This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH RFC 3/3] Hookize PROMOTE_FUNCTION_MODE with provision for HFmode
Eric Botcazou wrote:
>> I'd commit this to make it easier to grasp, do you agree?
> That's partly explained for POINTERS_EXTEND_UNSIGNED just below so I'd just
> mention ptr_mode and move Pmode in-between:
BTW, this means that my original patch, while conceptually not the best,
would not have caused code generation differences. It would have caused
*punsignedp not to be set when a pointer type and Pmode were passed to
sparc_promote_function_mode; but *punsignedp is never used by any caller
if there is no promotion to do.