[PATCH 0/3] Split off powerpcspe from rs6000 port

[Chris Johns]

On 17/05/2017 00:03, Joel Sherrill wrote:
> 
> 
> On Tue, May 16, 2017 at 8:48 AM, Sebastian Huber 
> <sebastian.huber at embedded-brains.de 
> <mailto:sebastian.huber at embedded-brains.de>> wrote:
> 
>     On 16/05/17 15:45, Joel Sherrill wrote:
> 
> 
> 
>         On Tue, May 16, 2017 at 8:28 AM, Sebastian Huber
>         <sebastian.huber at embedded-brains.de
>         <mailto:sebastian.huber at embedded-brains.de>
>         <mailto:sebastian.huber at embedded-brains.de
>         <mailto:sebastian.huber at embedded-brains.de>>> wrote:
> 
>              FYI
> 
>              I guess for RTEMS we should use "powerpcspe-rtems*-*". Is it
>              possible to use the RTEMS "powerpc" directories with such a
>              target? We had an "arm-rtemseabi*" maybe due to some
>              configure/automake limitations. So, maybe "powerpc-rtemsspe*"?
> 
> 
>         I personally think the GCC discussions which put SPE as part of
>         the OS name are
>         horribly incorrect. SPE is an architecture variant and the
>         pattern for configure
>         triples is very well defined. The pattern is
>         ARCHITECTURE-VENDOR-OS[version]
> 
>         I would prefer powerpcspe-rtemsVERSION.
> 
>         On the sharing the code issue, how much gets shared? How do you
>         envision
>         this impacting the RTEMS tree? Split the PowerPC port like GCC?
>         Or just
>         somehow magically build the same powerpc directories two
>         different ways?
> 
> 
>     The reason for this split in GCC is that IBM is no longer willing to
>     maintain the stuff from Freescale/NXP/Qualcomm. There is no reason
>     to do this split in RTEMS.
> 
> 
> I didn't either except for the purity of matching directories to tool 
> architectures.
> 
> So it will be configure magic to subset the BSPs I suppose.
> 

Does this mean we configure RTEMS with:

   configure --target=powerpcspe-rtems4.12

?

It would seem to create a new architecture and I suspect this will have 
a follow on effects with rtems_waf and the rtems-tools.

Chris