[CPU] Fix f16 pack rounding and SHORT_2 test input

Add round-to-nearest-even to the fast float16 pack path by folding a
0xFFF rounding bias into XMMF16PackLCPI0 and extracting bit 13 of the
source as the tie-breaker, matching the software fallback behavior.

Fix PACK_SHORT_2 test to use pre-biased float input (0x40400000 = 3.0)
as the hardware expects, rather than raw 0.0f which is out of range.

src/xenia/cpu/backend/x64/x64_emitter.cc

@@ -1152,8 +1152,8 @@ static const vec128_t xmm_consts[] = {
         */
     vec128q(0x7fe000007fe000ULL),
 
-    /* XMMF16PackLCPI0*/
-    vec128i(0x8000000),
+    /* XMMF16PackLCPI0 (bias + round-half-down) */
+    vec128i(0x8000FFF),
     /*XMMF16PackLCPI2*/
     vec128i(0x47ffe000),
     /*XMMF16PackLCPI3*/