mirror of
https://github.com/autc04/Retro68.git
synced 2024-11-18 11:08:26 +00:00
777 lines
46 KiB
C
777 lines
46 KiB
C
/* s390-opc.c -- S390 opcode list
|
|
Copyright (C) 2000-2017 Free Software Foundation, Inc.
|
|
Contributed by Martin Schwidefsky (schwidefsky@de.ibm.com).
|
|
|
|
This file is part of the GNU opcodes library.
|
|
|
|
This library is free software; you can redistribute it and/or modify
|
|
it under the terms of the GNU General Public License as published by
|
|
the Free Software Foundation; either version 3, or (at your option)
|
|
any later version.
|
|
|
|
It is distributed in the hope that it will be useful, but WITHOUT
|
|
ANY WARRANTY; without even the implied warranty of MERCHANTABILITY
|
|
or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public
|
|
License for more details.
|
|
|
|
You should have received a copy of the GNU General Public License
|
|
along with this file; see the file COPYING. If not, write to the
|
|
Free Software Foundation, 51 Franklin Street - Fifth Floor, Boston,
|
|
MA 02110-1301, USA. */
|
|
|
|
#include <stdio.h>
|
|
#include "ansidecl.h"
|
|
#include "opcode/s390.h"
|
|
|
|
/* This file holds the S390 opcode table. The opcode table
|
|
includes almost all of the extended instruction mnemonics. This
|
|
permits the disassembler to use them, and simplifies the assembler
|
|
logic, at the cost of increasing the table size. The table is
|
|
strictly constant data, so the compiler should be able to put it in
|
|
the .text section.
|
|
|
|
This file also holds the operand table. All knowledge about
|
|
inserting operands into instructions and vice-versa is kept in this
|
|
file. */
|
|
|
|
/* The operands table.
|
|
The fields are bits, shift, insert, extract, flags. */
|
|
|
|
const struct s390_operand s390_operands[] =
|
|
{
|
|
#define UNUSED 0
|
|
{ 0, 0, 0 }, /* Indicates the end of the operand list */
|
|
|
|
/* General purpose register operands. */
|
|
|
|
#define R_8 1 /* GPR starting at position 8 */
|
|
{ 4, 8, S390_OPERAND_GPR },
|
|
#define R_12 2 /* GPR starting at position 12 */
|
|
{ 4, 12, S390_OPERAND_GPR },
|
|
#define R_16 3 /* GPR starting at position 16 */
|
|
{ 4, 16, S390_OPERAND_GPR },
|
|
#define R_20 4 /* GPR starting at position 20 */
|
|
{ 4, 20, S390_OPERAND_GPR },
|
|
#define R_24 5 /* GPR starting at position 24 */
|
|
{ 4, 24, S390_OPERAND_GPR },
|
|
#define R_28 6 /* GPR starting at position 28 */
|
|
{ 4, 28, S390_OPERAND_GPR },
|
|
#define R_32 7 /* GPR starting at position 32 */
|
|
{ 4, 32, S390_OPERAND_GPR },
|
|
|
|
/* General purpose register pair operands. */
|
|
|
|
#define RE_8 8 /* GPR starting at position 8 */
|
|
{ 4, 8, S390_OPERAND_GPR | S390_OPERAND_REG_PAIR },
|
|
#define RE_12 9 /* GPR starting at position 12 */
|
|
{ 4, 12, S390_OPERAND_GPR | S390_OPERAND_REG_PAIR },
|
|
#define RE_16 10 /* GPR starting at position 16 */
|
|
{ 4, 16, S390_OPERAND_GPR | S390_OPERAND_REG_PAIR },
|
|
#define RE_20 11 /* GPR starting at position 20 */
|
|
{ 4, 20, S390_OPERAND_GPR | S390_OPERAND_REG_PAIR },
|
|
#define RE_24 12 /* GPR starting at position 24 */
|
|
{ 4, 24, S390_OPERAND_GPR | S390_OPERAND_REG_PAIR },
|
|
#define RE_28 13 /* GPR starting at position 28 */
|
|
{ 4, 28, S390_OPERAND_GPR | S390_OPERAND_REG_PAIR },
|
|
#define RE_32 14 /* GPR starting at position 32 */
|
|
{ 4, 32, S390_OPERAND_GPR | S390_OPERAND_REG_PAIR },
|
|
|
|
/* Floating point register operands. */
|
|
|
|
#define F_8 15 /* FPR starting at position 8 */
|
|
{ 4, 8, S390_OPERAND_FPR },
|
|
#define F_12 16 /* FPR starting at position 12 */
|
|
{ 4, 12, S390_OPERAND_FPR },
|
|
#define F_16 17 /* FPR starting at position 16 */
|
|
{ 4, 16, S390_OPERAND_FPR },
|
|
#define F_24 18 /* FPR starting at position 24 */
|
|
{ 4, 24, S390_OPERAND_FPR },
|
|
#define F_28 19 /* FPR starting at position 28 */
|
|
{ 4, 28, S390_OPERAND_FPR },
|
|
#define F_32 20 /* FPR starting at position 32 */
|
|
{ 4, 32, S390_OPERAND_FPR },
|
|
|
|
/* Floating point register pair operands. */
|
|
|
|
#define FE_8 21 /* FPR starting at position 8 */
|
|
{ 4, 8, S390_OPERAND_FPR | S390_OPERAND_REG_PAIR },
|
|
#define FE_12 22 /* FPR starting at position 12 */
|
|
{ 4, 12, S390_OPERAND_FPR | S390_OPERAND_REG_PAIR },
|
|
#define FE_16 23 /* FPR starting at position 16 */
|
|
{ 4, 16, S390_OPERAND_FPR | S390_OPERAND_REG_PAIR },
|
|
#define FE_24 24 /* FPR starting at position 24 */
|
|
{ 4, 24, S390_OPERAND_FPR | S390_OPERAND_REG_PAIR },
|
|
#define FE_28 25 /* FPR starting at position 28 */
|
|
{ 4, 28, S390_OPERAND_FPR | S390_OPERAND_REG_PAIR },
|
|
#define FE_32 26 /* FPR starting at position 32 */
|
|
{ 4, 32, S390_OPERAND_FPR | S390_OPERAND_REG_PAIR },
|
|
|
|
/* Vector register operands. */
|
|
|
|
/* For each of these operands and additional bit in the RXB operand is
|
|
needed. */
|
|
|
|
#define V_8 27 /* Vector reg. starting at position 8 */
|
|
{ 4, 8, S390_OPERAND_VR },
|
|
#define V_12 28 /* Vector reg. starting at position 12 */
|
|
{ 4, 12, S390_OPERAND_VR },
|
|
#define V_CP16_12 29 /* Vector reg. starting at position 12 */
|
|
{ 4, 12, S390_OPERAND_VR | S390_OPERAND_CP16 }, /* with a copy at pos 16 */
|
|
#define V_16 30 /* Vector reg. starting at position 16 */
|
|
{ 4, 16, S390_OPERAND_VR },
|
|
#define V_32 31 /* Vector reg. starting at position 32 */
|
|
{ 4, 32, S390_OPERAND_VR },
|
|
|
|
/* Access register operands. */
|
|
|
|
#define A_8 32 /* Access reg. starting at position 8 */
|
|
{ 4, 8, S390_OPERAND_AR },
|
|
#define A_12 33 /* Access reg. starting at position 12 */
|
|
{ 4, 12, S390_OPERAND_AR },
|
|
#define A_24 34 /* Access reg. starting at position 24 */
|
|
{ 4, 24, S390_OPERAND_AR },
|
|
#define A_28 35 /* Access reg. starting at position 28 */
|
|
{ 4, 28, S390_OPERAND_AR },
|
|
|
|
/* Control register operands. */
|
|
|
|
#define C_8 36 /* Control reg. starting at position 8 */
|
|
