qemu-e2k/target-ppc/op_helper.c
ths bfed01fc79 Clean up of some target specifics in exec.c/cpu-exec.c.
git-svn-id: svn://svn.savannah.nongnu.org/qemu/trunk@2936 c046a42c-6fe2-441c-8c8c-71466251a162
2007-06-03 17:44:37 +00:00

2608 lines
59 KiB
C

/*
* PowerPC emulation helpers for qemu.
*
* Copyright (c) 2003-2007 Jocelyn Mayer
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2 of the License, or (at your option) any later version.
*
* This library 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
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
#include "exec.h"
#include "op_helper.h"
#define MEMSUFFIX _raw
#include "op_helper.h"
#include "op_helper_mem.h"
#if !defined(CONFIG_USER_ONLY)
#define MEMSUFFIX _user
#include "op_helper.h"
#include "op_helper_mem.h"
#define MEMSUFFIX _kernel
#include "op_helper.h"
#include "op_helper_mem.h"
#endif
//#define DEBUG_OP
//#define DEBUG_EXCEPTIONS
//#define DEBUG_SOFTWARE_TLB
//#define FLUSH_ALL_TLBS
/*****************************************************************************/
/* Exceptions processing helpers */
void do_raise_exception_err (uint32_t exception, int error_code)
{
#if 0
printf("Raise exception %3x code : %d\n", exception, error_code);
#endif
switch (exception) {
case EXCP_PROGRAM:
if (error_code == EXCP_FP && msr_fe0 == 0 && msr_fe1 == 0)
return;
break;
default:
break;
}
env->exception_index = exception;
env->error_code = error_code;
cpu_loop_exit();
}
void do_raise_exception (uint32_t exception)
{
do_raise_exception_err(exception, 0);
}
void cpu_dump_EA (target_ulong EA);
void do_print_mem_EA (target_ulong EA)
{
cpu_dump_EA(EA);
}
/*****************************************************************************/
/* Registers load and stores */
void do_load_cr (void)
{
T0 = (env->crf[0] << 28) |
(env->crf[1] << 24) |
(env->crf[2] << 20) |
(env->crf[3] << 16) |
(env->crf[4] << 12) |
(env->crf[5] << 8) |
(env->crf[6] << 4) |
(env->crf[7] << 0);
}
void do_store_cr (uint32_t mask)
{
int i, sh;
for (i = 0, sh = 7; i < 8; i++, sh --) {
if (mask & (1 << sh))
env->crf[i] = (T0 >> (sh * 4)) & 0xFUL;
}
}
void do_load_xer (void)
{
T0 = (xer_so << XER_SO) |
(xer_ov << XER_OV) |
(xer_ca << XER_CA) |
(xer_bc << XER_BC) |
(xer_cmp << XER_CMP);
}
void do_store_xer (void)
{
xer_so = (T0 >> XER_SO) & 0x01;
xer_ov = (T0 >> XER_OV) & 0x01;
xer_ca = (T0 >> XER_CA) & 0x01;
xer_cmp = (T0 >> XER_CMP) & 0xFF;
xer_bc = (T0 >> XER_BC) & 0x7F;
}
void do_load_fpscr (void)
{
/* The 32 MSB of the target fpr are undefined.
* They'll be zero...
*/
union {
float64 d;
struct {
uint32_t u[2];
} s;
} u;
int i;
#if defined(WORDS_BIGENDIAN)
#define WORD0 0
#define WORD1 1
#else
#define WORD0 1
#define WORD1 0
#endif
u.s.u[WORD0] = 0;
u.s.u[WORD1] = 0;
for (i = 0; i < 8; i++)
u.s.u[WORD1] |= env->fpscr[i] << (4 * i);
FT0 = u.d;
}
void do_store_fpscr (uint32_t mask)
{
/*
* We use only the 32 LSB of the incoming fpr
*/
union {
double d;
struct {
uint32_t u[2];
} s;
} u;
int i, rnd_type;
u.d = FT0;
if (mask & 0x80)
env->fpscr[0] = (env->fpscr[0] & 0x9) | ((u.s.u[WORD1] >> 28) & ~0x9);
for (i = 1; i < 7; i++) {
if (mask & (1 << (7 - i)))
env->fpscr[i] = (u.s.u[WORD1] >> (4 * (7 - i))) & 0xF;
}
/* TODO: update FEX & VX */
/* Set rounding mode */
switch (env->fpscr[0] & 0x3) {
case 0:
/* Best approximation (round to nearest) */
rnd_type = float_round_nearest_even;
break;
case 1:
/* Smaller magnitude (round toward zero) */
rnd_type = float_round_to_zero;
break;
case 2:
/* Round toward +infinite */
rnd_type = float_round_up;
break;
default:
case 3:
/* Round toward -infinite */
rnd_type = float_round_down;
break;
}
set_float_rounding_mode(rnd_type, &env->fp_status);
}
target_ulong ppc_load_dump_spr (int sprn)
{
if (loglevel != 0) {
fprintf(logfile, "Read SPR %d %03x => " ADDRX "\n",
sprn, sprn, env->spr[sprn]);
}
return env->spr[sprn];
}
void ppc_store_dump_spr (int sprn, target_ulong val)
{
if (loglevel != 0) {
fprintf(logfile, "Write SPR %d %03x => " ADDRX " <= " ADDRX "\n",
sprn, sprn, env->spr[sprn], val);
}
env->spr[sprn] = val;
}
/*****************************************************************************/
/* Fixed point operations helpers */
#if defined(TARGET_PPC64)
static void add128 (uint64_t *plow, uint64_t *phigh, uint64_t a, uint64_t b)
{
*plow += a;
/* carry test */
if (*plow < a)
(*phigh)++;
*phigh += b;
}
static void neg128 (uint64_t *plow, uint64_t *phigh)
{
*plow = ~ *plow;
*phigh = ~ *phigh;
add128(plow, phigh, 1, 0);
}
static void mul64 (uint64_t *plow, uint64_t *phigh, uint64_t a, uint64_t b)
{
uint32_t a0, a1, b0, b1;
uint64_t v;
a0 = a;
a1 = a >> 32;
b0 = b;
b1 = b >> 32;
v = (uint64_t)a0 * (uint64_t)b0;
*plow = v;
*phigh = 0;
v = (uint64_t)a0 * (uint64_t)b1;
add128(plow, phigh, v << 32, v >> 32);
v = (uint64_t)a1 * (uint64_t)b0;
add128(plow, phigh, v << 32, v >> 32);
v = (uint64_t)a1 * (uint64_t)b1;
*phigh += v;
#if defined(DEBUG_MULDIV)
printf("mul: 0x%016llx * 0x%016llx = 0x%016llx%016llx\n",
a, b, *phigh, *plow);
#endif
}
void do_mul64 (uint64_t *plow, uint64_t *phigh)
{
mul64(plow, phigh, T0, T1);
}
static void imul64 (uint64_t *plow, uint64_t *phigh, int64_t a, int64_t b)
{
int sa, sb;
sa = (a < 0);
if (sa)
a = -a;
sb = (b < 0);
if (sb)
b = -b;
mul64(plow, phigh, a, b);
if (sa ^ sb) {
neg128(plow, phigh);
}
}
void do_imul64 (uint64_t *plow, uint64_t *phigh)
{
imul64(plow, phigh, T0, T1);
}
#endif
void do_adde (void)
{
T2 = T0;
T0 += T1 + xer_ca;
if (likely(!((uint32_t)T0 < (uint32_t)T2 ||
(xer_ca == 1 && (uint32_t)T0 == (uint32_t)T2)))) {
xer_ca = 0;
} else {
xer_ca = 1;
}
}
#if defined(TARGET_PPC64)
void do_adde_64 (void)
