This is the mail archive of the java-patches@gcc.gnu.org mailing list for the Java project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [patch] libffi darwin x86-32bit, fix return_sc testcase.


Andrew Haley wrote:
Richard Guenther writes:

 > We obviously disagree about what the ABI specifies.  r126480 made
 > the C/C++ frontend behavior the same as other frontends behavior
 > (for example Fortran behavior, which I explicitly checked).  As GCC
 > supports both promoting function return values or not in the
 > backends and the x86 backend explicitly does _not_ enable this
 > promotion it agrees with my reading of the ABI.

But the ABI isn't in doubt: it is required to

Functions pass all integer-valued arguments as words, expanding or
padding signed or unsigned bytes and halfwords as needed'

and this surely applies to return values as well as arguments passed to functions. It is utterly perverse to assume that return values are treated differently from arguments passed to functions. So if the back end isn't promoting return values then the back end is wrong.

I'm not in the position to judge about this, I can only read the ABI and I have to give my vote to aph, at least from what _I_ understand.
Maybe my understanding lacks some basic issues, in this case I'd like to be advised to understand it right. But until then, I suppose this patch is hold on ice, right?
This blocks my further work and I fear this issue might take some moons to be resolved/clarified.


Who has the ability to clarify this issue? Do we need to expand the to/cc list to the _right_ people? Who are the right people?

I'd like to have this clarified, even if I can't help that much in clarifying....

Thanks,
Andreas


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]