diff options
Diffstat (limited to 'tcg')
-rw-r--r-- | tcg/arm/tcg-target.c | 237 | ||||
-rw-r--r-- | tcg/ia64/tcg-target.c | 58 | ||||
-rw-r--r-- | tcg/mips/tcg-target.c | 158 |
3 files changed, 276 insertions, 177 deletions
diff --git a/tcg/arm/tcg-target.c b/tcg/arm/tcg-target.c index 4d59a6385..cf0ca3d0b 100644 --- a/tcg/arm/tcg-target.c +++ b/tcg/arm/tcg-target.c @@ -176,6 +176,13 @@ static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str) so don't use these. */ tcg_regset_reset_reg(ct->u.regs, TCG_REG_R0); tcg_regset_reset_reg(ct->u.regs, TCG_REG_R1); +#if defined(CONFIG_TCG_PASS_AREG0) && (TARGET_LONG_BITS == 64) + /* If we're passing env to the helper as r0 and need a regpair + * for the address then r2 will be overwritten as we're setting + * up the args to the helper. + */ + tcg_regset_reset_reg(ct->u.regs, TCG_REG_R2); +#endif #endif break; case 'L': @@ -197,6 +204,12 @@ static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str) use these. */ tcg_regset_reset_reg(ct->u.regs, TCG_REG_R0); tcg_regset_reset_reg(ct->u.regs, TCG_REG_R1); +#if defined(CONFIG_SOFTMMU) && \ + defined(CONFIG_TCG_PASS_AREG0) && (TARGET_LONG_BITS == 64) + /* Avoid clashes with registers being used for helper args */ + tcg_regset_reset_reg(ct->u.regs, TCG_REG_R2); + tcg_regset_reset_reg(ct->u.regs, TCG_REG_R3); +#endif break; /* qemu_st64 data_reg2 */ case 'S': @@ -210,6 +223,10 @@ static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str) #ifdef CONFIG_SOFTMMU /* r2 is still needed to load data_reg, so don't use it. */ tcg_regset_reset_reg(ct->u.regs, TCG_REG_R2); +#if defined(CONFIG_TCG_PASS_AREG0) && (TARGET_LONG_BITS == 64) + /* Avoid clashes with registers being used for helper args */ + tcg_regset_reset_reg(ct->u.regs, TCG_REG_R3); +#endif #endif break; @@ -388,6 +405,14 @@ static inline void tcg_out_dat_reg(TCGContext *s, (rn << 16) | (rd << 12) | shift | rm); } +static inline void tcg_out_mov_reg(TCGContext *s, int cond, int rd, int rm) +{ + /* Simple reg-reg move, optimising out the 'do nothing' case */ + if (rd != rm) { + tcg_out_dat_reg(s, cond, ARITH_MOV, rd, 0, rm, SHIFT_IMM_LSL(0)); + } +} + static inline void tcg_out_dat_reg2(TCGContext *s, int cond, int opc0, int opc1, int rd0, int rd1, int rn0, int rn1, int rm0, int rm1, int shift) @@ -966,6 +991,90 @@ static void *qemu_st_helpers[4] = { __stq_mmu, }; #endif + +/* Helper routines for marshalling helper function arguments into + * the correct registers and stack. + * argreg is where we want to put this argument, arg is the argument itself. + * Return value is the updated argreg ready for the next call. + * Note that argreg 0..3 is real registers, 4+ on stack. + * When we reach the first stacked argument, we allocate space for it + * and the following stacked arguments using "str r8, [sp, #-0x10]!". + * Following arguments are filled in with "str r8, [sp, #0xNN]". + * For more than 4 stacked arguments we'd need to know how much + * space to allocate when we pushed the first stacked argument. + * We don't need this, so don't implement it (and will assert if you try it.) + * + * We provide routines for