{
T2 = T0;
T0 += T1 + xer_ca;
if (likely(!((uint64_t)T0 < (uint64_t)T2 ||
(xer_ca == 1 && (uint64_t)T0 == (uint64_t)T2)))) {
xer_ca = 0;
} else {
xer_ca = 1;
}
}
#endif
void do_addmeo (void)
{
T1 = T0;
T0 += xer_ca + (-1);
if (likely(!((uint32_t)T1 &
((uint32_t)T1 ^ (uint32_t)T0) & (1UL << 31)))) {
xer_ov = 0;
} else {
xer_so = 1;
xer_ov = 1;
}
if (likely(T1 != 0))
xer_ca = 1;
}
#if defined(TARGET_PPC64)
void do_addmeo_64 (void)
{
T1 = T0;
T0 += xer_ca + (-1);
if (likely(!((uint64_t)T1 &
((uint64_t)T1 ^ (uint64_t)T0) & (1ULL << 63)))) {
xer_ov = 0;
} else {
xer_so = 1;
xer_ov = 1;
}
if (likely(T1 != 0))
xer_ca = 1;
}
#endif
void do_divwo (void)
{
if (likely(!(((int32_t)T0 == INT32_MIN && (int32_t)T1 == -1) ||
(int32_t)T1 == 0))) {
xer_ov = 0;
T0 = (int32_t)T0 / (int32_t)T1;
} else {
xer_so = 1;
xer_ov = 1;
T0 = (-1) * ((uint32_t)T0 >> 31);
}
}
#if defined(TARGET_PPC64)
void do_divdo (void)
{
if (likely(!(((int64_t)T0 == INT64_MIN && (int64_t)T1 == -1ULL) ||
(int64_t)T1 == 0))) {
xer_ov = 0;
T0 = (int64_t)T0 / (int64_t)T1;
} else {
xer_so = 1;
xer_ov = 1;
T0 = (-1ULL) * ((uint64_t)T0 >> 63);
}
}
#endif
void do_divwuo (void)
{
if (likely((uint32_t)T1 != 0)) {
xer_ov = 0;
T0 = (uint32_t)T0 / (uint32_t)T1;
} else {
xer_so = 1;
xer_ov = 1;
T0 = 0;
}
}
#if defined(TARGET_PPC64)
void do_divduo (void)
{
if (likely((uint64_t)T1 != 0)) {
xer_ov = 0;
T0 = (uint64_t)T0 / (uint64_t)T1;
} else {
xer_so = 1;
xer_ov = 1;
T0 = 0;
}
}
#endif
void do_mullwo (void)
{
int64_t res = (int64_t)T0 * (int64_t)T1;
if (likely((int32_t)res == res)) {
xer_ov = 0;
} else {
xer_ov = 1;
xer_so = 1;
}
T0 = (int32_t)res;
}
#if defined(TARGET_PPC64)
void do_mulldo (void)
{
int64_t th;
uint64_t tl;
do_imul64(&tl, &th);
if (likely(th == 0)) {
xer_ov = 0;
} else {
xer_ov = 1;
xer_so = 1;
}
T0 = (int64_t)tl;
}
#endif
void do_nego (void)
{
if (likely((int32_t)T0 != INT32_MIN)) {
xer_ov = 0;
T0 = -(int32_t)T0;
} else {
xer_ov = 1;
xer_so = 1;
}
}
#if defined(TARGET_PPC64)
void do_nego_64 (void)
{
if (likely((int64_t)T0 != INT64_MIN)) {
xer_ov = 0;
T0 = -(int64_t)T0;
} else {
xer_ov = 1;
xer_so = 1;
}
}
#endif
void do_subfe (void)
{
T0 = T1 + ~T0 + xer_ca;
if (likely((uint32_t)T0 >= (uint32_t)T1 &&
(xer_ca == 0 || (uint32_t)T0 != (uint32_t)T1))) {
xer_ca = 0;
} else {
xer_ca = 1;
}
}
#if defined(TARGET_PPC64)
void do_subfe_64 (void)
{
T0 = T1 + ~T0 + xer_ca;
if (likely((uint64_t)T0 >= (uint64_t)T1 &&
(xer_ca == 0 || (uint64_t)T0 != (uint64_t)T1))) {
xer_ca = 0;
} else {
xer_ca = 1;
}
}
#endif
void do_subfmeo (void)
{
T1 = T0;
T0 = ~T0 + xer_ca - 1;
if (likely(!((uint32_t)~T1 & ((uint32_t)~T1 ^ (uint32_t)T0) &
(1UL << 31)))) {
xer_ov = 0;
} else {
xer_so = 1;
xer_ov = 1;
}
if (likely((uint32_t)T1 != UINT32_MAX))
xer_ca = 1;
}
#if defined(TARGET_PPC64)
void do_subfmeo_64 (void)
{
T1 = T0;
T0 = ~T0 + xer_ca - 1;
if (likely(!((uint64_t)~T1 & ((uint64_t)~T1 ^ (uint64_t)T0) &
(1ULL << 63)))) {
xer_ov = 0;
} else {
xer_so = 1;
xer_ov = 1;
}
if (likely((uint64_t)T1 != UINT64_MAX))
xer_ca = 1;
}
#endif
void do_subfzeo (void)
{
T1 = T0;
T0 = ~T0 + xer_ca;
if (likely(!(((uint32_t)~T1 ^ UINT32_MAX) &
((uint32_t)(~T1) ^ (uint32_t)T0) & (1UL << 31)))) {
xer_ov = 0;
} else {
xer_ov = 1;
xer_so = 1;
}
if (likely((uint32_t)T0 >= (uint32_t)~T1)) {
xer_ca = 0;
} else {
xer_ca = 1;
}
}
#if defined(TARGET_PPC64)
void do_subfzeo_64 (void)
{
T1 = T0;
T0 = ~T0 + xer_ca;
if (likely(!(((uint64_t)~T1 ^ UINT64_MAX) &
((uint64_t)(~T1) ^ (uint64_t)T0) & (1ULL << 63)))) {
xer_ov = 0;
} else {
xer_ov = 1;
xer_so = 1;
}
if (likely((uint64_t)T0 >= (uint64_t)~T1)) {
xer_ca = 0;
} else {
xer_ca = 1;
}
}
#endif
/* shift right arithmetic helper */
void do_sraw (void)
{
int32_t ret;
if (likely(!(T1 & 0x20UL))) {
if (likely((uint32_t)T1 != 0)) {
ret = (int32_t)T0 >> (T1 & 0x1fUL);
if (likely(ret >= 0 || ((int32_t)T0 & ((1 << T1) - 1)) == 0)) {
xer_ca = 0;
} else {
xer_ca = 1;
}
} else {
ret = T0;
xer_ca = 0;
}
} else {
ret = (-1) * ((uint32_t)T0 >> 31);
if (likely(ret >= 0 || ((uint32_t)T0 & ~0x80000000UL) == 0)) {
xer_ca = 0;
} else {
xer_ca = 1;
}
}
T0 = ret;
}
#if defined(TARGET_PPC64)
void do_srad (void)
{
int64_t ret;
if (likely(!(T1 & 0x40UL))) {
if (likely((uint64_t)T1 != 0)) {
ret = (int64_t)T0 >> (T1 & 0x3FUL);
if (likely(ret >= 0 || ((int64_t)T0 & ((1 << T1) - 1)) == 0)) {
xer_ca = 0;
} else {
xer_ca = 1;
}
} else {
ret = T0;
xer_ca = 0;
}
} else {
ret = (-1) * ((uint64_t)T0 >> 63);
if (likely(ret >= 0 || ((uint64_t)T0 & ~0x8000000000000000ULL) == 0)) {
xer_ca = 0;
} else {
xer_ca = 1;
}
}
T0 = ret;
}
#endif
static inline int popcnt (uint32_t val)
{
int i;
for (i = 0; val != 0;)
val = val ^ (val - 1);
return i;
}
void do_popcntb (void)
{
uint32_t ret;
int i;
ret = 0;
for (i = 0; i < 32; i += 8)
ret |= popcnt((T0 >> i) & 0xFF) << i;
T0 = ret;
}
#if defined(TARGET_PPC64)
void do_popcntb_64 (void)
{
uint64_t ret;
int i;
ret = 0;
for (i = 0; i < 64; i += 8)
ret |= popcnt((T0 >> i) & 0xFF) << i;
T0 = ret;
}
#endif
/*****************************************************************************/
/* Floating point operations helpers */
void do_fctiw (void)
{
union {
double d;
uint64_t i;
} p;
p.i = float64_to_int32(FT0, &env->fp_status);
#if USE_PRECISE_EMULATION
/* XXX: higher bits are not supposed to be significant.
* to make tests easier, return the same as a real PowerPC 750 (aka G3)
*/
p.i |= 0xFFF80000ULL << 32;
#endif
FT0 = p.d;
}
void do_fctiwz (void)
{
union {
double d;
uint64_t i;
} p;
p.i = float64_to_int32_round_to_zero(FT0, &env->fp_status);
#if USE_PRECISE_EMULATION
/* XXX: higher bits are not supposed to be significant.
