[PATCH] x86: correct aad's operand type

["Jan Beulich" <JBeulich at novell dot com>]

aad performs an unsigned division, not a signed one, hence its operand
must
be an unsigned immediate. For consistency, also changing this for aam.

Built and tested on i686-pc-linux-gnu.

Jan

include/opcode/
2005-07-05  Jan Beulich  <jbeulich@novell.com>

	* i386.h (i386_optab): Operands of aam and aad are unsigned.

---
/home/jbeulich/src/binutils/mainline/2005-07-05/include/opcode/i386.h	2005-07-05
09:14:29.000000000 +0200
+++ 2005-07-05/include/opcode/i386.h	2005-07-05 14:29:14.686760872
+0200
@@ -295,9 +295,9 @@ static const template i386_optab[] =
 {"daa",	   0,	0x27, X,
CpuNo64,	NoSuf,			{ 0, 0, 0} },
 {"das",	   0,	0x2f, X,
CpuNo64,	NoSuf,			{ 0, 0, 0} },
 {"aad",	   0, 0xd50a, X,
CpuNo64,	NoSuf,			{ 0, 0, 0} },
-{"aad",	   1,   0xd5, X,
CpuNo64,	NoSuf,			{ Imm8S, 0, 0} },
+{"aad",	   1,   0xd5, X,
CpuNo64,	NoSuf,			{ Imm8, 0, 0} },
 {"aam",	   0, 0xd40a, X,
CpuNo64,	NoSuf,			{ 0, 0, 0} },
-{"aam",	   1,   0xd4, X,
CpuNo64,	NoSuf,			{ Imm8S, 0, 0} },
+{"aam",	   1,   0xd4, X,
CpuNo64,	NoSuf,			{ Imm8, 0, 0} },
 
 /* Conversion insns.  */
 /* Intel naming */

Attachment: binutils-mainline-x86-aad.patch
Description: Text document