arguments which are: immediate, 32 bit + * value in register, 16 and 8 bit values in register (which must be zero + * extended before use) and 64 bit value in a lo:hi register pair. + */ +#define DEFINE_TCG_OUT_ARG(NAME, ARGPARAM) \ + static TCGReg NAME(TCGContext *s, TCGReg argreg, ARGPARAM) \ + { \ + if (argreg < 4) { \ + TCG_OUT_ARG_GET_ARG(argreg); \ + } else if (argreg == 4) { \ + TCG_OUT_ARG_GET_ARG(TCG_REG_R8); \ + tcg_out32(s, (COND_AL << 28) | 0x052d8010); \ + } else { \ + assert(argreg < 8); \ + TCG_OUT_ARG_GET_ARG(TCG_REG_R8); \ + tcg_out32(s, (COND_AL << 28) | 0x058d8000 | (argreg - 4) * 4); \ + } \ + return argreg + 1; \ + } + +#define TCG_OUT_ARG_GET_ARG(A) tcg_out_dat_imm(s, COND_AL, ARITH_MOV, A, 0, arg) +DEFINE_TCG_OUT_ARG(tcg_out_arg_imm32, uint32_t arg) +#undef TCG_OUT_ARG_GET_ARG +#define TCG_OUT_ARG_GET_ARG(A) tcg_out_ext8u(s, COND_AL, A, arg) +DEFINE_TCG_OUT_ARG(tcg_out_arg_reg8, TCGReg arg) +#undef TCG_OUT_ARG_GET_ARG +#define TCG_OUT_ARG_GET_ARG(A) tcg_out_ext16u(s, COND_AL, A, arg) +DEFINE_TCG_OUT_ARG(tcg_out_arg_reg16, TCGReg arg) +#undef TCG_OUT_ARG_GET_ARG + +/* We don't use the macro for this one to avoid an unnecessary reg-reg + * move when storing to the stack. + */ +static TCGReg tcg_out_arg_reg32(TCGContext *s, TCGReg argreg, TCGReg arg) +{ + if (argreg < 4) { + tcg_out_mov_reg(s, COND_AL, argreg, arg); + } else if (argreg == 4) { + /* str arg, [sp, #-0x10]! */ + tcg_out32(s, (COND_AL << 28) | 0x052d0010 | (arg << 12)); + } else { + assert(argreg < 8); + /* str arg, [sp, #0xNN] */ + tcg_out32(s, (COND_AL << 28) | 0x058d0000 | + (arg << 12) | (argreg - 4) * 4); + } + return argreg + 1; +} + +static inline TCGReg tcg_out_arg_reg64(TCGContext *s, TCGReg argreg, + TCGReg arglo, TCGReg arghi) +{ + /* 64 bit arguments must go in even/odd register pairs + * and in 8-aligned stack slots. + */ + if (argreg & 1) { + argreg++; + } + argreg = tcg_out_arg_reg32(s, argreg, arglo); + argreg = tcg_out_arg_reg32(s, argreg, arghi); + return argreg; +} + +static inline void tcg_out_arg_stacktidy(TCGContext *s, TCGReg argreg) +{ + /* Output any necessary post-call cleanup of the stack */ + if (argreg > 4) { + tcg_out_dat_imm(s, COND_AL, ARITH_ADD, TCG_REG_R13, TCG_REG_R13, 0x10); + } +} + #endif #define TLB_SHIFT (CPU_TLB_ENTRY_BITS + CPU_TLB_BITS) @@ -975,6 +1084,7 @@ static inline void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, int opc) int addr_reg, data_reg, data_reg2, bswap; #ifdef CONFIG_SOFTMMU int mem_index, s_bits; + TCGReg argreg; # if TARGET_LONG_BITS == 64 int addr_reg2; # endif @@ -1088,31 +1198,22 @@ static inline void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, int opc) tcg_out_b_noaddr(s, COND_EQ); /* TODO: move this code to where the constants pool will be */ - if (addr_reg != TCG_REG_R0) { - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - TCG_REG_R0, 0, addr_reg, SHIFT_IMM_LSL(0)); - } -# if TARGET_LONG_BITS == 32 - tcg_out_dat_imm(s, COND_AL, ARITH_MOV, TCG_REG_R1, 0, mem_index); -# else - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - TCG_REG_R1, 0, addr_reg2, SHIFT_IMM_LSL(0)); - tcg_out_dat_imm(s, COND_AL, ARITH_MOV, TCG_REG_R2, 