* to make tests easier, return the same as a real PowerPC 750 (aka G3)
*/
p.i |= 0xFFF80000ULL << 32;
#endif
FT0 = p.d;
}
#if defined(TARGET_PPC64)
void do_fcfid (void)
{
union {
double d;
uint64_t i;
} p;
p.d = FT0;
FT0 = int64_to_float64(p.i, &env->fp_status);
}
void do_fctid (void)
{
union {
double d;
uint64_t i;
} p;
p.i = float64_to_int64(FT0, &env->fp_status);
FT0 = p.d;
}
void do_fctidz (void)
{
union {
double d;
uint64_t i;
} p;
p.i = float64_to_int64_round_to_zero(FT0, &env->fp_status);
FT0 = p.d;
}
#endif
#if USE_PRECISE_EMULATION
void do_fmadd (void)
{
#ifdef FLOAT128
float128 ft0_128, ft1_128;
ft0_128 = float64_to_float128(FT0, &env->fp_status);
ft1_128 = float64_to_float128(FT1, &env->fp_status);
ft0_128 = float128_mul(ft0_128, ft1_128, &env->fp_status);
ft1_128 = float64_to_float128(FT2, &env->fp_status);
ft0_128 = float128_add(ft0_128, ft1_128, &env->fp_status);
FT0 = float128_to_float64(ft0_128, &env->fp_status);
#else
/* This is OK on x86 hosts */
FT0 = (FT0 * FT1) + FT2;
#endif
}
void do_fmsub (void)
{
#ifdef FLOAT128
float128 ft0_128, ft1_128;
ft0_128 = float64_to_float128(FT0, &env->fp_status);
ft1_128 = float64_to_float128(FT1, &env->fp_status);
ft0_128 = float128_mul(ft0_128, ft1_128, &env->fp_status);
ft1_128 = float64_to_float128(FT2, &env->fp_status);
ft0_128 = float128_sub(ft0_128, ft1_128, &env->fp_status);
FT0 = float128_to_float64(ft0_128, &env->fp_status);
#else
/* This is OK on x86 hosts */
FT0 = (FT0 * FT1) - FT2;
#endif
}
#endif /* USE_PRECISE_EMULATION */
void do_fnmadd (void)
{
#if USE_PRECISE_EMULATION
#ifdef FLOAT128
float128 ft0_128, ft1_128;
ft0_128 = float64_to_float128(FT0, &env->fp_status);
ft1_128 = float64_to_float128(FT1, &env->fp_status);
ft0_128 = float128_mul(ft0_128, ft1_128, &env->fp_status);
ft1_128 = float64_to_float128(FT2, &env->fp_status);
ft0_128 = float128_add(ft0_128, ft1_128, &env->fp_status);
FT0 = float128_to_float64(ft0_128, &env->fp_status);
#else
/* This is OK on x86 hosts */
FT0 = (FT0 * FT1) + FT2;
#endif
#else
FT0 = float64_mul(FT0, FT1, &env->fp_status);
FT0 = float64_add(FT0, FT2, &env->fp_status);
#endif
if (likely(!isnan(FT0)))
FT0 = float64_chs(FT0);
}
void do_fnmsub (void)
{
#if USE_PRECISE_EMULATION
#ifdef FLOAT128
float128 ft0_128, ft1_128;
ft0_128 = float64_to_float128(FT0, &env->fp_status);
ft1_128 = float64_to_float128(FT1, &env->fp_status);
ft0_128 = float128_mul(ft0_128, ft1_128, &env->fp_status);
ft1_128 = float64_to_float128(FT2, &env->fp_status);
ft0_128 = float128_sub(ft0_128, ft1_128, &env->fp_status);
FT0 = float128_to_float64(ft0_128, &env->fp_status);
#else
/* This is OK on x86 hosts */
FT0 = (FT0 * FT1) - FT2;
#endif
#else
FT0 = float64_mul(FT0, FT1, &env->fp_status);
FT0 = float64_sub(FT0, FT2, &env->fp_status);
#endif
if (likely(!isnan(FT0)))
FT0 = float64_chs(FT0);
}
void do_fsqrt (void)
{
FT0 = float64_sqrt(FT0, &env->fp_status);
}
void do_fres (void)
{
union {
double d;
uint64_t i;
} p;
if (likely(isnormal(FT0))) {
#if USE_PRECISE_EMULATION
FT0 = float64_div(1.0, FT0, &env->fp_status);
FT0 = float64_to_float32(FT0, &env->fp_status);
#else
FT0 = float32_div(1.0, FT0, &env->fp_status);
#endif
} else {
p.d = FT0;
if (p.i == 0x8000000000000000ULL) {
p.i = 0xFFF0000000000000ULL;
} else if (p.i == 0x0000000000000000ULL) {
p.i = 0x7FF0000000000000ULL;
} else if (isnan(FT0)) {
p.i = 0x7FF8000000000000ULL;
} else if (FT0 < 0.0) {
p.i = 0x8000000000000000ULL;
} else {
p.i = 0x0000000000000000ULL;
}
FT0 = p.d;
}
}
void do_frsqrte (void)
{
union {
double d;
uint64_t i;
} p;
if (likely(isnormal(FT0) && FT0 > 0.0)) {
FT0 = float64_sqrt(FT0, &env->fp_status);
FT0 = float32_div(1.0, FT0, &env->fp_status);
} else {
p.d = FT0;
if (p.i == 0x8000000000000000ULL) {
p.i = 0xFFF0000000000000ULL;
} else if (p.i == 0x0000000000000000ULL) {
p.i = 0x7FF0000000000000ULL;
} else if (isnan(FT0)) {
if (!(p.i & 0x0008000000000000ULL))
p.i |= 0x000FFFFFFFFFFFFFULL;
} else if (FT0 < 0) {
p.i = 0x7FF8000000000000ULL;
} else {
p.i = 0x0000000000000000ULL;
}
FT0 = p.d;
}
}
void do_fsel (void)
{
if (FT0 >= 0)
FT0 = FT1;
else
FT0 = FT2;
}
void do_fcmpu (void)
{
if (likely(!isnan(FT0) && !isnan(FT1))) {
if (float64_lt(FT0, FT1, &env->fp_status)) {
T0 = 0x08UL;
} else if (!float64_le(FT0, FT1, &env->fp_status)) {
T0 = 0x04UL;
} else {
T0 = 0x02UL;
}
} else {
T0 = 0x01UL;
env->fpscr[4] |= 0x1;
env->fpscr[6] |= 0x1;
}
env->fpscr[3] = T0;
}
void do_fcmpo (void)
{
env->fpscr[4] &= ~0x1;
if (likely(!isnan(FT0) && !isnan(FT1))) {
if (float64_lt(FT0, FT1, &env->fp_status)) {
T0 = 0x08UL;
} else if (!float64_le(FT0, FT1, &env->fp_status)) {
T0 = 0x04UL;
} else {
T0 = 0x02UL;
}
} else {
T0 = 0x01UL;
env->fpscr[4] |= 0x1;
if (!float64_is_signaling_nan(FT0) || !float64_is_signaling_nan(FT1)) {
/* Quiet NaN case */
env->fpscr[6] |= 0x1;
if (!(env->fpscr[1] & 0x8))
env->fpscr[4] |= 0x8;
} else {
env->fpscr[4] |= 0x8;
}
}
env->fpscr[3] = T0;
}
#if !defined (CONFIG_USER_ONLY)
void cpu_dump_rfi (target_ulong RA, target_ulong msr);
void do_rfi (void)
{
#if defined(TARGET_PPC64)
if (env->spr[SPR_SRR1] & (1ULL << MSR_SF)) {
env->nip = (uint64_t)(env->spr[SPR_SRR0] & ~0x00000003);
do_store_msr(env, (uint64_t)(env->spr[SPR_SRR1] & ~0xFFFF0000UL));
} else {
env->nip = (uint32_t)(env->spr[SPR_SRR0] & ~0x00000003);
ppc_store_msr_32(env, (uint32_t)(env->spr[SPR_SRR1] & ~0xFFFF0000UL));
}
#else
env->nip = (uint32_t)(env->spr[SPR_SRR0] & ~0x00000003);
do_store_msr(env, (uint32_t)(env->spr[SPR_SRR1] & ~0xFFFF0000UL));
#endif
#if defined (DEBUG_OP)
cpu_dump_rfi(env->nip, do_load_msr(env));
#endif
env->interrupt_request |= CPU_INTERRUPT_EXITTB;
}
#if defined(TARGET_PPC64)
void do_rfid (void)
{
if (env->spr[SPR_SRR1] & (1ULL << MSR_SF)) {
env->nip = (uint64_t)(env->spr[SPR_SRR0] & ~0x00000003);
do_store_msr(env, (uint64_t)(env->spr[SPR_SRR1] & ~0xFFFF0000UL));
} else {
env->nip = (uint32_t)(env->spr[SPR_SRR0] & ~0x00000003);
do_store_msr(env, (uint32_t)(env->spr[SPR_SRR1] & ~0xFFFF0000UL));
}
#if defined (DEBUG_OP)
cpu_dump_rfi(env->nip, do_load_msr(env));
#endif
env->interrupt_request |= CPU_INTERRUPT_EXITTB;
}
#endif
#endif
void do_tw (int flags)
{
if (!likely(!(((int32_t)T0 < (int32_t)T1 && (flags & 0x10)) ||
((int32_t)T0 > (int32_t)T1 && (flags & 0x08)) ||
((int32_t)T0 == (int32_t)T1 && (flags & 0x04)) ||
((uint32_t)T0 < (uint32_t)T1 && (flags & 0x02)) ||
((uint32_t)T0 > (uint32_t)T1 && (flags & 0x01))))) {
do_raise_exception_err(EXCP_PROGRAM, EXCP_TRAP);
}
}
#if defined(TARGET_PPC64)
void do_td (int flags)
{
if (!likely(!(((int64_t)T0 < (int64_t)T1 && (flags & 0x10)) ||
((int64_t)T0 > (int64_t)T1 && (flags & 0x08)) ||
((int64_t)T0 == (int64_t)T1 && (flags & 0x04)) ||
((uint64_t)T0 < (uint64_t)T1 && (flags & 0x02)) ||
((uint64_t)T0 > (uint64_t)T1 && (flags & 0x01)))))
do_raise_exception_err(EXCP_PROGRAM, EXCP_TRAP);
}
#endif
/*****************************************************************************/
/* PowerPC 601 specific instructions (POWER bridge) */
void do_POWER_abso (void)
{
if ((uint32_t)T0 == INT32_MIN) {
T0 = INT32_MAX;
xer_ov = 1;
xer_so = 1;
} else {
T0 = -T0;
xer_ov = 0;
}
}
void do_POWER_clcs (void)
{
switch (T0) {
case 0x0CUL:
/* Instruction cache line size */
T0 = ICACHE_LINE_SIZE;
break;
case 0x0DUL:
/* Data cache line size */
T0 = DCACHE_LINE_SIZE;
break;
case 0x0EUL:
/* Minimum cache line size */
T0 = ICACHE_LINE_SIZE < DCACHE_LINE_SIZE ?
