target/avr: Add instruction translation - Branch Instructions

This includes:
    - RJMP, IJMP, EIJMP, JMP
    - RCALL, ICALL, EICALL, CALL
    - RET, RETI
    - CPSE, CP, CPC, CPI
    - SBRC, SBRS, SBIC, SBIS
    - BRBC, BRBS

Signed-off-by: Michael Rolnik <mrolnik@gmail.com>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Aleksandar Markovic <aleksandar.m.mail@gmail.com>
Tested-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Aleksandar Markovic <aleksandar.m.mail@gmail.com>
Signed-off-by: Thomas Huth <huth@tuxfamily.org>
Message-Id: <20200705140315.260514-13-huth@tuxfamily.org>
Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>

2 files changed, 576 insertions, 0 deletions

diff --git a/target/avr/insn.decode b/target/avr/insn.decode
index 94351ca787..851eddedc7 100644
--- a/target/avr/insn.decode
+++ b/target/avr/insn.decode
@@ -74,3 +74,36 @@ FMUL            0000 0011 0 ... 1 ...       @fmul
 FMULS           0000 0011 1 ... 0 ...       @fmul
 FMULSU          0000 0011 1 ... 1 ...       @fmul
 DES             1001 0100 imm:4 1011
+
+#
+# Branch Instructions
+#
+
+# The 22-bit immediate is partially in the opcode word,
+# and partially in the next.  Use append_16 to build the
+# complete 22-bit value.
+%imm_call       4:5 0:1                     !function=append_16
+
+@op_bit         .... .... . bit:3 ....
+@op_bit_imm     .... .. imm:s7 bit:3
+
+RJMP            1100 imm:s12
+IJMP            1001 0100 0000 1001
+EIJMP           1001 0100 0001 1001
+JMP             1001 010 ..... 110 .        imm=%imm_call
+RCALL           1101 imm:s12
+ICALL           1001 0101 0000 1001
+EICALL          1001 0101 0001 1001
+CALL            1001 010 ..... 111 .        imm=%imm_call
+RET             1001 0101 0000 1000
+RETI            1001 0101 0001 1000
+CPSE            0001 00 . ..... ....        @op_rd_rr
+CP              0001 01 . ..... ....        @op_rd_rr
+CPC             0000 01 . ..... ....        @op_rd_rr
+CPI             0011 .... .... ....         @op_rd_imm8
+SBRC            1111 110 rr:5 0 bit:3
+SBRS            1111 111 rr:5 0 bit:3
+SBIC            1001 1001 reg:5 bit:3
+SBIS            1001 1011 reg:5 bit:3
+BRBS            1111 00 ....... ...         @op_bit_imm
+BRBC            1111 01 ....... ...         @op_bit_imm
diff --git a/target/avr/translate.c b/target/avr/translate.c
index fae9ed742d..5581264045 100644
--- a/target/avr/translate.c
+++ b/target/avr/translate.c
@@ -144,6 +144,16 @@ static int to_regs_24_30_by_two(DisasContext *ctx, int indx)
 }
 
 
+static uint16_t next_word(DisasContext *ctx)
+{
+    return cpu_lduw_code(ctx->env, ctx->npc++ * 2);
+}
+
+static int append_16(DisasContext *ctx, int x)
+{
+    return x << 16 | next_word(ctx);
+}
+
 static bool avr_have_feature(DisasContext *ctx, int feature)
 {
     if (!avr_feature(ctx->env, feature)) {
@@ -960,3 +970,536 @@ static bool trans_DES(DisasContext *ctx, arg_DES *a)
 
     return true;
 }
+
+/*
+ * Branch Instructions
+ */
+static void gen_jmp_ez(DisasContext *ctx)