0, mem_index); -# endif + /* Note that this code relies on the constraints we set in arm_op_defs[] + * to ensure that later arguments are not passed to us in registers we + * trash by moving the earlier arguments into them. + */ + argreg = TCG_REG_R0; #ifdef CONFIG_TCG_PASS_AREG0 - /* XXX/FIXME: suboptimal and incorrect for 64 bit */ - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - tcg_target_call_iarg_regs[2], 0, - tcg_target_call_iarg_regs[1], SHIFT_IMM_LSL(0)); - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - tcg_target_call_iarg_regs[1], 0, - tcg_target_call_iarg_regs[0], SHIFT_IMM_LSL(0)); - - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - tcg_target_call_iarg_regs[0], 0, TCG_AREG0, - SHIFT_IMM_LSL(0)); + argreg = tcg_out_arg_reg32(s, argreg, TCG_AREG0); #endif +#if TARGET_LONG_BITS == 64 + argreg = tcg_out_arg_reg64(s, argreg, addr_reg, addr_reg2); +#else + argreg = tcg_out_arg_reg32(s, argreg, addr_reg); +#endif + argreg = tcg_out_arg_imm32(s, argreg, mem_index); tcg_out_call(s, (tcg_target_long) qemu_ld_helpers[s_bits]); + tcg_out_arg_stacktidy(s, argreg); switch (opc) { case 0 | 4: @@ -1211,6 +1312,7 @@ static inline void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, int opc) int addr_reg, data_reg, data_reg2, bswap; #ifdef CONFIG_SOFTMMU int mem_index, s_bits; + TCGReg argreg; # if TARGET_LONG_BITS == 64 int addr_reg2; # endif @@ -1314,89 +1416,38 @@ static inline void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, int opc) tcg_out_b_noaddr(s, COND_EQ); /* TODO: move this code to where the constants pool will be */ - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - TCG_REG_R0, 0, addr_reg, SHIFT_IMM_LSL(0)); -# if TARGET_LONG_BITS == 32 - switch (opc) { - case 0: - tcg_out_ext8u(s, COND_AL, TCG_REG_R1, data_reg); - tcg_out_dat_imm(s, COND_AL, ARITH_MOV, TCG_REG_R2, 0, mem_index); - break; - case 1: - tcg_out_ext16u(s, COND_AL, TCG_REG_R1, data_reg); - tcg_out_dat_imm(s, COND_AL, ARITH_MOV, TCG_REG_R2, 0, mem_index); - break; - case 2: - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - TCG_REG_R1, 0, data_reg, SHIFT_IMM_LSL(0)); - tcg_out_dat_imm(s, COND_AL, ARITH_MOV, TCG_REG_R2, 0, mem_index); - break; - case 3: - tcg_out_dat_imm(s, COND_AL, ARITH_MOV, TCG_REG_R8, 0, mem_index); - tcg_out32(s, (COND_AL << 28) | 0x052d8010); /* str r8, [sp, #-0x10]! */ - if (data_reg != TCG_REG_R2) { - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - TCG_REG_R2, 0, data_reg, SHIFT_IMM_LSL(0)); - } - if (data_reg2 != TCG_REG_R3) { - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - TCG_REG_R3, 0, data_reg2, SHIFT_IMM_LSL(0)); - } - break; - } -# else - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - TCG_REG_R1, 0, addr_reg2, SHIFT_IMM_LSL(0)); + /* Note that this code relies on the constraints we set in arm_op_defs[] + * to ensure that later arguments are not passed to us in registers we + * trash by moving the earlier arguments into them. + */ + argreg = TCG_REG_R0; +#ifdef CONFIG_TCG_PASS_AREG0 + argreg = tcg_out_arg_reg32(s, argreg, TCG_AREG0); +#endif +#if TARGET_LONG_BITS == 64 + argreg = tcg_out_arg_reg64(s, argreg, addr_reg, addr_reg2); +#else + argreg = tcg_out_arg_reg32(s, argreg, addr_reg); +#endif + switch (opc) { case 0: - tcg_out_ext8u(s, COND_AL, TCG_REG_R2, data_reg); - tcg_out_dat_imm(s, COND_AL, ARITH_MOV, TCG_REG_R3, 0, mem_index); + argreg = tcg_out_arg_reg8(s, argreg, data_reg); break; case 1: - tcg_out_ext16u(s, COND_AL, TCG_REG_R2, data_reg); - tcg_out_dat_imm(s, COND_AL, ARITH_MOV, TCG_REG_R3, 0, mem_index); + argreg = tcg_out_arg_reg16(s, argreg, data_reg); break; case 2: - if (data_reg != TCG_REG_R2) { - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - TCG_REG_R2, 0, data_reg, SHIFT_IMM_LSL(0)); - } - tcg_out_dat_imm(s, COND_AL, ARITH_MOV, TCG_REG_R3, 0, mem_index); + argreg = tcg_out_arg_reg32(s, argreg, data_reg); break; case 3: - tcg_out_dat_imm(s, COND_AL, ARITH_MOV, TCG_REG_R8, 0, mem_index); - tcg_out32(s, (COND_AL << 28) | 0x052d8010); /* str r8, [sp, #-0x10]! */ - if (data_reg != TCG_REG_R2) { - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - TCG_REG_R2, 0, data_reg, SHIFT_IMM_LSL(0)); - } - if (data_reg2 != TCG_REG_R3) { - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - TCG_REG_R3, 0, data_reg2, SHIFT_IMM_LSL(0)); - } + argreg = tcg_out_arg_reg64(s, argreg, data_reg, data_reg2); break; } -# endif - -#ifdef CONFIG_TCG_PASS_AREG0 - /* XXX/FIXME: suboptimal and incorrect for 64 bit */ - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - tcg_target_call_iarg_regs[3], 0, - tcg_target_call_iarg_regs[2], SHIFT_IMM_LSL(0)); - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - tcg_target_call_iarg_regs[2], 0, - tcg_target_call_iarg_regs[1], SHIFT_IMM_LSL(0)); - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - tcg_target_call_iarg_regs[1], 0, - tcg_target_call_iarg_regs[0], SHIFT_IMM_LSL(0)); - tcg_out_dat_reg(s, COND_AL, ARITH_MOV, - tcg_target_call_iarg_regs[0], 0, TCG_AREG0, - SHIFT_IMM_LSL(0)); -#endif + argreg = tcg_out_arg_imm32(s, argreg, mem_index); tcg_out_call(s, (tcg_target_long) qemu_st_helpers[s_bits]); - if (opc == 3) - tcg_out_dat_imm(s, COND_AL, ARITH_ADD, TCG_REG_R13, TCG_REG_R13, 0x10); + tcg_out_arg_stacktidy(s, argreg); reloc_pc24(label_ptr, (tcg_target_long)s->code_ptr); #else /* !CONFIG_SOFTMMU */ diff --git a/tcg/ia64/tcg-target.c b/tcg/ia64/tcg-target.c index e02dacc84..dc588dbea 100644 --- a/tcg/ia64/tcg-target.c +++ b/tcg/ia64/tcg-target.c @@ -107,7 +107,7 @@ enum { }; static const int tcg_target_reg_alloc_order[] = { - TCG_REG_R34, + TCG_REG_R33, TCG_REG_R35, TCG_REG_R36, TCG_REG_R37, @@ -1532,12 +1532,13 @@ static inline void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, int opc) } #ifdef CONFIG_TCG_PASS_AREG0 /* XXX/FIXME: suboptimal */ - tcg_out_mov(s, TCG_TYPE_I32, tcg_target_call_iarg_regs[2], - tcg_target_call_iarg_regs[1]); - tcg_out_mov(s, TCG_TYPE_TL, tcg_target_call_iarg_regs[1], - tcg_target_call_iarg_regs[0]); - tcg_out_mov(s, TCG_TYPE_PTR, tcg_target_call_iarg_regs[0], - TCG_AREG0); + tcg_out_bundle(s, mII, + tcg_opc_a5 (TCG_REG_P7, OPC_ADDL_A5, TCG_REG_R58, + mem_index, TCG_REG_R0), + tcg_opc_a4 (TCG_REG_P7, OPC_ADDS_A4, + TCG_REG_R57, 