ICACHE_LINE_SIZE : DCACHE_LINE_SIZE;
break;
case 0x0FUL:
/* Maximum cache line size */
T0 = ICACHE_LINE_SIZE > DCACHE_LINE_SIZE ?
ICACHE_LINE_SIZE : DCACHE_LINE_SIZE;
break;
default:
/* Undefined */
break;
}
}
void do_POWER_div (void)
{
uint64_t tmp;
if (((int32_t)T0 == INT32_MIN && (int32_t)T1 == -1) || (int32_t)T1 == 0) {
T0 = (long)((-1) * (T0 >> 31));
env->spr[SPR_MQ] = 0;
} else {
tmp = ((uint64_t)T0 << 32) | env->spr[SPR_MQ];
env->spr[SPR_MQ] = tmp % T1;
T0 = tmp / (int32_t)T1;
}
}
void do_POWER_divo (void)
{
int64_t tmp;
if (((int32_t)T0 == INT32_MIN && (int32_t)T1 == -1) || (int32_t)T1 == 0) {
T0 = (long)((-1) * (T0 >> 31));
env->spr[SPR_MQ] = 0;
xer_ov = 1;
xer_so = 1;
} else {
tmp = ((uint64_t)T0 << 32) | env->spr[SPR_MQ];
env->spr[SPR_MQ] = tmp % T1;
tmp /= (int32_t)T1;
if (tmp > (int64_t)INT32_MAX || tmp < (int64_t)INT32_MIN) {
xer_ov = 1;
xer_so = 1;
} else {
xer_ov = 0;
}
T0 = tmp;
}
}
void do_POWER_divs (void)
{
if (((int32_t)T0 == INT32_MIN && (int32_t)T1 == -1) || (int32_t)T1 == 0) {
T0 = (long)((-1) * (T0 >> 31));
env->spr[SPR_MQ] = 0;
} else {
env->spr[SPR_MQ] = T0 % T1;
T0 = (int32_t)T0 / (int32_t)T1;
}
}
void do_POWER_divso (void)
{
if (((int32_t)T0 == INT32_MIN && (int32_t)T1 == -1) || (int32_t)T1 == 0) {
T0 = (long)((-1) * (T0 >> 31));
env->spr[SPR_MQ] = 0;
xer_ov = 1;
xer_so = 1;
} else {
T0 = (int32_t)T0 / (int32_t)T1;
env->spr[SPR_MQ] = (int32_t)T0 % (int32_t)T1;
xer_ov = 0;
}
}
void do_POWER_dozo (void)
{
if ((int32_t)T1 > (int32_t)T0) {
T2 = T0;
T0 = T1 - T0;
if (((uint32_t)(~T2) ^ (uint32_t)T1 ^ UINT32_MAX) &
((uint32_t)(~T2) ^ (uint32_t)T0) & (1UL << 31)) {
xer_so = 1;
xer_ov = 1;
} else {
xer_ov = 0;
}
} else {
T0 = 0;
xer_ov = 0;
}
}
void do_POWER_maskg (void)
{
uint32_t ret;
if ((uint32_t)T0 == (uint32_t)(T1 + 1)) {
ret = -1;
} else {
ret = (((uint32_t)(-1)) >> ((uint32_t)T0)) ^
(((uint32_t)(-1) >> ((uint32_t)T1)) >> 1);
if ((uint32_t)T0 > (uint32_t)T1)
ret = ~ret;
}
T0 = ret;
}
void do_POWER_mulo (void)
{
uint64_t tmp;
tmp = (uint64_t)T0 * (uint64_t)T1;
env->spr[SPR_MQ] = tmp >> 32;
T0 = tmp;
if (tmp >> 32 != ((uint64_t)T0 >> 16) * ((uint64_t)T1 >> 16)) {
xer_ov = 1;
xer_so = 1;
} else {
xer_ov = 0;
}
}
#if !defined (CONFIG_USER_ONLY)
void do_POWER_rac (void)
{
#if 0
mmu_ctx_t ctx;
/* We don't have to generate many instances of this instruction,
* as rac is supervisor only.
*/
if (get_physical_address(env, &ctx, T0, 0, ACCESS_INT, 1) == 0)
T0 = ctx.raddr;
#endif
}
void do_POWER_rfsvc (void)
{
env->nip = env->lr & ~0x00000003UL;
T0 = env->ctr & 0x0000FFFFUL;
do_store_msr(env, T0);
#if defined (DEBUG_OP)
cpu_dump_rfi(env->nip, do_load_msr(env));
#endif
env->interrupt_request |= CPU_INTERRUPT_EXITTB;
}
/* PowerPC 601 BAT management helper */
void do_store_601_batu (int nr)
{
do_store_ibatu(env, nr, (uint32_t)T0);
env->DBAT[0][nr] = env->IBAT[0][nr];
env->DBAT[1][nr] = env->IBAT[1][nr];
}
#endif
/*****************************************************************************/
/* 602 specific instructions */
/* mfrom is the most crazy instruction ever seen, imho ! */
/* Real implementation uses a ROM table. Do the same */
#define USE_MFROM_ROM_TABLE
void do_op_602_mfrom (void)
{
if (likely(T0 < 602)) {
#if defined(USE_MFROM_ROM_TABLE)
#include "mfrom_table.c"
T0 = mfrom_ROM_table[T0];
#else
double d;
/* Extremly decomposed:
* -T0 / 256
* T0 = 256 * log10(10 + 1.0) + 0.5
*/
d = T0;
d = float64_div(d, 256, &env->fp_status);
d = float64_chs(d);
d = exp10(d); // XXX: use float emulation function
d = float64_add(d, 1.0, &env->fp_status);
d = log10(d); // XXX: use float emulation function
d = float64_mul(d, 256, &env->fp_status);
d = float64_add(d, 0.5, &env->fp_status);
T0 = float64_round_to_int(d, &env->fp_status);
#endif
} else {
T0 = 0;
}
}
/*****************************************************************************/
/* Embedded PowerPC specific helpers */
void do_405_check_ov (void)
{
if (likely((((uint32_t)T1 ^ (uint32_t)T2) >> 31) ||
!(((uint32_t)T0 ^ (uint32_t)T2) >> 31))) {
xer_ov = 0;
} else {
xer_ov = 1;
xer_so = 1;
}
}
void do_405_check_sat (void)
{
if (!likely((((uint32_t)T1 ^ (uint32_t)T2) >> 31) ||
!(((uint32_t)T0 ^ (uint32_t)T2) >> 31))) {
/* Saturate result */
if (T2 >> 31) {
T0 = INT32_MIN;
} else {
T0 = INT32_MAX;
}
}
}
#if !defined(CONFIG_USER_ONLY)
void do_40x_rfci (void)
{
env->nip = env->spr[SPR_40x_SRR2];
do_store_msr(env, env->spr[SPR_40x_SRR3] & ~0xFFFF0000);
#if defined (DEBUG_OP)
cpu_dump_rfi(env->nip, do_load_msr(env));
#endif
env->interrupt_request = CPU_INTERRUPT_EXITTB;
}
void do_rfci (void)
{
#if defined(TARGET_PPC64)
if (env->spr[SPR_BOOKE_CSRR1] & (1 << MSR_CM)) {
env->nip = (uint64_t)env->spr[SPR_BOOKE_CSRR0];
} else
#endif
{
env->nip = (uint32_t)env->spr[SPR_BOOKE_CSRR0];
}
do_store_msr(env, (uint32_t)env->spr[SPR_BOOKE_CSRR1] & ~0x3FFF0000);
#if defined (DEBUG_OP)
cpu_dump_rfi(env->nip, do_load_msr(env));
#endif
env->interrupt_request = CPU_INTERRUPT_EXITTB;
}
void do_rfdi (void)
{
#if defined(TARGET_PPC64)
if (env->spr[SPR_BOOKE_DSRR1] & (1 << MSR_CM)) {
env->nip = (uint64_t)env->spr[SPR_BOOKE_DSRR0];
} else
#endif
{
env->nip = (uint32_t)env->spr[SPR_BOOKE_DSRR0];
}
do_store_msr(env, (uint32_t)env->spr[SPR_BOOKE_DSRR1] & ~0x3FFF0000);
#if defined (DEBUG_OP)
cpu_dump_rfi(env->nip, do_load_msr(env));
#endif
env->interrupt_request = CPU_INTERRUPT_EXITTB;
}
void do_rfmci (void)
{
#if defined(TARGET_PPC64)
if (env->spr[SPR_BOOKE_MCSRR1] & (1 << MSR_CM)) {
env->nip = (uint64_t)env->spr[SPR_BOOKE_MCSRR0];
} else
#endif
{
env->nip = (uint32_t)env->spr[SPR_BOOKE_MCSRR0];
}
do_store_msr(env, (uint32_t)env->spr[SPR_BOOKE_MCSRR1] & ~0x3FFF0000);
#if defined (DEBUG_OP)
cpu_dump_rfi(env->nip, do_load_msr(env));
#endif
env->interrupt_request = CPU_INTERRUPT_EXITTB;
}
void do_load_dcr (void)
{
target_ulong val;
if (unlikely(env->dcr_env == NULL)) {
if (loglevel != 0) {
fprintf(logfile, "No DCR environment\n");
}
do_raise_exception_err(EXCP_PROGRAM, EXCP_INVAL | EXCP_INVAL_INVAL);
} else if (unlikely(ppc_dcr_read(env->dcr_env, T0, &val) != 0)) {
if (loglevel != 0) {
fprintf(logfile, "DCR read error %d %03x\n", (int)T0, (int)T0);
}
