qemu/target/openrisc/translate.c
Richard Henderson c7b6f54bf8 target/openrisc: Merge disas_openrisc_insn
Acked-by: Stafford Horne <shorne@gmail.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
2018-05-14 14:58:08 -07:00

1533 lines
42 KiB
C

/*
* OpenRISC translation
*
* Copyright (c) 2011-2012 Jia Liu <proljc@gmail.com>
* Feng Gao <gf91597@gmail.com>
*
* 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, see <http://www.gnu.org/licenses/>.
*/
#include "qemu/osdep.h"
#include "cpu.h"
#include "exec/exec-all.h"
#include "disas/disas.h"
#include "tcg-op.h"
#include "qemu-common.h"
#include "qemu/log.h"
#include "qemu/bitops.h"
#include "exec/cpu_ldst.h"
#include "exec/translator.h"
#include "exec/helper-proto.h"
#include "exec/helper-gen.h"
#include "exec/gen-icount.h"
#include "trace-tcg.h"
#include "exec/log.h"
#define LOG_DIS(str, ...) \
qemu_log_mask(CPU_LOG_TB_IN_ASM, "%08x: " str, dc->base.pc_next, \
## __VA_ARGS__)
/* is_jmp field values */
#define DISAS_JUMP DISAS_TARGET_0 /* only pc was modified dynamically */
#define DISAS_UPDATE DISAS_TARGET_1 /* cpu state was modified dynamically */
#define DISAS_TB_JUMP DISAS_TARGET_2 /* only pc was modified statically */
typedef struct DisasContext {
DisasContextBase base;
uint32_t mem_idx;
uint32_t tb_flags;
uint32_t delayed_branch;
} DisasContext;
/* Include the auto-generated decoder. */
#include "decode.inc.c"
static TCGv cpu_sr;
static TCGv cpu_R[32];
static TCGv cpu_R0;
static TCGv cpu_pc;
static TCGv jmp_pc; /* l.jr/l.jalr temp pc */
static TCGv cpu_ppc;
static TCGv cpu_sr_f; /* bf/bnf, F flag taken */
static TCGv cpu_sr_cy; /* carry (unsigned overflow) */
static TCGv cpu_sr_ov; /* signed overflow */
static TCGv cpu_lock_addr;
static TCGv cpu_lock_value;
static TCGv_i32 fpcsr;
static TCGv_i64 cpu_mac; /* MACHI:MACLO */
static TCGv_i32 cpu_dflag;
void openrisc_translate_init(void)
{
static const char * const regnames[] = {
"r0", "r1", "r2", "r3", "r4", "r5", "r6", "r7",
"r8", "r9", "r10", "r11", "r12", "r13", "r14", "r15",
"r16", "r17", "r18", "r19", "r20", "r21", "r22", "r23",
"r24", "r25", "r26", "r27", "r28", "r29", "r30", "r31",
};
int i;
cpu_sr = tcg_global_mem_new(cpu_env,
offsetof(CPUOpenRISCState, sr), "sr");
cpu_dflag = tcg_global_mem_new_i32(cpu_env,
offsetof(CPUOpenRISCState, dflag),
"dflag");
cpu_pc = tcg_global_mem_new(cpu_env,
offsetof(CPUOpenRISCState, pc), "pc");
cpu_ppc = tcg_global_mem_new(cpu_env,
offsetof(CPUOpenRISCState, ppc), "ppc");
jmp_pc = tcg_global_mem_new(cpu_env,
offsetof(CPUOpenRISCState, jmp_pc), "jmp_pc");
cpu_sr_f = tcg_global_mem_new(cpu_env,
offsetof(CPUOpenRISCState, sr_f), "sr_f");
cpu_sr_cy = tcg_global_mem_new(cpu_env,
offsetof(CPUOpenRISCState, sr_cy), "sr_cy");
cpu_sr_ov = tcg_global_mem_new(cpu_env,
offsetof(CPUOpenRISCState, sr_ov), "sr_ov");
cpu_lock_addr = tcg_global_mem_new(cpu_env,
offsetof(CPUOpenRISCState, lock_addr),
"lock_addr");
cpu_lock_value = tcg_global_mem_new(cpu_env,
offsetof(CPUOpenRISCState, lock_value),
"lock_value");
fpcsr = tcg_global_mem_new_i32(cpu_env,
offsetof(CPUOpenRISCState, fpcsr),
"fpcsr");
cpu_mac = tcg_global_mem_new_i64(cpu_env,
offsetof(CPUOpenRISCState, mac),
"mac");
for (i = 0; i < 32; i++) {
cpu_R[i] = tcg_global_mem_new(cpu_env,
offsetof(CPUOpenRISCState,
shadow_gpr[0][i]),
regnames[i]);
}
cpu_R0 = cpu_R[0];
}
static void gen_exception(DisasContext *dc, unsigned int excp)
{
TCGv_i32 tmp = tcg_const_i32(excp);
gen_helper_exception(cpu_env, tmp);
tcg_temp_free_i32(tmp);
}
static void gen_illegal_exception(DisasContext *dc)
{
tcg_gen_movi_tl(cpu_pc, dc->base.pc_next);
gen_exception(dc, EXCP_ILLEGAL);
dc->base.is_jmp = DISAS_NORETURN;
}
/* not used yet, open it when we need or64. */
/*#ifdef TARGET_OPENRISC64
static void check_ob64s(DisasContext *dc)
{
