K 10
svn:author
V 10
nwhitehorn
K 8
svn:date
V 27
2009-07-30T21:51:07.045490Z
K 7
svn:log
V 484
The kernel now tries (and fails, since there isn't one) to start init
after some more hacking with TOC pointers and with setfault().

More irritatingly, to deal with certain broken implementations of this
architecture (I'm looking at you, Cell) that only implement a 65-bit
virtual address space, I've had to reduce the accessible effective
address range to 45 bits. This needs a lot of reexamination, and fixing
it properly will probably involve some kind of global VSID hash table.

END