This is the mail archive of the gcc@gcc.gnu.org mailing list for the GCC 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: "return" expander


>>>>> "Momchil" == Momchil Velikov <velco@fadata.bg> writes:

>>>>> "Momchil" == Momchil Velikov <velco@fadata.bg> writes:
Momchil> The return insn on v850 is ``jmp [r31]''.  I have a case with a
Momchil> function where r31 is allocated and the jump at the end of the
Momchil> function jumps to nowhere.  I guess the ``return'' pattern should make
Momchil> ``r31'' live at the end so it is either saved/restored or not
Momchil> allocated. I've tried with

Momchil> What is the proper way to describe insn like these ?

Momchil> Looks like ``(define_expand "return" [(set (pc) (reg:SI 31))] ...)''
Momchil> did the trick ...

Nope, it didn't. I guess someone have to set ``SET_IS_RETURN_P (x) = 1''
somewhere. But where ?

~velco


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