diff options
author | Yang Zhang <yang.z.zhang@intel.com> | 2015-08-28 09:58:54 +0800 |
---|---|---|
committer | Yang Zhang <yang.z.zhang@intel.com> | 2015-09-01 12:44:00 +0800 |
commit | e44e3482bdb4d0ebde2d8b41830ac2cdb07948fb (patch) | |
tree | 66b09f592c55df2878107a468a91d21506104d3f /qemu/target-cris/translate.c | |
parent | 9ca8dbcc65cfc63d6f5ef3312a33184e1d726e00 (diff) |
Add qemu 2.4.0
Change-Id: Ic99cbad4b61f8b127b7dc74d04576c0bcbaaf4f5
Signed-off-by: Yang Zhang <yang.z.zhang@intel.com>
Diffstat (limited to 'qemu/target-cris/translate.c')
-rw-r--r-- | qemu/target-cris/translate.c | 3462 |
1 files changed, 3462 insertions, 0 deletions
diff --git a/qemu/target-cris/translate.c b/qemu/target-cris/translate.c new file mode 100644 index 000000000..3e59601eb --- /dev/null +++ b/qemu/target-cris/translate.c @@ -0,0 +1,3462 @@ +/* + * CRIS emulation for qemu: main translation routines. + * + * Copyright (c) 2008 AXIS Communications AB + * Written by Edgar E. Iglesias. + * + * 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/>. + */ + +/* + * FIXME: + * The condition code translation is in need of attention. + */ + +#include "cpu.h" +#include "disas/disas.h" +#include "tcg-op.h" +#include "exec/helper-proto.h" +#include "mmu.h" +#include "exec/cpu_ldst.h" +#include "crisv32-decode.h" + +#include "exec/helper-gen.h" + +#include "trace-tcg.h" + + +#define DISAS_CRIS 0 +#if DISAS_CRIS +# define LOG_DIS(...) qemu_log_mask(CPU_LOG_TB_IN_ASM, ## __VA_ARGS__) +#else +# define LOG_DIS(...) do { } while (0) +#endif + +#define D(x) +#define BUG() (gen_BUG(dc, __FILE__, __LINE__)) +#define BUG_ON(x) ({if (x) BUG();}) + +#define DISAS_SWI 5 + +/* Used by the decoder. */ +#define EXTRACT_FIELD(src, start, end) \ + (((src) >> start) & ((1 << (end - start + 1)) - 1)) + +#define CC_MASK_NZ 0xc +#define CC_MASK_NZV 0xe +#define CC_MASK_NZVC 0xf +#define CC_MASK_RNZV 0x10e + +static TCGv_ptr cpu_env; +static TCGv cpu_R[16]; +static TCGv cpu_PR[16]; +static TCGv cc_x; +static TCGv cc_src; +static TCGv cc_dest; +static TCGv cc_result; +static TCGv cc_op; +static TCGv cc_size; +static TCGv cc_mask; + +static TCGv env_btaken; +static TCGv env_btarget; +static TCGv env_pc; + +#include "exec/gen-icount.h" + +/* This is the state at translation time. */ +typedef struct DisasContext { + CRISCPU *cpu; + target_ulong pc, ppc; + + /* Decoder. */ + unsigned int (*decoder)(CPUCRISState *env, struct DisasContext *dc); + uint32_t ir; + uint32_t opcode; + unsigned int op1; + unsigned int op2; + unsigned int zsize, zzsize; + unsigned int mode; + unsigned int postinc; + + unsigned int size; + unsigned int src; + unsigned int dst; + unsigned int cond; + + int update_cc; + int cc_op; + int cc_size; + uint32_t cc_mask; + + int cc_size_uptodate; /* -1 invalid or last written value. */ + + int cc_x_uptodate; /* 1 - ccs, 2 - known | X_FLAG. 0 not uptodate. */ + int flags_uptodate; /* Wether or not $ccs is uptodate. */ + int flagx_known; /* Wether or not flags_x has the x flag known at + translation time. */ + int flags_x; + + int clear_x; /* Clear x after this insn? */ + int clear_prefix; /* Clear prefix after this insn? */ + int clear_locked_irq; /* Clear the irq lockout. */ + int cpustate_changed; + unsigned int tb_flags; /* tb dependent flags. */ + int is_jmp; + +#define JMP_NOJMP 0 +#define JMP_DIRECT 1 +#define JMP_DIRECT_CC 2 +#define JMP_INDIRECT 3 + int jmp; /* 0=nojmp, 1=direct, 2=indirect. */ + uint32_t jmp_pc; + + int delayed_branch; + + struct TranslationBlock *tb; + int singlestep_enabled; +} DisasContext; + +static void gen_BUG(DisasContext *dc, const char *file, int line) +{ + printf("BUG: pc=%x %s %d\n", dc->pc, file, line); + qemu_log("BUG: pc=%x %s %d\n", dc->pc, file, line); + cpu_abort(CPU(dc->cpu), "%s:%d\n", file, line); +} + +static const char *regnames[] = +{ + "$r0", "$r1", "$r2", "$r3", + "$r4", "$r5", "$r6", "$r7", + "$r8", "$r9", "$r10", "$r11", + "$r12", "$r13", "$sp", "$acr", +}; +static const char *pregnames[] = +{ + "$bz", "$vr", "$pid", "$srs", + "$wz", "$exs", "$eda", "$mof", + "$dz", "$ebp", "$erp", "$srp", + "$nrp", "$ccs", "$usp", "$spc", +}; + +/* We need this table to handle preg-moves with implicit width. */ +static int preg_sizes[] = { + 1, /* bz. */ + 1, /* vr. */ + 4, /* pid. */ + 1, /* srs. */ + 2, /* wz. */ + 4, 4, 4, + 4, 4, 4, 4, + 4, 4, 4, 4, +}; + +#define t_gen_mov_TN_env(tn, member) \ + tcg_gen_ld_tl(tn, cpu_env, offsetof(CPUCRISState, member)) +#define t_gen_mov_env_TN(member, tn) \ + tcg_gen_st_tl(tn, cpu_env, offsetof(CPUCRISState, member)) + +static inline void t_gen_mov_TN_preg(TCGv tn, int r) +{ + assert(r >= 0 && r <= 15); + if (r == PR_BZ || r == PR_WZ || r == PR_DZ) { + tcg_gen_mov_tl(tn, tcg_const_tl(0)); + } else if (r == PR_VR) { + tcg_gen_mov_tl(tn, tcg_const_tl(32)); + } else { + tcg_gen_mov_tl(tn, cpu_PR[r]); + } +} +static inline void t_gen_mov_preg_TN(DisasContext *dc, int r, TCGv tn) +{ + assert(r >= 0 && r <= 15); + if (r == PR_BZ || r == PR_WZ || r == PR_DZ) { + return; + } else if (r == PR_SRS) { + tcg_gen_andi_tl(cpu_PR[r], tn, 3); + } else { + if (r == PR_PID) { + gen_helper_tlb_flush_pid(cpu_env, tn); + } + if (dc->tb_flags & S_FLAG && r == PR_SPC) { + gen_helper_spc_write(cpu_env, tn); + } else if (r == PR_CCS) { + dc->cpustate_changed = 1; + } + tcg_gen_mov_tl(cpu_PR[r], tn); + } +} + +/* Sign extend at translation time. */ +static int sign_extend(unsigned int val, unsigned int width) +{ + int sval; + + /* LSL. */ + val <<= 31 - width; + sval = val; + /* ASR. */ + sval >>= 31 - width; + return sval; +} + +static int cris_fetch(CPUCRISState *env, DisasContext *dc, uint32_t addr, + unsigned int size, unsigned int sign) +{ + int r; + + switch (size) { + case 4: + { + r = cpu_ldl_code(env, addr); + break; + } + case 2: + { + if (sign) { + r = cpu_ldsw_code(env, addr); + } else { + r = cpu_lduw_code(env, addr); + } + break; + } + case 1: + { + if (sign) { + r = cpu_ldsb_code(env, addr); + } else { + r = cpu_ldub_code(env, addr); + } + break; + } + default: + cpu_abort(CPU(dc->cpu), "Invalid fetch size %d\n", size); + break; + } + return r; +} + +static void cris_lock_irq(DisasContext *dc) +{ + dc->clear_locked_irq = 0; + t_gen_mov_env_TN(locked_irq, tcg_const_tl(1)); +} + +static inline void t_gen_raise_exception(uint32_t index) +{ + TCGv_i32 tmp = tcg_const_i32(index); + gen_helper_raise_exception(cpu_env, tmp); + tcg_temp_free_i32(tmp); +} + +static void t_gen_lsl(TCGv d, TCGv a, TCGv b) +{ + TCGv t0, t_31; + + t0 = tcg_temp_new(); + t_31 = tcg_const_tl(31); + tcg_gen_shl_tl(d, a, b); + + tcg_gen_sub_tl(t0, t_31, b); + tcg_gen_sar_tl(t0, t0, t_31); + tcg_gen_and_tl(t0, t0, d); + tcg_gen_xor_tl(d, d, t0); + tcg_temp_free(t0); + tcg_temp_free(t_31); +} + +static void t_gen_lsr(TCGv d, TCGv a, TCGv b) +{ + TCGv t0, t_31; + + t0 = tcg_temp_new(); + t_31 = tcg_temp_new(); + tcg_gen_shr_tl(d, a, b); + + tcg_gen_movi_tl(t_31, 31); + tcg_gen_sub_tl(t0, t_31, b); + tcg_gen_sar_tl(t0, t0, t_31); + tcg_gen_and_tl(t0, t0, d); + tcg_gen_xor_tl(d, d, t0); + tcg_temp_free(t0); + tcg_temp_free(t_31); +} + +static void t_gen_asr(TCGv d, TCGv a, TCGv b) +{ + TCGv t0, t_31; + + t0 = tcg_temp_new(); + t_31 = tcg_temp_new(); + tcg_gen_sar_tl(d, a, b); + + tcg_gen_movi_tl(t_31, 31); + tcg_gen_sub_tl(t0, t_31, b); + tcg_gen_sar_tl(t0, t0, t_31); + tcg_gen_or_tl(d, d, t0); + tcg_temp_free(t0); + tcg_temp_free(t_31); +} + +static void t_gen_cris_dstep(TCGv d, TCGv a, TCGv b) +{ + TCGLabel *l1 = gen_new_label(); + + /* + * d <<= 1 + * if (d >= s) + * d -= s; + */ + tcg_gen_shli_tl(d, a, 1); + tcg_gen_brcond_tl(TCG_COND_LTU, d, b, l1); + tcg_gen_sub_tl(d, d, b); + gen_set_label(l1); +} + +static void t_gen_cris_mstep(TCGv d, TCGv a, TCGv b, TCGv ccs) +{ + TCGv t; + + /* + * d <<= 1 + * if (n) + * d += s; + */ + t = tcg_temp_new(); + tcg_gen_shli_tl(d, a, 1); + tcg_gen_shli_tl(t, ccs, 31 - 3); + tcg_gen_sari_tl(t, t, 31); + tcg_gen_and_tl(t, t, b); + tcg_gen_add_tl(d, d, t); + tcg_temp_free(t); +} + +/* Extended arithmetics on CRIS. */ +static inline void t_gen_add_flag(TCGv d, int flag) +{ + TCGv c; + + c = tcg_temp_new(); + t_gen_mov_TN_preg(c, PR_CCS); + /* Propagate carry into d. */ + tcg_gen_andi_tl(c, c, 1 << flag); + if (flag) { + tcg_gen_shri_tl(c, c, flag); + } + tcg_gen_add_tl(d, d, c); + tcg_temp_free(c); +} + +static inline void t_gen_addx_carry(DisasContext *dc, TCGv d) +{ + if (dc->flagx_known) { + if (dc->flags_x) { + TCGv c; + + c = tcg_temp_new(); + t_gen_mov_TN_preg(c, PR_CCS); + /* C flag is already at bit 0. */ + tcg_gen_andi_tl(c, c, C_FLAG); + tcg_gen_add_tl(d, d, c); + tcg_temp_free(c); + } + } else { + TCGv x, c; + + x = tcg_temp_new(); + c = tcg_temp_new(); + t_gen_mov_TN_preg(x, PR_CCS); + tcg_gen_mov_tl(c, x); + + /* Propagate carry into d if X is set. Branch free. */ + tcg_gen_andi_tl(c, c, C_FLAG); + tcg_gen_andi_tl(x, x, X_FLAG); + tcg_gen_shri_tl(x, x, 4); + + tcg_gen_and_tl(x, x, c); + tcg_gen_add_tl(d, d, x); + tcg_temp_free(x); + tcg_temp_free(c); + } +} + +static inline void t_gen_subx_carry(DisasContext *dc, TCGv d) +{ + if (dc->flagx_known) { + if (dc->flags_x) { + TCGv c; + + c = tcg_temp_new(); + t_gen_mov_TN_preg(c, PR_CCS); + /* C flag is already at bit 0. */ + tcg_gen_andi_tl(c, c, C_FLAG); + tcg_gen_sub_tl(d, d, c); + tcg_temp_free(c); + } + } else { + TCGv x, c; + + x = tcg_temp_new(); + c = tcg_temp_new(); + t_gen_mov_TN_preg(x, PR_CCS); + tcg_gen_mov_tl(c, x); + + /* Propagate carry into d if X is set. Branch free. */ + tcg_gen_andi_tl(c, c, C_FLAG); + tcg_gen_andi_tl(x, x, X_FLAG); + tcg_gen_shri_tl(x, x, 4); + + tcg_gen_and_tl(x, x, c); + tcg_gen_sub_tl(d, d, x); + tcg_temp_free(x); + tcg_temp_free(c); + } +} + +/* Swap the two bytes within each half word of the s operand. + T0 = ((T0 << 8) & 0xff00ff00) | ((T0 >> 8) & 0x00ff00ff) */ +static inline void t_gen_swapb(TCGv d, TCGv s) +{ + TCGv t, org_s; + + t = tcg_temp_new(); + org_s = tcg_temp_new(); + + /* d and s may refer to the same object. */ + tcg_gen_mov_tl(org_s, s); + tcg_gen_shli_tl(t, org_s, 8); + tcg_gen_andi_tl(d, t, 0xff00ff00); + tcg_gen_shri_tl(t, org_s, 8); + tcg_gen_andi_tl(t, t, 0x00ff00ff); + tcg_gen_or_tl(d, d, t); + tcg_temp_free(t); + tcg_temp_free(org_s); +} + +/* Swap the halfwords of the s operand. */ +static inline void t_gen_swapw(TCGv d, TCGv s) +{ + TCGv t; + /* d and s refer the same object. */ + t = tcg_temp_new(); + tcg_gen_mov_tl(t, s); + tcg_gen_shli_tl(d, t, 16); + tcg_gen_shri_tl(t, t, 16); + tcg_gen_or_tl(d, d, t); + tcg_temp_free(t); +} + +/* Reverse the within each byte. + T0 = (((T0 << 7) & 0x80808080) | + ((T0 << 5) & 0x40404040) | + ((T0 << 3) & 0x20202020) | + ((T0 << 1) & 0x10101010) | + ((T0 >> 1) & 0x08080808) | + ((T0 >> 3) & 0x04040404) | + ((T0 >> 5) & 0x02020202) | + ((T0 >> 7) & 0x01010101)); + */ +static inline void t_gen_swapr(TCGv d, TCGv s) +{ + struct { + int shift; /* LSL when positive, LSR when negative. */ + uint32_t mask; + } bitrev[] = { + {7, 0x80808080}, + {5, 0x40404040}, + {3, 0x20202020}, + {1, 0x10101010}, + {-1, 0x08080808}, + {-3, 0x04040404}, + {-5, 0x02020202}, + {-7, 0x01010101} + }; + int i; + TCGv t, org_s; + + /* d and s refer the same object. */ + t = tcg_temp_new(); + org_s = tcg_temp_new(); + tcg_gen_mov_tl(org_s, s); + + tcg_gen_shli_tl(t, org_s, bitrev[0].shift); + tcg_gen_andi_tl(d, t, bitrev[0].mask); + for (i = 1; i < ARRAY_SIZE(bitrev); i++) { + if (bitrev[i].shift >= 0) { + tcg_gen_shli_tl(t, org_s, bitrev[i].shift); + } else { + tcg_gen_shri_tl(t, org_s, -bitrev[i].shift); + } + tcg_gen_andi_tl(t, t, bitrev[i].mask); + tcg_gen_or_tl(d, d, t); + } + tcg_temp_free(t); + tcg_temp_free(org_s); +} + +static void t_gen_cc_jmp(TCGv pc_true, TCGv pc_false) +{ + TCGLabel *l1 = gen_new_label(); + + /* Conditional jmp. */ + tcg_gen_mov_tl(env_pc, pc_false); + tcg_gen_brcondi_tl(TCG_COND_EQ, env_btaken, 0, l1); + tcg_gen_mov_tl(env_pc, pc_true); + gen_set_label(l1); +} + +static void gen_goto_tb(DisasContext *dc, int n, target_ulong dest) +{ + TranslationBlock *tb; + tb = dc->tb; + if ((tb->pc & TARGET_PAGE_MASK) == (dest & TARGET_PAGE_MASK)) { + tcg_gen_goto_tb(n); + tcg_gen_movi_tl(env_pc, dest); + tcg_gen_exit_tb((uintptr_t)tb + n); + } else { + tcg_gen_movi_tl(env_pc, dest); + tcg_gen_exit_tb(0); + } +} + +static inline void cris_clear_x_flag(DisasContext *dc) +{ + if (dc->flagx_known && dc->flags_x) { + dc->flags_uptodate = 0; + } + + dc->flagx_known = 1; + dc->flags_x = 0; +} + +static void cris_flush_cc_state(DisasContext *dc) +{ + if (dc->cc_size_uptodate != dc->cc_size) { + tcg_gen_movi_tl(cc_size, dc->cc_size); + dc->cc_size_uptodate = dc->cc_size; + } + tcg_gen_movi_tl(cc_op, dc->cc_op); + tcg_gen_movi_tl(cc_mask, dc->cc_mask); +} + +static void cris_evaluate_flags(DisasContext *dc) +{ + if (dc->flags_uptodate) { + return; + } + + cris_flush_cc_state(dc); + + switch (dc->cc_op) { + case CC_OP_MCP: + gen_helper_evaluate_flags_mcp(cpu_PR[PR_CCS], cpu_env, + cpu_PR[PR_CCS], cc_src, + cc_dest, cc_result); + break; + case CC_OP_MULS: + gen_helper_evaluate_flags_muls(cpu_PR[PR_CCS], cpu_env, + cpu_PR[PR_CCS], cc_result, + cpu_PR[PR_MOF]); + break; + case CC_OP_MULU: + gen_helper_evaluate_flags_mulu(cpu_PR[PR_CCS], cpu_env, + cpu_PR[PR_CCS], cc_result, + cpu_PR[PR_MOF]); + break; + case CC_OP_MOVE: + case CC_OP_AND: + case CC_OP_OR: + case CC_OP_XOR: + case CC_OP_ASR: + case CC_OP_LSR: + case CC_OP_LSL: + switch (dc->cc_size) { + case 4: + gen_helper_evaluate_flags_move_4(cpu_PR[PR_CCS], + cpu_env, cpu_PR[PR_CCS], cc_result); + break; + case 2: + gen_helper_evaluate_flags_move_2(cpu_PR[PR_CCS], + cpu_env, cpu_PR[PR_CCS], cc_result); + break; + default: + gen_helper_evaluate_flags(cpu_env); + break; + } + break; + case CC_OP_FLAGS: + /* live. */ + break; + case CC_OP_SUB: + case CC_OP_CMP: + if (dc->cc_size == 4) { + gen_helper_evaluate_flags_sub_4(cpu_PR[PR_CCS], cpu_env, + cpu_PR[PR_CCS], cc_src, cc_dest, cc_result); + } else { + gen_helper_evaluate_flags(cpu_env); + } + + break; + default: + switch (dc->cc_size) { + case 4: + gen_helper_evaluate_flags_alu_4(cpu_PR[PR_CCS], cpu_env, + cpu_PR[PR_CCS], cc_src, cc_dest, cc_result); + break; + default: + gen_helper_evaluate_flags(cpu_env); + break; + } + break; + } + + if (dc->flagx_known) { + if (dc->flags_x) { + tcg_gen_ori_tl(cpu_PR[PR_CCS], cpu_PR[PR_CCS], X_FLAG); + } else if (dc->cc_op == CC_OP_FLAGS) { + tcg_gen_andi_tl(cpu_PR[PR_CCS], cpu_PR[PR_CCS], ~X_FLAG); + } + } + dc->flags_uptodate = 1; +} + +static void cris_cc_mask(DisasContext *dc, unsigned int mask) +{ + uint32_t ovl; + + if (!mask) { + dc->update_cc = 0; + return; + } + + /* Check if we need to evaluate the condition codes due to + CC overlaying. */ + ovl = (dc->cc_mask ^ mask) & ~mask; + if (ovl) { + /* TODO: optimize this case. It trigs all the time. */ + cris_evaluate_flags(dc); + } + dc->cc_mask = mask; + dc->update_cc = 1; +} + +static void cris_update_cc_op(DisasContext *dc, int op, int size) +{ + dc->cc_op = op; + dc->cc_size = size; + dc->flags_uptodate = 0; +} + +static inline void cris_update_cc_x(DisasContext *dc) +{ + /* Save the x flag state at the time of the cc snapshot. */ + if (dc->flagx_known) { + if (dc->cc_x_uptodate == (2 | dc->flags_x)) { + return; + } + tcg_gen_movi_tl(cc_x, dc->flags_x); + dc->cc_x_uptodate = 2 | dc->flags_x; + } else { + tcg_gen_andi_tl(cc_x, cpu_PR[PR_CCS], X_FLAG); + dc->cc_x_uptodate = 1; + } +} + +/* Update cc prior to executing ALU op. Needs source operands untouched. */ +static void cris_pre_alu_update_cc(DisasContext *dc, int op, + TCGv dst, TCGv src, int size) +{ + if (dc->update_cc) { + cris_update_cc_op(dc, op, size); + tcg_gen_mov_tl(cc_src, src); + + if (op != CC_OP_MOVE + && op != CC_OP_AND + && op != CC_OP_OR + && op != CC_OP_XOR + && op != CC_OP_ASR + && op != CC_OP_LSR + && op != CC_OP_LSL) { + tcg_gen_mov_tl(cc_dest, dst); + } + + cris_update_cc_x(dc); + } +} + +/* Update cc after executing ALU op. needs the result. */ +static inline void cris_update_result(DisasContext *dc, TCGv res) +{ + if (dc->update_cc) { + tcg_gen_mov_tl(cc_result, res); + } +} + +/* Returns one if the write back stage should execute. */ +static void cris_alu_op_exec(DisasContext *dc, int op, + TCGv dst, TCGv a, TCGv b, int size) +{ + /* Emit the ALU insns. */ + switch (op) { + case CC_OP_ADD: + tcg_gen_add_tl(dst, a, b); + /* Extended arithmetics. */ + t_gen_addx_carry(dc, dst); + break; + case CC_OP_ADDC: + tcg_gen_add_tl(dst, a, b); + t_gen_add_flag(dst, 0); /* C_FLAG. */ + break; + case CC_OP_MCP: + tcg_gen_add_tl(dst, a, b); + t_gen_add_flag(dst, 8); /* R_FLAG. */ + break; + case CC_OP_SUB: + tcg_gen_sub_tl(dst, a, b); + /* Extended arithmetics. */ + t_gen_subx_carry(dc, dst); + break; + case CC_OP_MOVE: + tcg_gen_mov_tl(dst, b); + break; + case CC_OP_OR: + tcg_gen_or_tl(dst, a, b); + break; + case CC_OP_AND: + tcg_gen_and_tl(dst, a, b); + break; + case CC_OP_XOR: + tcg_gen_xor_tl(dst, a, b); + break; + case CC_OP_LSL: + t_gen_lsl(dst, a, b); + break; + case CC_OP_LSR: + t_gen_lsr(dst, a, b); + break; + case CC_OP_ASR: + t_gen_asr(dst, a, b); + break; + case CC_OP_NEG: + tcg_gen_neg_tl(dst, b); + /* Extended arithmetics. */ + t_gen_subx_carry(dc, dst); + break; + case CC_OP_LZ: + gen_helper_lz(dst, b); + break; + case CC_OP_MULS: + tcg_gen_muls2_tl(dst, cpu_PR[PR_MOF], a, b); + break; + case CC_OP_MULU: + tcg_gen_mulu2_tl(dst, cpu_PR[PR_MOF], a, b); + break; + case CC_OP_DSTEP: + t_gen_cris_dstep(dst, a, b); + break; + case CC_OP_MSTEP: + t_gen_cris_mstep(dst, a, b, cpu_PR[PR_CCS]); + break; + case CC_OP_BOUND: + { + TCGLabel *l1 = gen_new_label(); + tcg_gen_mov_tl(dst, a); + tcg_gen_brcond_tl(TCG_COND_LEU, a, b, l1); + tcg_gen_mov_tl(dst, b); + gen_set_label(l1); + } + break; + case CC_OP_CMP: + tcg_gen_sub_tl(dst, a, b); + /* Extended arithmetics. */ + t_gen_subx_carry(dc, dst); + break; + default: + qemu_log("illegal ALU op.\n"); + BUG(); + break; + } + + if (size == 1) { + tcg_gen_andi_tl(dst, dst, 0xff); + } else if (size == 2) { + tcg_gen_andi_tl(dst, dst, 0xffff); + } +} + +static void cris_alu(DisasContext *dc, int op, + TCGv d, TCGv op_a, TCGv op_b, int size) +{ + TCGv tmp; + int writeback; + + writeback = 1; + + if (op == CC_OP_CMP) { + tmp = tcg_temp_new(); + writeback = 0; + } else if (size == 4) { + tmp = d; + writeback = 0; + } else { + tmp = tcg_temp_new(); + } + + + cris_pre_alu_update_cc(dc, op, op_a, op_b, size); + cris_alu_op_exec(dc, op, tmp, op_a, op_b, size); + cris_update_result(dc, tmp); + + /* Writeback. */ + if (writeback) { + if (size == 1) { + tcg_gen_andi_tl(d, d, ~0xff); + } else { + tcg_gen_andi_tl(d, d, ~0xffff); + } + tcg_gen_or_tl(d, d, tmp); + } + if (!TCGV_EQUAL(tmp, d)) { + tcg_temp_free(tmp); + } +} + +static int arith_cc(DisasContext *dc) +{ + if (dc->update_cc) { + switch (dc->cc_op) { + case CC_OP_ADDC: return 1; + case CC_OP_ADD: return 1; + case CC_OP_SUB: return 1; + case CC_OP_DSTEP: return 1; + case CC_OP_LSL: return 1; + case CC_OP_LSR: return 1; + case CC_OP_ASR: return 1; + case CC_OP_CMP: return 1; + case CC_OP_NEG: return 1; + case CC_OP_OR: return 1; + case CC_OP_AND: return 1; + case CC_OP_XOR: return 1; + case CC_OP_MULU: return 1; + case CC_OP_MULS: return 1; + default: + return 0; + } + } + return 0; +} + +static void gen_tst_cc (DisasContext *dc, TCGv cc, int cond) +{ + int arith_opt, move_opt; + + /* TODO: optimize more condition codes. */ + + /* + * If the flags are live, we've gotta look into the bits of CCS. + * Otherwise, if we just did an arithmetic operation we try to + * evaluate the condition code faster. + * + * When this function is done, T0 should be non-zero if the condition + * code is true. + */ + arith_opt = arith_cc(dc) && !dc->flags_uptodate; + move_opt = (dc->cc_op == CC_OP_MOVE); + switch (cond) { + case CC_EQ: + if ((arith_opt || move_opt) + && dc->cc_x_uptodate != (2 | X_FLAG)) { + tcg_gen_setcond_tl(TCG_COND_EQ, cc, + cc_result, tcg_const_tl(0)); + } else { + cris_evaluate_flags(dc); + tcg_gen_andi_tl(cc, + cpu_PR[PR_CCS], Z_FLAG); + } + break; + case