Need advice on building GDB for a large number of PowerPC variants.

[Anmol P. Paralkar]

Hello,

  Forwarding Michael Snyder's affirmation below. (For some reason, his
  response bounced off the mailing list server). Someone else might
  need to benefit from this, hence the forward.

Best Regards,
Anmol.

On Thu, 17 May 2007, Michael Snyder wrote:

> Hi Anmol,
>
> You're perfectly correct -- in fact, you might be even luckier in that
> in most respects, a native (ppc-linux) gdb should be able to support
> "target remote".  I can't immediately think of anything that should
> prevent debugging both remote and native programs with the same gdb.
>
> Hope you're well,
> Michael
>
>
> -----Original Message-----
> From: gdb-owner@sourceware.org on behalf of Anmol P. Paralkar
> Sent: Thu 5/17/2007 2:43 PM
> To: gdb@sourceware.org
> Subject: Need advice on building GDB for a large number of PowerPC variants.
>
> Hello,
>
>  I need to build GDB for a debugging a variety of PowerPC families.
>  The host environment is x86 plus a small subset of the PowerPC
>  variants. GDB will be used native as well as cross (remote) from
>  each host (x86 (cross-only) as well as PowerPC). A substantial
>  number of PowerPC flavours is desired to be targetted - the number
>  of families and the instances within each family all result in a
>  fairly large number of individual targets. Both the variations in
>  configurations: powerpc-linux-gnu as well as powerpc-eabi are needed
>  for almost each individual target.
>
>  I reckon that I only need build:
>
>  * GDB in both eabi and linux configurations per host; any desired
>    target can be debugged by doing a 'set architecture' for it.
>
>    That way the number of GDB's built is two per host.
>
>  as against:
>
>  * A GDB per (host, target) combination of interest.
>
>    Which will needless to say, will result in a large number of
>    GDB executables needed to be built.
>
>  Please could you tell me if my reasoning is correct or am I missing
>  something essential?
>
>  Thank you very much for your help.
>
> Best Regards,
> Anmol P. Paralkar
>
>