>1) /sys/arm32/arm32/genassym.c could not be compiled.
>I tracked this down to the renaming of the struct fp_state to fpe_sp_state in
>machine/fp.h which was not taken care of in machine/pcb.h.
>I'm quite sure that you have to change "struct fp_state pcb_fpstate" at the
end
>of the pcb struct declaration to "struct fpe_sp_state pcb_fpstate" because it
>compiles fine afterwards.
I found this as well. This has been fixed (your solution was correct) and
should be picked up by the next supscan

>2) Conflicting return types for htonl and ntohl in /sys/lib/libkern/htonl.c
and
>ntohl.c are reported.
>They are currently unsigned long, but are declared to u_int32_t in
>machine/endian.h. There is a declaration of in_addr_t anyway which is then
used
>to declare the return type, so I'm quite confident this should also be used in
>the abovementioned files.
>This is machine independant source. What is the appropriate way to report
this?
>A message to tech-kern, a send-pr?
>And don't tell me to post to current-users as usual. ;-)))

Yep an MI thing due to very recent changes. I saw changes to these files go
by on source-changes today so again this has been fixed.

>3) When linking, I get
>init_main.o(.text+0xac): undefined reference to 'mountroot'
>shell_shell.o(.text+0x15d4): undefined reference to 'checkinodes'

ok the checkinodes() thing has also been fixed (early hours of yesterday
morning). If you cannot wait just remove the references to checkinodes or
create a dummy function. checkinodes() is a debug function that should not be
called
from shell_shell.c but the refs go by me in the recent grand commital.

Do you have the line
option GENERIC
in your config file ? If not try adding it. The mountroot() stuff from stubs.c
should only be included if GENERIC is defined.

Cheers,
				Mark



-- 
Mark Brinicombe				amb@physig.ph.kcl.ac.uk
Research Associate			http://www.ph.kcl.ac.uk/~amb/
Department of Physics			tel: 0171 873 2894
King's College London			fax: 0171 873 2716