+{
+    tcg_gen_deposit_tl(cpu_pc, cpu_r[30], cpu_r[31], 8, 8);
+    tcg_gen_or_tl(cpu_pc, cpu_pc, cpu_eind);
+    ctx->bstate = DISAS_LOOKUP;
+}
+
+static void gen_jmp_z(DisasContext *ctx)
+{
+    tcg_gen_deposit_tl(cpu_pc, cpu_r[30], cpu_r[31], 8, 8);
+    ctx->bstate = DISAS_LOOKUP;
+}
+
+static void gen_push_ret(DisasContext *ctx, int ret)
+{
+    if (avr_feature(ctx->env, AVR_FEATURE_1_BYTE_PC)) {
+
+        TCGv t0 = tcg_const_i32((ret & 0x0000ff));
+
+        tcg_gen_qemu_st_tl(t0, cpu_sp, MMU_DATA_IDX, MO_UB);
+        tcg_gen_subi_tl(cpu_sp, cpu_sp, 1);
+
+        tcg_temp_free_i32(t0);
+    } else if (avr_feature(ctx->env, AVR_FEATURE_2_BYTE_PC)) {
+
+        TCGv t0 = tcg_const_i32((ret & 0x00ffff));
+
+        tcg_gen_subi_tl(cpu_sp, cpu_sp, 1);
+        tcg_gen_qemu_st_tl(t0, cpu_sp, MMU_DATA_IDX, MO_BEUW);
+        tcg_gen_subi_tl(cpu_sp, cpu_sp, 1);
+
+        tcg_temp_free_i32(t0);
+
+    } else if (avr_feature(ctx->env, AVR_FEATURE_3_BYTE_PC)) {
+
+        TCGv lo = tcg_const_i32((ret & 0x0000ff));
+        TCGv hi = tcg_const_i32((ret & 0xffff00) >> 8);
+
+        tcg_gen_qemu_st_tl(lo, cpu_sp, MMU_DATA_IDX, MO_UB);
+        tcg_gen_subi_tl(cpu_sp, cpu_sp, 2);
+        tcg_gen_qemu_st_tl(hi, cpu_sp, MMU_DATA_IDX, MO_BEUW);
+        tcg_gen_subi_tl(cpu_sp, cpu_sp, 1);
+
+        tcg_temp_free_i32(lo);
+        tcg_temp_free_i32(hi);
+    }
+}
+
+static void gen_pop_ret(DisasContext *ctx, TCGv ret)
+{
+    if (avr_feature(ctx->env, AVR_FEATURE_1_BYTE_PC)) {
+        tcg_gen_addi_tl(cpu_sp, cpu_sp, 1);
+        tcg_gen_qemu_ld_tl(ret, cpu_sp, MMU_DATA_IDX, MO_UB);
+    } else if (avr_feature(ctx->env, AVR_FEATURE_2_BYTE_PC)) {
+        tcg_gen_addi_tl(cpu_sp, cpu_sp, 1);
+        tcg_gen_qemu_ld_tl(ret, cpu_sp, MMU_DATA_IDX, MO_BEUW);
+        tcg_gen_addi_tl(cpu_sp, cpu_sp, 1);
+    } else if (avr_feature(ctx->env, AVR_FEATURE_3_BYTE_PC)) {
+        TCGv lo = tcg_temp_new_i32();
+        TCGv hi = tcg_temp_new_i32();
+
+        tcg_gen_addi_tl(cpu_sp, cpu_sp, 1);
+        tcg_gen_qemu_ld_tl(hi, cpu_sp, MMU_DATA_IDX, MO_BEUW);
+
+        tcg_gen_addi_tl(cpu_sp, cpu_sp, 2);
+        tcg_gen_qemu_ld_tl(lo, cpu_sp, MMU_DATA_IDX, MO_UB);
+
+        tcg_gen_deposit_tl(ret, lo, hi, 8, 16);
+
+        tcg_temp_free_i32(lo);
+        tcg_temp_free_i32(hi);
+    }
+}
+
+static void gen_goto_tb(DisasContext *ctx, int n, target_ulong dest)
+{
+    TranslationBlock *tb = ctx->tb;
+
+    if (ctx->singlestep == 0) {
+        tcg_gen_goto_tb(n);
+        tcg_gen_movi_i32(cpu_pc, dest);
+        tcg_gen_exit_tb(tb, n);
+    } else {
+        tcg_gen_movi_i32(cpu_pc, dest);
+        gen_helper_debug(cpu_env);
+        tcg_gen_exit_tb(NULL, 0);
+    }
+    ctx->bstate = DISAS_NORETURN;
+}
+
+/*
+ *  Relative jump to an address within PC - 2K +1 and PC + 2K (words). For
+ *  AVR microcontrollers with Program memory not exceeding 4K words (8KB) this
+ *  instruction can address the entire memory from every address location. See
+ *  also JMP.