{ 4, 8, S390_OPERAND_CR },
|
|
#define C_12 37 /* Control reg. starting at position 12 */
|
|
{ 4, 12, S390_OPERAND_CR },
|
|
|
|
/* Base register operands. */
|
|
|
|
#define B_16 38 /* Base register starting at position 16 */
|
|
{ 4, 16, S390_OPERAND_BASE | S390_OPERAND_GPR },
|
|
#define B_32 39 /* Base register starting at position 32 */
|
|
{ 4, 32, S390_OPERAND_BASE | S390_OPERAND_GPR },
|
|
|
|
#define X_12 40 /* Index register starting at position 12 */
|
|
{ 4, 12, S390_OPERAND_INDEX | S390_OPERAND_GPR },
|
|
|
|
#define VX_12 41 /* Vector index register starting at position 12 */
|
|
{ 4, 12, S390_OPERAND_INDEX | S390_OPERAND_VR },
|
|
|
|
/* Address displacement operands. */
|
|
|
|
#define D_20 42 /* Displacement starting at position 20 */
|
|
{ 12, 20, S390_OPERAND_DISP },
|
|
#define D_36 43 /* Displacement starting at position 36 */
|
|
{ 12, 36, S390_OPERAND_DISP },
|
|
#define D20_20 44 /* 20 bit displacement starting at 20 */
|
|
{ 20, 20, S390_OPERAND_DISP | S390_OPERAND_SIGNED },
|
|
|
|
/* Length operands. */
|
|
|
|
#define L4_8 45 /* 4 bit length starting at position 8 */
|
|
{ 4, 8, S390_OPERAND_LENGTH },
|
|
#define L4_12 46 /* 4 bit length starting at position 12 */
|
|
{ 4, 12, S390_OPERAND_LENGTH },
|
|
#define L8_8 47 /* 8 bit length starting at position 8 */
|
|
{ 8, 8, S390_OPERAND_LENGTH },
|
|
|
|
/* Signed immediate operands. */
|
|
|
|
#define I8_8 48 /* 8 bit signed value starting at 8 */
|
|
{ 8, 8, S390_OPERAND_SIGNED },
|
|
#define I8_32 49 /* 8 bit signed value starting at 32 */
|
|
{ 8, 32, S390_OPERAND_SIGNED },
|
|
#define I12_12 50 /* 12 bit signed value starting at 12 */
|
|
{ 12, 12, S390_OPERAND_SIGNED },
|
|
#define I16_16 51 /* 16 bit signed value starting at 16 */
|
|
{ 16, 16, S390_OPERAND_SIGNED },
|
|
#define I16_32 52 /* 16 bit signed value starting at 32 */
|
|
{ 16, 32, S390_OPERAND_SIGNED },
|
|
#define I24_24 53 /* 24 bit signed value starting at 24 */
|
|
{ 24, 24, S390_OPERAND_SIGNED },
|
|
#define I32_16 54 /* 32 bit signed value starting at 16 */
|
|
{ 32, 16, S390_OPERAND_SIGNED },
|
|
|
|
/* Unsigned immediate operands. */
|
|
|
|
#define U4_8 55 /* 4 bit unsigned value starting at 8 */
|
|
{ 4, 8, 0 },
|
|
#define U4_12 56 /* 4 bit unsigned value starting at 12 */
|
|
{ 4, 12, 0 },
|
|
#define U4_16 57 /* 4 bit unsigned value starting at 16 */
|
|
{ 4, 16, 0 },
|
|
#define U4_20 58 /* 4 bit unsigned value starting at 20 */
|
|
{ 4, 20, 0 },
|
|
#define U4_24 59 /* 4 bit unsigned value starting at 24 */
|
|
{ 4, 24, 0 },
|
|
#define U4_OR1_24 60 /* 4 bit unsigned value ORed with 1 */
|
|
{ 4, 24, S390_OPERAND_OR1 }, /* starting at 24 */
|
|
#define U4_OR2_24 61 /* 4 bit unsigned value ORed with 2 */
|
|
{ 4, 24, S390_OPERAND_OR2 }, /* starting at 24 */
|
|
#define U4_OR3_24 62 /* 4 bit unsigned value ORed with 3 */
|
|
{ 4, 24, S390_OPERAND_OR1 | S390_OPERAND_OR2 }, /* starting at 24 */
|
|
#define U4_28 63 /* 4 bit unsigned value starting at 28 */
|
|
{ 4, 28, 0 },
|
|
#define U4_OR8_28 64 /* 4 bit unsigned value ORed with 8 */
|
|
{ 4, 28, S390_OPERAND_OR8 }, /* starting at 28 */
|
|
#define U4_32 65 /* 4 bit unsigned value starting at 32 */
|
|
{ 4, 32, 0 },
|
|
#define U4_36 66 /* 4 bit unsigned value starting at 36 */
|
|
{ 4, 36, 0 },
|
|
#define U8_8 67 /* 8 bit unsigned value starting at 8 */
|
|
{ 8, 8, 0 },
|
|
#define U8_16 68 /* 8 bit unsigned value starting at 16 */
|
|
{ 8, 16, 0 },
|
|
#define U8_24 69 /* 8 bit unsigned value starting at 24 */
|
|
{ 8, 24, 0 },
|
|
#define U8_28 70 /* 8 bit unsigned value starting at 28 */
|
|
{ 8, 28, 0 },
|
|
#define U8_32 71 /* 8 bit unsigned value starting at 32 */
|
|
{ 8, 32, 0 },
|
|
#define U12_16 72 /* 12 bit unsigned value starting at 16 */
|
|
{ 12, 16, 0 },
|
|
#define U16_16 73 /* 16 bit unsigned value starting at 16 */
|
|
{ 16, 16, 0 },
|
|
#define U16_32 74 /* 16 bit unsigned value starting at 32 */
|
|
{ 16, 32, 0 },
|
|
#define U32_16 75 /* 32 bit unsigned value starting at 16 */
|
|
{ 32, 16, 0 },
|
|
|
|
/* PC-relative address operands. */
|
|
|
|
#define J12_12 76 /* 12 bit PC relative offset at 12 */
|
|
{ 12, 12, S390_OPERAND_PCREL },
|
|
#define J16_16 77 /* 16 bit PC relative offset at 16 */
|
|
{ 16, 16, S390_OPERAND_PCREL },
|
|
#define J16_32 78 /* 16 bit PC relative offset at 32 */
|
|
{ 16, 32, S390_OPERAND_PCREL },
|
|
#define J24_24 79 /* 24 bit PC relative offset at 24 */
|
|
{ 24, 24, S390_OPERAND_PCREL },
|
|
#define J32_16 80 /* 32 bit PC relative offset at 16 */
|
|
{ 32, 16, S390_OPERAND_PCREL },
|
|
|
|
};
|
|
|
|
|
|
/* Macros used to form opcodes. */
|
|
|
|
/* 8/16/48 bit opcodes. */
|
|
#define OP8(x) { x, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define OP16(x) { x >> 8, x & 255, 0x00, 0x00, 0x00, 0x00 }
|
|
#define OP32(x) { x >> 24, (x >> 16) & 255, (x >> 8) & 255, x & 255, \
|
|
0x00, 0x00 }
|
|
#define OP48(x) { x >> 40, (x >> 32) & 255, (x >> 24) & 255, \
|
|
(x >> 16) & 255, (x >> 8) & 255, x & 255}
|
|
|
|
/* The new format of the INSTR_x_y and MASK_x_y defines is based
|
|
on the following rules:
|
|
1) the middle part of the definition (x in INSTR_x_y) is the official
|
|
names of the instruction format that you can find in the principals
|
|
of operation.
|
|
2) the last part of the definition (y in INSTR_x_y) gives you an idea
|
|
which operands the binary represenation of the instruction has.
|
|
The meanings of the letters in y are:
|
|
a - access register
|
|
c - control register
|
|
d - displacement, 12 bit
|
|
f - floating pointer register
|
|
fe - fpr extended operand, a valid floating pointer register pair
|
|
i - signed integer, 4, 8, 16 or 32 bit
|
|
l - length, 4 or 8 bit
|
|
p - pc relative
|
|
r - general purpose register
|
|
re - gpr extended operand, a valid general purpose register pair
|
|
u - unsigned integer, 4, 8, 16 or 32 bit
|
|
m - mode field, 4 bit
|
|
0 - operand skipped.
