I have a bunch of files with unicode filenames that I'm trying to copy onto an MSDOS mount and what ends up happening is I get a bunch of invalid nodes that cannot be accessed in any way, including rm -rf /mymountpt/dir. This seems to be because mount_msdos is trying to be too smart for its own good and convert character sets in a way that apparently results in invalid filenames sometimes. I am using -L en_US.UTF-8 -D UTF-8 to mount the fs.
I've already had to wipe the FAT table, re-newfs and restore from backup on OpenBSD because there was no other way to get rid of the invalid nodes (the nodes created by DragonFly are undeleteable even on OpenBSD). As far as I'm aware, OpenBSD's mount_msdos treats each filename as a raw char *, to be passed through byte for byte without regard to charsets or code pages or what have you. This is exactly the behavior I'm looking for. So my question is, is there any way to force raw, unconverted filenames on DragonFly? I'd like to be able to use this thing on all my boxen without resorting to silliness like NFS exporting it from OpenBSD. Mike