+ */
+static bool trans_RJMP(DisasContext *ctx, arg_RJMP *a)
+{
+    int dst = ctx->npc + a->imm;
+
+    gen_goto_tb(ctx, 0, dst);
+
+    return true;
+}
+
+/*
+ *  Indirect jump to the address pointed to by the Z (16 bits) Pointer
+ *  Register in the Register File. The Z-pointer Register is 16 bits wide and
+ *  allows jump within the lowest 64K words (128KB) section of Program memory.
+ *  This instruction is not available in all devices. Refer to the device
+ *  specific instruction set summary.
+ */
+static bool trans_IJMP(DisasContext *ctx, arg_IJMP *a)
+{
+    if (!avr_have_feature(ctx, AVR_FEATURE_IJMP_ICALL)) {
+        return true;
+    }
+
+    gen_jmp_z(ctx);
+
+    return true;
+}
+
+/*
+ *  Indirect jump to the address pointed to by the Z (16 bits) Pointer
+ *  Register in the Register File and the EIND Register in the I/O space. This
+ *  instruction allows for indirect jumps to the entire 4M (words) Program
+ *  memory space. See also IJMP.  This instruction is not available in all
+ *  devices. Refer to the device specific instruction set summary.
+ */
+static bool trans_EIJMP(DisasContext *ctx, arg_EIJMP *a)
+{
+    if (!avr_have_feature(ctx, AVR_FEATURE_EIJMP_EICALL)) {
+        return true;
+    }
+
+    gen_jmp_ez(ctx);
+    return true;
+}
+
+/*
+ *  Jump to an address within the entire 4M (words) Program memory. See also
+ *  RJMP.  This instruction is not available in all devices. Refer to the device
+ *  specific instruction set summary.0
+ */
+static bool trans_JMP(DisasContext *ctx, arg_JMP *a)
+{
+    if (!avr_have_feature(ctx, AVR_FEATURE_JMP_CALL)) {
+        return true;
+    }
+
+    gen_goto_tb(ctx, 0, a->imm);
+
+    return true;
+}
+
+/*
+ *  Relative call to an address within PC - 2K + 1 and PC + 2K (words). The
+ *  return address (the instruction after the RCALL) is stored onto the Stack.
+ *  See also CALL. For AVR microcontrollers with Program memory not exceeding 4K
+ *  words (8KB) this instruction can address the entire memory from every
+ *  address location. The Stack Pointer uses a post-decrement scheme during
+ *  RCALL.
+ */
+static bool trans_RCALL(DisasContext *ctx, arg_RCALL *a)
+{
+    int ret = ctx->npc;
+    int dst = ctx->npc + a->imm;
+
+    gen_push_ret(ctx, ret);
+    gen_goto_tb(ctx, 0, dst);
+
+    return true;
+}
+
+/*
+ *  Calls to a subroutine within the entire 4M (words) Program memory. The
+ *  return address (to the instruction after the CALL) will be stored onto the
+ *  Stack. See also RCALL. The Stack Pointer uses a post-decrement scheme during
+ *  CALL.  This instruction is not available in all devices. Refer to the device
+ *  specific instruction set summary.