if (!(dc->flags & CPUCFGR_OB64S)) {
gen_illegal_exception(dc);
}
}
static void check_of64s(DisasContext *dc)
{
if (!(dc->flags & CPUCFGR_OF64S)) {
gen_illegal_exception(dc);
}
}
static void check_ov64s(DisasContext *dc)
{
if (!(dc->flags & CPUCFGR_OV64S)) {
gen_illegal_exception(dc);
}
}
#endif*/
/* We're about to write to REG. On the off-chance that the user is
writing to R0, re-instate the architectural register. */
#define check_r0_write(reg) \
do { \
if (unlikely(reg == 0)) { \
cpu_R[0] = cpu_R0; \
} \
} while (0)
static inline bool use_goto_tb(DisasContext *dc, target_ulong dest)
{
if (unlikely(dc->base.singlestep_enabled)) {
return false;
}
#ifndef CONFIG_USER_ONLY
return (dc->base.tb->pc & TARGET_PAGE_MASK) == (dest & TARGET_PAGE_MASK);
#else
return true;
#endif
}
static void gen_goto_tb(DisasContext *dc, int n, target_ulong dest)
{
if (use_goto_tb(dc, dest)) {
tcg_gen_movi_tl(cpu_pc, dest);
tcg_gen_goto_tb(n);
tcg_gen_exit_tb((uintptr_t)dc->base.tb + n);
} else {
tcg_gen_movi_tl(cpu_pc, dest);
if (dc->base.singlestep_enabled) {
gen_exception(dc, EXCP_DEBUG);
}
tcg_gen_exit_tb(0);
}
}
static void gen_ove_cy(DisasContext *dc)
{
if (dc->tb_flags & SR_OVE) {
gen_helper_ove_cy(cpu_env);
}
}
static void gen_ove_ov(DisasContext *dc)
{
if (dc->tb_flags & SR_OVE) {
gen_helper_ove_ov(cpu_env);
}
}
static void gen_ove_cyov(DisasContext *dc)
{
if (dc->tb_flags & SR_OVE) {
gen_helper_ove_cyov(cpu_env);
}
}
static void gen_add(DisasContext *dc, TCGv dest, TCGv srca, TCGv srcb)
{
TCGv t0 = tcg_const_tl(0);
TCGv res = tcg_temp_new();
tcg_gen_add2_tl(res, cpu_sr_cy, srca, t0, srcb, t0);
tcg_gen_xor_tl(cpu_sr_ov, srca, srcb);
tcg_gen_xor_tl(t0, res, srcb);
tcg_gen_andc_tl(cpu_sr_ov, t0, cpu_sr_ov);
tcg_temp_free(t0);
tcg_gen_mov_tl(dest, res);
tcg_temp_free(res);
gen_ove_cyov(dc);
}
static void gen_addc(DisasContext *dc, TCGv dest, TCGv srca, TCGv srcb)
{
TCGv t0 = tcg_const_tl(0);
TCGv res = tcg_temp_new();
tcg_gen_add2_tl(res, cpu_sr_cy, srca, t0, cpu_sr_cy, t0);
tcg_gen_add2_tl(res, cpu_sr_cy, res, cpu_sr_cy, srcb, t0);
tcg_gen_xor_tl(cpu_sr_ov, srca, srcb);
tcg_gen_xor_tl(t0, res, srcb);
tcg_gen_andc_tl(cpu_sr_ov, t0, cpu_sr_ov);
tcg_temp_free(t0);
tcg_gen_mov_tl(dest, res);
tcg_temp_free(res);
gen_ove_cyov(dc);
}
static void gen_sub(DisasContext *dc, TCGv dest, TCGv srca, TCGv srcb)
{
TCGv res = tcg_temp_new();
tcg_gen_sub_tl(res, srca, srcb);
tcg_gen_xor_tl(cpu_sr_cy, srca, srcb);
tcg_gen_xor_tl(cpu_sr_ov, res, srcb);
tcg_gen_and_tl(cpu_sr_ov, cpu_sr_ov, cpu_sr_cy);
tcg_gen_setcond_tl(TCG_COND_LTU, cpu_sr_cy, srca, srcb);
tcg_gen_mov_tl(dest, res);
tcg_temp_free(res);
gen_ove_cyov(dc);
}
static void gen_mul(DisasContext *dc, TCGv dest, TCGv srca, TCGv srcb)
{
TCGv t0 = tcg_temp_new();
tcg_gen_muls2_tl(dest, cpu_sr_ov, srca, srcb);
tcg_gen_sari_tl(t0, dest, TARGET_LONG_BITS - 1);
tcg_gen_setcond_tl(TCG_COND_NE, cpu_sr_ov, cpu_sr_ov, t0);
tcg_temp_free(t0);
tcg_gen_neg_tl(cpu_sr_ov, cpu_sr_ov);
gen_ove_ov(dc);
}
static void gen_mulu(DisasContext *dc, TCGv dest, TCGv srca, TCGv srcb)
{
tcg_gen_muls2_tl(dest, cpu_sr_cy, srca, srcb);
tcg_gen_setcondi_tl(TCG_COND_NE, cpu_sr_cy, cpu_sr_cy, 0);
gen_ove_cy(dc);
}
static void gen_div(DisasContext *dc, TCGv dest, TCGv srca, TCGv srcb)
{
TCGv t0 = tcg_temp_new();
tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_sr_ov, srcb, 0);
/* The result of divide-by-zero is undefined.
Supress the host-side exception by dividing by 1. */
tcg_gen_or_tl(t0, srcb, cpu_sr_ov);
tcg_gen_div_tl(dest, srca, t0);
tcg_temp_free(t0);
tcg_gen_neg_tl(cpu_sr_ov, cpu_sr_ov);
gen_ove_ov(dc);
}
static void gen_divu(DisasContext *dc, TCGv dest, TCGv srca, TCGv srcb)
{
TCGv t0 = tcg_temp_new();
tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_sr_cy, srcb, 0);
/* The result of divide-by-zero is undefined.