CC_NE: + if ((arith_opt || move_opt) + && dc->cc_x_uptodate != (2 | X_FLAG)) { + tcg_gen_mov_tl(cc, cc_result); + } else { + cris_evaluate_flags(dc); + tcg_gen_xori_tl(cc, cpu_PR[PR_CCS], + Z_FLAG); + tcg_gen_andi_tl(cc, cc, Z_FLAG); + } + break; + case CC_CS: + cris_evaluate_flags(dc); + tcg_gen_andi_tl(cc, cpu_PR[PR_CCS], C_FLAG); + break; + case CC_CC: + cris_evaluate_flags(dc); + tcg_gen_xori_tl(cc, cpu_PR[PR_CCS], C_FLAG); + tcg_gen_andi_tl(cc, cc, C_FLAG); + break; + case CC_VS: + cris_evaluate_flags(dc); + tcg_gen_andi_tl(cc, cpu_PR[PR_CCS], V_FLAG); + break; + case CC_VC: + cris_evaluate_flags(dc); + tcg_gen_xori_tl(cc, cpu_PR[PR_CCS], + V_FLAG); + tcg_gen_andi_tl(cc, cc, V_FLAG); + break; + case CC_PL: + if (arith_opt || move_opt) { + int bits = 31; + + if (dc->cc_size == 1) { + bits = 7; + } else if (dc->cc_size == 2) { + bits = 15; + } + + tcg_gen_shri_tl(cc, cc_result, bits); + tcg_gen_xori_tl(cc, cc, 1); + } else { + cris_evaluate_flags(dc); + tcg_gen_xori_tl(cc, cpu_PR[PR_CCS], + N_FLAG); + tcg_gen_andi_tl(cc, cc, N_FLAG); + } + break; + case CC_MI: + if (arith_opt || move_opt) { + int bits = 31; + + if (dc->cc_size == 1) { + bits = 7; + } else if (dc->cc_size == 2) { + bits = 15; + } + + tcg_gen_shri_tl(cc, cc_result, bits); + tcg_gen_andi_tl(cc, cc, 1); + } else { + cris_evaluate_flags(dc); + tcg_gen_andi_tl(cc, cpu_PR[PR_CCS], + N_FLAG); + } + break; + case CC_LS: + cris_evaluate_flags(dc); + tcg_gen_andi_tl(cc, cpu_PR[PR_CCS], + C_FLAG | Z_FLAG); + break; + case CC_HI: + cris_evaluate_flags(dc); + { + TCGv tmp; + + tmp = tcg_temp_new(); + tcg_gen_xori_tl(tmp, cpu_PR[PR_CCS], + C_FLAG | Z_FLAG); + /* Overlay the C flag on top of the Z. */ + tcg_gen_shli_tl(cc, tmp, 2); + tcg_gen_and_tl(cc, tmp, cc); + tcg_gen_andi_tl(cc, cc, Z_FLAG); + + tcg_temp_free(tmp); + } + break; + case CC_GE: + cris_evaluate_flags(dc); + /* Overlay the V flag on top of the N. */ + tcg_gen_shli_tl(cc, cpu_PR[PR_CCS], 2); + tcg_gen_xor_tl(cc, + cpu_PR[PR_CCS], cc); + tcg_gen_andi_tl(cc, cc, N_FLAG); + tcg_gen_xori_tl(cc, cc, N_FLAG); + break; + case CC_LT: + cris_evaluate_flags(dc); + /* Overlay the V flag on top of the N. */ + tcg_gen_shli_tl(cc, cpu_PR[PR_CCS], 2); + tcg_gen_xor_tl(cc, + cpu_PR[PR_CCS], cc); + tcg_gen_andi_tl(cc, cc, N_FLAG); + break; + case CC_GT: + cris_evaluate_flags(dc); + { + TCGv n, z; + + n = tcg_temp_new(); + z = tcg_temp_new(); + + /* To avoid a shift we overlay everything on + the V flag. */ + tcg_gen_shri_tl(n, cpu_PR[PR_CCS], 2); + tcg_gen_shri_tl(z, cpu_PR[PR_CCS], 1); + /* invert Z. */ + tcg_gen_xori_tl(z, z, 2); + + tcg_gen_xor_tl(n, n, cpu_PR[PR_CCS]); + tcg_gen_xori_tl(n, n, 2); + tcg_gen_and_tl(cc, z, n); + tcg_gen_andi_tl(cc, cc, 2); + + tcg_temp_free(n); + tcg_temp_free(z); + } + break; + case CC_LE: + cris_evaluate_flags(dc); + { + TCGv n, z; + + n = tcg_temp_new(); + z = tcg_temp_new(); + + /* To avoid a shift we overlay everything on + the V flag. */ + tcg_gen_shri_tl(n, cpu_PR[PR_CCS], 2); + tcg_gen_shri_tl(z, cpu_PR[PR_CCS], 1); + + tcg_gen_xor_tl(n, n, cpu_PR[PR_CCS]); + tcg_gen_or_tl(cc, z, n); + tcg_gen_andi_tl(cc, cc, 2); + + tcg_temp_free(n); + tcg_temp_free(z); + } + break; + case CC_P: + cris_evaluate_flags(dc); + tcg_gen_andi_tl(cc, cpu_PR[PR_CCS], P_FLAG); + break; + case CC_A: + tcg_gen_movi_tl(cc, 1); + break; + default: + BUG(); + break; + }; +} + +static void cris_store_direct_jmp(DisasContext *dc) +{ + /* Store the direct jmp state into the cpu-state. */ + if (dc->jmp == JMP_DIRECT || dc->jmp == JMP_DIRECT_CC) { + if (dc->jmp == JMP_DIRECT) { + tcg_gen_movi_tl(env_btaken, 1); + } + tcg_gen_movi_tl(env_btarget, dc->jmp_pc); + dc->jmp = JMP_INDIRECT; + } +} + +static void cris_prepare_cc_branch (DisasContext *dc, + int offset, int cond) +{ + /* This helps us re-schedule the micro-code to insns in delay-slots + before the actual jump. */ + dc->delayed_branch = 2; + dc->jmp = JMP_DIRECT_CC; + dc->jmp_pc = dc->pc + offset; + + gen_tst_cc(dc, env_btaken, cond); + tcg_gen_movi_tl(env_btarget, dc->jmp_pc); +} + + +/* jumps, when the dest is in a live reg for example. Direct should be set + when the dest addr is constant to allow tb chaining. */ +static inline void cris_prepare_jmp (DisasContext *dc, unsigned int type) +{ + /* This helps us re-schedule the micro-code to insns in delay-slots + before the actual jump. */ + dc->delayed_branch = 2; + dc->jmp = type; + if (type == JMP_INDIRECT) { + tcg_gen_movi_tl(env_btaken, 1); + } +} + +static void gen_load64(DisasContext *dc, TCGv_i64 dst, TCGv addr) +{ + int mem_index = cpu_mmu_index(&dc->cpu->env); + + /* If we get a fault on a delayslot we must keep the jmp state in + the cpu-state to be able to re-execute the jmp. */ + if (dc->delayed_branch == 1) { + cris_store_direct_jmp(dc); + } + + tcg_gen_qemu_ld_i64(dst, addr, mem_index, MO_TEQ); +} + +static void gen_load(DisasContext *dc, TCGv dst, TCGv addr, + unsigned int size, int sign) +{ + int mem_index = cpu_mmu_index(&dc->cpu->env); + + /* If we get a fault on a delayslot we must keep the jmp state in + the cpu-state to be able to re-execute the jmp. */ + if (dc->delayed_branch == 1) { + cris_store_direct_jmp(dc); + } + + tcg_gen_qemu_ld_tl(dst, addr, mem_index, + MO_TE + ctz32(size) + (sign ? MO_SIGN : 0)); +} + +static void gen_store (DisasContext *dc, TCGv addr, TCGv val, + unsigned int size) +{ + int mem_index = cpu_mmu_index(&dc->cpu->env); + + /* If we get a fault on a delayslot we must keep the jmp state in + the cpu-state to be able to re-execute the jmp. */ + if (dc->delayed_branch == 1) { + cris_store_direct_jmp(dc); + } + + + /* Conditional writes. We only support the kind were X and P are known + at translation time. */ + if (dc->flagx_known && dc->flags_x && (dc->tb_flags & P_FLAG)) { + dc->postinc = 0; + cris_evaluate_flags(dc); + tcg_gen_ori_tl(cpu_PR[PR_CCS], cpu_PR[PR_CCS], C_FLAG); + return; + } + + tcg_gen_qemu_st_tl(val, addr, mem_index, MO_TE + ctz32(size)); + + if (dc->flagx_known && dc->flags_x) { + cris_evaluate_flags(dc); + tcg_gen_andi_tl(cpu_PR[PR_CCS], cpu_PR[PR_CCS], ~C_FLAG); + } +} + +static inline void t_gen_sext(TCGv d, TCGv s, int size) +{ + if (size == 1) { + tcg_gen_ext8s_i32(d, s); + } else if (size == 2) { + tcg_gen_ext16s_i32(d, s); + } else if (!TCGV_EQUAL(d, s)) { + tcg_gen_mov_tl(d, s); + } +} + +static inline void t_gen_zext(TCGv d, TCGv s, int size) +{ + if (size == 1) { + tcg_gen_ext8u_i32(d, s); + } else if (size == 2) { + tcg_gen_ext16u_i32(d, s); + } else if (!TCGV_EQUAL(d, s)) { + tcg_gen_mov_tl(d, s); + } +} + +#if DISAS_CRIS +static char memsize_char(int size) +{ + switch (size) { + case 1: return 'b'; break; + case 2: return 'w'; break; + case 4: return 'd'; break; + default: + return 'x'; + break; + } +} +#endif + +static inline unsigned int memsize_z(DisasContext *dc) +{ + return dc->zsize + 1; +} + +static inline unsigned int memsize_zz(DisasContext *dc) +{ + switch (dc->zzsize) { + case 0: return 1; + case 1: return 2; + default: + return 4; + } +} + +static inline void do_postinc (DisasContext *dc, int size) +{ + if (dc->postinc) { + tcg_gen_addi_tl(cpu_R[dc->op1], cpu_R[dc->op1], size); + } +} + +static inline void dec_prep_move_r(DisasContext *dc, int rs, int rd, + int size, int s_ext, TCGv dst) +{ + if (s_ext) { + t_gen_sext(dst, cpu_R[rs], size); + } else { + t_gen_zext(dst, cpu_R[rs], size); + } +} + +/* Prepare T0 and T1 for a register alu operation. + s_ext decides if the operand1 should be sign-extended or zero-extended when + needed. */ +static void dec_prep_alu_r(DisasContext *dc, int rs, int rd, + int size, int s_ext, TCGv dst, TCGv src) +{ + dec_prep_move_r(dc, rs, rd, size, s_ext, src); + + if (s_ext) { + t_gen_sext(dst, cpu_R[rd], size); + } else { + t_gen_zext(dst, cpu_R[rd], size); + } +} + +static int dec_prep_move_m(CPUCRISState *env, DisasContext *dc, + int s_ext, int memsize, TCGv dst) +{ + unsigned int rs; + uint32_t imm; + int is_imm; + int insn_len = 2; + + rs = dc->op1; + is_imm = rs == 15 && dc->postinc; + + /* Load [$rs] onto T1. */ + if (is_imm) { + insn_len = 2 + memsize; + if (memsize == 1) { + insn_len++; + } + + imm = cris_fetch(env, dc, dc->pc + 2, memsize, s_ext); + tcg_gen_movi_tl(dst, imm); + dc->postinc = 0; + } else { + cris_flush_cc_state(dc); + gen_load(dc, dst, cpu_R[rs], memsize, 0); + if (s_ext) { + t_gen_sext(dst, dst, memsize); + } else { + t_gen_zext(dst, dst, memsize); + } + } + return insn_len; +} + +/* Prepare T0 and T1 for a memory + alu operation. + s_ext decides if the operand1 should be sign-extended or zero-extended when + needed. */ +static int dec_prep_alu_m(CPUCRISState *env, DisasContext *dc, + int s_ext, int memsize, TCGv dst, TCGv src) +{ + int insn_len; + + insn_len = dec_prep_move_m(env, dc, s_ext, memsize, src); + tcg_gen_mov_tl(dst, cpu_R[dc->op2]); + return insn_len; +} + +#if DISAS_CRIS +static const char *cc_name(int cc) +{ + static const char *cc_names[16] = { + "cc", "cs", "ne", "eq", "vc", "vs", "pl", "mi", + "ls", "hi", "ge", "lt", "gt", "le", "a", "p" + }; + assert(cc < 16); + return cc_names[cc]; +} +#endif + +/* Start of insn decoders. */ + +static int dec_bccq(CPUCRISState *env, DisasContext *dc) +{ + int32_t offset; + int sign; + uint32_t cond = dc->op2; + + offset = EXTRACT_FIELD(dc->ir, 1, 7); + sign = EXTRACT_FIELD(dc->ir, 0, 0); + + offset *= 2; + offset |= sign << 8; + offset = sign_extend(offset, 8); + + LOG_DIS("b%s %x\n", cc_name(cond), dc->pc + offset); + + /* op2 holds the condition-code. */ + cris_cc_mask(dc, 0); + cris_prepare_cc_branch(dc, offset, cond); + return 2; +} +static int dec_addoq(CPUCRISState *env, DisasContext *dc) +{ + int32_t imm; + + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 7); + imm = sign_extend(dc->op1, 7); + + LOG_DIS("addoq %d, $r%u\n", imm, dc->op2); + cris_cc_mask(dc, 0); + /* Fetch register operand, */ + tcg_gen_addi_tl(cpu_R[R_ACR], cpu_R[dc->op2], imm); + + return 2; +} +static int dec_addq(CPUCRISState *env, DisasContext *dc) +{ + LOG_DIS("addq %u, $r%u\n", dc->op1, dc->op2); + + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 5); + + cris_cc_mask(dc, CC_MASK_NZVC); + + cris_alu(dc, CC_OP_ADD, + cpu_R[dc->op2], cpu_R[dc->op2], tcg_const_tl(dc->op1), 4); + return 2; +} +static int dec_moveq(CPUCRISState *env, DisasContext *dc) +{ + uint32_t imm; + + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 5); + imm = sign_extend(dc->op1, 5); + LOG_DIS("moveq %d, $r%u\n", imm, dc->op2); + + tcg_gen_movi_tl(cpu_R[dc->op2], imm); + return 2; +} +static int dec_subq(CPUCRISState *env, DisasContext *dc) +{ + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 5); + + LOG_DIS("subq %u, $r%u\n", dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_SUB, + cpu_R[dc->op2], cpu_R[dc->op2], tcg_const_tl(dc->op1), 4); + return 2; +} +static int dec_cmpq(CPUCRISState *env, DisasContext *dc) +{ + uint32_t imm; + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 5); + imm = sign_extend(dc->op1, 5); + + LOG_DIS("cmpq %d, $r%d\n", imm, dc->op2); + cris_cc_mask(dc, CC_MASK_NZVC); + + cris_alu(dc, CC_OP_CMP, + cpu_R[dc->op2], cpu_R[dc->op2], tcg_const_tl(imm), 4); + return 2; +} +static int dec_andq(CPUCRISState *env, DisasContext *dc) +{ + uint32_t imm; + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 5); + imm = sign_extend(dc->op1, 5); + + LOG_DIS("andq %d, $r%d\n", imm, dc->op2); + cris_cc_mask(dc, CC_MASK_NZ); + + cris_alu(dc, CC_OP_AND, + cpu_R[dc->op2], cpu_R[dc->op2], tcg_const_tl(imm), 4); + return 2; +} +static int dec_orq(CPUCRISState *env, DisasContext *dc) +{ + uint32_t imm; + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 5); + imm = sign_extend(dc->op1, 5); + LOG_DIS("orq %d, $r%d\n", imm, dc->op2); + cris_cc_mask(dc, CC_MASK_NZ); + + cris_alu(dc, CC_OP_OR, + cpu_R[dc->op2], cpu_R[dc->op2], tcg_const_tl(imm), 4); + return 2; +} +static int dec_btstq(CPUCRISState *env, DisasContext *dc) +{ + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 4); + LOG_DIS("btstq %u, $r%d\n", dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZ); + cris_evaluate_flags(dc); + gen_helper_btst(cpu_PR[PR_CCS], cpu_env, cpu_R[dc->op2], + tcg_const_tl(dc->op1), cpu_PR[PR_CCS]); + cris_alu(dc, CC_OP_MOVE, + cpu_R[dc->op2], cpu_R[dc->op2], cpu_R[dc->op2], 4); + cris_update_cc_op(dc, CC_OP_FLAGS, 4); + dc->flags_uptodate = 1; + return 2; +} +static int dec_asrq(CPUCRISState *env, DisasContext *dc) +{ + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 4); + LOG_DIS("asrq %u, $r%d\n", dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZ); + + tcg_gen_sari_tl(cpu_R[dc->op2], cpu_R[dc->op2], dc->op1); + cris_alu(dc, CC_OP_MOVE, + cpu_R[dc->op2], + cpu_R[dc->op2], cpu_R[dc->op2], 4); + return 2; +} +static int dec_lslq(CPUCRISState *env, DisasContext *dc) +{ + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 4); + LOG_DIS("lslq %u, $r%d\n", dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZ); + + tcg_gen_shli_tl(cpu_R[dc->op2], cpu_R[dc->op2], dc->op1); + + cris_alu(dc, CC_OP_MOVE, + cpu_R[dc->op2], + cpu_R[dc->op2], cpu_R[dc->op2], 4); + return 2; +} +static int dec_lsrq(CPUCRISState *env, DisasContext *dc) +{ + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 4); + LOG_DIS("lsrq %u, $r%d\n", dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZ); + + tcg_gen_shri_tl(cpu_R[dc->op2], cpu_R[dc->op2], dc->op1); + cris_alu(dc, CC_OP_MOVE, + cpu_R[dc->op2], + cpu_R[dc->op2], cpu_R[dc->op2], 4); + return 2; +} + +static int dec_move_r(CPUCRISState *env, DisasContext *dc) +{ + int size = memsize_zz(dc); + + LOG_DIS("move.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZ); + if (size == 4) { + dec_prep_move_r(dc, dc->op1, dc->op2, size, 0, cpu_R[dc->op2]); + cris_cc_mask(dc, CC_MASK_NZ); + cris_update_cc_op(dc, CC_OP_MOVE, 4); + cris_update_cc_x(dc); + cris_update_result(dc, cpu_R[dc->op2]); + } else { + TCGv t0; + + t0 = tcg_temp_new(); + dec_prep_move_r(dc, dc->op1, dc->op2, size, 0, t0); + cris_alu(dc, CC_OP_MOVE, + cpu_R[dc->op2], + cpu_R[dc->op2], t0, size); + tcg_temp_free(t0); + } + return 2; +} + +static int dec_scc_r(CPUCRISState *env, DisasContext *dc) +{ + int cond = dc->op2; + + LOG_DIS("s%s $r%u\n", + cc_name(cond), dc->op1); + + if (cond != CC_A) { + TCGLabel *l1 = gen_new_label(); + gen_tst_cc(dc, cpu_R[dc->op1], cond); + tcg_gen_brcondi_tl(TCG_COND_EQ, cpu_R[dc->op1], 0, l1); + tcg_gen_movi_tl(cpu_R[dc->op1], 1); + gen_set_label(l1); + } else { + tcg_gen_movi_tl(cpu_R[dc->op1], 1); + } + + cris_cc_mask(dc, 0); + return 2; +} + +static inline void cris_alu_alloc_temps(DisasContext *dc, int size, TCGv *t) +{ + if (size == 4) { + t[0] = cpu_R[dc->op2]; + t[1] = cpu_R[dc->op1]; + } else { + t[0] = tcg_temp_new(); + t[1] = tcg_temp_new(); + } +} + +static inline void cris_alu_free_temps(DisasContext *dc, int size, TCGv *t) +{ + if (size != 4) { + tcg_temp_free(t[0]); + tcg_temp_free(t[1]); + } +} + +static int dec_and_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + + LOG_DIS("and.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZ); + + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 0, t[0], t[1]); + cris_alu(dc, CC_OP_AND, cpu_R[dc->op2], t[0], t[1], size); + cris_alu_free_temps(dc, size, t); + return 2; +} + +static int dec_lz_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + LOG_DIS("lz $r%u, $r%u\n", + dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZ); + t0 = tcg_temp_new(); + dec_prep_alu_r(dc, dc->op1, dc->op2, 4, 0, cpu_R[dc->op2], t0); + cris_alu(dc, CC_OP_LZ, cpu_R[dc->op2], cpu_R[dc->op2], t0, 4); + tcg_temp_free(t0); + return 2; +} + +static int dec_lsl_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + + LOG_DIS("lsl.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 0, t[0], t[1]); + tcg_gen_andi_tl(t[1], t[1], 63); + cris_alu(dc, CC_OP_LSL, cpu_R[dc->op2], t[0], t[1], size); + cris_alu_alloc_temps(dc, size, t); + return 2; +} + +static int dec_lsr_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + + LOG_DIS("lsr.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 0, t[0], t[1]); + tcg_gen_andi_tl(t[1], t[1], 63); + cris_alu(dc, CC_OP_LSR, cpu_R[dc->op2], t[0], t[1], size); + cris_alu_free_temps(dc, size, t); + return 2; +} + +static int dec_asr_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + + LOG_DIS("asr.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 1, t[0], t[1]); + tcg_gen_andi_tl(t[1], t[1], 63); + cris_alu(dc, CC_OP_ASR, cpu_R[dc->op2], t[0], t[1], size); + cris_alu_free_temps(dc, size, t); + return 2; +} + +static int dec_muls_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + + LOG_DIS("muls.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZV); + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 1, t[0], t[1]); + + cris_alu(dc, CC_OP_MULS, cpu_R[dc->op2], t[0], t[1], 4); + cris_alu_free_temps(dc, size, t); + return 2; +} + +static int dec_mulu_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + + LOG_DIS("mulu.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZV); + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 0, t[0], t[1]); + + cris_alu(dc, CC_OP_MULU, cpu_R[dc->op2], t[0], t[1], 4); + cris_alu_alloc_temps(dc, size, t); + return 2; +} + + +static int dec_dstep_r(CPUCRISState *env, DisasContext *dc) +{ + LOG_DIS("dstep $r%u, $r%u\n", dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu(dc, CC_OP_DSTEP, + cpu_R[dc->op2], cpu_R[dc->op2], cpu_R[dc->op1], 4); + return 2; +} + +static int dec_xor_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + LOG_DIS("xor.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + BUG_ON(size != 4); /* xor is dword. */ + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 0, t[0], t[1]); + + cris_alu(dc, CC_OP_XOR, cpu_R[dc->op2], t[0], t[1], 4); + cris_alu_free_temps(dc, size, t); + return 2; +} + +static int dec_bound_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv l0; + int size = memsize_zz(dc); + LOG_DIS("bound.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZ); + l0 = tcg_temp_local_new(); + dec_prep_move_r(dc, dc->op1, dc->op2, size, 0, l0); + cris_alu(dc, CC_OP_BOUND, cpu_R[dc->op2], cpu_R[dc->op2], l0, 4); + tcg_temp_free(l0); + return 2; +} + +static int dec_cmp_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + LOG_DIS("cmp.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 0, t[0], t[1]); + + cris_alu(dc, CC_OP_CMP, cpu_R[dc->op2], t[0], t[1], size); + cris_alu_free_temps(dc, size, t); + return 2; +} + +static int dec_abs_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + + LOG_DIS("abs $r%u, $r%u\n", + dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZ); + + t0 = tcg_temp_new(); + tcg_gen_sari_tl(t0, cpu_R[dc->op1], 31); + tcg_gen_xor_tl(cpu_R[dc->op2], cpu_R[dc->op1], t0); + tcg_gen_sub_tl(cpu_R[dc->op2], cpu_R[dc->op2], t0); + tcg_temp_free(t0); + + cris_alu(dc, CC_OP_MOVE, + cpu_R[dc->op2], cpu_R[dc->op2], cpu_R[dc->op2], 4); + return 2; +} + +static int dec_add_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + LOG_DIS("add.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 0, t[0], t[1]); + + cris_alu(dc, CC_OP_ADD, cpu_R[dc->op2], t[0], t[1], size); + cris_alu_free_temps(dc, size, t); + return 2; +} + +static int dec_addc_r(CPUCRISState *env, DisasContext *dc) +{ + LOG_DIS("addc $r%u, $r%u\n", + dc->op1, dc->op2); + cris_evaluate_flags(dc); + /* Set for this insn. */ + dc->flagx_known = 1; + dc->flags_x = X_FLAG; + + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_ADDC, + cpu_R[dc->op2], cpu_R[dc->op2], cpu_R[dc->op1], 4); + return 2; +} + +static int dec_mcp_r(CPUCRISState *env, DisasContext *dc) +{ + LOG_DIS("mcp $p%u, $r%u\n", + dc->op2, dc->op1); + cris_evaluate_flags(dc); + cris_cc_mask(dc, CC_MASK_RNZV); + cris_alu(dc, CC_OP_MCP, + cpu_R[dc->op1], cpu_R[dc->op1], cpu_PR[dc->op2], 4); + return 2; +} + +#if DISAS_CRIS +static char * swapmode_name(int mode, char *modename) { + int i = 0; + if (mode & 8) { + modename[i++] = 'n'; + } + if (mode & 4) { + modename[i++] = 'w'; + } + if (mode & 2) { + modename[i++] = 'b'; + } + if (mode & 1) { + modename[i++] = 'r'; + } + modename[i++] = 0; + return modename; +} +#endif + +static int dec_swap_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; +#if DISAS_CRIS + char modename[4]; +#endif + LOG_DIS("swap%s $r%u\n", + swapmode_name(dc->op2, modename), dc->op1); + + cris_cc_mask(dc, CC_MASK_NZ); + t0 = tcg_temp_new(); + tcg_gen_mov_tl(t0, cpu_R[dc->op1]); + if (dc->op2 & 8) { + tcg_gen_not_tl(t0, t0); + } + if (dc->op2 & 4) { + t_gen_swapw(t0, t0); + } + if (dc->op2 & 2) { + t_gen_swapb(t0, t0); + } + if (dc->op2 & 1) { + t_gen_swapr(t0, t0); + } + cris_alu(dc, CC_OP_MOVE, cpu_R[dc->op1], cpu_R[dc->op1], t0, 4); + tcg_temp_free(t0); + return 2; +} + +static int dec_or_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + LOG_DIS("or.