This is the mail archive of the
gdb-patches@sourceware.org
mailing list for the GDB project.
Re: [commit] Remove config/m68k/tm-monitor.h
- From: Daniel Jacobowitz <drow at false dot org>
- To: gdb-patches at sourceware dot org
- Cc: Nathan Sidwell <nathan at codesourcery dot com>
- Date: Thu, 12 Apr 2007 19:15:28 -0400
- Subject: Re: [commit] Remove config/m68k/tm-monitor.h
- References: <200704122305.l3CN5S9S032295@d12av02.megacenter.de.ibm.com>
On Fri, Apr 13, 2007 at 01:05:28AM +0200, Ulrich Weigand wrote:
> #define DECR_PC_AFTER_BREAK 0
> *does* change the default behaviour of
> set_gdbarch_decr_pc_after_break (gdbarch, 2);
> set in m68k_gdbarch_init. That default is used for the Linux
> and *BSD targets. The patch moves the call from m68k_gdbarch_init
> to m68kbsd_init_abi and m68k_linux_init_abi. Then, the #define
> becomes superfluous as well.
>
> The patch should not change the behaviour on any m68k target.
Thanks.
We have an unsubmitted m68k-like target (Innovasic's "fido" processor)
which used to override this in its own tm.h, because using JTAG to
talk to it results in the architecture-standard behavior (i.e. GDB has
to take care of DECR_PC_AFTER_BREAK itself, i.e. should be 2). Does
anyone currently use the m68k-elf targets to talk to something where
the setting of 0 is correct - some monitor, maybe?
Hmm, it occurs to me I should check our ColdFire JTAG port too. Maybe
it's a difference between the two JTAG units.
--
Daniel Jacobowitz
CodeSourcery