do_raise_exception_err(EXCP_PROGRAM, EXCP_INVAL | EXCP_PRIV_REG);
} else {
T0 = val;
}
}
void do_store_dcr (void)
{
if (unlikely(env->dcr_env == NULL)) {
if (loglevel != 0) {
fprintf(logfile, "No DCR environment\n");
}
do_raise_exception_err(EXCP_PROGRAM, EXCP_INVAL | EXCP_INVAL_INVAL);
} else if (unlikely(ppc_dcr_write(env->dcr_env, T0, T1) != 0)) {
if (loglevel != 0) {
fprintf(logfile, "DCR write error %d %03x\n", (int)T0, (int)T0);
}
do_raise_exception_err(EXCP_PROGRAM, EXCP_INVAL | EXCP_PRIV_REG);
}
}
void do_load_403_pb (int num)
{
T0 = env->pb[num];
}
void do_store_403_pb (int num)
{
if (likely(env->pb[num] != T0)) {
env->pb[num] = T0;
/* Should be optimized */
tlb_flush(env, 1);
}
}
#endif
/* 440 specific */
void do_440_dlmzb (void)
{
target_ulong mask;
int i;
i = 1;
for (mask = 0xFF000000; mask != 0; mask = mask >> 8) {
if ((T0 & mask) == 0)
goto done;
i++;
}
for (mask = 0xFF000000; mask != 0; mask = mask >> 8) {
if ((T1 & mask) == 0)
break;
i++;
}
done:
T0 = i;
}
#if defined(TARGET_PPCEMB)
/* SPE extension helpers */
/* Use a table to make this quicker */
static uint8_t hbrev[16] = {
0x0, 0x8, 0x4, 0xC, 0x2, 0xA, 0x6, 0xE,
0x1, 0x9, 0x5, 0xD, 0x3, 0xB, 0x7, 0xF,
};
static inline uint8_t byte_reverse (uint8_t val)
{
return hbrev[val >> 4] | (hbrev[val & 0xF] << 4);
}
static inline uint32_t word_reverse (uint32_t val)
{
return byte_reverse(val >> 24) | (byte_reverse(val >> 16) << 8) |
(byte_reverse(val >> 8) << 16) | (byte_reverse(val) << 24);
}
#define MASKBITS 16 // Random value - to be fixed
void do_brinc (void)
{
uint32_t a, b, d, mask;
mask = (uint32_t)(-1UL) >> MASKBITS;
b = T1_64 & mask;
a = T0_64 & mask;
d = word_reverse(1 + word_reverse(a | ~mask));
T0_64 = (T0_64 & ~mask) | (d & mask);
}
#define DO_SPE_OP2(name) \
void do_ev##name (void) \
{ \
T0_64 = ((uint64_t)_do_e##name(T0_64 >> 32, T1_64 >> 32) << 32) | \
(uint64_t)_do_e##name(T0_64, T1_64); \
}
#define DO_SPE_OP1(name) \
void do_ev##name (void) \
{ \
T0_64 = ((uint64_t)_do_e##name(T0_64 >> 32) << 32) | \
(uint64_t)_do_e##name(T0_64); \
}
/* Fixed-point vector arithmetic */
static inline uint32_t _do_eabs (uint32_t val)
{
if (val != 0x80000000)
val &= ~0x80000000;
return val;
}
static inline uint32_t _do_eaddw (uint32_t op1, uint32_t op2)
{
return op1 + op2;
}
static inline int _do_ecntlsw (uint32_t val)
{
if (val & 0x80000000)
return _do_cntlzw(~val);
else
return _do_cntlzw(val);
}
static inline int _do_ecntlzw (uint32_t val)
{
return _do_cntlzw(val);
}
static inline uint32_t _do_eneg (uint32_t val)
{
if (val != 0x80000000)
val ^= 0x80000000;
return val;
}
static inline uint32_t _do_erlw (uint32_t op1, uint32_t op2)
{
return rotl32(op1, op2);
}
static inline uint32_t _do_erndw (uint32_t val)
{
return (val + 0x000080000000) & 0xFFFF0000;
}
static inline uint32_t _do_eslw (uint32_t op1, uint32_t op2)
{
/* No error here: 6 bits are used */
return op1 << (op2 & 0x3F);
}
static inline int32_t _do_esrws (int32_t op1, uint32_t op2)
{
/* No error here: 6 bits are used */
return op1 >> (op2 & 0x3F);
}
static inline uint32_t _do_esrwu (uint32_t op1, uint32_t op2)
{
/* No error here: 6 bits are used */
return op1 >> (op2 & 0x3F);
}
static inline uint32_t _do_esubfw (uint32_t op1, uint32_t op2)
{
return op2 - op1;
}
/* evabs */
DO_SPE_OP1(abs);
/* evaddw */
DO_SPE_OP2(addw);
/* evcntlsw */
DO_SPE_OP1(cntlsw);
/* evcntlzw */
DO_SPE_OP1(cntlzw);
/* evneg */
DO_SPE_OP1(neg);
/* evrlw */
DO_SPE_OP2(rlw);
/* evrnd */
DO_SPE_OP1(rndw);
/* evslw */
DO_SPE_OP2(slw);
/* evsrws */
DO_SPE_OP2(srws);
/* evsrwu */
DO_SPE_OP2(srwu);
/* evsubfw */
DO_SPE_OP2(subfw);
/* evsel is a little bit more complicated... */
static inline uint32_t _do_esel (uint32_t op1, uint32_t op2, int n)
{
if (n)
return op1;
else
return op2;
}
void do_evsel (void)
{
T0_64 = ((uint64_t)_do_esel(T0_64 >> 32, T1_64 >> 32, T0 >> 3) << 32) |
(uint64_t)_do_esel(T0_64, T1_64, (T0 >> 2) & 1);
}
/* Fixed-point vector comparisons */
#define DO_SPE_CMP(name) \
void do_ev##name (void) \
{ \
T0 = _do_evcmp_merge((uint64_t)_do_e##name(T0_64 >> 32, \
T1_64 >> 32) << 32, \
_do_e##name(T0_64, T1_64)); \
}
static inline uint32_t _do_evcmp_merge (int t0, int t1)
{
return (t0 << 3) | (t1 << 2) | ((t0 | t1) << 1) | (t0 & t1);
}
static inline int _do_ecmpeq (uint32_t op1, uint32_t op2)
{
return op1 == op2 ? 1 : 0;
}
static inline int _do_ecmpgts (int32_t op1, int32_t op2)
{
return op1 > op2 ? 1 : 0;
}
static inline int _do_ecmpgtu (uint32_t op1, uint32_t op2)
{
return op1 > op2 ? 1 : 0;
}
static inline int _do_ecmplts (int32_t op1, int32_t op2)
{
return op1 < op2 ? 1 : 0;
}
static inline int _do_ecmpltu (uint32_t op1, uint32_t op2)
{
return op1 < op2 ? 1 : 0;
}
/* evcmpeq */
DO_SPE_CMP(cmpeq);
/* evcmpgts */
DO_SPE_CMP(cmpgts);
/* evcmpgtu */
DO_SPE_CMP(cmpgtu);
/* evcmplts */
DO_SPE_CMP(cmplts);
/* evcmpltu */
DO_SPE_CMP(cmpltu);
/* Single precision floating-point conversions from/to integer */
static inline uint32_t _do_efscfsi (int32_t val)
{
union {
uint32_t u;
float32 f;
} u;
u.f = int32_to_float32(val, &env->spe_status);
return u.u;
}
static inline uint32_t _do_efscfui (uint32_t val)
{
union {
uint32_t u;
float32 f;
} u;
u.f = uint32_to_float32(val, &env->spe_status);
return u.u;
}
static inline int32_t _do_efsctsi (uint32_t val)
{
union {
int32_t u;
float32 f;
} u;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
return float32_to_int32(u.f, &env->spe_status);
}
static inline uint32_t _do_efsctui (uint32_t val)
{
union {
int32_t u;
float32 f;
} u;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
return float32_to_uint32(u.f, &env->spe_status);
}
static inline int32_t _do_efsctsiz (uint32_t val)
{
union {
int32_t u;
float32 f;
} u;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
return float32_to_int32_round_to_zero(u.f, &env->spe_status);
}
static inline uint32_t _do_efsctuiz (uint32_t val)
{
union {
int32_t u;
float32 f;
} u;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
return float32_to_uint32_round_to_zero(u.f, &env->spe_status);
}
void do_efscfsi (void)
{
T0_64 = _do_efscfsi(T0_64);
}
void do_efscfui (void)
{
T0_64 = _do_efscfui(T0_64);
}
void do_efsctsi (void)
{
T0_64 = _do_efsctsi(T0_64);