|
|
The order of the letters reflects the layout of the format in
|
|
storage and not the order of the paramaters of the instructions.
|
|
The use of the letters is not a 100% match with the PoP but it is
|
|
quite close.
|
|
|
|
For example the instruction "mvo" is defined in the PoP as follows:
|
|
|
|
MVO D1(L1,B1),D2(L2,B2) [SS]
|
|
|
|
--------------------------------------
|
|
| 'F1' | L1 | L2 | B1 | D1 | B2 | D2 |
|
|
--------------------------------------
|
|
0 8 12 16 20 32 36
|
|
|
|
The instruction format is: INSTR_SS_LLRDRD / MASK_SS_LLRDRD. */
|
|
|
|
#define INSTR_E 2, { 0,0,0,0,0,0 } /* e.g. pr */
|
|
#define INSTR_IE_UU 4, { U4_24,U4_28,0,0,0,0 } /* e.g. niai */
|
|
#define INSTR_MII_UPP 6, { U4_8,J12_12,J24_24 } /* e.g. bprp */
|
|
#define INSTR_RIE_RRP 6, { R_8,R_12,J16_16,0,0,0 } /* e.g. brxhg */
|
|
#define INSTR_RIE_RRPU 6, { R_8,R_12,U4_32,J16_16,0,0 } /* e.g. crj */
|
|
#define INSTR_RIE_RRP0 6, { R_8,R_12,J16_16,0,0,0 } /* e.g. cgrjne */
|
|
#define INSTR_RIE_RRI0 6, { R_8,R_12,I16_16,0,0,0 } /* e.g. ahik */
|
|
#define INSTR_RIE_RUPI 6, { R_8,I8_32,U4_12,J16_16,0,0 } /* e.g. cij */
|
|
#define INSTR_RIE_R0PI 6, { R_8,I8_32,J16_16,0,0,0 } /* e.g. cijne */
|
|
#define INSTR_RIE_RUPU 6, { R_8,U8_32,U4_12,J16_16,0,0 } /* e.g. clij */
|
|
#define INSTR_RIE_R0PU 6, { R_8,U8_32,J16_16,0,0,0 } /* e.g. clijne */
|
|
#define INSTR_RIE_R0IU 6, { R_8,I16_16,U4_32,0,0,0 } /* e.g. cit */
|
|
#define INSTR_RIE_R0I0 6, { R_8,I16_16,0,0,0,0 } /* e.g. citne */
|
|
#define INSTR_RIE_R0UU 6, { R_8,U16_16,U4_32,0,0,0 } /* e.g. clfit */
|
|
#define INSTR_RIE_R0U0 6, { R_8,U16_16,0,0,0,0 } /* e.g. clfitne */
|
|
#define INSTR_RIE_RUI0 6, { R_8,I16_16,U4_12,0,0,0 } /* e.g. lochi */
|
|
#define INSTR_RIE_RRUUU 6, { R_8,R_12,U8_16,U8_24,U8_32,0 } /* e.g. rnsbg */
|
|
#define INSTR_RIL_0P 6, { J32_16,0,0,0,0 } /* e.g. jg */
|
|
#define INSTR_RIL_RP 6, { R_8,J32_16,0,0,0,0 } /* e.g. brasl */
|
|
#define INSTR_RIL_UP 6, { U4_8,J32_16,0,0,0,0 } /* e.g. brcl */
|
|
#define INSTR_RIL_RI 6, { R_8,I32_16,0,0,0,0 } /* e.g. afi */
|
|
#define INSTR_RIL_RU 6, { R_8,U32_16,0,0,0,0 } /* e.g. alfi */
|
|
#define INSTR_RI_0P 4, { J16_16,0,0,0,0,0 } /* e.g. j */
|
|
#define INSTR_RI_RI 4, { R_8,I16_16,0,0,0,0 } /* e.g. ahi */
|
|
#define INSTR_RI_RP 4, { R_8,J16_16,0,0,0,0 } /* e.g. brct */
|
|
#define INSTR_RI_RU 4, { R_8,U16_16,0,0,0,0 } /* e.g. tml */
|
|
#define INSTR_RI_UP 4, { U4_8,J16_16,0,0,0,0 } /* e.g. brc */
|
|
#define INSTR_RIS_RURDI 6, { R_8,I8_32,U4_12,D_20,B_16,0 } /* e.g. cib */
|
|
#define INSTR_RIS_R0RDI 6, { R_8,I8_32,D_20,B_16,0,0 } /* e.g. cibne */
|
|
#define INSTR_RIS_RURDU 6, { R_8,U8_32,U4_12,D_20,B_16,0 } /* e.g. clib */
|
|
#define INSTR_RIS_R0RDU 6, { R_8,U8_32,D_20,B_16,0,0 } /* e.g. clibne*/
|
|
#define INSTR_RRE_00 4, { 0,0,0,0,0,0 } /* e.g. palb */
|
|
#define INSTR_RRE_0R 4, { R_28,0,0,0,0,0 } /* e.g. tb */
|
|
#define INSTR_RRE_AA 4, { A_24,A_28,0,0,0,0 } /* e.g. cpya */
|
|
#define INSTR_RRE_AR 4, { A_24,R_28,0,0,0,0 } /* e.g. sar */
|
|
#define INSTR_RRE_F0 4, { F_24,0,0,0,0,0 } /* e.g. lzer */
|
|
#define INSTR_RRE_FE0 4, { FE_24,0,0,0,0,0 } /* e.g. lzxr */
|
|
#define INSTR_RRE_FF 4, { F_24,F_28,0,0,0,0 } /* e.g. debr */
|
|
#define INSTR_RRE_FEF 4, { FE_24,F_28,0,0,0,0 } /* e.g. lxdbr */
|
|
#define INSTR_RRE_FFE 4, { F_24,FE_28,0,0,0,0 } /* e.g. lexr */
|
|
#define INSTR_RRE_FEFE 4, { FE_24,FE_28,0,0,0,0 } /* e.g. dxr */
|
|
#define INSTR_RRE_R0 4, { R_24,0,0,0,0,0 } /* e.g. ipm */
|
|
#define INSTR_RRE_RA 4, { R_24,A_28,0,0,0,0 } /* e.g. ear */
|
|
#define INSTR_RRE_RF 4, { R_24,F_28,0,0,0,0 } /* e.g. lgdr */
|
|
#define INSTR_RRE_RFE 4, { R_24,FE_28,0,0,0,0 } /* e.g. csxtr */
|
|
#define INSTR_RRE_RR 4, { R_24,R_28,0,0,0,0 } /* e.g. lura */
|
|
#define INSTR_RRE_RER 4, { RE_24,R_28,0,0,0,0 } /* e.g. tre */
|
|
#define INSTR_RRE_RERE 4, { RE_24,RE_28,0,0,0,0 } /* e.g. cuse */
|
|
#define INSTR_RRE_FR 4, { F_24,R_28,0,0,0,0 } /* e.g. ldgr */
|
|
#define INSTR_RRE_FER 4, { FE_24,R_28,0,0,0,0 } /* e.g. cxfbr */
|
|
#define INSTR_RRF_F0FF 4, { F_16,F_24,F_28,0,0,0 } /* e.g. madbr */
|
|
#define INSTR_RRF_FE0FF 4, { F_16,F_24,F_28,0,0,0 } /* e.g. myr */
|
|
#define INSTR_RRF_F0FF2 4, { F_24,F_16,F_28,0,0,0 } /* e.g. cpsdr */
|
|
#define INSTR_RRF_F0FR 4, { F_24,F_16,R_28,0,0,0 } /* e.g. iedtr */
|
|
#define INSTR_RRF_FE0FER 4, { FE_24,FE_16,R_28,0,0,0 } /* e.g. iextr */
|
|
#define INSTR_RRF_FUFF 4, { F_24,F_16,F_28,U4_20,0,0 } /* e.g. didbr */
|
|
#define INSTR_RRF_FEUFEFE 4, { FE_24,FE_16,FE_28,U4_20,0,0 } /* e.g. qaxtr */
|
|
#define INSTR_RRF_FUFF2 4, { F_24,F_28,F_16,U4_20,0,0 } /* e.g. adtra */
|
|
#define INSTR_RRF_FEUFEFE2 4, { FE_24,FE_28,FE_16,U4_20,0,0 } /* e.g. axtra */
|
|
#define INSTR_RRF_RURR 4, { R_24,R_28,R_16,U4_20,0,0 } /* e.g. ipte */
|
|
#define INSTR_RRF_RURR2 4, { R_24,R_16,R_28,U4_20,0,0 } /* e.g. lptea */
|
|
#define INSTR_RRF_R0RR 4, { R_24,R_16,R_28,0,0,0 } /* e.g. idte */
|
|
#define INSTR_RRF_R0RR2 4, { R_24,R_28,R_16,0,0,0 } /* e.g. ark */
|
|
#define INSTR_RRF_U0FF 4, { F_24,U4_16,F_28,0,0,0 } /* e.g. fidbr */
|
|
#define INSTR_RRF_U0FEFE 4, { FE_24,U4_16,FE_28,0,0,0 } /* e.g. fixbr */
|
|
#define INSTR_RRF_U0RF 4, { R_24,U4_16,F_28,0,0,0 } /* e.g. cfebr */
|
|
#define INSTR_RRF_U0RFE 4, { R_24,U4_16,FE_28,0,0,0 } /* e.g. cfxbr */
|
|
#define INSTR_RRF_UUFF 4, { F_24,U4_16,F_28,U4_20,0,0 } /* e.g. fidtr */
|
|
#define INSTR_RRF_UUFFE 4, { F_24,U4_16,FE_28,U4_20,0,0 } /* e.g. ldxtr */
|
|
