RFC: Support non-standard extension (call via casted function pointer)

[Jeff Law]

On 01/27/2016 02:40 PM, Michael Karcher wrote:
> On 27.01.2016 11:33, Richard Biener wrote:
>> On Tue, Jan 26, 2016 at 9:54 PM, Michael Karcher
>> <debian@mkarcher.dialup.fu-berlin.de> wrote:
>>> On 26.01.2016 21:47, Richard Biener wrote:
>>>
>>>> I would still prefer the more obvious approach of using the target hook transition.
>>> I intended to express in my last email: Me too, and I will prepare a
>>> patch with a target hook if there is consensus that this patch can be
>>> accepted. I do not want the current patch committed as-is, I just posted
>>> it to get the idea across and start discussion.
>> Yes, I think such patch would be accepted.
> Great, that is at least one "yes". I still see a lot of doubt on
> adjusting the m68k backend to not disturb the call of a function
> declared with a wrong prototype when casted to the right type. This
> includes Thorsten Otto, who suggests that the m68k backend is the wrong
> place to patch, and Andreas Schwab (I took this discussion part off
> gcc@gcc.gnu.org, probably a bad idea in hindsight) who points out that
> ghc is lying to the compiler (I agree with that) and thus the compiler
> does not need to be made compatible with that lie. I don't know the
> process of patch acceptence in the gcc project, but one positive vote
> against two doubtful or negative votes is not what I would call "consensus".
Given overall contribution history within the GCC project, Richi and 
myself carry more weight than Thorsten & Andreas.

jeff