}
void do_efsctui (void)
{
T0_64 = _do_efsctui(T0_64);
}
void do_efsctsiz (void)
{
T0_64 = _do_efsctsiz(T0_64);
}
void do_efsctuiz (void)
{
T0_64 = _do_efsctuiz(T0_64);
}
/* Single precision floating-point conversion to/from fractional */
static inline uint32_t _do_efscfsf (uint32_t val)
{
union {
uint32_t u;
float32 f;
} u;
float32 tmp;
u.f = int32_to_float32(val, &env->spe_status);
tmp = int64_to_float32(1ULL << 32, &env->spe_status);
u.f = float32_div(u.f, tmp, &env->spe_status);
return u.u;
}
static inline uint32_t _do_efscfuf (uint32_t val)
{
union {
uint32_t u;
float32 f;
} u;
float32 tmp;
u.f = uint32_to_float32(val, &env->spe_status);
tmp = uint64_to_float32(1ULL << 32, &env->spe_status);
u.f = float32_div(u.f, tmp, &env->spe_status);
return u.u;
}
static inline int32_t _do_efsctsf (uint32_t val)
{
union {
int32_t u;
float32 f;
} u;
float32 tmp;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
tmp = uint64_to_float32(1ULL << 32, &env->spe_status);
u.f = float32_mul(u.f, tmp, &env->spe_status);
return float32_to_int32(u.f, &env->spe_status);
}
static inline uint32_t _do_efsctuf (uint32_t val)
{
union {
int32_t u;
float32 f;
} u;
float32 tmp;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
tmp = uint64_to_float32(1ULL << 32, &env->spe_status);
u.f = float32_mul(u.f, tmp, &env->spe_status);
return float32_to_uint32(u.f, &env->spe_status);
}
static inline int32_t _do_efsctsfz (uint32_t val)
{
union {
int32_t u;
float32 f;
} u;
float32 tmp;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
tmp = uint64_to_float32(1ULL << 32, &env->spe_status);
u.f = float32_mul(u.f, tmp, &env->spe_status);
return float32_to_int32_round_to_zero(u.f, &env->spe_status);
}
static inline uint32_t _do_efsctufz (uint32_t val)
{
union {
int32_t u;
float32 f;
} u;
float32 tmp;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
tmp = uint64_to_float32(1ULL << 32, &env->spe_status);
u.f = float32_mul(u.f, tmp, &env->spe_status);
return float32_to_uint32_round_to_zero(u.f, &env->spe_status);
}
void do_efscfsf (void)
{
T0_64 = _do_efscfsf(T0_64);
}
void do_efscfuf (void)
{
T0_64 = _do_efscfuf(T0_64);
}
void do_efsctsf (void)
{
T0_64 = _do_efsctsf(T0_64);
}
void do_efsctuf (void)
{
T0_64 = _do_efsctuf(T0_64);
}
void do_efsctsfz (void)
{
T0_64 = _do_efsctsfz(T0_64);
}
void do_efsctufz (void)
{
T0_64 = _do_efsctufz(T0_64);
}
/* Double precision floating point helpers */
static inline int _do_efdcmplt (uint64_t op1, uint64_t op2)
{
/* XXX: TODO: test special values (NaN, infinites, ...) */
return _do_efdtstlt(op1, op2);
}
static inline int _do_efdcmpgt (uint64_t op1, uint64_t op2)
{
/* XXX: TODO: test special values (NaN, infinites, ...) */
return _do_efdtstgt(op1, op2);
}
static inline int _do_efdcmpeq (uint64_t op1, uint64_t op2)
{
/* XXX: TODO: test special values (NaN, infinites, ...) */
return _do_efdtsteq(op1, op2);
}
void do_efdcmplt (void)
{
T0 = _do_efdcmplt(T0_64, T1_64);
}
void do_efdcmpgt (void)
{
T0 = _do_efdcmpgt(T0_64, T1_64);
}
void do_efdcmpeq (void)
{
T0 = _do_efdcmpeq(T0_64, T1_64);
}
/* Double precision floating-point conversion to/from integer */
static inline uint64_t _do_efdcfsi (int64_t val)
{
union {
uint64_t u;
float64 f;
} u;
u.f = int64_to_float64(val, &env->spe_status);
return u.u;
}
static inline uint64_t _do_efdcfui (uint64_t val)
{
union {
uint64_t u;
float64 f;
} u;
u.f = uint64_to_float64(val, &env->spe_status);
return u.u;
}
static inline int64_t _do_efdctsi (uint64_t val)
{
union {
int64_t u;
float64 f;
} u;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
return float64_to_int64(u.f, &env->spe_status);
}
static inline uint64_t _do_efdctui (uint64_t val)
{
union {
int64_t u;
float64 f;
} u;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
return float64_to_uint64(u.f, &env->spe_status);
}
static inline int64_t _do_efdctsiz (uint64_t val)
{
union {
int64_t u;
float64 f;
} u;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
return float64_to_int64_round_to_zero(u.f, &env->spe_status);
}
static inline uint64_t _do_efdctuiz (uint64_t val)
{
union {
int64_t u;
float64 f;
} u;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
return float64_to_uint64_round_to_zero(u.f, &env->spe_status);
}
void do_efdcfsi (void)
{
T0_64 = _do_efdcfsi(T0_64);
}
void do_efdcfui (void)
{
T0_64 = _do_efdcfui(T0_64);
}
void do_efdctsi (void)
{
T0_64 = _do_efdctsi(T0_64);
}
void do_efdctui (void)
{
T0_64 = _do_efdctui(T0_64);
}
void do_efdctsiz (void)
{
T0_64 = _do_efdctsiz(T0_64);
}
void do_efdctuiz (void)
{
T0_64 = _do_efdctuiz(T0_64);
}
/* Double precision floating-point conversion to/from fractional */
static inline uint64_t _do_efdcfsf (int64_t val)
{
union {
uint64_t u;
float64 f;
} u;
float64 tmp;
u.f = int32_to_float64(val, &env->spe_status);
tmp = int64_to_float64(1ULL << 32, &env->spe_status);
u.f = float64_div(u.f, tmp, &env->spe_status);
return u.u;
}
static inline uint64_t _do_efdcfuf (uint64_t val)
{
union {
uint64_t u;
float64 f;
} u;
float64 tmp;
u.f = uint32_to_float64(val, &env->spe_status);
tmp = int64_to_float64(1ULL << 32, &env->spe_status);
u.f = float64_div(u.f, tmp, &env->spe_status);
return u.u;
}
static inline int64_t _do_efdctsf (uint64_t val)
{
union {
int64_t u;
float64 f;
} u;
float64 tmp;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
tmp = uint64_to_float64(1ULL << 32, &env->spe_status);
u.f = float64_mul(u.f, tmp, &env->spe_status);
return float64_to_int32(u.f, &env->spe_status);
}
static inline uint64_t _do_efdctuf (uint64_t val)
{
union {
int64_t u;
float64 f;
} u;
float64 tmp;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
tmp = uint64_to_float64(1ULL << 32, &env->spe_status);
u.f = float64_mul(u.f, tmp, &env->spe_status);
return float64_to_uint32(u.f, &env->spe_status);
}
static inline int64_t _do_efdctsfz (uint64_t val)
{
union {
int64_t u;
float64 f;
} u;
float64 tmp;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
tmp = uint64_to_float64(1ULL << 32, &env->spe_status);
u.f = float64_mul(u.f, tmp, &env->spe_status);
return float64_to_int32_round_to_zero(u.f, &env->spe_status);
}
static inline uint64_t _do_efdctufz (uint64_t val)
{
union {
