Re: Another patch to aix-thread.c

[Kevin Buettner <kevinb at redhat dot com>]

On Jul 22,  3:43pm, Andrew Cagney wrote:

> > +  regcache_collect (regno,     iar);
> > +  regcache_collect (regno + 1, msr);
> 
> Suggest ``regno + 0''.  Otherwize gdb_indent.sh will mess up your nice 
> alignment.

I agree that this is a good idea.  However, I'm about to make some
changes which will obviate the use of regno + N.  I.e, I'm about to
transform the above into the following:

	struct gdbarch_tdep *tdep = gdbarch_tdep (current_gdbarch);
	...
	regcache_collect (PC_REGNUM, iar);
	regcache_collect (tdep->ppc_ps_regnum, msr);
	...

This will also eliminate the use of FIRST_UISA_SP_REGNUM from the
file.  While I'm at it, I'll also look at eliminating this macro
from rs6000-nat.c which means that it can be eliminated from GDB
entirely.

(It bothered me that aix-thread.c was making assumptions about the
order of the registers.  Also, it was more difficult to verify the
correctness by having expressions of the form "regno + small_offset".
I was having to look elsewhere to verify that the specified offset
was correct.)

Kevin