[LLVMbugs] [Bug 8747] [x86] error in definition of SSE 4.1 round_ss intrinsic?

bugzilla-daemon at llvm.org bugzilla-daemon at llvm.org
Mon Dec 6 14:36:46 PST 2010


Eric Christopher <echristo at gmail.com> changed:

           What    |Removed                     |Added
             Status|ASSIGNED                    |RESOLVED
         Resolution|                            |INVALID

--- Comment #8 from Eric Christopher <echristo at gmail.com> 2010-12-06 16:36:45 CST ---
Pretty sure this is correct. The llvm intrinsic is designed to be similar to
the _mm_round_ss api intrinsic which is:

__m128 _mm_round_ss (__m128 a, __m128 b, const int c)

>From some documentation:

A 128-bit parameter. The lowest 32 bits are the result of the rounding function
on b0. The higher order 96 bits are copied directly from input parameter a. The
return value is described by the following equations:

r0 = RND(b0)
r1 = a1
r2 = a2
r3 = a3

Yeah, it's weird. No, it really doesn't make any sense. Apparently it's
intentional though.

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.

More information about the llvm-bugs mailing list