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 0, t[0], t[1]); + cris_alu(dc, CC_OP_OR, cpu_R[dc->op2], t[0], t[1], size); + cris_alu_free_temps(dc, size, t); + return 2; +} + +static int dec_addi_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + LOG_DIS("addi.%c $r%u, $r%u\n", + memsize_char(memsize_zz(dc)), dc->op2, dc->op1); + cris_cc_mask(dc, 0); + t0 = tcg_temp_new(); + tcg_gen_shl_tl(t0, cpu_R[dc->op2], tcg_const_tl(dc->zzsize)); + tcg_gen_add_tl(cpu_R[dc->op1], cpu_R[dc->op1], t0); + tcg_temp_free(t0); + return 2; +} + +static int dec_addi_acr(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + LOG_DIS("addi.%c $r%u, $r%u, $acr\n", + memsize_char(memsize_zz(dc)), dc->op2, dc->op1); + cris_cc_mask(dc, 0); + t0 = tcg_temp_new(); + tcg_gen_shl_tl(t0, cpu_R[dc->op2], tcg_const_tl(dc->zzsize)); + tcg_gen_add_tl(cpu_R[R_ACR], cpu_R[dc->op1], t0); + tcg_temp_free(t0); + return 2; +} + +static int dec_neg_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + LOG_DIS("neg.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 0, t[0], t[1]); + + cris_alu(dc, CC_OP_NEG, cpu_R[dc->op2], t[0], t[1], size); + cris_alu_free_temps(dc, size, t); + return 2; +} + +static int dec_btst_r(CPUCRISState *env, DisasContext *dc) +{ + LOG_DIS("btst $r%u, $r%u\n", + dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZ); + cris_evaluate_flags(dc); + gen_helper_btst(cpu_PR[PR_CCS], cpu_env, cpu_R[dc->op2], + cpu_R[dc->op1], cpu_PR[PR_CCS]); + cris_alu(dc, CC_OP_MOVE, cpu_R[dc->op2], + cpu_R[dc->op2], cpu_R[dc->op2], 4); + cris_update_cc_op(dc, CC_OP_FLAGS, 4); + dc->flags_uptodate = 1; + return 2; +} + +static int dec_sub_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int size = memsize_zz(dc); + LOG_DIS("sub.%c $r%u, $r%u\n", + memsize_char(size), dc->op1, dc->op2); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu_alloc_temps(dc, size, t); + dec_prep_alu_r(dc, dc->op1, dc->op2, size, 0, t[0], t[1]); + cris_alu(dc, CC_OP_SUB, cpu_R[dc->op2], t[0], t[1], size); + cris_alu_free_temps(dc, size, t); + return 2; +} + +/* Zero extension. From size to dword. */ +static int dec_movu_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + int size = memsize_z(dc); + LOG_DIS("movu.%c $r%u, $r%u\n", + memsize_char(size), + dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZ); + t0 = tcg_temp_new(); + dec_prep_move_r(dc, dc->op1, dc->op2, size, 0, t0); + cris_alu(dc, CC_OP_MOVE, cpu_R[dc->op2], cpu_R[dc->op2], t0, 4); + tcg_temp_free(t0); + return 2; +} + +/* Sign extension. From size to dword. */ +static int dec_movs_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + int size = memsize_z(dc); + LOG_DIS("movs.%c $r%u, $r%u\n", + memsize_char(size), + dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZ); + t0 = tcg_temp_new(); + /* Size can only be qi or hi. */ + t_gen_sext(t0, cpu_R[dc->op1], size); + cris_alu(dc, CC_OP_MOVE, + cpu_R[dc->op2], cpu_R[dc->op1], t0, 4); + tcg_temp_free(t0); + return 2; +} + +/* zero extension. From size to dword. */ +static int dec_addu_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + int size = memsize_z(dc); + LOG_DIS("addu.%c $r%u, $r%u\n", + memsize_char(size), + dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZVC); + t0 = tcg_temp_new(); + /* Size can only be qi or hi. */ + t_gen_zext(t0, cpu_R[dc->op1], size); + cris_alu(dc, CC_OP_ADD, cpu_R[dc->op2], cpu_R[dc->op2], t0, 4); + tcg_temp_free(t0); + return 2; +} + +/* Sign extension. From size to dword. */ +static int dec_adds_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + int size = memsize_z(dc); + LOG_DIS("adds.%c $r%u, $r%u\n", + memsize_char(size), + dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZVC); + t0 = tcg_temp_new(); + /* Size can only be qi or hi. */ + t_gen_sext(t0, cpu_R[dc->op1], size); + cris_alu(dc, CC_OP_ADD, + cpu_R[dc->op2], cpu_R[dc->op2], t0, 4); + tcg_temp_free(t0); + return 2; +} + +/* Zero extension. From size to dword. */ +static int dec_subu_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + int size = memsize_z(dc); + LOG_DIS("subu.%c $r%u, $r%u\n", + memsize_char(size), + dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZVC); + t0 = tcg_temp_new(); + /* Size can only be qi or hi. */ + t_gen_zext(t0, cpu_R[dc->op1], size); + cris_alu(dc, CC_OP_SUB, + cpu_R[dc->op2], cpu_R[dc->op2], t0, 4); + tcg_temp_free(t0); + return 2; +} + +/* Sign extension. From size to dword. */ +static int dec_subs_r(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + int size = memsize_z(dc); + LOG_DIS("subs.%c $r%u, $r%u\n", + memsize_char(size), + dc->op1, dc->op2); + + cris_cc_mask(dc, CC_MASK_NZVC); + t0 = tcg_temp_new(); + /* Size can only be qi or hi. */ + t_gen_sext(t0, cpu_R[dc->op1], size); + cris_alu(dc, CC_OP_SUB, + cpu_R[dc->op2], cpu_R[dc->op2], t0, 4); + tcg_temp_free(t0); + return 2; +} + +static int dec_setclrf(CPUCRISState *env, DisasContext *dc) +{ + uint32_t flags; + int set = (~dc->opcode >> 2) & 1; + + + flags = (EXTRACT_FIELD(dc->ir, 12, 15) << 4) + | EXTRACT_FIELD(dc->ir, 0, 3); + if (set && flags == 0) { + LOG_DIS("nop\n"); + return 2; + } else if (!set && (flags & 0x20)) { + LOG_DIS("di\n"); + } else { + LOG_DIS("%sf %x\n", set ? "set" : "clr", flags); + } + + /* User space is not allowed to touch these. Silently ignore. */ + if (dc->tb_flags & U_FLAG) { + flags &= ~(S_FLAG | I_FLAG | U_FLAG); + } + + if (flags & X_FLAG) { + dc->flagx_known = 1; + if (set) { + dc->flags_x = X_FLAG; + } else { + dc->flags_x = 0; + } + } + + /* Break the TB if any of the SPI flag changes. */ + if (flags & (P_FLAG | S_FLAG)) { + tcg_gen_movi_tl(env_pc, dc->pc + 2); + dc->is_jmp = DISAS_UPDATE; + dc->cpustate_changed = 1; + } + + /* For the I flag, only act on posedge. */ + if ((flags & I_FLAG)) { + tcg_gen_movi_tl(env_pc, dc->pc + 2); + dc->is_jmp = DISAS_UPDATE; + dc->cpustate_changed = 1; + } + + + /* Simply decode the flags. */ + cris_evaluate_flags(dc); + cris_update_cc_op(dc, CC_OP_FLAGS, 4); + cris_update_cc_x(dc); + tcg_gen_movi_tl(cc_op, dc->cc_op); + + if (set) { + if (!(dc->tb_flags & U_FLAG) && (flags & U_FLAG)) { + /* Enter user mode. */ + t_gen_mov_env_TN(ksp, cpu_R[R_SP]); + tcg_gen_mov_tl(cpu_R[R_SP], cpu_PR[PR_USP]); + dc->cpustate_changed = 1; + } + tcg_gen_ori_tl(cpu_PR[PR_CCS], cpu_PR[PR_CCS], flags); + } else { + tcg_gen_andi_tl(cpu_PR[PR_CCS], cpu_PR[PR_CCS], ~flags); + } + + dc->flags_uptodate = 1; + dc->clear_x = 0; + return 2; +} + +static int dec_move_rs(CPUCRISState *env, DisasContext *dc) +{ + LOG_DIS("move $r%u, $s%u\n", dc->op1, dc->op2); + cris_cc_mask(dc, 0); + gen_helper_movl_sreg_reg(cpu_env, tcg_const_tl(dc->op2), + tcg_const_tl(dc->op1)); + return 2; +} +static int dec_move_sr(CPUCRISState *env, DisasContext *dc) +{ + LOG_DIS("move $s%u, $r%u\n", dc->op2, dc->op1); + cris_cc_mask(dc, 0); + gen_helper_movl_reg_sreg(cpu_env, tcg_const_tl(dc->op1), + tcg_const_tl(dc->op2)); + return 2; +} + +static int dec_move_rp(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + LOG_DIS("move $r%u, $p%u\n", dc->op1, dc->op2); + cris_cc_mask(dc, 0); + + t[0] = tcg_temp_new(); + if (dc->op2 == PR_CCS) { + cris_evaluate_flags(dc); + tcg_gen_mov_tl(t[0], cpu_R[dc->op1]); + if (dc->tb_flags & U_FLAG) { + t[1] = tcg_temp_new(); + /* User space is not allowed to touch all flags. */ + tcg_gen_andi_tl(t[0], t[0], 0x39f); + tcg_gen_andi_tl(t[1], cpu_PR[PR_CCS], ~0x39f); + tcg_gen_or_tl(t[0], t[1], t[0]); + tcg_temp_free(t[1]); + } + } else { + tcg_gen_mov_tl(t[0], cpu_R[dc->op1]); + } + + t_gen_mov_preg_TN(dc, dc->op2, t[0]); + if (dc->op2 == PR_CCS) { + cris_update_cc_op(dc, CC_OP_FLAGS, 4); + dc->flags_uptodate = 1; + } + tcg_temp_free(t[0]); + return 2; +} +static int dec_move_pr(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + LOG_DIS("move $p%u, $r%u\n", dc->op2, dc->op1); + cris_cc_mask(dc, 0); + + if (dc->op2 == PR_CCS) { + cris_evaluate_flags(dc); + } + + if (dc->op2 == PR_DZ) { + tcg_gen_movi_tl(cpu_R[dc->op1], 0); + } else { + t0 = tcg_temp_new(); + t_gen_mov_TN_preg(t0, dc->op2); + cris_alu(dc, CC_OP_MOVE, + cpu_R[dc->op1], cpu_R[dc->op1], t0, + preg_sizes[dc->op2]); + tcg_temp_free(t0); + } + return 2; +} + +static int dec_move_mr(CPUCRISState *env, DisasContext *dc) +{ + int memsize = memsize_zz(dc); + int insn_len; + LOG_DIS("move.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + if (memsize == 4) { + insn_len = dec_prep_move_m(env, dc, 0, 4, cpu_R[dc->op2]); + cris_cc_mask(dc, CC_MASK_NZ); + cris_update_cc_op(dc, CC_OP_MOVE, 4); + cris_update_cc_x(dc); + cris_update_result(dc, cpu_R[dc->op2]); + } else { + TCGv t0; + + t0 = tcg_temp_new(); + insn_len = dec_prep_move_m(env, dc, 0, memsize, t0); + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu(dc, CC_OP_MOVE, + cpu_R[dc->op2], cpu_R[dc->op2], t0, memsize); + tcg_temp_free(t0); + } + do_postinc(dc, memsize); + return insn_len; +} + +static inline void cris_alu_m_alloc_temps(TCGv *t) +{ + t[0] = tcg_temp_new(); + t[1] = tcg_temp_new(); +} + +static inline void cris_alu_m_free_temps(TCGv *t) +{ + tcg_temp_free(t[0]); + tcg_temp_free(t[1]); +} + +static int dec_movs_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_z(dc); + int insn_len; + LOG_DIS("movs.