Supress the host-side exception by dividing by 1. */
tcg_gen_or_tl(t0, srcb, cpu_sr_cy);
tcg_gen_divu_tl(dest, srca, t0);
tcg_temp_free(t0);
gen_ove_cy(dc);
}
static void gen_muld(DisasContext *dc, TCGv srca, TCGv srcb)
{
TCGv_i64 t1 = tcg_temp_new_i64();
TCGv_i64 t2 = tcg_temp_new_i64();
tcg_gen_ext_tl_i64(t1, srca);
tcg_gen_ext_tl_i64(t2, srcb);
if (TARGET_LONG_BITS == 32) {
tcg_gen_mul_i64(cpu_mac, t1, t2);
tcg_gen_movi_tl(cpu_sr_ov, 0);
} else {
TCGv_i64 high = tcg_temp_new_i64();
tcg_gen_muls2_i64(cpu_mac, high, t1, t2);
tcg_gen_sari_i64(t1, cpu_mac, 63);
tcg_gen_setcond_i64(TCG_COND_NE, t1, t1, high);
tcg_temp_free_i64(high);
tcg_gen_trunc_i64_tl(cpu_sr_ov, t1);
tcg_gen_neg_tl(cpu_sr_ov, cpu_sr_ov);
gen_ove_ov(dc);
}
tcg_temp_free_i64(t1);
tcg_temp_free_i64(t2);
}
static void gen_muldu(DisasContext *dc, TCGv srca, TCGv srcb)
{
TCGv_i64 t1 = tcg_temp_new_i64();
TCGv_i64 t2 = tcg_temp_new_i64();
tcg_gen_extu_tl_i64(t1, srca);
tcg_gen_extu_tl_i64(t2, srcb);
if (TARGET_LONG_BITS == 32) {
tcg_gen_mul_i64(cpu_mac, t1, t2);
tcg_gen_movi_tl(cpu_sr_cy, 0);
} else {
TCGv_i64 high = tcg_temp_new_i64();
tcg_gen_mulu2_i64(cpu_mac, high, t1, t2);
tcg_gen_setcondi_i64(TCG_COND_NE, high, high, 0);
tcg_gen_trunc_i64_tl(cpu_sr_cy, high);
tcg_temp_free_i64(high);
gen_ove_cy(dc);
}
tcg_temp_free_i64(t1);
tcg_temp_free_i64(t2);
}
static void gen_mac(DisasContext *dc, TCGv srca, TCGv srcb)
{
TCGv_i64 t1 = tcg_temp_new_i64();
TCGv_i64 t2 = tcg_temp_new_i64();
tcg_gen_ext_tl_i64(t1, srca);
tcg_gen_ext_tl_i64(t2, srcb);
tcg_gen_mul_i64(t1, t1, t2);
/* Note that overflow is only computed during addition stage. */
tcg_gen_xor_i64(t2, cpu_mac, t1);
tcg_gen_add_i64(cpu_mac, cpu_mac, t1);
tcg_gen_xor_i64(t1, t1, cpu_mac);
tcg_gen_andc_i64(t1, t1, t2);
tcg_temp_free_i64(t2);
#if TARGET_LONG_BITS == 32
tcg_gen_extrh_i64_i32(cpu_sr_ov, t1);
#else
tcg_gen_mov_i64(cpu_sr_ov, t1);
#endif
tcg_temp_free_i64(t1);
gen_ove_ov(dc);
}
static void gen_macu(DisasContext *dc, TCGv srca, TCGv srcb)
{
TCGv_i64 t1 = tcg_temp_new_i64();
TCGv_i64 t2 = tcg_temp_new_i64();
tcg_gen_extu_tl_i64(t1, srca);
tcg_gen_extu_tl_i64(t2, srcb);
tcg_gen_mul_i64(t1, t1, t2);
tcg_temp_free_i64(t2);
/* Note that overflow is only computed during addition stage. */
tcg_gen_add_i64(cpu_mac, cpu_mac, t1);
tcg_gen_setcond_i64(TCG_COND_LTU, t1, cpu_mac, t1);
tcg_gen_trunc_i64_tl(cpu_sr_cy, t1);
tcg_temp_free_i64(t1);
gen_ove_cy(dc);
}
static void gen_msb(DisasContext *dc, TCGv srca, TCGv srcb)
{
TCGv_i64 t1 = tcg_temp_new_i64();
TCGv_i64 t2 = tcg_temp_new_i64();
tcg_gen_ext_tl_i64(t1, srca);
tcg_gen_ext_tl_i64(t2, srcb);
tcg_gen_mul_i64(t1, t1, t2);
/* Note that overflow is only computed during subtraction stage. */
tcg_gen_xor_i64(t2, cpu_mac, t1);
tcg_gen_sub_i64(cpu_mac, cpu_mac, t1);
tcg_gen_xor_i64(t1, t1, cpu_mac);
tcg_gen_and_i64(t1, t1, t2);
tcg_temp_free_i64(t2);
#if TARGET_LONG_BITS == 32
tcg_gen_extrh_i64_i32(cpu_sr_ov, t1);
#else
tcg_gen_mov_i64(cpu_sr_ov, t1);
#endif
tcg_temp_free_i64(t1);
gen_ove_ov(dc);
}
static void gen_msbu(DisasContext *dc, TCGv srca, TCGv srcb)
{
TCGv_i64 t1 = tcg_temp_new_i64();
TCGv_i64 t2 = tcg_temp_new_i64();
tcg_gen_extu_tl_i64(t1, srca);
tcg_gen_extu_tl_i64(t2, srcb);
tcg_gen_mul_i64(t1, t1, t2);
/* Note that overflow is only computed during subtraction stage. */
tcg_gen_setcond_i64(TCG_COND_LTU, t2, cpu_mac, t1);
tcg_gen_sub_i64(cpu_mac, cpu_mac, t1);
tcg_gen_trunc_i64_tl(cpu_sr_cy, t2);
tcg_temp_free_i64(t2);
tcg_temp_free_i64(t1);
gen_ove_cy(dc);
}
static bool trans_l_add(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.add r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
gen_add(dc, cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_addc(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.addc r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
gen_addc(dc, cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sub(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.sub r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
gen_sub(dc, cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_and(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.and r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
tcg_gen_and_tl(cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_or(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.or r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
tcg_gen_or_tl(cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_xor(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.xor r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
tcg_gen_xor_tl(cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sll(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.sll r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
tcg_gen_shl_tl(cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_srl(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.srl r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
tcg_gen_shr_tl(cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sra(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.sra r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
tcg_gen_sar_tl(cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_ror(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.ror r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
tcg_gen_rotr_tl(cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_exths(DisasContext *dc, arg_da *a, uint32_t insn)
{
LOG_DIS("l.exths r%d, r%d\n", a->d, a->a);
check_r0_write(a->d);
tcg_gen_ext16s_tl(cpu_R[a->d], cpu_R[a->a]);
return true;
}
static bool trans_l_extbs(DisasContext *dc, arg_da *a, uint32_t insn)
{
LOG_DIS("l.extbs r%d, r%d\n", a->d, a->a);
check_r0_write(a->d);
tcg_gen_ext8s_tl(cpu_R[a->d], cpu_R[a->a]);
return true;
}
static bool trans_l_exthz(DisasContext *dc, arg_da *a, uint32_t insn)
{
LOG_DIS("l.exthz r%d, r%d\n", a->d, a->a);
check_r0_write(a->d);
