Two questions about what seems to be a deviation in the qemu-implementation from PowerPC specification:
1. In comparison operations (eg. PPC_OP(cmp)), why does qemu not set the SO bit (which will be the fourth bit) of T0? The spec says that the xer_so bit is copied into the fourth bit of the CRF register.
2. In set_Rc0 operations (eg. PPC_OP(set_Rc0)), why is xer_ov used instead of xer_so? The spec says that a copy of xer_so is used.