#define INSTR_RRF_UUFEFE 4, { FE_24,U4_16,FE_28,U4_20,0,0 } /* e.g. fixtr */
|
|
#define INSTR_RRF_0UFF 4, { F_24,F_28,U4_20,0,0,0 } /* e.g. ldetr */
|
|
#define INSTR_RRF_0UFEF 4, { FE_24,F_28,U4_20,0,0,0 } /* e.g. lxdtr */
|
|
#define INSTR_RRF_FFRU 4, { F_24,F_16,R_28,U4_20,0,0 } /* e.g. rrdtr */
|
|
#define INSTR_RRF_FEFERU 4, { FE_24,FE_16,R_28,U4_20,0,0 } /* e.g. rrxtr */
|
|
#define INSTR_RRF_U0RR 4, { R_24,R_28,U4_16,0,0,0 } /* e.g. sske */
|
|
#define INSTR_RRF_U0RER 4, { RE_24,R_28,U4_16,0,0,0 } /* e.g. trte */
|
|
#define INSTR_RRF_U0RERE 4, { RE_24,RE_28,U4_16,0,0,0 } /* e.g. cu24 */
|
|
#define INSTR_RRF_00RR 4, { R_24,R_28,0,0,0,0 } /* e.g. clrtne */
|
|
#define INSTR_RRF_0URF 4, { R_24,F_28,U4_20,0,0,0 } /* e.g. csdtr */
|
|
#define INSTR_RRF_0UREFE 4, { RE_24,FE_28,U4_20,0,0,0 } /* e.g. csxtr */
|
|
#define INSTR_RRF_UUFR 4, { F_24,U4_16,R_28,U4_20,0,0 } /* e.g. cdgtra */
|
|
#define INSTR_RRF_UUFER 4, { FE_24,U4_16,R_28,U4_20,0,0 } /* e.g. cxfbra */
|
|
#define INSTR_RRF_UURF 4, { R_24,U4_16,F_28,U4_20,0,0 } /* e.g. cgdtra */
|
|
#define INSTR_RRF_UURFE 4, { R_24,U4_16,FE_28,U4_20,0,0 } /* e.g. cfxbra */
|
|
#define INSTR_RR_0R 2, { R_12, 0,0,0,0,0 } /* e.g. br */
|
|
#define INSTR_RR_FF 2, { F_8,F_12,0,0,0,0 } /* e.g. adr */
|
|
#define INSTR_RR_FEF 2, { FE_8,F_12,0,0,0,0 } /* e.g. mxdr */
|
|
#define INSTR_RR_FFE 2, { F_8,FE_12,0,0,0,0 } /* e.g. ldxr */
|
|
#define INSTR_RR_FEFE 2, { FE_8,FE_12,0,0,0,0 } /* e.g. axr */
|
|
#define INSTR_RR_R0 2, { R_8, 0,0,0,0,0 } /* e.g. spm */
|
|
#define INSTR_RR_RR 2, { R_8,R_12,0,0,0,0 } /* e.g. lr */
|
|
#define INSTR_RR_RER 2, { RE_8,R_12,0,0,0,0 } /* e.g. dr */
|
|
#define INSTR_RR_U0 2, { U8_8, 0,0,0,0,0 } /* e.g. svc */
|
|
#define INSTR_RR_UR 2, { U4_8,R_12,0,0,0,0 } /* e.g. bcr */
|
|
#define INSTR_RRR_F0FF 4, { F_24,F_28,F_16,0,0,0 } /* e.g. ddtr */
|
|
#define INSTR_RRR_FE0FEFE 4, { FE_24,FE_28,FE_16,0,0,0 } /* e.g. axtr */
|
|
#define INSTR_RRS_RRRDU 6, { R_8,R_12,U4_32,D_20,B_16 } /* e.g. crb */
|
|
#define INSTR_RRS_RRRD0 6, { R_8,R_12,D_20,B_16,0 } /* e.g. crbne */
|
|
#define INSTR_RSE_RRRD 6, { R_8,R_12,D_20,B_16,0,0 } /* e.g. lmh */
|
|
#define INSTR_RSE_RERERD 6, { RE_8,RE_12,D_20,B_16,0,0 } /* e.g. mvclu */
|
|
#define INSTR_RSE_CCRD 6, { C_8,C_12,D_20,B_16,0,0 } /* e.g. stctg */
|
|
#define INSTR_RSE_RURD 6, { R_8,U4_12,D_20,B_16,0,0 } /* e.g. icmh */
|
|
#define INSTR_RSL_R0RD 6, { D_20,L4_8,B_16,0,0,0 } /* e.g. tp */
|
|
#define INSTR_RSL_LRDFU 6, { F_32,D_20,L8_8,B_16,U4_36,0 } /* e.g. cdzt */
|
|
#define INSTR_RSL_LRDFEU 6, { FE_32,D_20,L8_8,B_16,U4_36,0 } /* e.g. cxzt */
|
|
#define INSTR_RSI_RRP 4, { R_8,R_12,J16_16,0,0,0 } /* e.g. brxh */
|
|
#define INSTR_RSY_RRRD 6, { R_8,R_12,D20_20,B_16,0,0 } /* e.g. stmy */
|
|
#define INSTR_RSY_RERERD 6, { RE_8,RE_12,D20_20,B_16,0,0 } /* e.g. cdsy */
|
|
#define INSTR_RSY_RURD 6, { R_8,U4_12,D20_20,B_16,0,0 } /* e.g. icmh */
|
|
#define INSTR_RSY_RURD2 6, { R_8,D20_20,B_16,U4_12,0,0 } /* e.g. loc */
|
|
#define INSTR_RSY_R0RD 6, { R_8,D20_20,B_16,0,0,0 } /* e.g. locne */
|
|
#define INSTR_RSY_AARD 6, { A_8,A_12,D20_20,B_16,0,0 } /* e.g. lamy */
|
|
#define INSTR_RSY_CCRD 6, { C_8,C_12,D20_20,B_16,0,0 } /* e.g. stctg */
|
|
#define INSTR_RS_AARD 4, { A_8,A_12,D_20,B_16,0,0 } /* e.g. lam */
|
|
#define INSTR_RS_CCRD 4, { C_8,C_12,D_20,B_16,0,0 } /* e.g. lctl */
|
|
#define INSTR_RS_R0RD 4, { R_8,D_20,B_16,0,0,0 } /* e.g. sll */
|
|
#define INSTR_RS_RE0RD 4, { RE_8,D_20,B_16,0,0,0 } /* e.g. slda */
|
|
#define INSTR_RS_RRRD 4, { R_8,R_12,D_20,B_16,0,0 } /* e.g. cs */
|
|
#define INSTR_RS_RERERD 4, { RE_8,RE_12,D_20,B_16,0,0 } /* e.g. cds */
|
|
#define INSTR_RS_RURD 4, { R_8,U4_12,D_20,B_16,0,0 } /* e.g. icm */
|
|
#define INSTR_RXE_FRRD 6, { F_8,D_20,X_12,B_16,0,0 } /* e.g. adb */
|
|
#define INSTR_RXE_FERRD 6, { FE_8,D_20,X_12,B_16,0,0 } /* e.g. lxdb */
|
|
#define INSTR_RXE_RRRD 6, { R_8,D_20,X_12,B_16,0,0 } /* e.g. lg */
|
|
#define INSTR_RXE_RRRDU 6, { R_8,D_20,X_12,B_16,U4_32,0 } /* e.g. lcbb */
|
|
#define INSTR_RXE_RERRD 6, { RE_8,D_20,X_12,B_16,0,0 } /* e.g. dsg */
|
|
#define INSTR_RXF_FRRDF 6, { F_32,F_8,D_20,X_12,B_16,0 } /* e.g. madb */
|
|
#define INSTR_RXF_FRRDFE 6, { FE_32,F_8,D_20,X_12,B_16,0 } /* e.g. my */
|
|
#define INSTR_RXF_FERRDFE 6, { FE_32,FE_8,D_20,X_12,B_16,0 } /* e.g. slxt */
|
|
#define INSTR_RXF_RRRDR 6, { R_32,R_8,D_20,X_12,B_16,0 } /* e.g. .insn */
|
|
#define INSTR_RXY_RRRD 6, { R_8,D20_20,X_12,B_16,0,0 } /* e.g. ly */
|
|
#define INSTR_RXY_RERRD 6, { RE_8,D20_20,X_12,B_16,0,0 } /* e.g. dsg */
|
|
#define INSTR_RXY_FRRD 6, { F_8,D20_20,X_12,B_16,0,0 } /* e.g. ley */
|
|
#define INSTR_RXY_URRD 6, { U4_8,D20_20,X_12,B_16,0,0 } /* e.g. pfd */
|
|
#define INSTR_RXY_0RRD 6, { D20_20,X_12,B_16,0,0 } /* e.g. bic */
|
|
#define INSTR_RX_0RRD 4, { D_20,X_12,B_16,0,0,0 } /* e.g. be */
|
|
#define INSTR_RX_FRRD 4, { F_8,D_20,X_12,B_16,0,0 } /* e.g. ae */
|
|
#define INSTR_RX_FERRD 4, { FE_8,D_20,X_12,B_16,0,0 } /* e.g. mxd */
|
|
#define INSTR_RX_RRRD 4, { R_8,D_20,X_12,B_16,0,0 } /* e.g. l */
|
|
#define INSTR_RX_RERRD 4, { RE_8,D_20,X_12,B_16,0,0 } /* e.g. d */
|
|
#define INSTR_RX_URRD 4, { U4_8,D_20,X_12,B_16,0,0 } /* e.g. bc */
|
|
#define INSTR_SI_URD 4, { D_20,B_16,U8_8,0,0,0 } /* e.g. cli */
|
|
#define INSTR_SIY_URD 6, { D20_20,B_16,U8_8,0,0,0 } /* e.g. tmy */
|
|
#define INSTR_SIY_IRD 6, { D20_20,B_16,I8_8,0,0,0 } /* e.g. asi */
|
|
#define INSTR_SIL_RDI 6, { D_20,B_16,I16_32,0,0,0 } /* e.g. chhsi */
|
|