tcg_gen_ext16u_tl(cpu_R[a->d], cpu_R[a->a]);
return true;
}
static bool trans_l_extbz(DisasContext *dc, arg_da *a, uint32_t insn)
{
LOG_DIS("l.extbz r%d, r%d\n", a->d, a->a);
check_r0_write(a->d);
tcg_gen_ext8u_tl(cpu_R[a->d], cpu_R[a->a]);
return true;
}
static bool trans_l_cmov(DisasContext *dc, arg_dab *a, uint32_t insn)
{
TCGv zero;
LOG_DIS("l.cmov r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
zero = tcg_const_tl(0);
tcg_gen_movcond_tl(TCG_COND_NE, cpu_R[a->d], cpu_sr_f, zero,
cpu_R[a->a], cpu_R[a->b]);
tcg_temp_free(zero);
return true;
}
static bool trans_l_ff1(DisasContext *dc, arg_da *a, uint32_t insn)
{
LOG_DIS("l.ff1 r%d, r%d\n", a->d, a->a);
check_r0_write(a->d);
tcg_gen_ctzi_tl(cpu_R[a->d], cpu_R[a->a], -1);
tcg_gen_addi_tl(cpu_R[a->d], cpu_R[a->d], 1);
return true;
}
static bool trans_l_fl1(DisasContext *dc, arg_da *a, uint32_t insn)
{
LOG_DIS("l.fl1 r%d, r%d\n", a->d, a->a);
check_r0_write(a->d);
tcg_gen_clzi_tl(cpu_R[a->d], cpu_R[a->a], TARGET_LONG_BITS);
tcg_gen_subfi_tl(cpu_R[a->d], TARGET_LONG_BITS, cpu_R[a->d]);
return true;
}
static bool trans_l_mul(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.mul r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
gen_mul(dc, cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_mulu(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.mulu r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
gen_mulu(dc, cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_div(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.div r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
gen_div(dc, cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_divu(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("l.divu r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
gen_divu(dc, cpu_R[a->d], cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_muld(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("l.muld r%d, r%d\n", a->a, a->b);
gen_muld(dc, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_muldu(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("l.muldu r%d, r%d\n", a->a, a->b);
gen_muldu(dc, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_j(DisasContext *dc, arg_l_j *a, uint32_t insn)
{
target_ulong tmp_pc = dc->base.pc_next + a->n * 4;
LOG_DIS("l.j %d\n", a->n);
tcg_gen_movi_tl(jmp_pc, tmp_pc);
dc->delayed_branch = 2;
return true;
}
static bool trans_l_jal(DisasContext *dc, arg_l_jal *a, uint32_t insn)
{
target_ulong tmp_pc = dc->base.pc_next + a->n * 4;
target_ulong ret_pc = dc->base.pc_next + 8;
LOG_DIS("l.jal %d\n", a->n);
tcg_gen_movi_tl(cpu_R[9], ret_pc);
/* Optimize jal being used to load the PC for PIC. */
if (tmp_pc != ret_pc) {
tcg_gen_movi_tl(jmp_pc, tmp_pc);
dc->delayed_branch = 2;
}
return true;
}
static void do_bf(DisasContext *dc, arg_l_bf *a, TCGCond cond)
{
target_ulong tmp_pc = dc->base.pc_next + a->n * 4;
TCGv t_next = tcg_const_tl(dc->base.pc_next + 8);
TCGv t_true = tcg_const_tl(tmp_pc);
TCGv t_zero = tcg_const_tl(0);
tcg_gen_movcond_tl(cond, jmp_pc, cpu_sr_f, t_zero, t_true, t_next);
tcg_temp_free(t_next);
tcg_temp_free(t_true);
tcg_temp_free(t_zero);
dc->delayed_branch = 2;
}
static bool trans_l_bf(DisasContext *dc, arg_l_bf *a, uint32_t insn)
{
LOG_DIS("l.bf %d\n", a->n);
do_bf(dc, a, TCG_COND_NE);
return true;
}
static bool trans_l_bnf(DisasContext *dc, arg_l_bf *a, uint32_t insn)
{
LOG_DIS("l.bnf %d\n", a->n);
do_bf(dc, a, TCG_COND_EQ);
return true;
}
static bool trans_l_jr(DisasContext *dc, arg_l_jr *a, uint32_t insn)
{
LOG_DIS("l.jr r%d\n", a->b);
tcg_gen_mov_tl(jmp_pc, cpu_R[a->b]);
dc->delayed_branch = 2;
return true;
}
static bool trans_l_jalr(DisasContext *dc, arg_l_jalr *a, uint32_t insn)
{
LOG_DIS("l.jalr r%d\n", a->b);
tcg_gen_mov_tl(jmp_pc, cpu_R[a->b]);
tcg_gen_movi_tl(cpu_R[9], dc->base.pc_next + 8);
dc->delayed_branch = 2;
return true;
}
static bool trans_l_lwa(DisasContext *dc, arg_load *a, uint32_t insn)
{
TCGv ea;
LOG_DIS("l.lwa r%d, r%d, %d\n", a->d, a->a, a->i);
check_r0_write(a->d);
ea = tcg_temp_new();
tcg_gen_addi_tl(ea, cpu_R[a->a], a->i);
tcg_gen_qemu_ld_tl(cpu_R[a->d], ea, dc->mem_idx, MO_TEUL);
tcg_gen_mov_tl(cpu_lock_addr, ea);
tcg_gen_mov_tl(cpu_lock_value, cpu_R[a->d]);
tcg_temp_free(ea);
return true;
}
static void do_load(DisasContext *dc, arg_load *a, TCGMemOp mop)
{
TCGv ea;
check_r0_write(a->d);
ea = tcg_temp_new();
tcg_gen_addi_tl(ea, cpu_R[a->a], a->i);
tcg_gen_qemu_ld_tl(cpu_R[a->d], ea, dc->mem_idx, mop);
tcg_temp_free(ea);
}
static bool trans_l_lwz(DisasContext *dc, arg_load *a, uint32_t insn)
{
LOG_DIS("l.lwz r%d, r%d, %d\n", a->d, a->a, a->i);
do_load(dc, a, MO_TEUL);
return true;
}
static bool trans_l_lws(DisasContext *dc, arg_load *a, uint32_t insn)
{
LOG_DIS("l.lws r%d, r%d, %d\n", a->d, a->a, a->i);
do_load(dc, a, MO_TESL);
return true;
}
static bool trans_l_lbz(DisasContext *dc, arg_load *a, uint32_t insn)
{
LOG_DIS("l.lbz r%d, r%d, %d\n", a->d, a->a, a->i);
do_load(dc, a, MO_UB);
return true;
}
static bool trans_l_lbs(DisasContext *dc, arg_load *a, uint32_t insn)
{
LOG_DIS("l.lbs r%d, r%d, %d\n", a->d, a->a, a->i);
do_load(dc, a, MO_SB);
return true;
}
static bool trans_l_lhz(DisasContext *dc, arg_load *a, uint32_t insn)
{
LOG_DIS("l.lhz r%d, r%d, %d\n", a->d, a->a, a->i);
do_load(dc, a, MO_TEUW);
return true;
}
static bool trans_l_lhs(DisasContext *dc, arg_load *a, uint32_t insn)
{
LOG_DIS("l.lhs r%d, r%d, %d\n", a->d, a->a, a->i);
do_load(dc, a, MO_TESW);
return true;
}
static bool trans_l_swa(DisasContext *dc, arg_store *a, uint32_t insn)
{
TCGv ea, val;
TCGLabel *lab_fail, *lab_done;
LOG_DIS("l.swa r%d, r%d, %d\n", a->a, a->b, a->i);
ea = tcg_temp_new();
tcg_gen_addi_tl(ea, cpu_R[a->a], a->i);
/* For TB_FLAGS_R0_0, the branch below invalidates the temporary assigned
to cpu_R[0]. Since l.swa is quite often immediately followed by a
branch, don't bother reallocating; finish the TB using the "real" R0.