int64_t u;
float64 f;
} u;
float64 tmp;
u.u = val;
/* NaN are not treated the same way IEEE 754 does */
if (unlikely(isnan(u.f)))
return 0;
tmp = uint64_to_float64(1ULL << 32, &env->spe_status);
u.f = float64_mul(u.f, tmp, &env->spe_status);
return float64_to_uint32_round_to_zero(u.f, &env->spe_status);
}
void do_efdcfsf (void)
{
T0_64 = _do_efdcfsf(T0_64);
}
void do_efdcfuf (void)
{
T0_64 = _do_efdcfuf(T0_64);
}
void do_efdctsf (void)
{
T0_64 = _do_efdctsf(T0_64);
}
void do_efdctuf (void)
{
T0_64 = _do_efdctuf(T0_64);
}
void do_efdctsfz (void)
{
T0_64 = _do_efdctsfz(T0_64);
}
void do_efdctufz (void)
{
T0_64 = _do_efdctufz(T0_64);
}
/* Floating point conversion between single and double precision */
static inline uint32_t _do_efscfd (uint64_t val)
{
union {
uint64_t u;
float64 f;
} u1;
union {
uint32_t u;
float32 f;
} u2;
u1.u = val;
u2.f = float64_to_float32(u1.f, &env->spe_status);
return u2.u;
}
static inline uint64_t _do_efdcfs (uint32_t val)
{
union {
uint64_t u;
float64 f;
} u2;
union {
uint32_t u;
float32 f;
} u1;
u1.u = val;
u2.f = float32_to_float64(u1.f, &env->spe_status);
return u2.u;
}
void do_efscfd (void)
{
T0_64 = _do_efscfd(T0_64);
}
void do_efdcfs (void)
{
T0_64 = _do_efdcfs(T0_64);
}
/* Single precision fixed-point vector arithmetic */
/* evfsabs */
DO_SPE_OP1(fsabs);
/* evfsnabs */
DO_SPE_OP1(fsnabs);
/* evfsneg */
DO_SPE_OP1(fsneg);
/* evfsadd */
DO_SPE_OP2(fsadd);
/* evfssub */
DO_SPE_OP2(fssub);
/* evfsmul */
DO_SPE_OP2(fsmul);
/* evfsdiv */
DO_SPE_OP2(fsdiv);
/* Single-precision floating-point comparisons */
static inline int _do_efscmplt (uint32_t op1, uint32_t op2)
{
/* XXX: TODO: test special values (NaN, infinites, ...) */
return _do_efststlt(op1, op2);
}
static inline int _do_efscmpgt (uint32_t op1, uint32_t op2)
{
/* XXX: TODO: test special values (NaN, infinites, ...) */
return _do_efststgt(op1, op2);
}
static inline int _do_efscmpeq (uint32_t op1, uint32_t op2)
{
/* XXX: TODO: test special values (NaN, infinites, ...) */
return _do_efststeq(op1, op2);
}
void do_efscmplt (void)
{
T0 = _do_efscmplt(T0_64, T1_64);
}
void do_efscmpgt (void)
{
T0 = _do_efscmpgt(T0_64, T1_64);
}
void do_efscmpeq (void)
{
T0 = _do_efscmpeq(T0_64, T1_64);
}
/* Single-precision floating-point vector comparisons */
/* evfscmplt */
DO_SPE_CMP(fscmplt);
/* evfscmpgt */
DO_SPE_CMP(fscmpgt);
/* evfscmpeq */
DO_SPE_CMP(fscmpeq);
/* evfststlt */
DO_SPE_CMP(fststlt);
/* evfststgt */
DO_SPE_CMP(fststgt);
/* evfststeq */
DO_SPE_CMP(fststeq);
/* Single-precision floating-point vector conversions */
/* evfscfsi */
DO_SPE_OP1(fscfsi);
/* evfscfui */
DO_SPE_OP1(fscfui);
/* evfscfuf */
DO_SPE_OP1(fscfuf);
/* evfscfsf */
DO_SPE_OP1(fscfsf);
/* evfsctsi */
DO_SPE_OP1(fsctsi);
/* evfsctui */
DO_SPE_OP1(fsctui);
/* evfsctsiz */
DO_SPE_OP1(fsctsiz);
/* evfsctuiz */
DO_SPE_OP1(fsctuiz);
/* evfsctsf */
DO_SPE_OP1(fsctsf);
/* evfsctuf */
DO_SPE_OP1(fsctuf);
#endif /* defined(TARGET_PPCEMB) */
/*****************************************************************************/
/* Softmmu support */
#if !defined (CONFIG_USER_ONLY)
#define MMUSUFFIX _mmu
#define GETPC() (__builtin_return_address(0))
#define SHIFT 0
#include "softmmu_template.h"
#define SHIFT 1
#include "softmmu_template.h"
#define SHIFT 2
#include "softmmu_template.h"
#define SHIFT 3
#include "softmmu_template.h"
/* try to fill the TLB and return an exception if error. If retaddr is
NULL, it means that the function was called in C code (i.e. not
from generated code or from helper.c) */
/* XXX: fix it to restore all registers */
void tlb_fill (target_ulong addr, int is_write, int is_user, void *retaddr)
{
TranslationBlock *tb;
CPUState *saved_env;
target_phys_addr_t pc;
int ret;
/* XXX: hack to restore env in all cases, even if not called from
generated code */
saved_env = env;
env = cpu_single_env;
ret = cpu_ppc_handle_mmu_fault(env, addr, is_write, is_user, 1);
if (unlikely(ret != 0)) {
if (likely(retaddr)) {
/* now we have a real cpu fault */
pc = (target_phys_addr_t)retaddr;
tb = tb_find_pc(pc);
if (likely(tb)) {
/* the PC is inside the translated code. It means that we have
a virtual CPU fault */
cpu_restore_state(tb, env, pc, NULL);
}
}
do_raise_exception_err(env->exception_index, env->error_code);
}
env = saved_env;
}
/* TLB invalidation helpers */
void do_tlbia (void)
{
ppc_tlb_invalidate_all(env);
}
void do_tlbie (void)
{
T0 = (uint32_t)T0;
#if !defined(FLUSH_ALL_TLBS)
if (unlikely(PPC_MMU(env) == PPC_FLAGS_MMU_SOFT_6xx)) {
ppc6xx_tlb_invalidate_virt(env, T0 & TARGET_PAGE_MASK, 0);
if (env->id_tlbs == 1)
ppc6xx_tlb_invalidate_virt(env, T0 & TARGET_PAGE_MASK, 1);
} else if (unlikely(PPC_MMU(env) == PPC_FLAGS_MMU_SOFT_4xx)) {
/* XXX: TODO */
#if 0
ppcbooke_tlb_invalidate_virt(env, T0 & TARGET_PAGE_MASK,
env->spr[SPR_BOOKE_PID]);
#endif
} else {
/* tlbie invalidate TLBs for all segments */
T0 &= TARGET_PAGE_MASK;
T0 &= ~((target_ulong)-1 << 28);
/* XXX: this case should be optimized,
* giving a mask to tlb_flush_page
*/
tlb_flush_page(env, T0 | (0x0 << 28));
tlb_flush_page(env, T0 | (0x1 << 28));
tlb_flush_page(env, T0 | (0x2 << 28));
tlb_flush_page(env, T0 | (0x3 << 28));
tlb_flush_page(env, T0 | (0x4 << 28));
tlb_flush_page(env, T0 | (0x5 << 28));
tlb_flush_page(env, T0 | (0x6 << 28));
tlb_flush_page(env, T0 | (0x7 << 28));
tlb_flush_page(env, T0 | (0x8 << 28));
tlb_flush_page(env, T0 | (0x9 << 28));
tlb_flush_page(env, T0 | (0xA << 28));
tlb_flush_page(env, T0 | (0xB << 28));
tlb_flush_page(env, T0 | (0xC << 28));
tlb_flush_page(env, T0 | (0xD << 28));
tlb_flush_page(env, T0 | (0xE << 28));
tlb_flush_page(env, T0 | (0xF << 28));
}
#else
do_tlbia();
#endif
}
#if defined(TARGET_PPC64)
void do_tlbie_64 (void)
{
T0 = (uint64_t)T0;
#if !defined(FLUSH_ALL_TLBS)
if (unlikely(PPC_MMU(env) == PPC_FLAGS_MMU_SOFT_6xx)) {
ppc6xx_tlb_invalidate_virt(env, T0 & TARGET_PAGE_MASK, 0);
if (env->id_tlbs == 1)
ppc6xx_tlb_invalidate_virt(env, T0 & TARGET_PAGE_MASK, 1);
} else if (unlikely(PPC_MMU(env) == PPC_FLAGS_MMU_SOFT_4xx)) {
/* XXX: TODO */
#if 0
ppcbooke_tlb_invalidate_virt(env, T0 & TARGET_PAGE_MASK,
env->spr[SPR_BOOKE_PID]);
#endif
} else {