0, TCG_REG_R56), + tcg_opc_a4 (TCG_REG_P7, OPC_ADDS_A4, + TCG_REG_R56, 0, TCG_AREG0)); #endif if (!bswap || s_bits == 0) { tcg_out_bundle(s, miB, @@ -1659,15 +1660,21 @@ static inline void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, int opc) #ifdef CONFIG_TCG_PASS_AREG0 /* XXX/FIXME: suboptimal */ - tcg_out_mov(s, TCG_TYPE_I32, tcg_target_call_iarg_regs[3], - tcg_target_call_iarg_regs[2]); - tcg_out_mov(s, TCG_TYPE_I64, tcg_target_call_iarg_regs[2], - tcg_target_call_iarg_regs[1]); - tcg_out_mov(s, TCG_TYPE_TL, tcg_target_call_iarg_regs[1], - tcg_target_call_iarg_regs[0]); - tcg_out_mov(s, TCG_TYPE_PTR, tcg_target_call_iarg_regs[0], - TCG_AREG0); -#endif + tcg_out_bundle(s, mII, + tcg_opc_a5 (TCG_REG_P7, OPC_ADDL_A5, TCG_REG_R59, + mem_index, TCG_REG_R0), + tcg_opc_a4 (TCG_REG_P7, OPC_ADDS_A4, + TCG_REG_R58, 0, TCG_REG_R57), + tcg_opc_a4 (TCG_REG_P7, OPC_ADDS_A4, + TCG_REG_R57, 0, TCG_REG_R56)); + tcg_out_bundle(s, miB, + tcg_opc_m4 (TCG_REG_P6, opc_st_m4[opc], + data_reg, TCG_REG_R3), + tcg_opc_a4 (TCG_REG_P7, OPC_ADDS_A4, + TCG_REG_R56, 0, TCG_AREG0), + tcg_opc_b5 (TCG_REG_P7, OPC_BR_CALL_SPTK_MANY_B5, + TCG_REG_B0, TCG_REG_B6)); +#else tcg_out_bundle(s, miB, tcg_opc_m4 (TCG_REG_P6, opc_st_m4[opc], data_reg, TCG_REG_R3), @@ -1675,6 +1682,7 @@ static inline void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, int opc) mem_index, TCG_REG_R0), tcg_opc_b5 (TCG_REG_P7, OPC_BR_CALL_SPTK_MANY_B5, TCG_REG_B0, TCG_REG_B6)); +#endif } #else /* !CONFIG_SOFTMMU */ @@ -2314,13 +2322,13 @@ static void tcg_target_qemu_prologue(TCGContext *s) s->code_ptr += 16; /* skip GP */ /* prologue */ - tcg_out_bundle(s, mII, + tcg_out_bundle(s, miI, tcg_opc_m34(TCG_REG_P0, OPC_ALLOC_M34, - TCG_REG_R33, 32, 24, 0), + TCG_REG_R34, 32, 24, 0), + tcg_opc_a4 (TCG_REG_P0, OPC_ADDS_A4, + TCG_AREG0, 0, TCG_REG_R32), tcg_opc_i21(TCG_REG_P0, OPC_MOV_I21, - TCG_REG_B6, TCG_REG_R33, 0), - tcg_opc_i22(TCG_REG_P0, OPC_MOV_I22, - TCG_REG_R32, TCG_REG_B0)); + TCG_REG_B6, TCG_REG_R33, 0)); /* ??? If GUEST_BASE < 0x200000, we could load the register via an ADDL in the M slot of the next bundle. */ @@ -2335,9 +2343,9 @@ static void tcg_target_qemu_prologue(TCGContext *s) tcg_out_bundle(s, miB, tcg_opc_a4 (TCG_REG_P0, OPC_ADDS_A4, - TCG_AREG0, 0, TCG_REG_R32), - tcg_opc_a4 (TCG_REG_P0, OPC_ADDS_A4, TCG_REG_R12, -frame_size, TCG_REG_R12), + tcg_opc_i22(TCG_REG_P0, OPC_MOV_I22, + TCG_REG_R32, TCG_REG_B0), tcg_opc_b4 (TCG_REG_P0, OPC_BR_SPTK_MANY_B4, TCG_REG_B6)); /* epilogue */ @@ -2351,7 +2359,7 @@ static void tcg_target_qemu_prologue(TCGContext *s) tcg_out_bundle(s, miB, tcg_opc_m48(TCG_REG_P0, OPC_NOP_M48, 0), tcg_opc_i26(TCG_REG_P0, OPC_MOV_I_I26, - TCG_REG_PFS, TCG_REG_R33), + TCG_REG_PFS, TCG_REG_R34), tcg_opc_b4 (TCG_REG_P0, OPC_BR_RET_SPTK_MANY_B4, TCG_REG_B0)); } @@ -2403,7 +2411,7 @@ static void tcg_target_init(TCGContext *s) tcg_regset_set_reg(s->reserved_regs, TCG_REG_R12); /* stack pointer */ tcg_regset_set_reg(s->reserved_regs, TCG_REG_R13); /* thread pointer */ tcg_regset_set_reg(s->reserved_regs, TCG_REG_R32); /* return address */ - tcg_regset_set_reg(s->reserved_regs, TCG_REG_R33); /* PFS */ + tcg_regset_set_reg(s->reserved_regs, TCG_REG_R34); /* PFS */ /* The following 3 are not in use, are call-saved, but *not* saved by the prologue. Therefore we cannot use them without modifying diff --git a/tcg/mips/tcg-target.c b/tcg/mips/tcg-target.c index 393ba07f2..1006e2800 100644 --- a/tcg/mips/tcg-target.c +++ b/tcg/mips/tcg-target.c @@ -217,6 +217,9 @@ static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str) tcg_regset_set(ct->u.regs, 0xffffffff); #if defined(CONFIG_SOFTMMU) tcg_regset_reset_reg(ct->u.regs, TCG_REG_A0); +# if defined(CONFIG_TCG_PASS_AREG0) && (TARGET_LONG_BITS == 64) + tcg_regset_reset_reg(ct->u.regs, TCG_REG_A2); +# endif #endif break; case 'S': /* qemu_st constraint */ @@ -224,10 +227,14 @@ static int target_parse_constraint(TCGArgConstraint *ct, const char **pct_str) tcg_regset_set(ct->u.regs, 0xffffffff); tcg_regset_reset_reg(ct->u.regs, TCG_REG_A0); #if defined(CONFIG_SOFTMMU) -# if TARGET_LONG_BITS == 64 +# if (defined(CONFIG_TCG_PASS_AREG0) && TARGET_LONG_BITS == 32) || \ + (!defined(CONFIG_TCG_PASS_AREG0) && TARGET_LONG_BITS == 64) tcg_regset_reset_reg(ct->u.regs, TCG_REG_A1); # endif tcg_regset_reset_reg(ct->u.regs, TCG_REG_A2); +# if defined(CONFIG_TCG_PASS_AREG0) && TARGET_LONG_BITS == 64 + tcg_regset_reset_reg(ct->u.regs, TCG_REG_A3); +# endif #endif break; case 'I': @@ -382,7 +389,10 @@ static inline void tcg_out_nop(TCGContext *s) static inline void tcg_out_mov(TCGContext *s, TCGType type, TCGReg ret, TCGReg arg) { - tcg_out_opc_reg(s, OPC_ADDU, ret, arg, TCG_REG_ZERO); + /* Simple reg-reg move, optimising out the 'do nothing' case */ + if (ret != arg) { + tcg_out_opc_reg(s, OPC_ADDU, ret, arg, TCG_REG_ZERO); + } } static inline void tcg_out_movi(TCGContext *s, TCGType type, @@ -503,6 +513,67 @@ static inline void tcg_out_addi(TCGContext *s, int reg, tcg_target_long val) } } +/* Helper routines for marshalling helper function arguments into + * the correct registers and stack. + * arg_num is where we want to put this argument, and is updated to be ready + * for the next call. arg is the argument itself. Note that arg_num 0..3 is + * real registers, 4+ on stack. + * + * We provide routines for arguments which are: immediate, 32 bit + * value in register, 16 and 8 bit values in register (which must be zero + * extended before use) and 64 bit value in a lo:hi register pair. + */ +#define DEFINE_TCG_OUT_CALL_IARG(NAME, ARGPARAM) \ + static inline void NAME(TCGContext *s, int *arg_num, ARGPARAM) \ + { \ + if (*arg_num < 4) { \ + DEFINE_TCG_OUT_CALL_IARG_GET_ARG(tcg_target_call_iarg_regs[*arg_num]); \ + } else { \ + DEFINE_TCG_OUT_CALL_IARG_GET_ARG(TCG_REG_AT); \ + tcg_out_st(s, TCG_TYPE_I32, TCG_REG_AT, TCG_REG_SP, 4 * (*arg_num)); \ + } \ + (*arg_num)++; \ +} +#define DEFINE_TCG_OUT_CALL_IARG_GET_ARG(A) \ + tcg_out_opc_imm(s, OPC_ANDI, A, arg, 0xff); +DEFINE_TCG_OUT_CALL_IARG(tcg_out_call_iarg_reg8, TCGReg arg) +#undef DEFINE_TCG_OUT_CALL_IARG_GET_ARG +#define DEFINE_TCG_OUT_CALL_IARG_GET_ARG(A) \ + tcg_out_opc_imm(s, OPC_ANDI, A, arg, 0xffff); +DEFINE_TCG_OUT_CALL_IARG(tcg_out_call_iarg_reg16, TCGReg arg) +#undef