This also takes care of RB input across the branch. */
cpu_R[0] = cpu_R0;
lab_fail = gen_new_label();
lab_done = gen_new_label();
tcg_gen_brcond_tl(TCG_COND_NE, ea, cpu_lock_addr, lab_fail);
tcg_temp_free(ea);
val = tcg_temp_new();
tcg_gen_atomic_cmpxchg_tl(val, cpu_lock_addr, cpu_lock_value,
cpu_R[a->b], dc->mem_idx, MO_TEUL);
tcg_gen_setcond_tl(TCG_COND_EQ, cpu_sr_f, val, cpu_lock_value);
tcg_temp_free(val);
tcg_gen_br(lab_done);
gen_set_label(lab_fail);
tcg_gen_movi_tl(cpu_sr_f, 0);
gen_set_label(lab_done);
tcg_gen_movi_tl(cpu_lock_addr, -1);
return true;
}
static void do_store(DisasContext *dc, arg_store *a, TCGMemOp mop)
{
TCGv t0 = tcg_temp_new();
tcg_gen_addi_tl(t0, cpu_R[a->a], a->i);
tcg_gen_qemu_st_tl(cpu_R[a->b], t0, dc->mem_idx, mop);
tcg_temp_free(t0);
}
static bool trans_l_sw(DisasContext *dc, arg_store *a, uint32_t insn)
{
LOG_DIS("l.sw r%d, r%d, %d\n", a->a, a->b, a->i);
do_store(dc, a, MO_TEUL);
return true;
}
static bool trans_l_sb(DisasContext *dc, arg_store *a, uint32_t insn)
{
LOG_DIS("l.sb r%d, r%d, %d\n", a->a, a->b, a->i);
do_store(dc, a, MO_UB);
return true;
}
static bool trans_l_sh(DisasContext *dc, arg_store *a, uint32_t insn)
{
LOG_DIS("l.sh r%d, r%d, %d\n", a->a, a->b, a->i);
do_store(dc, a, MO_TEUW);
return true;
}
static bool trans_l_nop(DisasContext *dc, arg_l_nop *a, uint32_t insn)
{
LOG_DIS("l.nop %d\n", a->k);
return true;
}
static bool trans_l_addi(DisasContext *dc, arg_rri *a, uint32_t insn)
{
TCGv t0;
LOG_DIS("l.addi r%d, r%d, %d\n", a->d, a->a, a->i);
check_r0_write(a->d);
t0 = tcg_const_tl(a->i);
gen_add(dc, cpu_R[a->d], cpu_R[a->a], t0);
tcg_temp_free(t0);
return true;
}
static bool trans_l_addic(DisasContext *dc, arg_rri *a, uint32_t insn)
{
TCGv t0;
LOG_DIS("l.addic r%d, r%d, %d\n", a->d, a->a, a->i);
check_r0_write(a->d);
t0 = tcg_const_tl(a->i);
gen_addc(dc, cpu_R[a->d], cpu_R[a->a], t0);
tcg_temp_free(t0);
return true;
}
static bool trans_l_muli(DisasContext *dc, arg_rri *a, uint32_t insn)
{
TCGv t0;
LOG_DIS("l.muli r%d, r%d, %d\n", a->d, a->a, a->i);
check_r0_write(a->d);
t0 = tcg_const_tl(a->i);
gen_mul(dc, cpu_R[a->d], cpu_R[a->a], t0);
tcg_temp_free(t0);
return true;
}
static bool trans_l_maci(DisasContext *dc, arg_l_maci *a, uint32_t insn)
{
TCGv t0;
LOG_DIS("l.maci r%d, %d\n", a->a, a->i);
t0 = tcg_const_tl(a->i);
gen_mac(dc, cpu_R[a->a], t0);
tcg_temp_free(t0);
return true;
}
static bool trans_l_andi(DisasContext *dc, arg_rrk *a, uint32_t insn)
{
LOG_DIS("l.andi r%d, r%d, %d\n", a->d, a->a, a->k);
check_r0_write(a->d);
tcg_gen_andi_tl(cpu_R[a->d], cpu_R[a->a], a->k);
return true;
}
static bool trans_l_ori(DisasContext *dc, arg_rrk *a, uint32_t insn)
{
LOG_DIS("l.ori r%d, r%d, %d\n", a->d, a->a, a->k);
check_r0_write(a->d);
tcg_gen_ori_tl(cpu_R[a->d], cpu_R[a->a], a->k);
return true;
}
static bool trans_l_xori(DisasContext *dc, arg_rri *a, uint32_t insn)
{
LOG_DIS("l.xori r%d, r%d, %d\n", a->d, a->a, a->i);
check_r0_write(a->d);
tcg_gen_xori_tl(cpu_R[a->d], cpu_R[a->a], a->i);
return true;
}
static bool trans_l_mfspr(DisasContext *dc, arg_l_mfspr *a, uint32_t insn)
{
LOG_DIS("l.mfspr r%d, r%d, %d\n", a->d, a->a, a->k);
check_r0_write(a->d);
#ifdef CONFIG_USER_ONLY
gen_illegal_exception(dc);
#else
if (dc->mem_idx == MMU_USER_IDX) {
gen_illegal_exception(dc);
} else {
TCGv_i32 ti = tcg_const_i32(a->k);
gen_helper_mfspr(cpu_R[a->d], cpu_env, cpu_R[a->d], cpu_R[a->a], ti);
tcg_temp_free_i32(ti);
}
#endif
return true;
}
static bool trans_l_mtspr(DisasContext *dc, arg_l_mtspr *a, uint32_t insn)
{
LOG_DIS("l.mtspr r%d, r%d, %d\n", a->a, a->b, a->k);
#ifdef CONFIG_USER_ONLY
gen_illegal_exception(dc);
#else
if (dc->mem_idx == MMU_USER_IDX) {
gen_illegal_exception(dc);
} else {
TCGv_i32 ti = tcg_const_i32(a->k);
gen_helper_mtspr(cpu_env, cpu_R[a->a], cpu_R[a->b], ti);
tcg_temp_free_i32(ti);
}
#endif
return true;
}
static bool trans_l_mac(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("l.mac r%d, r%d\n", a->a, a->b);
gen_mac(dc, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_msb(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("l.msb r%d, r%d\n", a->a, a->b);
gen_msb(dc, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_macu(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("l.mac r%d, r%d\n", a->a, a->b);
gen_macu(dc, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_msbu(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("l.msb r%d, r%d\n", a->a, a->b);
gen_msbu(dc, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_slli(DisasContext *dc, arg_dal *a, uint32_t insn)
{
LOG_DIS("l.slli r%d, r%d, %d\n", a->d, a->a, a->l);