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + /* sign extend. */ + insn_len = dec_prep_alu_m(env, dc, 1, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu(dc, CC_OP_MOVE, + cpu_R[dc->op2], cpu_R[dc->op2], t[1], 4); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_addu_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_z(dc); + int insn_len; + LOG_DIS("addu.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + /* sign extend. */ + insn_len = dec_prep_alu_m(env, dc, 0, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_ADD, + cpu_R[dc->op2], cpu_R[dc->op2], t[1], 4); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_adds_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_z(dc); + int insn_len; + LOG_DIS("adds.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + /* sign extend. */ + insn_len = dec_prep_alu_m(env, dc, 1, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_ADD, cpu_R[dc->op2], cpu_R[dc->op2], t[1], 4); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_subu_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_z(dc); + int insn_len; + LOG_DIS("subu.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + /* sign extend. */ + insn_len = dec_prep_alu_m(env, dc, 0, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_SUB, cpu_R[dc->op2], cpu_R[dc->op2], t[1], 4); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_subs_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_z(dc); + int insn_len; + LOG_DIS("subs.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + /* sign extend. */ + insn_len = dec_prep_alu_m(env, dc, 1, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_SUB, cpu_R[dc->op2], cpu_R[dc->op2], t[1], 4); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_movu_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_z(dc); + int insn_len; + + LOG_DIS("movu.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 0, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu(dc, CC_OP_MOVE, cpu_R[dc->op2], cpu_R[dc->op2], t[1], 4); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_cmpu_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_z(dc); + int insn_len; + LOG_DIS("cmpu.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 0, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_CMP, cpu_R[dc->op2], cpu_R[dc->op2], t[1], 4); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_cmps_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_z(dc); + int insn_len; + LOG_DIS("cmps.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 1, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_CMP, + cpu_R[dc->op2], cpu_R[dc->op2], t[1], + memsize_zz(dc)); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_cmp_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_zz(dc); + int insn_len; + LOG_DIS("cmp.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 0, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_CMP, + cpu_R[dc->op2], cpu_R[dc->op2], t[1], + memsize_zz(dc)); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_test_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_zz(dc); + int insn_len; + LOG_DIS("test.%c [$r%u%s] op2=%x\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_evaluate_flags(dc); + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 0, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZ); + tcg_gen_andi_tl(cpu_PR[PR_CCS], cpu_PR[PR_CCS], ~3); + + cris_alu(dc, CC_OP_CMP, + cpu_R[dc->op2], t[1], tcg_const_tl(0), memsize_zz(dc)); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_and_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_zz(dc); + int insn_len; + LOG_DIS("and.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 0, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu(dc, CC_OP_AND, cpu_R[dc->op2], t[0], t[1], memsize_zz(dc)); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_add_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_zz(dc); + int insn_len; + LOG_DIS("add.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 0, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_ADD, + cpu_R[dc->op2], t[0], t[1], memsize_zz(dc)); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_addo_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_zz(dc); + int insn_len; + LOG_DIS("add.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 1, memsize, t[0], t[1]); + cris_cc_mask(dc, 0); + cris_alu(dc, CC_OP_ADD, cpu_R[R_ACR], t[0], t[1], 4); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_bound_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv l[2]; + int memsize = memsize_zz(dc); + int insn_len; + LOG_DIS("bound.%c [$r%u%s, $r%u\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + l[0] = tcg_temp_local_new(); + l[1] = tcg_temp_local_new(); + insn_len = dec_prep_alu_m(env, dc, 0, memsize, l[0], l[1]); + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu(dc, CC_OP_BOUND, cpu_R[dc->op2], l[0], l[1], 4); + do_postinc(dc, memsize); + tcg_temp_free(l[0]); + tcg_temp_free(l[1]); + return insn_len; +} + +static int dec_addc_mr(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int insn_len = 2; + LOG_DIS("addc [$r%u%s, $r%u\n", + dc->op1, dc->postinc ? "+]" : "]", + dc->op2); + + cris_evaluate_flags(dc); + + /* Set for this insn. */ + dc->flagx_known = 1; + dc->flags_x = X_FLAG; + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 0, 4, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_ADDC, cpu_R[dc->op2], t[0], t[1], 4); + do_postinc(dc, 4); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_sub_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_zz(dc); + int insn_len; + LOG_DIS("sub.%c [$r%u%s, $r%u ir=%x zz=%x\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2, dc->ir, dc->zzsize); + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 0, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZVC); + cris_alu(dc, CC_OP_SUB, cpu_R[dc->op2], t[0], t[1], memsize); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_or_m(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_zz(dc); + int insn_len; + LOG_DIS("or.%c [$r%u%s, $r%u pc=%x\n", + memsize_char(memsize), + dc->op1, dc->postinc ? "+]" : "]", + dc->op2, dc->pc); + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 0, memsize, t[0], t[1]); + cris_cc_mask(dc, CC_MASK_NZ); + cris_alu(dc, CC_OP_OR, + cpu_R[dc->op2], t[0], t[1], memsize_zz(dc)); + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_move_mp(CPUCRISState *env, DisasContext *dc) +{ + TCGv t[2]; + int memsize = memsize_zz(dc); + int insn_len = 2; + + LOG_DIS("move.%c [$r%u%s, $p%u\n", + memsize_char(memsize), + dc->op1, + dc->postinc ? "+]" : "]", + dc->op2); + + cris_alu_m_alloc_temps(t); + insn_len = dec_prep_alu_m(env, dc, 0, memsize, t[0], t[1]); + cris_cc_mask(dc, 0); + if (dc->op2 == PR_CCS) { + cris_evaluate_flags(dc); + if (dc->tb_flags & U_FLAG) { + /* User space is not allowed to touch all flags. */ + tcg_gen_andi_tl(t[1], t[1], 0x39f); + tcg_gen_andi_tl(t[0], cpu_PR[PR_CCS], ~0x39f); + tcg_gen_or_tl(t[1], t[0], t[1]); + } + } + + t_gen_mov_preg_TN(dc, dc->op2, t[1]); + + do_postinc(dc, memsize); + cris_alu_m_free_temps(t); + return insn_len; +} + +static int dec_move_pm(CPUCRISState *env, DisasContext *dc) +{ + TCGv t0; + int memsize; + + memsize = preg_sizes[dc->op2]; + + LOG_DIS("move.%c $p%u, [$r%u%s\n", + memsize_char(memsize), + dc->op2, dc->op1, dc->postinc ? "+]" : "]"); + + /* prepare store. Address in T0, value in T1. */ + if (dc->op2 == PR_CCS) { + cris_evaluate_flags(dc); + } + t0 = tcg_temp_new(); + t_gen_mov_TN_preg(t0, dc->op2); + cris_flush_cc_state(dc); + gen_store(dc, cpu_R[dc->op1], t0, memsize); + tcg_temp_free(t0); + + cris_cc_mask(dc, 0); + if (dc->postinc) { + tcg_gen_addi_tl(cpu_R[dc->op1], cpu_R[dc->op1], memsize); + } + return 2; +} + +static int dec_movem_mr(CPUCRISState *env, DisasContext *dc) +{ + TCGv_i64 tmp[16]; + TCGv tmp32; + TCGv addr; + int i; + int nr = dc->op2 + 1; + + LOG_DIS("movem [$r%u%s, $r%u\n", dc->op1, + dc->postinc ? "+]" : "]", dc->op2); + + addr = tcg_temp_new(); + /* There are probably better ways of doing this. */ + cris_flush_cc_state(dc); + for (i = 0; i < (nr >> 1); i++) { + tmp[i] = tcg_temp_new_i64(); + tcg_gen_addi_tl(addr, cpu_R[dc->op1], i * 8); + gen_load64(dc, tmp[i], addr); + } + if (nr & 1) { + tmp32 = tcg_temp_new_i32(); + tcg_gen_addi_tl(addr, cpu_R[dc->op1], i * 8); + gen_load(dc, tmp32, addr, 4, 0); + } else { + TCGV_UNUSED(tmp32); + } + tcg_temp_free(addr); + + for (i = 0; i < (nr >> 1); i++) { + tcg_gen_trunc_i64_i32(cpu_R[i * 2], tmp[i]); + tcg_gen_shri_i64(tmp[i], tmp[i], 32); + tcg_gen_trunc_i64_i32(cpu_R[i * 2 + 1], tmp[i]); + tcg_temp_free_i64(tmp[i]); + } + if (nr & 1) { + tcg_gen_mov_tl(cpu_R[dc->op2], tmp32); + tcg_temp_free(tmp32); + } + + /* writeback the updated pointer value. */ + if (dc->postinc) { + tcg_gen_addi_tl(cpu_R[dc->op1], cpu_R[dc->op1], nr * 4); + } + + /* gen_load might want to evaluate the previous insns flags. */ + cris_cc_mask(dc, 0); + return 2; +} + +static int dec_movem_rm(CPUCRISState *env, DisasContext *dc) +{ + TCGv tmp; + TCGv addr; + int i; + + LOG_DIS("movem $r%u, [$r%u%s\n", dc->op2, dc->op1, + dc->postinc ? "+]" : "]"); + + cris_flush_cc_state(dc); + + tmp = tcg_temp_new(); + addr = tcg_temp_new(); + tcg_gen_movi_tl(tmp, 4); + tcg_gen_mov_tl(addr, cpu_R[dc->op1]); + for (i = 0; i <= dc->op2; i++) { + /* Displace addr. */ + /* Perform the store. */ + gen_store(dc, addr, cpu_R[i], 4); + tcg_gen_add_tl(addr, addr, tmp); + } + if (dc->postinc) { + tcg_gen_mov_tl(cpu_R[dc->op1], addr); + } + cris_cc_mask(dc, 0); + tcg_temp_free(tmp); + tcg_temp_free(addr); + return 2; +} + +static int dec_move_rm(CPUCRISState *env, DisasContext *dc) +{ + int memsize; + + memsize = memsize_zz(dc); + + LOG_DIS("move.