#define INSTR_SIL_RDU 6, { D_20,B_16,U16_32,0,0,0 } /* e.g. clfhsi */
|
|
#define INSTR_SMI_U0RDP 6, { U4_8,J16_32,D_20,B_16,0,0 } /* e.g. bpp */
|
|
#define INSTR_SSE_RDRD 6, { D_20,B_16,D_36,B_32,0,0 } /* e.g. mvcdk */
|
|
#define INSTR_SS_L0RDRD 6, { D_20,L8_8,B_16,D_36,B_32,0 } /* e.g. mvc */
|
|
#define INSTR_SS_L2RDRD 6, { D_20,B_16,D_36,L8_8,B_32,0 } /* e.g. pka */
|
|
#define INSTR_SS_LIRDRD 6, { D_20,L4_8,B_16,D_36,B_32,U4_12 } /* e.g. srp */
|
|
#define INSTR_SS_LLRDRD 6, { D_20,L4_8,B_16,D_36,L4_12,B_32 } /* e.g. pack */
|
|
#define INSTR_SS_RRRDRD 6, { D_20,R_8,B_16,D_36,B_32,R_12 } /* e.g. mvck */
|
|
#define INSTR_SS_RRRDRD2 6, { R_8,D_20,B_16,R_12,D_36,B_32 } /* e.g. plo */
|
|
#define INSTR_SS_RRRDRD3 6, { R_8,R_12,D_20,B_16,D_36,B_32 } /* e.g. lmd */
|
|
#define INSTR_SSF_RRDRD 6, { D_20,B_16,D_36,B_32,R_8,0 } /* e.g. mvcos */
|
|
#define INSTR_SSF_RERDRD2 6, { RE_8,D_20,B_16,D_36,B_32,0 } /* e.g. lpd */
|
|
#define INSTR_S_00 4, { 0,0,0,0,0,0 } /* e.g. hsch */
|
|
#define INSTR_S_RD 4, { D_20,B_16,0,0,0,0 } /* e.g. lpsw */
|
|
#define INSTR_VRV_VVXRDU 6, { V_8,D_20,VX_12,B_16,U4_32,0 } /* e.g. vgef */
|
|
#define INSTR_VRI_V0U 6, { V_8,U16_16,0,0,0,0 } /* e.g. vgbm */
|
|
#define INSTR_VRI_V 6, { V_8,0,0,0,0,0 } /* e.g. vzero */
|
|
#define INSTR_VRI_V0UUU 6, { V_8,U8_16,U8_24,U4_32,0,0 } /* e.g. vgm */
|
|
#define INSTR_VRI_V0UU 6, { V_8,U8_16,U8_24,0,0,0 } /* e.g. vgmb */
|
|
#define INSTR_VRI_V0UU2 6, { V_8,U16_16,U4_32,0,0,0 } /* e.g. vlip */
|
|
#define INSTR_VRI_VVUU 6, { V_8,V_12,U16_16,U4_32,0,0 } /* e.g. vrep */
|
|
#define INSTR_VRI_VVU 6, { V_8,V_12,U16_16,0,0,0 } /* e.g. vrepb */
|
|
#define INSTR_VRI_VVU2 6, { V_8,V_12,U12_16,0,0,0 } /* e.g. vftcidb */
|
|
#define INSTR_VRI_V0IU 6, { V_8,I16_16,U4_32,0,0,0 } /* e.g. vrepi */
|
|
#define INSTR_VRI_V0I 6, { V_8,I16_16,0,0,0,0 } /* e.g. vrepib */
|
|
#define INSTR_VRI_VVV0UU 6, { V_8,V_12,V_16,U8_24,U4_32,0 } /* e.g. verim */
|
|
#define INSTR_VRI_VVV0UU2 6, { V_8,V_12,V_16,U8_28,U4_24,0 } /* e.g. vap */
|
|
#define INSTR_VRI_VVV0U 6, { V_8,V_12,V_16,U8_24,0,0 } /* e.g. verimb*/
|
|
#define INSTR_VRI_VVUUU 6, { V_8,V_12,U12_16,U4_32,U4_28,0 } /* e.g. vftci */
|
|
#define INSTR_VRI_VVUUU2 6, { V_8,V_12,U8_28,U8_16,U4_24,0 } /* e.g. vpsop */
|
|
#define INSTR_VRI_VR0UU 6, { V_8,R_12,U8_28,U4_24,0,0 } /* e.g. vcvd */
|
|
#define INSTR_VRX_VRRD 6, { V_8,D_20,X_12,B_16,0,0 } /* e.g. vl */
|
|
#define INSTR_VRX_VV 6, { V_8,V_12,0,0,0,0 } /* e.g. vlr */
|
|
#define INSTR_VRX_VRRDU 6, { V_8,D_20,X_12,B_16,U4_32,0 } /* e.g. vlrep */
|
|
#define INSTR_VRS_RVRDU 6, { R_8,V_12,D_20,B_16,U4_32,0 } /* e.g. vlgv */
|
|
#define INSTR_VRS_RVRD 6, { R_8,V_12,D_20,B_16,0,0 } /* e.g. vlgvb */
|
|
#define INSTR_VRS_VVRDU 6, { V_8,V_12,D_20,B_16,U4_32,0 } /* e.g. verll */
|
|
#define INSTR_VRS_VVRD 6, { V_8,V_12,D_20,B_16,0,0 } /* e.g. vlm */
|
|
#define INSTR_VRS_VRRDU 6, { V_8,R_12,D_20,B_16,U4_32,0 } /* e.g. vlvg */
|
|
#define INSTR_VRS_VRRD 6, { V_8,R_12,D_20,B_16,0,0 } /* e.g. vlvgb */
|
|
#define INSTR_VRS_RRDV 6, { V_32,R_12,D_20,B_16,0,0 } /* e.g. vlrlr */
|
|
#define INSTR_VRR_0V 6, { V_12,0,0,0,0,0 } /* e.g. vtp */
|
|
#define INSTR_VRR_VRR 6, { V_8,R_12,R_16,0,0,0 } /* e.g. vlvgp */
|
|
#define INSTR_VRR_VVV0U 6, { V_8,V_12,V_16,U4_32,0,0 } /* e.g. vmrh */
|
|
#define INSTR_VRR_VVV0U0 6, { V_8,V_12,V_16,U4_24,0,0 } /* e.g. vfaeb */
|
|
#define INSTR_VRR_VVV0U1 6, { V_8,V_12,V_16,U4_OR1_24,0,0 } /* e.g. vfaebs*/
|
|
#define INSTR_VRR_VVV0U2 6, { V_8,V_12,V_16,U4_OR2_24,0,0 } /* e.g. vfaezb*/
|
|
#define INSTR_VRR_VVV0U3 6, { V_8,V_12,V_16,U4_OR3_24,0,0 } /* e.g. vfaezbs*/
|
|
#define INSTR_VRR_VVV 6, { V_8,V_12,V_16,0,0,0 } /* e.g. vmrhb */
|
|
#define INSTR_VRR_VVV2 6, { V_8,V_CP16_12,0,0,0,0 } /* e.g. vnot */
|
|
#define INSTR_VRR_VV0U 6, { V_8,V_12,U4_32,0,0,0 } /* e.g. vseg */
|
|
#define INSTR_VRR_VV0U2 6, { V_8,V_12,U4_24,0,0,0 } /* e.g. vistrb*/
|
|
#define INSTR_VRR_VV0UU 6, { V_8,V_12,U4_28,U4_24,0,0 } /* e.g. vcdgb */
|
|
#define INSTR_VRR_VV0UU2 6, { V_8,V_12,U4_32,U4_28,0,0 } /* e.g. wfc */
|
|
#define INSTR_VRR_VV0UU8 6, { V_8,V_12,U4_OR8_28,U4_24,0,0 } /* e.g. wcdgb */
|
|
#define INSTR_VRR_VV 6, { V_8,V_12,0,0,0,0 } /* e.g. vsegb */
|
|
#define INSTR_VRR_VVVUU0V 6, { V_8,V_12,V_16,V_32,U4_20,U4_24 } /* e.g. vstrc */
|
|
#define INSTR_VRR_VVVU0V 6, { V_8,V_12,V_16,V_32,U4_20,0 } /* e.g. vac */
|
|
#define INSTR_VRR_VVVU0VB 6, { V_8,V_12,V_16,V_32,U4_24,0 } /* e.g. vstrcb*/
|
|
#define INSTR_VRR_VVVU0VB1 6, { V_8,V_12,V_16,V_32,U4_OR1_24,0 } /* e.g. vstrcbs*/
|
|
#define INSTR_VRR_VVVU0VB2 6, { V_8,V_12,V_16,V_32,U4_OR2_24,0 } /* e.g. vstrczb*/
|
|
#define INSTR_VRR_VVVU0VB3 6, { V_8,V_12,V_16,V_32,U4_OR3_24,0 } /* e.g. vstrczbs*/
|
|
#define INSTR_VRR_VVV0V 6, { V_8,V_12,V_16,V_32,0,0 } /* e.g. vacq */
|
|
#define INSTR_VRR_VVV0U0U 6, { V_8,V_12,V_16,U4_32,U4_24,0 } /* e.g. vfae */
|
|
#define INSTR_VRR_VVVV 6, { V_8,V_12,V_16,V_32,0,0 } /* e.g. vfmadb*/
|
|
#define INSTR_VRR_VVV0UUU 6, { V_8,V_12,V_16,U4_32,U4_28,U4_24 }/* e.g. vfch */
|
|
#define INSTR_VRR_VVV0UU 6, { V_8,V_12,V_16,U4_32,U4_28,0 } /* e.g. vfa */
|
|
#define INSTR_VRR_VV0UUU 6, { V_8,V_12,U4_32,U4_28,U4_24,0 } /* e.g. vcdg */
|
|
#define INSTR_VRR_VVVU0UV 6, { V_8,V_12,V_16,V_32,U4_28,U4_20 } /* e.g. vfma */
|
|
#define INSTR_VRR_VV0U0U 6, { V_8,V_12,U4_32,U4_24,0,0 } /* e.g. vistr */
|
|
#define INSTR_VRR_0VV0U 6, { V_12,V_16,U4_24,0,0,0 } /* e.g. vcp */
|
|
#define INSTR_VRR_RV0U 6, { R_8,V_12,U4_24,0,0,0 } /* e.g. vcvb */
|
|