DEFINE_TCG_OUT_CALL_IARG_GET_ARG +#define DEFINE_TCG_OUT_CALL_IARG_GET_ARG(A) \ + tcg_out_movi(s, TCG_TYPE_I32, A, arg); +DEFINE_TCG_OUT_CALL_IARG(tcg_out_call_iarg_imm32, uint32_t arg) +#undef DEFINE_TCG_OUT_CALL_IARG_GET_ARG + +/* We don't use the macro for this one to avoid an unnecessary reg-reg + move when storing to the stack. */ +static inline void tcg_out_call_iarg_reg32(TCGContext *s, int *arg_num, + TCGReg arg) +{ + if (*arg_num < 4) { + tcg_out_mov(s, TCG_TYPE_I32, tcg_target_call_iarg_regs[*arg_num], arg); + } else { + tcg_out_st(s, TCG_TYPE_I32, arg, TCG_REG_SP, 4 * (*arg_num)); + } + (*arg_num)++; +} + +static inline void tcg_out_call_iarg_reg64(TCGContext *s, int *arg_num, + TCGReg arg_low, TCGReg arg_high) +{ + (*arg_num) = (*arg_num + 1) & ~1; + +#if defined(TCG_TARGET_WORDS_BIGENDIAN) + tcg_out_call_iarg_reg32(s, arg_num, arg_high); + tcg_out_call_iarg_reg32(s, arg_num, arg_low); +#else + tcg_out_call_iarg_reg32(s, arg_num, arg_low); + tcg_out_call_iarg_reg32(s, arg_num, arg_high); +#endif +} + static void tcg_out_brcond(TCGContext *s, TCGCond cond, int arg1, int arg2, int label_index) { @@ -792,18 +863,18 @@ static void *qemu_st_helpers[4] = { static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, int opc) { - int addr_regl, addr_reg1, addr_meml; + int addr_regl, addr_meml; int data_regl, data_regh, data_reg1, data_reg2; int mem_index, s_bits; #if defined(CONFIG_SOFTMMU) void *label1_ptr, *label2_ptr; - int sp_args; + int arg_num; #endif #if TARGET_LONG_BITS == 64 # if defined(CONFIG_SOFTMMU) uint8_t *label3_ptr; # endif - int addr_regh, addr_reg2, addr_memh; + int addr_regh, addr_memh; #endif data_regl = *args++; if (opc == 3) @@ -831,18 +902,13 @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, } #if TARGET_LONG_BITS == 64 # if defined(TCG_TARGET_WORDS_BIGENDIAN) - addr_reg1 = addr_regh; - addr_reg2 = addr_regl; addr_memh = 0; addr_meml = 4; # else - addr_reg1 = addr_regl; - addr_reg2 = addr_regh; addr_memh = 4; addr_meml = 0; # endif #else - addr_reg1 = addr_regl; addr_meml = 0; #endif @@ -875,22 +941,17 @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, # endif /* slow path */ - sp_args = TCG_REG_A0; - tcg_out_mov(s, TCG_TYPE_I32, sp_args++, addr_reg1); + arg_num = 0; +# ifdef CONFIG_TCG_PASS_AREG0 + tcg_out_call_iarg_reg32(s, &arg_num, TCG_AREG0); +# endif # if TARGET_LONG_BITS == 64 - tcg_out_mov(s, TCG_TYPE_I32, sp_args++, addr_reg2); + tcg_out_call_iarg_reg64(s, &arg_num, addr_regl, addr_regh); +# else + tcg_out_call_iarg_reg32(s, &arg_num, addr_regl); # endif - tcg_out_movi(s, TCG_TYPE_I32, sp_args++, mem_index); + tcg_out_call_iarg_imm32(s, &arg_num, mem_index); tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_T9, (tcg_target_long)qemu_ld_helpers[s_bits]); -#ifdef CONFIG_TCG_PASS_AREG0 - /* XXX/FIXME: suboptimal and incorrect for 64 on 32 bit */ - tcg_out_mov(s, TCG_TYPE_I32, tcg_target_call_iarg_regs[2], - tcg_target_call_iarg_regs[1]); - tcg_out_mov(s, TCG_TYPE_TL, tcg_target_call_iarg_regs[1], - tcg_target_call_iarg_regs[0]); - tcg_out_mov(s, TCG_TYPE_PTR, tcg_target_call_iarg_regs[0], - TCG_AREG0); -#endif