check_r0_write(a->d);
tcg_gen_shli_tl(cpu_R[a->d], cpu_R[a->a], a->l & (TARGET_LONG_BITS - 1));
return true;
}
static bool trans_l_srli(DisasContext *dc, arg_dal *a, uint32_t insn)
{
LOG_DIS("l.srli r%d, r%d, %d\n", a->d, a->a, a->l);
check_r0_write(a->d);
tcg_gen_shri_tl(cpu_R[a->d], cpu_R[a->a], a->l & (TARGET_LONG_BITS - 1));
return true;
}
static bool trans_l_srai(DisasContext *dc, arg_dal *a, uint32_t insn)
{
LOG_DIS("l.srai r%d, r%d, %d\n", a->d, a->a, a->l);
check_r0_write(a->d);
tcg_gen_sari_tl(cpu_R[a->d], cpu_R[a->a], a->l & (TARGET_LONG_BITS - 1));
return true;
}
static bool trans_l_rori(DisasContext *dc, arg_dal *a, uint32_t insn)
{
LOG_DIS("l.rori r%d, r%d, %d\n", a->d, a->a, a->l);
check_r0_write(a->d);
tcg_gen_rotri_tl(cpu_R[a->d], cpu_R[a->a], a->l & (TARGET_LONG_BITS - 1));
return true;
}
static bool trans_l_movhi(DisasContext *dc, arg_l_movhi *a, uint32_t insn)
{
LOG_DIS("l.movhi r%d, %d\n", a->d, a->k);
check_r0_write(a->d);
tcg_gen_movi_tl(cpu_R[a->d], a->k << 16);
return true;
}
static bool trans_l_macrc(DisasContext *dc, arg_l_macrc *a, uint32_t insn)
{
LOG_DIS("l.macrc r%d\n", a->d);
check_r0_write(a->d);
tcg_gen_trunc_i64_tl(cpu_R[a->d], cpu_mac);
tcg_gen_movi_i64(cpu_mac, 0);
return true;
}
static bool trans_l_sfeq(DisasContext *dc, arg_ab *a, TCGCond cond)
{
LOG_DIS("l.sfeq r%d, r%d\n", a->a, a->b);
tcg_gen_setcond_tl(TCG_COND_EQ, cpu_sr_f, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sfne(DisasContext *dc, arg_ab *a, TCGCond cond)
{
LOG_DIS("l.sfne r%d, r%d\n", a->a, a->b);
tcg_gen_setcond_tl(TCG_COND_NE, cpu_sr_f, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sfgtu(DisasContext *dc, arg_ab *a, TCGCond cond)
{
LOG_DIS("l.sfgtu r%d, r%d\n", a->a, a->b);
tcg_gen_setcond_tl(TCG_COND_GTU, cpu_sr_f, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sfgeu(DisasContext *dc, arg_ab *a, TCGCond cond)
{
LOG_DIS("l.sfgeu r%d, r%d\n", a->a, a->b);
tcg_gen_setcond_tl(TCG_COND_GEU, cpu_sr_f, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sfltu(DisasContext *dc, arg_ab *a, TCGCond cond)
{
LOG_DIS("l.sfltu r%d, r%d\n", a->a, a->b);
tcg_gen_setcond_tl(TCG_COND_LTU, cpu_sr_f, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sfleu(DisasContext *dc, arg_ab *a, TCGCond cond)
{
LOG_DIS("l.sfleu r%d, r%d\n", a->a, a->b);
tcg_gen_setcond_tl(TCG_COND_LEU, cpu_sr_f, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sfgts(DisasContext *dc, arg_ab *a, TCGCond cond)
{
LOG_DIS("l.sfgts r%d, r%d\n", a->a, a->b);
tcg_gen_setcond_tl(TCG_COND_GT, cpu_sr_f, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sfges(DisasContext *dc, arg_ab *a, TCGCond cond)
{
LOG_DIS("l.sfges r%d, r%d\n", a->a, a->b);
tcg_gen_setcond_tl(TCG_COND_GE, cpu_sr_f, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sflts(DisasContext *dc, arg_ab *a, TCGCond cond)
{
LOG_DIS("l.sflts r%d, r%d\n", a->a, a->b);
tcg_gen_setcond_tl(TCG_COND_LT, cpu_sr_f, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sfles(DisasContext *dc, arg_ab *a, TCGCond cond)
{
LOG_DIS("l.sfles r%d, r%d\n", a->a, a->b);
tcg_gen_setcond_tl(TCG_COND_LE, cpu_sr_f, cpu_R[a->a], cpu_R[a->b]);
return true;
}
static bool trans_l_sfeqi(DisasContext *dc, arg_ai *a, TCGCond cond)
{
LOG_DIS("l.sfeqi r%d, %d\n", a->a, a->i);
tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_sr_f, cpu_R[a->a], a->i);
return true;
}
static bool trans_l_sfnei(DisasContext *dc, arg_ai *a, TCGCond cond)
{
LOG_DIS("l.sfnei r%d, %d\n", a->a, a->i);
tcg_gen_setcondi_tl(TCG_COND_NE, cpu_sr_f, cpu_R[a->a], a->i);
return true;
}
static bool trans_l_sfgtui(DisasContext *dc, arg_ai *a, TCGCond cond)
{
LOG_DIS("l.sfgtui r%d, %d\n", a->a, a->i);
tcg_gen_setcondi_tl(TCG_COND_GTU, cpu_sr_f, cpu_R[a->a], a->i);
return true;
}
static bool trans_l_sfgeui(DisasContext *dc, arg_ai *a, TCGCond cond)
{
LOG_DIS("l.sfgeui r%d, %d\n", a->a, a->i);
tcg_gen_setcondi_tl(TCG_COND_GEU, cpu_sr_f, cpu_R[a->a], a->i);
return true;
}
static bool trans_l_sfltui(DisasContext *dc, arg_ai *a, TCGCond cond)
{
LOG_DIS("l.sfltui r%d, %d\n", a->a, a->i);
tcg_gen_setcondi_tl(TCG_COND_LTU, cpu_sr_f, cpu_R[a->a], a->i);
return true;
}
static bool trans_l_sfleui(DisasContext *dc, arg_ai *a, TCGCond cond)
{
LOG_DIS("l.sfleui r%d, %d\n", a->a, a->i);
tcg_gen_setcondi_tl(TCG_COND_LEU, cpu_sr_f, cpu_R[a->a], a->i);
return true;
}
static bool trans_l_sfgtsi(DisasContext *dc, arg_ai *a, TCGCond cond)
{
LOG_DIS("l.sfgtsi r%d, %d\n", a->a, a->i);
tcg_gen_setcondi_tl(TCG_COND_GT, cpu_sr_f, cpu_R[a->a], a->i);
return true;
}
static bool trans_l_sfgesi(DisasContext *dc, arg_ai *a, TCGCond cond)
