rsb problem with python

[Chris Johns]

On 31/5/18 7:32 am, Slide wrote:
> 
> That's something I would like to know as well. When something fails after
> building GCC, it would be nice to not have to rebuild GCC the next time. I don't
> know enough about the RTEMS build process at this point to be able to help you
> on that part.
> 

This topic appears from time to time, mostly around a period of change in the
RSB or a host that requires fixes. This post is a previous response:

https://lists.rtems.org/pipermail/users/2015-March/028656.html

The key issue is the integrity of the build if stages can be skipped. I would
like to be able to add support for managing library dependencies where possible
but it would need funding. It is complicated to do because it links internal
functionality of a package to the external RSB wrapping of that package. For
example consider GDB and Python which is a constant source of frustration, if
the RSB is taught to find Python 2.7 dev libraries and headers and GDB changes
to support Python 2.7 and Python 3 we need to know this, know when the change
happens and manage the relationship. If we do not we create a new set of
problems for users.

Improving the documentation would help and for that I ask users to consider
helping out because my perspective is often not helpful as it assumes too much.

Chris