%c $r%u, [$r%u]\n", + memsize_char(memsize), dc->op2, dc->op1); + + /* prepare store. */ + cris_flush_cc_state(dc); + gen_store(dc, cpu_R[dc->op1], cpu_R[dc->op2], memsize); + + if (dc->postinc) { + tcg_gen_addi_tl(cpu_R[dc->op1], cpu_R[dc->op1], memsize); + } + cris_cc_mask(dc, 0); + return 2; +} + +static int dec_lapcq(CPUCRISState *env, DisasContext *dc) +{ + LOG_DIS("lapcq %x, $r%u\n", + dc->pc + dc->op1*2, dc->op2); + cris_cc_mask(dc, 0); + tcg_gen_movi_tl(cpu_R[dc->op2], dc->pc + dc->op1 * 2); + return 2; +} + +static int dec_lapc_im(CPUCRISState *env, DisasContext *dc) +{ + unsigned int rd; + int32_t imm; + int32_t pc; + + rd = dc->op2; + + cris_cc_mask(dc, 0); + imm = cris_fetch(env, dc, dc->pc + 2, 4, 0); + LOG_DIS("lapc 0x%x, $r%u\n", imm + dc->pc, dc->op2); + + pc = dc->pc; + pc += imm; + tcg_gen_movi_tl(cpu_R[rd], pc); + return 6; +} + +/* Jump to special reg. */ +static int dec_jump_p(CPUCRISState *env, DisasContext *dc) +{ + LOG_DIS("jump $p%u\n", dc->op2); + + if (dc->op2 == PR_CCS) { + cris_evaluate_flags(dc); + } + t_gen_mov_TN_preg(env_btarget, dc->op2); + /* rete will often have low bit set to indicate delayslot. */ + tcg_gen_andi_tl(env_btarget, env_btarget, ~1); + cris_cc_mask(dc, 0); + cris_prepare_jmp(dc, JMP_INDIRECT); + return 2; +} + +/* Jump and save. */ +static int dec_jas_r(CPUCRISState *env, DisasContext *dc) +{ + LOG_DIS("jas $r%u, $p%u\n", dc->op1, dc->op2); + cris_cc_mask(dc, 0); + /* Store the return address in Pd. */ + tcg_gen_mov_tl(env_btarget, cpu_R[dc->op1]); + if (dc->op2 > 15) { + abort(); + } + t_gen_mov_preg_TN(dc, dc->op2, tcg_const_tl(dc->pc + 4)); + + cris_prepare_jmp(dc, JMP_INDIRECT); + return 2; +} + +static int dec_jas_im(CPUCRISState *env, DisasContext *dc) +{ + uint32_t imm; + + imm = cris_fetch(env, dc, dc->pc + 2, 4, 0); + + LOG_DIS("jas 0x%x\n", imm); + cris_cc_mask(dc, 0); + /* Store the return address in Pd. */ + t_gen_mov_preg_TN(dc, dc->op2, tcg_const_tl(dc->pc + 8)); + + dc->jmp_pc = imm; + cris_prepare_jmp(dc, JMP_DIRECT); + return 6; +} + +static int dec_jasc_im(CPUCRISState *env, DisasContext *dc) +{ + uint32_t imm; + + imm = cris_fetch(env, dc, dc->pc + 2, 4, 0); + + LOG_DIS("jasc 0x%x\n", imm); + cris_cc_mask(dc, 0); + /* Store the return address in Pd. */ + t_gen_mov_preg_TN(dc, dc->op2, tcg_const_tl(dc->pc + 8 + 4)); + + dc->jmp_pc = imm; + cris_prepare_jmp(dc, JMP_DIRECT); + return 6; +} + +static int dec_jasc_r(CPUCRISState *env, DisasContext *dc) +{ + LOG_DIS("jasc_r $r%u, $p%u\n", dc->op1, dc->op2); + cris_cc_mask(dc, 0); + /* Store the return address in Pd. */ + tcg_gen_mov_tl(env_btarget, cpu_R[dc->op1]); + t_gen_mov_preg_TN(dc, dc->op2, tcg_const_tl(dc->pc + 4 + 4)); + cris_prepare_jmp(dc, JMP_INDIRECT); + return 2; +} + +static int dec_bcc_im(CPUCRISState *env, DisasContext *dc) +{ + int32_t offset; + uint32_t cond = dc->op2; + + offset = cris_fetch(env, dc, dc->pc + 2, 2, 1); + + LOG_DIS("b%s %d pc=%x dst=%x\n", + cc_name(cond), offset, + dc->pc, dc->pc + offset); + + cris_cc_mask(dc, 0); + /* op2 holds the condition-code. */ + cris_prepare_cc_branch(dc, offset, cond); + return 4; +} + +static int dec_bas_im(CPUCRISState *env, DisasContext *dc) +{ + int32_t simm; + + simm = cris_fetch(env, dc, dc->pc + 2, 4, 0); + + LOG_DIS("bas 0x%x, $p%u\n", dc->pc + simm, dc->op2); + cris_cc_mask(dc, 0); + /* Store the return address in Pd. */ + t_gen_mov_preg_TN(dc, dc->op2, tcg_const_tl(dc->pc + 8)); + + dc->jmp_pc = dc->pc + simm; + cris_prepare_jmp(dc, JMP_DIRECT); + return 6; +} + +static int dec_basc_im(CPUCRISState *env, DisasContext *dc) +{ + int32_t simm; + simm = cris_fetch(env, dc, dc->pc + 2, 4, 0); + + LOG_DIS("basc 0x%x, $p%u\n", dc->pc + simm, dc->op2); + cris_cc_mask(dc, 0); + /* Store the return address in Pd. */ + t_gen_mov_preg_TN(dc, dc->op2, tcg_const_tl(dc->pc + 12)); + + dc->jmp_pc = dc->pc + simm; + cris_prepare_jmp(dc, JMP_DIRECT); + return 6; +} + +static int dec_rfe_etc(CPUCRISState *env, DisasContext *dc) +{ + cris_cc_mask(dc, 0); + + if (dc->op2 == 15) { + tcg_gen_st_i32(tcg_const_i32(1), cpu_env, + -offsetof(CRISCPU, env) + offsetof(CPUState, halted)); + tcg_gen_movi_tl(env_pc, dc->pc + 2); + t_gen_raise_exception(EXCP_HLT); + return 2; + } + + switch (dc->op2 & 7) { + case 2: + /* rfe. */ + LOG_DIS("rfe\n"); + cris_evaluate_flags(dc); + gen_helper_rfe(cpu_env); + dc->is_jmp = DISAS_UPDATE; + break; + case 5: + /* rfn. */ + LOG_DIS("rfn\n"); + cris_evaluate_flags(dc); + gen_helper_rfn(cpu_env); + dc->is_jmp = DISAS_UPDATE; + break; + case 6: + LOG_DIS("break %d\n", dc->op1); + cris_evaluate_flags(dc); + /* break. */ + tcg_gen_movi_tl(env_pc, dc->pc + 2); + + /* Breaks start at 16 in the exception vector. */ + t_gen_mov_env_TN(trap_vector, + tcg_const_tl(dc->op1 + 16)); + t_gen_raise_exception(EXCP_BREAK); + dc->is_jmp = DISAS_UPDATE; + break; + default: + printf("op2=%x\n", dc->op2); + BUG(); + break; + + } + return 2; +} + +static int dec_ftag_fidx_d_m(CPUCRISState *env, DisasContext *dc) +{ + return 2; +} + +static int dec_ftag_fidx_i_m(CPUCRISState *env, DisasContext *dc) +{ + return 2; +} + +static int dec_null(CPUCRISState *env, DisasContext *dc) +{ + printf("unknown insn pc=%x opc=%x op1=%x op2=%x\n", + dc->pc, dc->opcode, dc->op1, dc->op2); + fflush(NULL); + BUG(); + return 2; +} + +static struct decoder_info { + struct { + uint32_t bits; + uint32_t mask; + }; + int (*dec)(CPUCRISState *env, DisasContext *dc); +} decinfo[] = { + /* Order matters here. */ + {DEC_MOVEQ, dec_moveq}, + {DEC_BTSTQ, dec_btstq}, + {DEC_CMPQ, dec_cmpq}, + {DEC_ADDOQ, dec_addoq}, + {DEC_ADDQ, dec_addq}, + {DEC_SUBQ, dec_subq}, + {DEC_ANDQ, dec_andq}, + {DEC_ORQ, dec_orq}, + {DEC_ASRQ, dec_asrq}, + {DEC_LSLQ, dec_lslq}, + {DEC_LSRQ, dec_lsrq}, + {DEC_BCCQ, dec_bccq}, + + {DEC_BCC_IM, dec_bcc_im}, + {DEC_JAS_IM, dec_jas_im}, + {DEC_JAS_R, dec_jas_r}, + {DEC_JASC_IM, dec_jasc_im}, + {DEC_JASC_R, dec_jasc_r}, + {DEC_BAS_IM, dec_bas_im}, + {DEC_BASC_IM, dec_basc_im}, + {DEC_JUMP_P, dec_jump_p}, + {DEC_LAPC_IM, dec_lapc_im}, + {DEC_LAPCQ, dec_lapcq}, + + {DEC_RFE_ETC, dec_rfe_etc}, + {DEC_ADDC_MR, dec_addc_mr}, + + {DEC_MOVE_MP, dec_move_mp}, + {DEC_MOVE_PM, dec_move_pm}, + {DEC_MOVEM_MR, dec_movem_mr}, + {DEC_MOVEM_RM, dec_movem_rm}, + {DEC_MOVE_PR, dec_move_pr}, + {DEC_SCC_R, dec_scc_r}, + {DEC_SETF, dec_setclrf}, + {DEC_CLEARF, dec_setclrf}, + + {DEC_MOVE_SR, dec_move_sr}, + {DEC_MOVE_RP, dec_move_rp}, + {DEC_SWAP_R, dec_swap_r}, + {DEC_ABS_R, dec_abs_r}, + {DEC_LZ_R, dec_lz_r}, + {DEC_MOVE_RS, dec_move_rs}, + {DEC_BTST_R, dec_btst_r}, + {DEC_ADDC_R, dec_addc_r}, + + {DEC_DSTEP_R, dec_dstep_r}, + {DEC_XOR_R, dec_xor_r}, + {DEC_MCP_R, dec_mcp_r}, + {DEC_CMP_R, dec_cmp_r}, + + {DEC_ADDI_R, dec_addi_r}, + {DEC_ADDI_ACR, dec_addi_acr}, + + {DEC_ADD_R, dec_add_r}, + {DEC_SUB_R, dec_sub_r}, + + {DEC_ADDU_R, dec_addu_r}, + {DEC_ADDS_R, dec_adds_r}, + {DEC_SUBU_R, dec_subu_r}, + {DEC_SUBS_R, dec_subs_r}, + {DEC_LSL_R, dec_lsl_r}, + + {DEC_AND_R, dec_and_r}, + {DEC_OR_R, dec_or_r}, + {DEC_BOUND_R, dec_bound_r}, + {DEC_ASR_R, dec_asr_r}, + {DEC_LSR_R, dec_lsr_r}, + + {DEC_MOVU_R, dec_movu_r}, + {DEC_MOVS_R, dec_movs_r}, + {DEC_NEG_R, dec_neg_r}, + {DEC_MOVE_R, dec_move_r}, + + {DEC_FTAG_FIDX_I_M, dec_ftag_fidx_i_m}, + {DEC_FTAG_FIDX_D_M, dec_ftag_fidx_d_m}, + + {DEC_MULS_R, dec_muls_r}, + {DEC_MULU_R, dec_mulu_r}, + + {DEC_ADDU_M, dec_addu_m}, + {DEC_ADDS_M, dec_adds_m}, + {DEC_SUBU_M, dec_subu_m}, + {DEC_SUBS_M, dec_subs_m}, + + {DEC_CMPU_M, dec_cmpu_m}, + {DEC_CMPS_M, dec_cmps_m}, + {DEC_MOVU_M, dec_movu_m}, + {DEC_MOVS_M, dec_movs_m}, + + {DEC_CMP_M, dec_cmp_m}, + {DEC_ADDO_M, dec_addo_m}, + {DEC_BOUND_M, dec_bound_m}, + {DEC_ADD_M, dec_add_m}, + {DEC_SUB_M, dec_sub_m}, + {DEC_AND_M, dec_and_m}, + {DEC_OR_M, dec_or_m}, + {DEC_MOVE_RM, dec_move_rm}, + {DEC_TEST_M, dec_test_m}, + {DEC_MOVE_MR, dec_move_mr}, + + {{0, 0}, dec_null} +}; + +static unsigned int crisv32_decoder(CPUCRISState *env, DisasContext *dc) +{ + int insn_len = 2; + int i; + + if (unlikely(qemu_loglevel_mask(CPU_LOG_TB_OP | CPU_LOG_TB_OP_OPT))) { + tcg_gen_debug_insn_start(dc->pc); + } + + /* Load a halfword onto the instruction register. */ + dc->ir = cris_fetch(env, dc, dc->pc, 2, 0); + + /* Now decode it. */ + dc->opcode = EXTRACT_FIELD(dc->ir, 4, 11); + dc->op1 = EXTRACT_FIELD(dc->ir, 0, 3); + dc->op2 = EXTRACT_FIELD(dc->ir, 12, 15); + dc->zsize = EXTRACT_FIELD(dc->ir, 4, 4); + dc->zzsize = EXTRACT_FIELD(dc->ir, 4, 5); + dc->postinc = EXTRACT_FIELD(dc->ir, 10, 10); + + /* Large switch for all insns. */ + for (i = 0; i < ARRAY_SIZE(decinfo); i++) { + if ((dc->opcode & decinfo[i].mask) == decinfo[i].bits) { + insn_len = decinfo[i].dec(env, dc); + break; + } + } + +#if !defined(CONFIG_USER_ONLY) + /* Single-stepping ? */ + if (dc->tb_flags & S_FLAG) { + TCGLabel *l1 = gen_new_label(); + tcg_gen_brcondi_tl(TCG_COND_NE, cpu_PR[PR_SPC], dc->pc, l1); + /* We treat SPC as a break with an odd trap vector. */ + cris_evaluate_flags(dc); + t_gen_mov_env_TN(trap_vector, tcg_const_tl(3)); + tcg_gen_movi_tl(env_pc, dc->pc + insn_len); + tcg_gen_movi_tl(cpu_PR[PR_SPC], dc->pc + insn_len); + t_gen_raise_exception(EXCP_BREAK); + gen_set_label(l1); + } +#endif + return insn_len; +} + +static void check_breakpoint(CPUCRISState *env, DisasContext *dc) +{ + CPUState *cs = CPU(cris_env_get_cpu(env)); + CPUBreakpoint *bp; + + if (unlikely(!QTAILQ_EMPTY(&cs->breakpoints))) { + QTAILQ_FOREACH(bp, &cs->breakpoints, entry) { + if (bp->pc == dc->pc) { + cris_evaluate_flags(dc); + tcg_gen_movi_tl(env_pc, dc->pc); + t_gen_raise_exception(EXCP_DEBUG); + dc->is_jmp = DISAS_UPDATE; + } + } + } +} + +#include "translate_v10.c" + +/* + * Delay slots on QEMU/CRIS. + * + * If an exception hits on a delayslot, the core will let ERP (the Exception + * Return Pointer) point to the branch (the previous) insn and set the lsb to + * to give SW a hint that the exception actually hit on the dslot. + * + * CRIS expects all PC addresses to be 16-bit aligned. The lsb is ignored by + * the core and any jmp to an odd addresses will mask off that lsb. It is + * simply there to let sw know there was an exception on a dslot. + * + * When the software returns from an exception, the branch will re-execute. + * On QEMU care