Enable X86 maskAll instruction pattern for 32 bit JVM.

--N: o284 XorVMask === _ o310 o282 [[o311 ]] #vectormask[16]:{byte}

   --N: o310 VectorLoadMask === _ o309 [[o284 ]] #vectormask[16]:{byte}
   KREG 0 KREG
   KREG_K1 0 KREG_K1
   KREG_K2 0 KREG_K2
   KREG_K3 0 KREG_K3
   KREG_K4 0 KREG_K4
   KREG_K5 0 KREG_K5
   KREG_K6 0 KREG_K6
   KREG_K7 0 KREG_K7

   --N: o282 VectorMaskCast === _ o261 [[o284 ]] #vectormask[16]:{byte}
   VEC 0 vmaskcast
   LEGVEC 100 MoveVec2Leg

      --N: o261 ReplicateB === _ o260 [[o312 o282 ]] #vectorx[16]:{byte}
      RXMM0 0 RXMM0
      VEC 0 VEC
      LEGVEC 0 LEGVEC
      VECX 0 VECX
      LEGVECX 0 LEGVECX


X86 backend implementation does not enable maskAll instruction patterns for 32 bit JVM, due to which operations fall backs over replicateB operation which broadcasts the mask value in a vector and after unboxing-boxing optimization this vector eventually reaches to XorVMask which cribs since one operand is held in opmask register and other in vector.