#define INSTR_VSI_URDV 6, { V_32,D_20,B_16,U8_8,0,0 } /* e.g. vlrl */
|
|
|
|
#define MASK_E { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_IE_UU { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_MII_UPP { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RIE_RRP { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RIE_RRPU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RIE_RRP0 { 0xff, 0x00, 0x00, 0x00, 0xf0, 0xff }
|
|
#define MASK_RIE_RRI0 { 0xff, 0x00, 0x00, 0x00, 0xf0, 0xff }
|
|
#define MASK_RIE_RUPI { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RIE_R0PI { 0xff, 0x0f, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RIE_RUPU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RIE_R0PU { 0xff, 0x0f, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RIE_R0IU { 0xff, 0x0f, 0x00, 0x00, 0x0f, 0xff }
|
|
#define MASK_RIE_R0I0 { 0xff, 0x0f, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RIE_R0UU { 0xff, 0x0f, 0x00, 0x00, 0x0f, 0xff }
|
|
#define MASK_RIE_R0U0 { 0xff, 0x0f, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RIE_RUI0 { 0xff, 0x00, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RIE_RRUUU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RIL_0P { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RIL_RP { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RIL_UP { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RIL_RI { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RIL_RU { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RI_0P { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RI_RI { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RI_RP { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RI_RU { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RI_UP { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RIS_RURDI { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RIS_R0RDI { 0xff, 0x0f, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RIS_RURDU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RIS_R0RDU { 0xff, 0x0f, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RRE_00 { 0xff, 0xff, 0xff, 0xff, 0x00, 0x00 }
|
|
#define MASK_RRE_0R { 0xff, 0xff, 0xff, 0xf0, 0x00, 0x00 }
|
|
#define MASK_RRE_AA { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_AR { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_F0 { 0xff, 0xff, 0xff, 0x0f, 0x00, 0x00 }
|
|
#define MASK_RRE_FE0 { 0xff, 0xff, 0xff, 0x0f, 0x00, 0x00 }
|
|
#define MASK_RRE_FF { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_FEF { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_FFE { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_FEFE { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_R0 { 0xff, 0xff, 0xff, 0x0f, 0x00, 0x00 }
|
|
#define MASK_RRE_RA { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_RF { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_RFE { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_RR { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_RER { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_RERE { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_FR { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRE_FER { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_F0FF { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_FE0FF { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_F0FF2 { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_F0FR { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_FE0FER { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_FUFF { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_FEUFEFE { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_FUFF2 { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_FEUFEFE2 { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_RURR { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_RURR2 { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_R0RR { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_R0RR2 { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_U0FF { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_U0FEFE { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_U0RF { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_U0RFE { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_UUFF { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_UUFFE { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_UUFEFE { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_0UFF { 0xff, 0xff, 0xf0, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_0UFEF { 0xff, 0xff, 0xf0, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_FFRU { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_FEFERU { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_U0RR { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_U0RER { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_U0RERE { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_00RR { 0xff, 0xff, 0xff, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_0URF { 0xff, 0xff, 0xf0, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_0UREFE { 0xff, 0xff, 0xf0, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_UUFR { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_UUFER { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_UURF { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRF_UURFE { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RR_0R { 0xff, 0xf0, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RR_FF { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RR_FEF { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RR_FFE { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RR_FEFE { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RR_R0 { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RR_RR { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RR_RER { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RR_U0 { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RR_UR { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRR_F0FF { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRR_FE0FEFE { 0xff, 0xff, 0x0f, 0x00, 0x00, 0x00 }
