I need some advice about bootldr's "nand force_erase" function

I had a previously working 2.6.25.2 root partition with 7 bad blocks, yaffs cvs 05/03/09, to be struck by thousands of bad block and other assorted messages upon the first and subsequent mountings using 2.6.29-rc7 yaffs 03-05/03/09 (I will post a detailed message to yaffs ML later). I've tried to recover using rm -rf /mnt/root, including using earlier kernels/yaffs, but rm -rf repeatedly segfaults.

Is bootldr's nand force_erase an option (given the wiki warning about it) to aid return of the system to some sensible number of bad blocks, if so how do I use it?

With thanks,

Jim

hw: Balloon 3.21 E1 Engineering build Samsung Nand K9F2G08UOM 256MB

wiki warning:

"If force_erase is specified then the bad block markers are erased too. Don't do this unless you know what you are doing. It can be necessary when doing yaffs development and you have eroneously marked a load a blocks bad."

http://www.balloonboard.org/balloonwiki/BootLoaderInfo