This is the mail archive of the
gdb-patches@sourceware.cygnus.com
mailing list for the GDB project.
Re: RFC: Patch to arm-tdep.c -- Register flavors
- To: Scott Bambrough <scottb at netwinder dot org>
- Subject: Re: RFC: Patch to arm-tdep.c -- Register flavors
- From: Richard Earnshaw <rearnsha at arm dot com>
- Date: Wed, 16 Feb 2000 15:41:03 +0000
- Cc: Nick Clifton <nickc at cygnus dot com>, "fnasser at cygnus dot com" <fnasser at cygnus dot com>, GDB Patches Mail List <gdb-patches at sourceware dot cygnus dot com>
- Cc: rearnsha at arm dot com
- Organization: ARM Ltd.
- Reply-To: rearnsha at arm dot com
> > And finally, the ps register name is wrong, especially in the A{T}PCS area
> > but also more generally; the name is "cpsr" in all ARM documentation (the
> > "ps" name comes from when the processor flags weren't really a separate
> > register -- they were in the unused parts of r15 -- so there was no name).
>
> Richard is correct here, in arm-linux-nat.c, I initialize the PS register with
> the contents of CPSR if arm_apcs_32 is set, or with the contents of the PC if
> not. Perhaps we should set the register name based on this information as well.
This is one of the few cases where I would just delete the old name. Old
processors never had a specific flags register, so the 'ps' name was
'invented' by the original porter of gdb. Since then the cpsr name has
come into use and I doubt that even users debugging in 26-bit mode would
object violently to this name changing (maybe we can support "ps" as an
alias when parsing user input, but I see no point in ever printing
anything other than "cpsr".
R.