https://gcc.gnu.org/bugzilla/show_bug.cgi?id=121470
Segher Boessenkool <segher at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Status|RESOLVED |REOPENED
Resolution|DUPLICATE |---
--- Comment #5 from Segher Boessenkool <segher at gcc dot gnu.org> ---
(In reply to Andrew Pinski from comment #4)
> yes it is a dup.
> Did you read comment #1:
Yes, I did read. Please insult yourself. Thanks.
> It believe that the problem is in expand_asm_stmt ().
>
>
> The real issue is on rtl const_int are VOIDmode so always sign extended from
> the precision of the type.
>
> *** This bug has been marked as a duplicate of bug 85344 ***
32768 with "the usual promotions" is 32768, not -32768.
This is NOT A DUP (not of 85344, anyway).
It isn't RTL yet before expand starts, of course, and thins are perfectly
fine there. Somehow a value 32768 (not "unsigned short" there btw, just an
integer constant, earlier passes have eradicated the cast already!) is expanded
as -32768 here.
