Re: [commit] gdb_byte regcache.[hc]

[Michael Snyder <msnyder at redhat dot com>]

Mark Kettenis wrote:

> I know it's a bit late (just returned from vacation in .ca ;-), but
> there is a nasty problem with the register cache stuff in this patch
> :-(.  Many bits of native code pass native data types to
> regcache_raw_write, and this now raises a compiler warning.  I'm not
> sure how to solve this; adding casts all over the code is really,
> really ugly, and I really, really don't want to do that.  The best
> thing I can come up with is that we revert the regcache_raw_xxx
> interfaces back to using `void *', whereas we keep the
> regcache_cooked_xxx interface as they are now (using `gdb_byte *').

I'm coming to this discussion even later than you, Mark.
I've gone back thru the archive to catch up, but I'm still
not clear on what is the purpose of gdb_byte (other than
to avoid using bfd_byte).

I thought it was to refer to target data, so we could
differentiate between host bytes and target bytes.
That would make sense.  I'm not convinced that replacing
all uses of void* makes sense.