This is the mail archive of the
gdb-patches@sources.redhat.com
mailing list for the GDB project.
Re: [Patch 5/7]: 68HC11 port of gdb (sim-overlap address maps)
- To: Stephane Carrez <Stephane dot Carrez at free dot fr>
- Subject: Re: [Patch 5/7]: 68HC11 port of gdb (sim-overlap address maps)
- From: Andrew Cagney <ac131313 at cygnus dot com>
- Date: Thu, 27 Jul 2000 21:43:15 +1000
- CC: gdb-patches at sourceware dot cygnus dot com
- References: <39568A64.830EC800@free.fr>
Stephane Carrez wrote:
>
> Hi!
>
> The following patch adds the support to overlap some address spaces.
> I need this for 68hc11 because the IO is mapped over the normal memory.
> The internal RAM is mapped in the same way. Both may be moved at any
> 4K boundary. The EEPROM can be unmapped, showing up what is underneath it.
>
> Each device has an overlap flag that tells whether overlapping is
> allowed or not. The default is to disable that. This is the current
> behavior.
>
> When overlapping is allowed, we just record the mapping rather than
> rejecting it.
I'm not sure I understand the change. The memory model has the idea of
layers. Overlapping memory can normally be handled by that. The memory
would be set up as:
first layer: eeprom
second layer: memory
is this in addition to the layers?
Andrew
> The ChangeLog is for gdb/sim/common
>
> 2000-06-25 Stephane Carrez <Stephane.Carrez@worldnet.fr>
>
> * sim-core.c (sim_core_map_attach): Don't raise an error if
> the device allows overlapping of address maps.
> * hw-device.h (struct hw): Added the boolean 'overlap_mode_hw'
> to enable/disable overlapping of address maps. Overlapping is
> necessary for 68Hc11. The default is to disable overlapping
> (previous behavior).
> * dv-core.c (dv_core_attach_address_callback): Don't abort if
> the address space refers to the IO memory.
>