{
LOG_DIS("l.sfgesi r%d, %d\n", a->a, a->i);
tcg_gen_setcondi_tl(TCG_COND_GE, cpu_sr_f, cpu_R[a->a], a->i);
return true;
}
static bool trans_l_sfltsi(DisasContext *dc, arg_ai *a, TCGCond cond)
{
LOG_DIS("l.sfltsi r%d, %d\n", a->a, a->i);
tcg_gen_setcondi_tl(TCG_COND_LT, cpu_sr_f, cpu_R[a->a], a->i);
return true;
}
static bool trans_l_sflesi(DisasContext *dc, arg_ai *a, TCGCond cond)
{
LOG_DIS("l.sflesi r%d, %d\n", a->a, a->i);
tcg_gen_setcondi_tl(TCG_COND_LE, cpu_sr_f, cpu_R[a->a], a->i);
return true;
}
static bool trans_l_sys(DisasContext *dc, arg_l_sys *a, uint32_t insn)
{
LOG_DIS("l.sys %d\n", a->k);
tcg_gen_movi_tl(cpu_pc, dc->base.pc_next);
gen_exception(dc, EXCP_SYSCALL);
dc->base.is_jmp = DISAS_NORETURN;
return true;
}
static bool trans_l_trap(DisasContext *dc, arg_l_trap *a, uint32_t insn)
{
LOG_DIS("l.trap %d\n", a->k);
tcg_gen_movi_tl(cpu_pc, dc->base.pc_next);
gen_exception(dc, EXCP_TRAP);
dc->base.is_jmp = DISAS_NORETURN;
return true;
}
static bool trans_l_msync(DisasContext *dc, arg_l_msync *a, uint32_t insn)
{
LOG_DIS("l.msync\n");
tcg_gen_mb(TCG_MO_ALL);
return true;
}
static bool trans_l_psync(DisasContext *dc, arg_l_psync *a, uint32_t insn)
{
LOG_DIS("l.psync\n");
return true;
}
static bool trans_l_csync(DisasContext *dc, arg_l_csync *a, uint32_t insn)
{
LOG_DIS("l.csync\n");
return true;
}
static bool trans_l_rfe(DisasContext *dc, arg_l_rfe *a, uint32_t insn)
{
LOG_DIS("l.rfe\n");
#ifdef CONFIG_USER_ONLY
gen_illegal_exception(dc);
#else
if (dc->mem_idx == MMU_USER_IDX) {
gen_illegal_exception(dc);
} else {
gen_helper_rfe(cpu_env);
dc->base.is_jmp = DISAS_UPDATE;
}
#endif
return true;
}
static void do_fp2(DisasContext *dc, arg_da *a,
void (*fn)(TCGv, TCGv_env, TCGv))
{
check_r0_write(a->d);
fn(cpu_R[a->d], cpu_env, cpu_R[a->a]);
gen_helper_update_fpcsr(cpu_env);
}
static void do_fp3(DisasContext *dc, arg_dab *a,
void (*fn)(TCGv, TCGv_env, TCGv, TCGv))
{
check_r0_write(a->d);
fn(cpu_R[a->d], cpu_env, cpu_R[a->a], cpu_R[a->b]);
gen_helper_update_fpcsr(cpu_env);
}
static void do_fpcmp(DisasContext *dc, arg_ab *a,
void (*fn)(TCGv, TCGv_env, TCGv, TCGv),
bool inv, bool swap)
{
if (swap) {
fn(cpu_sr_f, cpu_env, cpu_R[a->b], cpu_R[a->a]);
} else {
fn(cpu_sr_f, cpu_env, cpu_R[a->a], cpu_R[a->b]);
}
if (inv) {
tcg_gen_xori_tl(cpu_sr_f, cpu_sr_f, 1);
}
gen_helper_update_fpcsr(cpu_env);
}
static bool trans_lf_add_s(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("lf.add.s r%d, r%d, r%d\n", a->d, a->a, a->b);
do_fp3(dc, a, gen_helper_float_add_s);
return true;
}
static bool trans_lf_sub_s(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("lf.sub.s r%d, r%d, r%d\n", a->d, a->a, a->b);
do_fp3(dc, a, gen_helper_float_sub_s);
return true;
}
static bool trans_lf_mul_s(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("lf.mul.s r%d, r%d, r%d\n", a->d, a->a, a->b);
do_fp3(dc, a, gen_helper_float_mul_s);
return true;
}
static bool trans_lf_div_s(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("lf.div.s r%d, r%d, r%d\n", a->d, a->a, a->b);
do_fp3(dc, a, gen_helper_float_div_s);
return true;
}
static bool trans_lf_rem_s(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("lf.rem.s r%d, r%d, r%d\n", a->d, a->a, a->b);
do_fp3(dc, a, gen_helper_float_rem_s);
return true;
}
static bool trans_lf_itof_s(DisasContext *dc, arg_da *a, uint32_t insn)
{
LOG_DIS("lf.itof.s r%d, r%d\n", a->d, a->a);
do_fp2(dc, a, gen_helper_itofs);
return true;
}
static bool trans_lf_ftoi_s(DisasContext *dc, arg_da *a, uint32_t insn)
{
LOG_DIS("lf.ftoi.s r%d, r%d\n", a->d, a->a);
do_fp2(dc, a, gen_helper_ftois);
return true;
}
static bool trans_lf_madd_s(DisasContext *dc, arg_dab *a, uint32_t insn)
{
LOG_DIS("lf.madd.s r%d, r%d, r%d\n", a->d, a->a, a->b);
check_r0_write(a->d);
gen_helper_float_madd_s(cpu_R[a->d], cpu_env, cpu_R[a->d],
cpu_R[a->a], cpu_R[a->b]);
gen_helper_update_fpcsr(cpu_env);
return true;
}
static bool trans_lf_sfeq_s(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("lf.sfeq.s r%d, r%d\n", a->a, a->b);
do_fpcmp(dc, a, gen_helper_float_eq_s, false, false);
return true;
}
static bool trans_lf_sfne_s(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("lf.sfne.s r%d, r%d\n", a->a, a->b);
do_fpcmp(dc, a, gen_helper_float_eq_s, true, false);
return true;
}
static bool trans_lf_sfgt_s(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("lf.sfgt.s r%d, r%d\n", a->a, a->b);
do_fpcmp(dc, a, gen_helper_float_lt_s, false, true);
return true;
}
static bool trans_lf_sfge_s(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("lf.sfge.s r%d, r%d\n", a->a, a->b);