|
|
#define MASK_RRS_RRRDU { 0xff, 0x00, 0x00, 0x00, 0x0f, 0xff }
|
|
#define MASK_RRS_RRRD0 { 0xff, 0x00, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RSE_RRRD { 0xff, 0x00, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RSE_RERERD { 0xff, 0x00, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RSE_CCRD { 0xff, 0x00, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RSE_RURD { 0xff, 0x00, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RSL_R0RD { 0xff, 0x0f, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RSL_LRDFU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RSL_LRDFEU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RSI_RRP { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RS_AARD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RS_CCRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RS_R0RD { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RS_RE0RD { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RS_RRRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RS_RERERD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RS_RURD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RSY_RRRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RSY_RERERD { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RSY_RURD { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RSY_RURD2 { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RSY_R0RD { 0xff, 0x0f, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RSY_AARD { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RSY_CCRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RXE_FRRD { 0xff, 0x00, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RXE_FERRD { 0xff, 0x00, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RXE_RRRD { 0xff, 0x00, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RXE_RRRDU { 0xff, 0x00, 0x00, 0x00, 0x0f, 0xff }
|
|
#define MASK_RXE_RERRD { 0xff, 0x00, 0x00, 0x00, 0xff, 0xff }
|
|
#define MASK_RXF_FRRDF { 0xff, 0x00, 0x00, 0x00, 0x0f, 0xff }
|
|
#define MASK_RXF_FRRDFE { 0xff, 0x00, 0x00, 0x00, 0x0f, 0xff }
|
|
#define MASK_RXF_FERRDFE { 0xff, 0x00, 0x00, 0x00, 0x0f, 0xff }
|
|
#define MASK_RXF_RRRDR { 0xff, 0x00, 0x00, 0x00, 0x0f, 0xff }
|
|
#define MASK_RXY_RRRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RXY_RERRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RXY_FRRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RXY_URRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RXY_0RRD { 0xff, 0xf0, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_RX_0RRD { 0xff, 0xf0, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RX_FRRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RX_FERRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RX_RRRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RX_RERRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_RX_URRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SI_URD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SIY_URD { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_SIY_IRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_SIL_RDI { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SIL_RDU { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SMI_U0RDP { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SSE_RDRD { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SS_L0RDRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SS_L2RDRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SS_LIRDRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SS_LLRDRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SS_RRRDRD { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SS_RRRDRD2 { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SS_RRRDRD3 { 0xff, 0x00, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SSF_RRDRD { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_SSF_RERDRD2 { 0xff, 0x0f, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_S_00 { 0xff, 0xff, 0xff, 0xff, 0x00, 0x00 }
|
|
#define MASK_S_RD { 0xff, 0xff, 0x00, 0x00, 0x00, 0x00 }
|
|
#define MASK_VRV_VVXRDU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRI_V0U { 0xff, 0x0f, 0x00, 0x00, 0xf0, 0xff }
|
|
#define MASK_VRI_V { 0xff, 0x0f, 0xff, 0xff, 0xf0, 0xff }
|
|
#define MASK_VRI_V0UUU { 0xff, 0x0f, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRI_V0UU { 0xff, 0x0f, 0x00, 0x00, 0xf0, 0xff }
|
|
#define MASK_VRI_V0UU2 { 0xff, 0x0f, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRI_VVUU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRI_VVU { 0xff, 0x00, 0x00, 0x00, 0xf0, 0xff }
|
|
#define MASK_VRI_VVU2 { 0xff, 0x00, 0x00, 0x0f, 0xf0, 0xff }
|
|
#define MASK_VRI_V0IU { 0xff, 0x0f, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRI_V0I { 0xff, 0x0f, 0x00, 0x00, 0xf0, 0xff }
|
|
