Strategy for packaging an ARM Cortex-M toolchain

[Ralf Corsepius]

On 05/24/2012 03:21 PM, Rob Spanton wrote:
> Hi Ralf,
>
> I wrote:
>> So is it best to attempt to get one arm-binutils package and remove
>> redundancy, or is it going to be more productive to just put up with
>> the redundancy for now?
>
> Ralf wrote:
>> No, this will hardly work and would be a nightmare to maintain.
>
> I had guessed that binutils didn't care what the ABI was.
True, AFAICS.

However, consider that
- binutils is only a comparativly small part of a target's toolchain.
- a target's binutils may require target-specific patching.
- there can be implicit couplings/incompatibilities between a target's 
binutils and other components of a cross-toolchain.

> Maybe I'm
> wrong about that, or is there something else that I'm missing that'd
> prevent this from working?

You are not necessarily wrong. I agree, using a "combined binutils" is 
possibile in many cases (But not always).

  It's just that, when taking into account that using a "combo" is close 
to impossible for other components of a cross-toolchain (esp. GCC), 
trying use "combined binutils" is simplier and "not worth the effort", IMO.

Ralf