On Thursday 08 December 2005 13:52, Charles Manning wrote:
> The original checkin had some kernel compilation issues which are now
> fixed.
>

The wide tnode support builds the memory map and choks it up into pieces which 
are 32-bit aligned and treats this as pointers.

See 
http://www.aleph1.co.uk/cgi-bin/viewcvs.cgi/yaffs2/yaffs_guts.c?r1=1.23&r2=1.24 
around line 576.

This code makes the assumption that it is OK to have a pointer on any 
arbitrary 32-bit boundary. This should be OK for ARM, but is it OK for all 
CPUs?