#define MASK_VRI_VVV0UU { 0xff, 0x00, 0x0f, 0x00, 0x00, 0xff }
|
|
#define MASK_VRI_VVV0UU2 { 0xff, 0x00, 0x0f, 0x00, 0x00, 0xff }
|
|
#define MASK_VRI_VVV0U { 0xff, 0x00, 0x0f, 0x00, 0xf0, 0xff }
|
|
#define MASK_VRI_VVUUU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRI_VVUUU2 { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRI_VR0UU { 0xff, 0x00, 0xff, 0x00, 0x00, 0xff }
|
|
#define MASK_VRX_VRRD { 0xff, 0x00, 0x00, 0x00, 0xf0, 0xff }
|
|
#define MASK_VRX_VV { 0xff, 0x00, 0xff, 0xff, 0xf0, 0xff }
|
|
#define MASK_VRX_VRRDU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRS_RVRDU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRS_RVRD { 0xff, 0x00, 0x00, 0x00, 0xf0, 0xff }
|
|
#define MASK_VRS_VVRDU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRS_VVRD { 0xff, 0x00, 0x00, 0x00, 0xf0, 0xff }
|
|
#define MASK_VRS_VRRDU { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRS_VRRD { 0xff, 0x00, 0x00, 0x00, 0xf0, 0xff }
|
|
#define MASK_VRS_RRDV { 0xff, 0xf0, 0x00, 0x00, 0x00, 0xff }
|
|
#define MASK_VRR_0V { 0xff, 0xf0, 0xff, 0xff, 0xf0, 0xff }
|
|
#define MASK_VRR_VRR { 0xff, 0x00, 0x0f, 0xff, 0xf0, 0xff }
|
|
#define MASK_VRR_VVV0U { 0xff, 0x00, 0x0f, 0xff, 0x00, 0xff }
|
|
#define MASK_VRR_VVV0U0 { 0xff, 0x00, 0x0f, 0x0f, 0xf0, 0xff }
|
|
#define MASK_VRR_VVV0U1 { 0xff, 0x00, 0x0f, 0x1f, 0xf0, 0xff }
|
|
#define MASK_VRR_VVV0U2 { 0xff, 0x00, 0x0f, 0x2f, 0xf0, 0xff }
|
|
#define MASK_VRR_VVV0U3 { 0xff, 0x00, 0x0f, 0x3f, 0xf0, 0xff }
|
|
#define MASK_VRR_VVV { 0xff, 0x00, 0x0f, 0xff, 0xf0, 0xff }
|
|
#define MASK_VRR_VVV2 { 0xff, 0x00, 0x0f, 0xff, 0xf0, 0xff }
|
|
#define MASK_VRR_VVV0V { 0xff, 0x00, 0x0f, 0xff, 0x00, 0xff }
|
|
#define MASK_VRR_VV0U { 0xff, 0x00, 0xff, 0xff, 0x00, 0xff }
|
|
#define MASK_VRR_VV0U2 { 0xff, 0x00, 0xff, 0x0f, 0xf0, 0xff }
|
|
#define MASK_VRR_VV0UU { 0xff, 0x00, 0xff, 0x00, 0xf0, 0xff }
|
|
#define MASK_VRR_VV0UU2 { 0xff, 0x00, 0xff, 0xf0, 0x00, 0xff }
|
|
#define MASK_VRR_VV0UU8 { 0xff, 0x00, 0xff, 0x08, 0xf0, 0xff }
|
|
#define MASK_VRR_VV { 0xff, 0x00, 0xff, 0xff, 0xf0, 0xff }
|
|
#define MASK_VRR_VVVUU0V { 0xff, 0x00, 0x00, 0x0f, 0x00, 0xff }
|
|
#define MASK_VRR_VVVU0V { 0xff, 0x00, 0x00, 0xff, 0x00, 0xff }
|
|
#define MASK_VRR_VVVU0VB { 0xff, 0x00, 0x0f, 0x0f, 0x00, 0xff }
|
|
#define MASK_VRR_VVVU0VB1 { 0xff, 0x00, 0x0f, 0x1f, 0x00, 0xff }
|
|
#define MASK_VRR_VVVU0VB2 { 0xff, 0x00, 0x0f, 0x2f, 0x00, 0xff }
|
|
#define MASK_VRR_VVVU0VB3 { 0xff, 0x00, 0x0f, 0x3f, 0x00, 0xff }
|
|
#define MASK_VRR_VVV0U0U { 0xff, 0x00, 0x0f, 0x0f, 0x00, 0xff }
|
|
#define MASK_VRR_VVVV { 0xff, 0x00, 0x0f, 0xff, 0x00, 0xff }
|
|
#define MASK_VRR_VVV0UUU { 0xff, 0x00, 0x0f, 0x00, 0x00, 0xff }
|
|
#define MASK_VRR_VVV0UU { 0xff, 0x00, 0x0f, 0xf0, 0x00, 0xff }
|
|
#define MASK_VRR_VV0UUU { 0xff, 0x00, 0xff, 0x00, 0x00, 0xff }
|
|
#define MASK_VRR_VVVU0UV { 0xff, 0x00, 0x00, 0xf0, 0x00, 0xff }
|
|
#define MASK_VRR_VV0U0U { 0xff, 0x00, 0xff, 0x0f, 0x00, 0xff }
|
|
#define MASK_VRR_0VV0U { 0xff, 0xf0, 0x0f, 0x0f, 0xf0, 0xff }
|
|
#define MASK_VRR_RV0U { 0xff, 0x00, 0xff, 0x0f, 0xf0, 0xff }
|
|
#define MASK_VSI_URDV { 0xff, 0x00, 0x00, 0x00, 0x00, 0xff }
|
|
|
|
|
|
/* The opcode formats table (blueprints for .insn pseudo mnemonic). */
|
|
|
|
const struct s390_opcode s390_opformats[] =
|
|
{
|
|
{ "e", OP8(0x00LL), MASK_E, INSTR_E, 3, 0 ,0 },
|
|
{ "ri", OP8(0x00LL), MASK_RI_RI, INSTR_RI_RI, 3, 0 ,0 },
|
|
{ "rie", OP8(0x00LL), MASK_RIE_RRP, INSTR_RIE_RRP, 3, 0 ,0 },
|
|
{ "ril", OP8(0x00LL), MASK_RIL_RP, INSTR_RIL_RP, 3, 0 ,0 },
|
|
{ "rilu", OP8(0x00LL), MASK_RIL_RU, INSTR_RIL_RU, 3, 0 ,0 },
|
|
{ "ris", OP8(0x00LL), MASK_RIS_RURDI, INSTR_RIS_RURDI, 3, 6 ,0 },
|
|
{ "rr", OP8(0x00LL), MASK_RR_RR, INSTR_RR_RR, 3, 0 ,0 },
|
|
{ "rre", OP8(0x00LL), MASK_RRE_RR, INSTR_RRE_RR, 3, 0 ,0 },
|
|
{ "rrf", OP8(0x00LL), MASK_RRF_RURR, INSTR_RRF_RURR, 3, 0 ,0 },
|
|
{ "rrs", OP8(0x00LL), MASK_RRS_RRRDU, INSTR_RRS_RRRDU, 3, 6 ,0 },
|
|
{ "rs", OP8(0x00LL), MASK_RS_RRRD, INSTR_RS_RRRD, 3, 0 ,0 },
|
|
{ "rse", OP8(0x00LL), MASK_RSE_RRRD, INSTR_RSE_RRRD, 3, 0 ,0 },
|
|
{ "rsi", OP8(0x00LL), MASK_RSI_RRP, INSTR_RSI_RRP, 3, 0 ,0 },
|
|
{ "rsy", OP8(0x00LL), MASK_RSY_RRRD, INSTR_RSY_RRRD, 3, 3 ,0 },
|
|
{ "rx", OP8(0x00LL), MASK_RX_RRRD, INSTR_RX_RRRD, 3, 0 ,0 },
|
|
{ "rxe", OP8(0x00LL), MASK_RXE_RRRD, INSTR_RXE_RRRD, 3, 0 ,0 },
|
|
{ "rxf", OP8(0x00LL), MASK_RXF_RRRDR, INSTR_RXF_RRRDR, 3, 0 ,0 },
|
|
{ "rxy", OP8(0x00LL), MASK_RXY_RRRD, INSTR_RXY_RRRD, 3, 3 ,0 },
|
|
{ "s", OP8(0x00LL), MASK_S_RD, INSTR_S_RD, 3, 0 ,0 },
|
|
{ "si", OP8(0x00LL), MASK_SI_URD, INSTR_SI_URD, 3, 0 ,0 },
|
|
{ "siy", OP8(0x00LL), MASK_SIY_URD, INSTR_SIY_URD, 3, 3 ,0 },
|
|
{ "sil", OP8(0x00LL), MASK_SIL_RDI, INSTR_SIL_RDI, 3, 6 ,0 },
|
|
{ "ss", OP8(0x00LL), MASK_SS_RRRDRD, INSTR_SS_RRRDRD, 3, 0 ,0 },
|
|
{ "sse", OP8(0x00LL), MASK_SSE_RDRD, INSTR_SSE_RDRD, 3, 0 ,0 },
|
|
{ "ssf", OP8(0x00LL), MASK_SSF_RRDRD, INSTR_SSF_RRDRD, 3, 0 ,0 },
|
|
{ "vrv", OP8(0x00LL), MASK_VRV_VVXRDU, INSTR_VRV_VVXRDU, 3, 9 ,0 },
|
|
{ "vri", OP8(0x00LL), MASK_VRI_VVUUU, INSTR_VRI_VVUUU, 3, 9 ,0 },
|
|
{ "vrx", OP8(0x00LL), MASK_VRX_VRRDU, INSTR_VRX_VRRDU, 3, 9 ,0 },
|
|
{ "vrs", OP8(0x00LL), MASK_VRS_RVRDU, INSTR_VRS_RVRDU, 3, 9 ,0 },
|
|
{ "vrr", OP8(0x00LL), MASK_VRR_VVV0UUU, INSTR_VRR_VVV0UUU, 3, 9 ,0 },
|
|
{ "vsi", OP8(0x00LL), MASK_VSI_URDV, INSTR_VSI_URDV, 3, 10 ,0 },
|
|
};
|
|
|
|
const int s390_num_opformats =
|
|
sizeof (s390_opformats) / sizeof (s390_opformats[0]);
|
|
|
|
#include "s390-opc.tab"
|