tcg_out_opc_reg(s, OPC_JALR, TCG_REG_RA, TCG_REG_T9, 0); tcg_out_nop(s); @@ -991,18 +1052,18 @@ static void tcg_out_qemu_ld(TCGContext *s, const TCGArg *args, static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, int opc) { - int addr_regl, addr_reg1, addr_meml; + int addr_regl, addr_meml; int data_regl, data_regh, data_reg1, data_reg2; int mem_index, s_bits; #if defined(CONFIG_SOFTMMU) uint8_t *label1_ptr, *label2_ptr; - int sp_args; + int arg_num; #endif #if TARGET_LONG_BITS == 64 # if defined(CONFIG_SOFTMMU) uint8_t *label3_ptr; # endif - int addr_regh, addr_reg2, addr_memh; + int addr_regh, addr_memh; #endif data_regl = *args++; @@ -1024,18 +1085,13 @@ static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, #if TARGET_LONG_BITS == 64 addr_regh = *args++; # if defined(TCG_TARGET_WORDS_BIGENDIAN) - addr_reg1 = addr_regh; - addr_reg2 = addr_regl; addr_memh = 0; addr_meml = 4; # else - addr_reg1 = addr_regl; - addr_reg2 = addr_regh; addr_memh = 4; addr_meml = 0; # endif #else - addr_reg1 = addr_regl; addr_meml = 0; #endif mem_index = *args; @@ -1070,49 +1126,33 @@ static void tcg_out_qemu_st(TCGContext *s, const TCGArg *args, # endif /* slow path */ - sp_args = TCG_REG_A0; - tcg_out_mov(s, TCG_TYPE_I32, sp_args++, addr_reg1); + arg_num = 0; +# ifdef CONFIG_TCG_PASS_AREG0 + tcg_out_call_iarg_reg32(s, &arg_num, TCG_AREG0); +# endif # if TARGET_LONG_BITS == 64 - tcg_out_mov(s, TCG_TYPE_I32, sp_args++, addr_reg2); + tcg_out_call_iarg_reg64(s, &arg_num, addr_regl, addr_regh); +# else + tcg_out_call_iarg_reg32(s, &arg_num, addr_regl); # endif switch(opc) { case 0: - tcg_out_opc_imm(s, OPC_ANDI, sp_args++, data_reg1, 0xff); + tcg_out_call_iarg_reg8(s, &arg_num, data_regl); break; case 1: - tcg_out_opc_imm(s, OPC_ANDI, sp_args++, data_reg1, 0xffff); + tcg_out_call_iarg_reg16(s, &arg_num, data_regl); break; case 2: - tcg_out_mov(s, TCG_TYPE_I32, sp_args++, data_reg1); + tcg_out_call_iarg_reg32(s, &arg_num, data_regl); break; case 3: - sp_args = (sp_args + 1) & ~1; - tcg_out_mov(s, TCG_TYPE_I32, sp_args++, data_reg1); - tcg_out_mov(s, TCG_TYPE_I32, sp_args++, data_reg2); + tcg_out_call_iarg_reg64(s, &arg_num, data_regl, data_regh); break; default: tcg_abort(); } - if (sp_args > TCG_REG_A3) { - /* Push mem_index on the stack */ - tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_AT, mem_index); - tcg_out_st(s, TCG_TYPE_I32, TCG_REG_AT, TCG_REG_SP, 16); - } else { - tcg_out_movi(s, TCG_TYPE_I32, sp_args, mem_index); - } - + tcg_out_call_iarg_imm32(s, &arg_num, mem_index); tcg_out_movi(s, TCG_TYPE_I32, TCG_REG_T9, (tcg_target_long)qemu_st_helpers[s_bits]); -#ifdef CONFIG_TCG_PASS_AREG0 - /* XXX/FIXME: suboptimal and incorrect for 64 on 32 bit */ - tcg_out_mov(s, TCG_TYPE_I32, tcg_target_call_iarg_regs[3], - tcg_target_call_iarg_regs[2]); - tcg_out_mov(s, TCG_TYPE_I64, tcg_target_call_iarg_regs[2], - tcg_target_call_iarg_regs[1]); - tcg_out_mov(s, TCG_TYPE_TL, tcg_target_call_iarg_regs[1], - tcg_target_call_iarg_regs[0]); - tcg_out_mov(s, TCG_TYPE_PTR, tcg_target_call_iarg_regs[0], - TCG_AREG0); -#endif tcg_out_opc_reg(s, OPC_JALR, TCG_REG_RA, TCG_REG_T9, 0); tcg_out_nop(s); |