into the yaffs documentation. > > Marty