/* tlbie invalidate TLBs for all segments
* As we have 2^36 segments, invalidate all qemu TLBs
*/
#if 0
T0 &= TARGET_PAGE_MASK;
T0 &= ~((target_ulong)-1 << 28);
/* XXX: this case should be optimized,
* giving a mask to tlb_flush_page
*/
tlb_flush_page(env, T0 | (0x0 << 28));
tlb_flush_page(env, T0 | (0x1 << 28));
tlb_flush_page(env, T0 | (0x2 << 28));
tlb_flush_page(env, T0 | (0x3 << 28));
tlb_flush_page(env, T0 | (0x4 << 28));
tlb_flush_page(env, T0 | (0x5 << 28));
tlb_flush_page(env, T0 | (0x6 << 28));
tlb_flush_page(env, T0 | (0x7 << 28));
tlb_flush_page(env, T0 | (0x8 << 28));
tlb_flush_page(env, T0 | (0x9 << 28));
tlb_flush_page(env, T0 | (0xA << 28));
tlb_flush_page(env, T0 | (0xB << 28));
tlb_flush_page(env, T0 | (0xC << 28));
tlb_flush_page(env, T0 | (0xD << 28));
tlb_flush_page(env, T0 | (0xE << 28));
tlb_flush_page(env, T0 | (0xF << 28));
#else
tlb_flush(env, 1);
#endif
}
#else
do_tlbia();
#endif
}
#endif
#if defined(TARGET_PPC64)
void do_slbia (void)
{
/* XXX: TODO */
tlb_flush(env, 1);
}
void do_slbie (void)
{
/* XXX: TODO */
tlb_flush(env, 1);
}
#endif
/* Software driven TLBs management */
/* PowerPC 602/603 software TLB load instructions helpers */
void do_load_6xx_tlb (int is_code)
{
target_ulong RPN, CMP, EPN;
int way;
RPN = env->spr[SPR_RPA];
if (is_code) {
CMP = env->spr[SPR_ICMP];
EPN = env->spr[SPR_IMISS];
} else {
CMP = env->spr[SPR_DCMP];
EPN = env->spr[SPR_DMISS];
}
way = (env->spr[SPR_SRR1] >> 17) & 1;
#if defined (DEBUG_SOFTWARE_TLB)
if (loglevel != 0) {
fprintf(logfile, "%s: EPN %08lx %08lx PTE0 %08lx PTE1 %08lx way %d\n",
__func__, (unsigned long)T0, (unsigned long)EPN,
(unsigned long)CMP, (unsigned long)RPN, way);
}
#endif
/* Store this TLB */
ppc6xx_tlb_store(env, (uint32_t)(T0 & TARGET_PAGE_MASK),
way, is_code, CMP, RPN);
}
static target_ulong booke_tlb_to_page_size (int size)
{
return 1024 << (2 * size);
}
static int booke_page_size_to_tlb (target_ulong page_size)
{
int size;
switch (page_size) {
case 0x00000400UL:
size = 0x0;
break;
case 0x00001000UL:
size = 0x1;
break;
case 0x00004000UL:
size = 0x2;
break;
case 0x00010000UL:
size = 0x3;
break;
case 0x00040000UL:
size = 0x4;
break;
case 0x00100000UL:
size = 0x5;
break;
case 0x00400000UL:
size = 0x6;
break;
case 0x01000000UL:
size = 0x7;
break;
case 0x04000000UL:
size = 0x8;
break;
case 0x10000000UL:
size = 0x9;
break;
case 0x40000000UL:
size = 0xA;
break;
#if defined (TARGET_PPC64)
case 0x000100000000ULL:
size = 0xB;
break;
case 0x000400000000ULL:
size = 0xC;
break;
case 0x001000000000ULL:
size = 0xD;
break;
case 0x004000000000ULL:
size = 0xE;
break;
case 0x010000000000ULL:
size = 0xF;
break;
#endif
default:
size = -1;
break;
}
return size;
}
/* Helpers for 4xx TLB management */
void do_4xx_tlbre_lo (void)
{
ppcemb_tlb_t *tlb;
int size;
T0 &= 0x3F;
tlb = &env->tlb[T0].tlbe;
T0 = tlb->EPN;
if (tlb->prot & PAGE_VALID)
T0 |= 0x400;
size = booke_page_size_to_tlb(tlb->size);
if (size < 0 || size > 0x7)
size = 1;
T0 |= size << 7;
env->spr[SPR_40x_PID] = tlb->PID;
}
void do_4xx_tlbre_hi (void)
{
ppcemb_tlb_t *tlb;
T0 &= 0x3F;
tlb = &env->tlb[T0].tlbe;
T0 = tlb->RPN;
if (tlb->prot & PAGE_EXEC)
T0 |= 0x200;
if (tlb->prot & PAGE_WRITE)
T0 |= 0x100;
}
void do_4xx_tlbsx (void)
{
T0 = ppcemb_tlb_search(env, T0);
}
void do_4xx_tlbsx_ (void)
{
int tmp = xer_ov;
T0 = ppcemb_tlb_search(env, T0);
if (T0 != -1)
tmp |= 0x02;
env->crf[0] = tmp;
}
void do_4xx_tlbwe_hi (void)
{
ppcemb_tlb_t *tlb;
target_ulong page, end;
#if defined (DEBUG_SOFTWARE_TLB)
if (loglevel != 0) {
fprintf(logfile, "%s T0 " REGX " T1 " REGX "\n", __func__, T0, T1);
}
#endif
T0 &= 0x3F;
tlb = &env->tlb[T0].tlbe;
/* Invalidate previous TLB (if it's valid) */
if (tlb->prot & PAGE_VALID) {
end = tlb->EPN + tlb->size;
#if defined (DEBUG_SOFTWARE_TLB)
if (loglevel != 0) {
fprintf(logfile, "%s: invalidate old TLB %d start " ADDRX
" end " ADDRX "\n", __func__, (int)T0, tlb->EPN, end);
}
#endif
for (page = tlb->EPN; page < end; page += TARGET_PAGE_SIZE)
tlb_flush_page(env, page);
}
tlb->size = booke_tlb_to_page_size((T1 >> 7) & 0x7);
/* We cannot handle TLB size < TARGET_PAGE_SIZE.
* If this ever occurs, one should use the ppcemb target instead
* of the ppc or ppc64 one
*/
if ((T1 & 0x40) && tlb->size < TARGET_PAGE_SIZE) {
cpu_abort(env, "TLB size %u < %u are not supported (%d)\n",
tlb->size, TARGET_PAGE_SIZE, (int)((T1 >> 7) & 0x7));
}
tlb->EPN = (T1 & 0xFFFFFC00) & ~(tlb->size - 1);
if (T1 & 0x40)
tlb->prot |= PAGE_VALID;
else
tlb->prot &= ~PAGE_VALID;
if (T1 & 0x20) {
/* XXX: TO BE FIXED */
cpu_abort(env, "Little-endian TLB entries are not supported by now\n");
}
tlb->PID = env->spr[SPR_40x_PID]; /* PID */
tlb->attr = T1 & 0xFF;
#if defined (DEBUG_SOFTWARE_TLB)
if (loglevel != 0) {
fprintf(logfile, "%s: set up TLB %d RPN " PADDRX " EPN " ADDRX
" size " ADDRX " prot %c%c%c%c PID %d\n", __func__,
(int)T0, tlb->RPN, tlb->EPN, tlb->size,
tlb->prot & PAGE_READ ? 'r' : '-',
tlb->prot & PAGE_WRITE ? 'w' : '-',
tlb->prot & PAGE_EXEC ? 'x' : '-',
tlb->prot & PAGE_VALID ? 'v' : '-', (int)tlb->PID);
}
#endif
/* Invalidate new TLB (if valid) */
if (tlb->prot & PAGE_VALID) {
end = tlb->EPN + tlb->size;
#if defined (DEBUG_SOFTWARE_TLB)
if (loglevel != 0) {
fprintf(logfile, "%s: invalidate TLB %d start " ADDRX
" end " ADDRX "\n", __func__, (int)T0, tlb->EPN, end);
}
#endif
for (page = tlb->EPN; page < end; page += TARGET_PAGE_SIZE)
tlb_flush_page(env, page);
}
}
void do_4xx_tlbwe_lo (void)
{
ppcemb_tlb_t *tlb;
#if defined (DEBUG_SOFTWARE_TLB)
if (loglevel != 0) {
fprintf(logfile, "%s T0 " REGX " T1 " REGX "\n", __func__, T0, T1);
}
#endif
T0 &= 0x3F;
tlb = &env->tlb[T0].tlbe;
tlb->RPN = T1 & 0xFFFFFC00;
tlb->prot = PAGE_READ;
if (T1 & 0x200)
tlb->prot |= PAGE_EXEC;
if (T1 & 0x100)
tlb->prot |= PAGE_WRITE;
#if defined (DEBUG_SOFTWARE_TLB)
if (loglevel != 0) {
fprintf(logfile, "%s: set up TLB %d RPN " PADDRX " EPN " ADDRX
" size " ADDRX " prot %c%c%c%c PID %d\n", __func__,
(int)T0, tlb->RPN, tlb->EPN, tlb->size,
tlb->prot & PAGE_READ ? 'r' : '-',
tlb->prot & PAGE_WRITE ? 'w' : '-',
tlb->prot & PAGE_EXEC ? 'x' : '-',
tlb->prot & PAGE_VALID ? 'v' : '-', (int)tlb->PID);
}
#endif
}
#endif /* !CONFIG_USER_ONLY */