+ */
+static bool trans_ICALL(DisasContext *ctx, arg_ICALL *a)
+{
+    if (!avr_have_feature(ctx, AVR_FEATURE_IJMP_ICALL)) {
+        return true;
+    }
+
+    int ret = ctx->npc;
+
+    gen_push_ret(ctx, ret);
+    gen_jmp_z(ctx);
+
+    return true;
+}
+
+/*
+ *  Indirect call of a subroutine pointed to by the Z (16 bits) Pointer
+ *  Register in the Register File and the EIND Register in the I/O space. This
+ *  instruction allows for indirect calls to the entire 4M (words) Program
+ *  memory space. See also ICALL. The Stack Pointer uses a post-decrement scheme
+ *  during EICALL.  This instruction is not available in all devices. Refer to
+ *  the device specific instruction set summary.
+ */
+static bool trans_EICALL(DisasContext *ctx, arg_EICALL *a)
+{
+    if (!avr_have_feature(ctx, AVR_FEATURE_EIJMP_EICALL)) {
+        return true;
+    }
+
+    int ret = ctx->npc;
+
+    gen_push_ret(ctx, ret);
+    gen_jmp_ez(ctx);
+    return true;
+}
+
+/*
+ *  Calls to a subroutine within the entire Program memory. The return
+ *  address (to the instruction after the CALL) will be stored onto the Stack.
+ *  (See also RCALL). The Stack Pointer uses a post-decrement scheme during
+ *  CALL.  This instruction is not available in all devices. Refer to the device
+ *  specific instruction set summary.
+ */
+static bool trans_CALL(DisasContext *ctx, arg_CALL *a)
+{
+    if (!avr_have_feature(ctx, AVR_FEATURE_JMP_CALL)) {
+        return true;
+    }
+
+    int Imm = a->imm;
+    int ret = ctx->npc;
+
+    gen_push_ret(ctx, ret);
+    gen_goto_tb(ctx, 0, Imm);
+
+    return true;
+}
+
+/*
+ *  Returns from subroutine. The return address is loaded from the STACK.
+ *  The Stack Pointer uses a preincrement scheme during RET.
+ */
+static bool trans_RET(DisasContext *ctx, arg_RET *a)
+{
+    gen_pop_ret(ctx, cpu_pc);
+
+    ctx->bstate = DISAS_LOOKUP;
+    return true;
+}
+
+/*
+ *  Returns from interrupt. The return address is loaded from the STACK and
+ *  the Global Interrupt Flag is set.  Note that the Status Register is not
+ *  automatically stored when entering an interrupt routine, and it is not
+ *  restored when returning from an interrupt routine. This must be handled by
+ *  the application program. The Stack Pointer uses a pre-increment scheme
+ *  during RETI.
+ */
+static bool trans_RETI(DisasContext *ctx, arg_RETI *a)
+{
+    gen_pop_ret(ctx, cpu_pc);
+    tcg_gen_movi_tl(cpu_If, 1);
+
+    /* Need to return to main loop to re-evaluate interrupts.  */
+    ctx->bstate = DISAS_EXIT;
+    return true;
+}
+
+/*
+ *  This instruction performs a compare between two registers Rd and Rr, and
+ *  skips the next instruction if Rd = Rr.
+ */
+static bool trans_CPSE(DisasContext *ctx, arg_CPSE *a)
+{
+    ctx->skip_cond = TCG_COND_EQ;
+    ctx->skip_var0 = cpu_r[a->rd];
+    ctx->skip_var1 = cpu_r[a->rr];
+    return true;
+}
+
+/*
+ *  This instruction performs a compare between two registers Rd and Rr.
+ *  None of the registers are changed. All conditional branches can be used
+ *  after this instruction.
