Hi,
  At last I am please to announce the first release of GDB for RiscBSD.
There is a gdb set in the 1.2-beta sets directory.

This is an alpha version so I expect there still to be bugs. It is based on
GDB 4.16 with an experimental arm-arm-netbsd configuration.

REQUIREMENTS: bsd-4554 or later kernel

Kernels released before version 4554 do not have userland breakpoint support
and do not properly support the ptrace(PT_[SG]ETFPREGS, ...) calls.

This means that if you run GDB I would expect GDB to die with a bus error.

with bsd-4554 or later (hence the release of the latest kernel) all
GDB features should work ...

Both integer and fp registers may be examined and changed.

Breakpoints should work fine (not extensively test, but they work in all my
test cases).

Single stepping is implemented by breakpointing the next instruction as there
is no CPU single step facility. This appears to be working fine but there is
always the possibility of a bug in calculating the PC following the next
instruction how ever it gets it right in all the test cases i.e. condition
execution, branches, DATA OPs with pc destination LDR pc, and LDM rx,{...pc}
that I have tried.

One thing that does NOT properly work with single stepping are signal handlers.
If the instruction being single stepped generates a SEGV say and an SEGV
handler is installed or an SIGALRM is generated say, then the signal handler
will not be single stepped. This is because the single step code cannot
breakpoint the signal handlers installed. However if this is a problem then you
should install a breakpoint manually in the signal handler.

Other than that you should now be able to debug code compiled with -g or
examine the coredumps produced after SEGV's etc.

If people find any problem please let me know.

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