On Wed, Dec 17, 2008 at 05:02:57PM +0000, Al Viro wrote:BTW, after taking a look at the remaing asm-*...
* frv, m32r and h8300 should be as trivial as git mv
Yep - I already have positive feedback from arch maintainers that
they will fix it in the -rc1 timeframe.
* m68k has a bit of extra PITA - we need to clean up two non-trivial
includes from m68knommu (pci.h and setup.h) first, then it's a matter
of git mv, adding arch/m68k/include to search path on m68knommu and
removal of dummy headers from the latter.
I took a different route here and unified the headers for m68k and m68knommu
under arch/m68k/include/asm
It was based on a script originally developed by Arnd Bergmann.
Geert already acked it, and I expect Greg to apply it in the -rc1 timeframe.
http://marc.info/?l=linux-kbuild&m=122927123805002&w=2 has Geert's ack,
but original patch did not hit marc?!?
* xtensa is already moved in linux-nextYes.
* arm seems to be lacking the final push for few remaining headers; presumablyrmk knows it and takes care.
will happen eventually
* mn10300 is playing with symlinks - the last one to do that for headers,Thanks Al.
AFAICS.
I've thrown together a patch series on top of linux-next doing m68k and
mn10300. *WARNING* - it's completely untested, basically about half an
hour diversion from sifting through the mailbox for missed VFS patches.
And that half and hour did *not* include anything resembling building
the resulting tree. So take that with a trainload of salt.
I will take a closer look at the mn10300 bits and send them to David
in due time.
Sam