do_fpcmp(dc, a, gen_helper_float_le_s, false, true);
return true;
}
static bool trans_lf_sflt_s(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("lf.sflt.s r%d, r%d\n", a->a, a->b);
do_fpcmp(dc, a, gen_helper_float_lt_s, false, false);
return true;
}
static bool trans_lf_sfle_s(DisasContext *dc, arg_ab *a, uint32_t insn)
{
LOG_DIS("lf.sfle.s r%d, r%d\n", a->a, a->b);
do_fpcmp(dc, a, gen_helper_float_le_s, false, false);
return true;
}
static void openrisc_tr_init_disas_context(DisasContextBase *dcb, CPUState *cs)
{
DisasContext *dc = container_of(dcb, DisasContext, base);
CPUOpenRISCState *env = cs->env_ptr;
int bound;
dc->mem_idx = cpu_mmu_index(env, false);
dc->tb_flags = dc->base.tb->flags;
dc->delayed_branch = (dc->tb_flags & TB_FLAGS_DFLAG) != 0;
bound = -(dc->base.pc_first | TARGET_PAGE_MASK) / 4;
dc->base.max_insns = MIN(dc->base.max_insns, bound);
}
static void openrisc_tr_tb_start(DisasContextBase *db, CPUState *cs)
{
DisasContext *dc = container_of(db, DisasContext, base);
/* Allow the TCG optimizer to see that R0 == 0,
when it's true, which is the common case. */
if (dc->tb_flags & TB_FLAGS_R0_0) {
cpu_R[0] = tcg_const_tl(0);
} else {
cpu_R[0] = cpu_R0;
}
}
static void openrisc_tr_insn_start(DisasContextBase *dcbase, CPUState *cs)
{
DisasContext *dc = container_of(dcbase, DisasContext, base);
tcg_gen_insn_start(dc->base.pc_next, (dc->delayed_branch ? 1 : 0)
| (dc->base.num_insns > 1 ? 2 : 0));
}
static bool openrisc_tr_breakpoint_check(DisasContextBase *dcbase, CPUState *cs,
const CPUBreakpoint *bp)
{
DisasContext *dc = container_of(dcbase, DisasContext, base);
tcg_gen_movi_tl(cpu_pc, dc->base.pc_next);
gen_exception(dc, EXCP_DEBUG);
dc->base.is_jmp = DISAS_NORETURN;
/* The address covered by the breakpoint must be included in
[tb->pc, tb->pc + tb->size) in order to for it to be
properly cleared -- thus we increment the PC here so that
the logic setting tb->size below does the right thing. */
dc->base.pc_next += 4;
return true;
}
static void openrisc_tr_translate_insn(DisasContextBase *dcbase, CPUState *cs)
{
DisasContext *dc = container_of(dcbase, DisasContext, base);
OpenRISCCPU *cpu = OPENRISC_CPU(cs);
uint32_t insn = cpu_ldl_code(&cpu->env, dc->base.pc_next);
if (!decode(dc, insn)) {
gen_illegal_exception(dc);
}
dc->base.pc_next += 4;
/* delay slot */
if (dc->delayed_branch) {
dc->delayed_branch--;
if (!dc->delayed_branch) {
tcg_gen_mov_tl(cpu_pc, jmp_pc);
tcg_gen_discard_tl(jmp_pc);
dc->base.is_jmp = DISAS_UPDATE;
return;
}
}
}
static void openrisc_tr_tb_stop(DisasContextBase *dcbase, CPUState *cs)
{
DisasContext *dc = container_of(dcbase, DisasContext, base);
if ((dc->tb_flags & TB_FLAGS_DFLAG ? 1 : 0) != (dc->delayed_branch != 0)) {
tcg_gen_movi_i32(cpu_dflag, dc->delayed_branch != 0);
}
tcg_gen_movi_tl(cpu_ppc, dc->base.pc_next - 4);
if (dc->base.is_jmp == DISAS_NEXT) {
dc->base.is_jmp = DISAS_UPDATE;
tcg_gen_movi_tl(cpu_pc, dc->base.pc_next);
}
if (unlikely(dc->base.singlestep_enabled)) {
gen_exception(dc, EXCP_DEBUG);
} else {
switch (dc->base.is_jmp) {
case DISAS_TOO_MANY:
gen_goto_tb(dc, 0, dc->base.pc_next);
break;
case DISAS_NORETURN:
case DISAS_JUMP:
case DISAS_TB_JUMP:
break;
case DISAS_UPDATE:
/* indicate that the hash table must be used
to find the next TB */
tcg_gen_exit_tb(0);
break;
default:
g_assert_not_reached();
}
}
}
static void openrisc_tr_disas_log(const DisasContextBase *dcbase, CPUState *cs)
{
DisasContext *s = container_of(dcbase, DisasContext, base);
qemu_log("IN: %s\n", lookup_symbol(s->base.pc_first));
log_target_disas(cs, s->base.pc_first, s->base.tb->size);
}
static const TranslatorOps openrisc_tr_ops = {
.init_disas_context = openrisc_tr_init_disas_context,
.tb_start = openrisc_tr_tb_start,
.insn_start = openrisc_tr_insn_start,
.breakpoint_check = openrisc_tr_breakpoint_check,
.translate_insn = openrisc_tr_translate_insn,
.tb_stop = openrisc_tr_tb_stop,
.disas_log = openrisc_tr_disas_log,
};
void gen_intermediate_code(CPUState *cs, struct TranslationBlock *tb)
{
DisasContext ctx;
translator_loop(&openrisc_tr_ops, &ctx.base, cs, tb);
}
void openrisc_cpu_dump_state(CPUState *cs, FILE *f,
fprintf_function cpu_fprintf,
int flags)
{
OpenRISCCPU *cpu = OPENRISC_CPU(cs);
CPUOpenRISCState *env = &cpu->env;
int i;
cpu_fprintf(f, "PC=%08x\n", env->pc);
for (i = 0; i < 32; ++i) {
cpu_fprintf(f, "R%02d=%08x%c", i, cpu_get_gpr(env, i),
(i % 4) == 3 ? '\n' : ' ');
}
}
void restore_state_to_opc(CPUOpenRISCState *env, TranslationBlock *tb,
target_ulong *data)
{
env->pc = data[0];
env->dflag = data[1] & 1;
if (data[1] & 2) {
env->ppc = env->pc - 4;
}
}