+ */
+static bool trans_CP(DisasContext *ctx, arg_CP *a)
+{
+    TCGv Rd = cpu_r[a->rd];
+    TCGv Rr = cpu_r[a->rr];
+    TCGv R = tcg_temp_new_i32();
+
+    tcg_gen_sub_tl(R, Rd, Rr); /* R = Rd - Rr */
+    tcg_gen_andi_tl(R, R, 0xff); /* make it 8 bits */
+
+    /* update status register */
+    gen_sub_CHf(R, Rd, Rr);
+    gen_sub_Vf(R, Rd, Rr);
+    gen_ZNSf(R);
+
+    tcg_temp_free_i32(R);
+
+    return true;
+}
+
+/*
+ *  This instruction performs a compare between two registers Rd and Rr and
+ *  also takes into account the previous carry. None of the registers are
+ *  changed. All conditional branches can be used after this instruction.
+ */
+static bool trans_CPC(DisasContext *ctx, arg_CPC *a)
+{
+    TCGv Rd = cpu_r[a->rd];
+    TCGv Rr = cpu_r[a->rr];
+    TCGv R = tcg_temp_new_i32();
+    TCGv zero = tcg_const_i32(0);
+
+    tcg_gen_sub_tl(R, Rd, Rr); /* R = Rd - Rr - Cf */
+    tcg_gen_sub_tl(R, R, cpu_Cf);
+    tcg_gen_andi_tl(R, R, 0xff); /* make it 8 bits */
+    /* update status register */
+    gen_sub_CHf(R, Rd, Rr);
+    gen_sub_Vf(R, Rd, Rr);
+    gen_NSf(R);
+
+    /*
+     * Previous value remains unchanged when the result is zero;
+     * cleared otherwise.
+     */
+    tcg_gen_movcond_tl(TCG_COND_EQ, cpu_Zf, R, zero, cpu_Zf, zero);
+
+    tcg_temp_free_i32(zero);
+    tcg_temp_free_i32(R);
+
+    return true;
+}
+
+/*
+ *  This instruction performs a compare between register Rd and a constant.
+ *  The register is not changed. All conditional branches can be used after this
+ *  instruction.
+ */
+static bool trans_CPI(DisasContext *ctx, arg_CPI *a)
+{
+    TCGv Rd = cpu_r[a->rd];
+    int Imm = a->imm;
+    TCGv Rr = tcg_const_i32(Imm);
+    TCGv R = tcg_temp_new_i32();
+
+    tcg_gen_sub_tl(R, Rd, Rr); /* R = Rd - Rr */
+    tcg_gen_andi_tl(R, R, 0xff); /* make it 8 bits */
+
+    /* update status register */
+    gen_sub_CHf(R, Rd, Rr);
+    gen_sub_Vf(R, Rd, Rr);
+    gen_ZNSf(R);
+
+    tcg_temp_free_i32(R);
+    tcg_temp_free_i32(Rr);
+
+    return true;
+}
+
+/*
+ *  This instruction tests a single bit in a register and skips the next
+ *  instruction if the bit is cleared.
+ */
+static bool trans_SBRC(DisasContext *ctx, arg_SBRC *a)
+{
+    TCGv Rr = cpu_r[a->rr];
+
+    ctx->skip_cond = TCG_COND_EQ;
+    ctx->skip_var0 = tcg_temp_new();
+    ctx->free_skip_var0 = true;
+
+    tcg_gen_andi_tl(ctx->skip_var0, Rr, 1 << a->bit);
+    return true;
+}
+
+/*
+ *  This instruction tests a single bit in a register and skips the next
+ *  instruction if the bit is set.
+ */
+static bool trans_SBRS(DisasContext *ctx, arg_SBRS *a)
+{
+    TCGv Rr = cpu_r[a->rr];
+
+    ctx->skip_cond = TCG_COND_NE;
+    ctx->skip_var0 = tcg_temp_new();
+    ctx->free_skip_var0 = true;
+
+    tcg_gen_andi_tl(ctx->skip_var0, Rr, 1 << a->bit);
+    return true;
+}
+
+/*
+ *  This instruction tests a single bit in an I/O Register and skips the
+ *  next instruction if the bit is cleared. This instruction operates on the
+ *  lower 32 I/O Registers -- addresses 0-31.
