Hi Rudolf, probably it's a stupid question, sorry. Do you have enabled in the kernel config the MSDOS partition tables? # # Partition Types # CONFIG_PARTITION_ADVANCED=y # CONFIG_ACORN_PARTITION is not set # CONFIG_OSF_PARTITION is not set # CONFIG_AMIGA_PARTITION is not set # CONFIG_ATARI_PARTITION is not set # CONFIG_MAC_PARTITION is not set CONFIG_MSDOS_PARTITION=y What is the output of /proc/partitions? I use the systemace driver with the FAT driver to access the compact flash and it works fine. In my system with 2 partitions (first one MSDOS, second one ext2), I enabled DEVFS and partition statistics too. # cat /proc/partitions major minor #blocks name rio rmerge rsect ruse wio wmerge wsect wuse running use aveq 254 0 125440 xsysace/disc0/disc 35 6 45 320 1 0 1 0 0 320 320 254 1 15717 xsysace/disc0/part1 34 3 37 260 1 0 1 0 0 260 260 254 2 109616 xsysace/disc0/part2 0 0 0 0 0 0 0 0 0 0 0 I hope it can help you. Cheers, Claudio Lanconelli Rudolf Usselmann wrote: > Trust me I have dug very deep in to it. It is not an endianes > problem. It reeds the MBR, and then goes to sector 233. There > is NO reference to sector 233 in the MBR (or at least I have > not found it). > VFAT and FAT are identical up to all of this. VFAT includes > long file names ... unwillingly I have been learning a lot > about the structure of MBR, and FAT file system ... > > Are you mounting the correct device ? If thereare partitions > > you need to mount the partition (ie., /dev/sad1 for partition 1), > Yes I am. I have a ton of debug messages, first from the low > level SD/MMC interface routines, than from the block driver, > and now I am adding stuff to the FAT FS driver as well. > Everything propagates correctly, and I can create "fake" cards > that will work (single partition, single FAT). > Trust me there is something funny about the FAT Drier in uCLinux ... > Thanks, > rudi |