needs to be taken when a branch+delayslot sequence is broken + * and the branch and delayslot dont share pages. + * + * The TB contaning the branch insn will set up env->btarget and evaluate + * env->btaken. When the translation loop exits we will note that the branch + * sequence is broken and let env->dslot be the size of the branch insn (those + * vary in length). + * + * The TB contaning the delayslot will have the PC of its real insn (i.e no lsb + * set). It will also expect to have env->dslot setup with the size of the + * delay slot so that env->pc - env->dslot point to the branch insn. This TB + * will execute the dslot and take the branch, either to btarget or just one + * insn ahead. + * + * When exceptions occur, we check for env->dslot in do_interrupt to detect + * broken branch sequences and setup $erp accordingly (i.e let it point to the + * branch and set lsb). Then env->dslot gets cleared so that the exception + * handler can enter. When returning from exceptions (jump $erp) the lsb gets + * masked off and we will reexecute the branch insn. + * + */ + +/* generate intermediate code for basic block 'tb'. */ +static inline void +gen_intermediate_code_internal(CRISCPU *cpu, TranslationBlock *tb, + bool search_pc) +{ + CPUState *cs = CPU(cpu); + CPUCRISState *env = &cpu->env; + uint32_t pc_start; + unsigned int insn_len; + int j, lj; + struct DisasContext ctx; + struct DisasContext *dc = &ctx; + uint32_t next_page_start; + target_ulong npc; + int num_insns; + int max_insns; + + if (env->pregs[PR_VR] == 32) { + dc->decoder = crisv32_decoder; + dc->clear_locked_irq = 0; + } else { + dc->decoder = crisv10_decoder; + dc->clear_locked_irq = 1; + } + + /* Odd PC indicates that branch is rexecuting due to exception in the + * delayslot, like in real hw. + */ + pc_start = tb->pc & ~1; + dc->cpu = cpu; + dc->tb = tb; + + dc->is_jmp = DISAS_NEXT; + dc->ppc = pc_start; + dc->pc = pc_start; + dc->singlestep_enabled = cs->singlestep_enabled; + dc->flags_uptodate = 1; + dc->flagx_known = 1; + dc->flags_x = tb->flags & X_FLAG; + dc->cc_x_uptodate = 0; + dc->cc_mask = 0; + dc->update_cc = 0; + dc->clear_prefix = 0; + + cris_update_cc_op(dc, CC_OP_FLAGS, 4); + dc->cc_size_uptodate = -1; + + /* Decode TB flags. */ + dc->tb_flags = tb->flags & (S_FLAG | P_FLAG | U_FLAG \ + | X_FLAG | PFIX_FLAG); + dc->delayed_branch = !!(tb->flags & 7); + if (dc->delayed_branch) { + dc->jmp = JMP_INDIRECT; + } else { + dc->jmp = JMP_NOJMP; + } + + dc->cpustate_changed = 0; + + if (qemu_loglevel_mask(CPU_LOG_TB_IN_ASM)) { + qemu_log( + "srch=%d pc=%x %x flg=%" PRIx64 " bt=%x ds=%u ccs=%x\n" + "pid=%x usp=%x\n" + "%x.%x.%x.%x\n" + "%x.%x.%x.%x\n" + "%x.%x.%x.%x\n" + "%x.%x.%x.%x\n", + search_pc, dc->pc, dc->ppc, + (uint64_t)tb->flags, + env->btarget, (unsigned)tb->flags & 7, + env->pregs[PR_CCS], + env->pregs[PR_PID], env->pregs[PR_USP], + env->regs[0], env->regs[1], env->regs[2], env->regs[3], + env->regs[4], env->regs[5], env->regs[6], env->regs[7], + env->regs[8], env->regs[9], + env->regs[10], env->regs[11], + env->regs[12], env->regs[13], + env->regs[14], env->regs[15]); + qemu_log("--------------\n"); + qemu_log("IN: %s\n", lookup_symbol(pc_start)); + } + + next_page_start = (pc_start & TARGET_PAGE_MASK) + TARGET_PAGE_SIZE; + lj = -1; + num_insns = 0; + max_insns = tb->cflags & CF_COUNT_MASK; + if (max_insns == 0) { + max_insns = CF_COUNT_MASK; + } + + gen_tb_start(tb); + do { + check_breakpoint(env, dc); + + if (search_pc) { + j = tcg_op_buf_count(); + if (lj < j) { + lj++; + while (lj < j) { + tcg_ctx.gen_opc_instr_start[lj++] = 0; + } + } + if (dc->delayed_branch == 1) { + tcg_ctx.gen_opc_pc[lj] = dc->ppc | 1; + } else { + tcg_ctx.gen_opc_pc[lj] = dc->pc; + } + tcg_ctx.gen_opc_instr_start[lj] = 1; + tcg_ctx.gen_opc_icount[lj] = num_insns; + } + + /* Pretty disas. */ + LOG_DIS("%8.8x:\t", dc->pc); + + if (num_insns + 1 == max_insns && (tb->cflags & CF_LAST_IO)) { + gen_io_start(); + } + dc->clear_x = 1; + + insn_len = dc->decoder(env, dc); + dc->ppc = dc->pc; + dc->pc += insn_len; + if (dc->clear_x) { + cris_clear_x_flag(dc); + } + + num_insns++; + /* Check for delayed branches here. If we do it before + actually generating any host code, the simulator will just + loop doing nothing for on this program location. */ + if (dc->delayed_branch) { + dc->delayed_branch--; + if (dc->delayed_branch == 0) { + if (tb->flags & 7) { + t_gen_mov_env_TN(dslot, tcg_const_tl(0)); + } + if (dc->cpustate_changed || !dc->flagx_known + || (dc->flags_x != (tb->flags & X_FLAG))) { + cris_store_direct_jmp(dc); + } + + if (dc->clear_locked_irq) { + dc->clear_locked_irq = 0; + t_gen_mov_env_TN(locked_irq, tcg_const_tl(0)); + } + + if (dc->jmp == JMP_DIRECT_CC) { + TCGLabel *l1 = gen_new_label(); + cris_evaluate_flags(dc); + + /* Conditional jmp. */ + tcg_gen_brcondi_tl(TCG_COND_EQ, + env_btaken, 0, l1); + gen_goto_tb(dc, 1, dc->jmp_pc); + gen_set_label(l1); + gen_goto_tb(dc, 0, dc->pc); + dc->is_jmp = DISAS_TB_JUMP; + dc->jmp = JMP_NOJMP; + } else if (dc->jmp == JMP_DIRECT) { + cris_evaluate_flags(dc); + gen_goto_tb(dc, 0, dc->jmp_pc); + dc->is_jmp = DISAS_TB_JUMP; + dc->jmp = JMP_NOJMP; + } else { + t_gen_cc_jmp(env_btarget, tcg_const_tl(dc->pc)); + dc->is_jmp = DISAS_JUMP; + } + break; + } + } + + /* If we are rexecuting a branch due to exceptions on + delay slots dont break. */ + if (!(tb->pc & 1) && cs->singlestep_enabled) { + break; + } + } while (!dc->is_jmp && !dc->cpustate_changed + && !tcg_op_buf_full() + && !singlestep + && (dc->pc < next_page_start) + && num_insns < max_insns); + + if (dc->clear_locked_irq) { + t_gen_mov_env_TN(locked_irq, tcg_const_tl(0)); + } + + npc = dc->pc; + + if (tb->cflags & CF_LAST_IO) + gen_io_end(); + /* Force an update if the per-tb cpu state has changed. */ + if (dc->is_jmp == DISAS_NEXT + && (dc->cpustate_changed || !dc->flagx_known + || (dc->flags_x != (tb->flags & X_FLAG)))) { + dc->is_jmp = DISAS_UPDATE; + tcg_gen_movi_tl(env_pc, npc); + } + /* Broken branch+delayslot sequence. */ + if (dc->delayed_branch == 1) { + /* Set env->dslot to the size of the branch insn. */ + t_gen_mov_env_TN(dslot, tcg_const_tl(dc->pc - dc->ppc)); + cris_store_direct_jmp(dc); + } + + cris_evaluate_flags(dc); + + if (unlikely(cs->singlestep_enabled)) { + if (dc->is_jmp == DISAS_NEXT) { + tcg_gen_movi_tl(env_pc, npc); + } + t_gen_raise_exception(EXCP_DEBUG); + } else { + switch (dc->is_jmp) { + case DISAS_NEXT: + gen_goto_tb(dc, 1, npc); + break; + default: + case DISAS_JUMP: + case DISAS_UPDATE: + /* indicate that the hash table must be used + to find the next TB */ + tcg_gen_exit_tb(0); + break; + case DISAS_SWI: + case DISAS_TB_JUMP: + /* nothing more to generate */ + break; + } + } + gen_tb_end(tb, num_insns); + + if (search_pc) { + j = tcg_op_buf_count(); + lj++; + while (lj <= j) { + tcg_ctx.gen_opc_instr_start[lj++] = 0; + } + } else { + tb->size = dc->pc - pc_start; + tb->icount = num_insns; + } + +#ifdef DEBUG_DISAS +#if !DISAS_CRIS + if (qemu_loglevel_mask(CPU_LOG_TB_IN_ASM)) { + log_target_disas(cs, pc_start, dc->pc - pc_start, + env->pregs[PR_VR]); + qemu_log("\nisize=%d osize=%d\n", + dc->pc - pc_start, tcg_op_buf_count()); + } +#endif +#endif +} + +void gen_intermediate_code (CPUCRISState *env, struct TranslationBlock *tb) +{ + gen_intermediate_code_internal(cris_env_get_cpu(env), tb, false); +} + +void gen_intermediate_code_pc (CPUCRISState *env, struct TranslationBlock *tb) +{ + gen_intermediate_code_internal(cris_env_get_cpu(env), tb, true); +} + +void cris_cpu_dump_state(CPUState *cs, FILE *f, fprintf_function cpu_fprintf, + int flags) +{ + CRISCPU *cpu = CRIS_CPU(cs); + CPUCRISState *env = &cpu->env; + int i; + uint32_t srs; + + if (!env || !f) { + return; + } + + cpu_fprintf(f, "PC=%x CCS=%x btaken=%d btarget=%x\n" + "cc_op=%d cc_src=%d cc_dest=%d cc_result=%x cc_mask=%x\n", + env->pc, env->pregs[PR_CCS], env->btaken, env->btarget, + env->cc_op, + env->cc_src, env->cc_dest, env->cc_result, env->cc_mask); + + + for (i = 0; i < 16; i++) { + cpu_fprintf(f, "%s=%8.8x ", regnames[i], env->regs[i]); + if ((i + 1) % 4 == 0) { + cpu_fprintf(f, "\n"); + } + } + cpu_fprintf(f, "\nspecial regs:\n"); + for (i = 0; i < 16; i++) { + cpu_fprintf(f, "%s=%8.8x ", pregnames[i], env->pregs[i]); + if ((i + 1) % 4 == 0) { + cpu_fprintf(f, "\n"); + } + } + srs = env->pregs[PR_SRS]; + cpu_fprintf(f, "\nsupport function regs bank %x:\n", srs); + if (srs < ARRAY_SIZE(env->sregs)) { + for (i = 0; i < 16; i++) { + cpu_fprintf(f, "s%2.2d=%8.8x ", + i, env->sregs[srs][i]); + if ((i + 1) % 4 == 0) { + cpu_fprintf(f, "\n"); + } + } + } + cpu_fprintf(f, "\n\n"); + +} + +void cris_initialize_tcg(void) +{ + int i; + + cpu_env = tcg_global_reg_new_ptr(TCG_AREG0, "env"); + cc_x = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, cc_x), "cc_x"); + cc_src = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, cc_src), "cc_src"); + cc_dest = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, cc_dest), + "cc_dest"); + cc_result = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, cc_result), + "cc_result"); + cc_op = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, cc_op), "cc_op"); + cc_size = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, cc_size), + "cc_size"); + cc_mask = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, cc_mask), + "cc_mask"); + + env_pc = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, pc), + "pc"); + env_btarget = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, btarget), + "btarget"); + env_btaken = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, btaken), + "btaken"); + for (i = 0; i < 16; i++) { + cpu_R[i] = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, regs[i]), + regnames[i]); + } + for (i = 0; i < 16; i++) { + cpu_PR[i] = tcg_global_mem_new(TCG_AREG0, + offsetof(CPUCRISState, pregs[i]), + pregnames[i]); + } +} + +void restore_state_to_opc(CPUCRISState *env, TranslationBlock *tb, int pc_pos) +{ + env->pc = tcg_ctx.gen_opc_pc[pc_pos]; +} |