[LLVMbugs] [Bug 4833] Information about the sign is lost for some NEON intrinsics

[bugzilla-daemon at cs.uiuc.edu]

http://llvm.org/bugs/show_bug.cgi?id=4833


Bob Wilson <bob.wilson at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |INVALID




--- Comment #1 from Bob Wilson <bob.wilson at apple.com>  2009-08-31 13:17:47 ---
This code (from the gcc testsuite) is a bad test.  Only the low byte of the
result is relevant (and not even that since the value is dead), so it doesn't
matter whether llvm selects vmov.s8 or vmov.u8 in this case.  I tried changing
the code to return the value as a signed int, and we correctly generate
vmov.s8.  I don't think there is a bug here.


-- 
Configure bugmail: http://llvm.org/bugs/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.