On 08/09/16 14:46, Jan Beulich wrote: > --- a/xen/arch/x86/x86_emulate/x86_emulate.c > +++ b/xen/arch/x86/x86_emulate/x86_emulate.c > @@ -223,7 +223,7 @@ static const opcode_desc_t twobyte_table > /* 0x70 - 0x7F */ > SrcImmByte|ModRM, SrcImmByte|ModRM, SrcImmByte|ModRM, SrcImmByte|ModRM, > ModRM, ModRM, ModRM, ImplicitOps, > - ModRM, ModRM, 0, 0, ModRM, ModRM, ModRM, ImplicitOps|ModRM, > + ModRM, ModRM, 0, 0, ModRM, ModRM, ImplicitOps|ModRM, ImplicitOps|ModRM, > /* 0x80 - 0x87 */ > DstImplicit|SrcImm, DstImplicit|SrcImm, > DstImplicit|SrcImm, DstImplicit|SrcImm, > @@ -2291,6 +2291,10 @@ x86_decode( > return X86EMUL_UNHANDLEABLE; > } > > + if ( op_bytes == 2 && > + (ctxt->opcode & X86EMUL_OPC_PFX_MASK) == X86EMUL_OPC_66(0, 0) ) > + op_bytes = 4;
What is this change for? I presume it is to undo the effect of the operand size override prefix when we have decided that the prefix actually had an alternate meaning? If so, can we have a comment to this effect? Everything else looks ok. ~Andrew _______________________________________________ Xen-devel mailing list Xen-devel@lists.xen.org https://lists.xen.org/xen-devel