+ */
+static bool trans_SBIC(DisasContext *ctx, arg_SBIC *a)
+{
+    TCGv temp = tcg_const_i32(a->reg);
+
+    gen_helper_inb(temp, cpu_env, temp);
+    tcg_gen_andi_tl(temp, temp, 1 << a->bit);
+    ctx->skip_cond = TCG_COND_EQ;
+    ctx->skip_var0 = temp;
+    ctx->free_skip_var0 = true;
+
+    return true;
+}
+
+/*
+ *  This instruction tests a single bit in an I/O Register and skips the
+ *  next instruction if the bit is set. This instruction operates on the lower
+ *  32 I/O Registers -- addresses 0-31.
+ */
+static bool trans_SBIS(DisasContext *ctx, arg_SBIS *a)
+{
+    TCGv temp = tcg_const_i32(a->reg);
+
+    gen_helper_inb(temp, cpu_env, temp);
+    tcg_gen_andi_tl(temp, temp, 1 << a->bit);
+    ctx->skip_cond = TCG_COND_NE;
+    ctx->skip_var0 = temp;
+    ctx->free_skip_var0 = true;
+
+    return true;
+}
+
+/*
+ *  Conditional relative branch. Tests a single bit in SREG and branches
+ *  relatively to PC if the bit is cleared. This instruction branches relatively
+ *  to PC in either direction (PC - 63 < = destination <= PC + 64). The
+ *  parameter k is the offset from PC and is represented in two's complement
+ *  form.
+ */
+static bool trans_BRBC(DisasContext *ctx, arg_BRBC *a)
+{
+    TCGLabel *not_taken = gen_new_label();
+
+    TCGv var;
+
+    switch (a->bit) {
+    case 0x00:
+        var = cpu_Cf;
+        break;
+    case 0x01:
+        var = cpu_Zf;
+        break;
+    case 0x02:
+        var = cpu_Nf;
+        break;
+    case 0x03:
+        var = cpu_Vf;
+        break;
+    case 0x04:
+        var = cpu_Sf;
+        break;
+    case 0x05:
+        var = cpu_Hf;
+        break;
+    case 0x06:
+        var = cpu_Tf;
+        break;
+    case 0x07:
+        var = cpu_If;
+        break;
+    default:
+        g_assert_not_reached();
+    }
+
+    tcg_gen_brcondi_i32(TCG_COND_NE, var, 0, not_taken);
+    gen_goto_tb(ctx, 0, ctx->npc + a->imm);
+    gen_set_label(not_taken);
+
+    ctx->bstate = DISAS_CHAIN;
+    return true;
+}
+
+/*
+ *  Conditional relative branch. Tests a single bit in SREG and branches
+ *  relatively to PC if the bit is set. This instruction branches relatively to
+ *  PC in either direction (PC - 63 < = destination <= PC + 64). The parameter k
+ *  is the offset from PC and is represented in two's complement form.
+ */
+static bool trans_BRBS(DisasContext *ctx, arg_BRBS *a)
+{
+    TCGLabel *not_taken = gen_new_label();
+
+    TCGv var;
+
+    switch (a->bit) {
+    case 0x00:
+        var = cpu_Cf;
+        break;
+    case 0x01:
+        var = cpu_Zf;
+        break;
+    case 0x02:
+        var = cpu_Nf;
+        break;
+    case 0x03:
+        var = cpu_Vf;
+        break;
+    case 0x04:
+        var = cpu_Sf;
+        break;
+    case 0x05:
+        var = cpu_Hf;
+        break;
+    case 0x06:
+        var = cpu_Tf;
+        break;
+    case 0x07:
+        var = cpu_If;
+        break;
+    default:
+        g_assert_not_reached();
+    }
+
+    tcg_gen_brcondi_i32(TCG_COND_EQ, var, 0, not_taken);
+    gen_goto_tb(ctx, 0, ctx->npc + a->imm);
+    gen_set_label(not_taken);
+
+    ctx->bstate = DISAS_CHAIN;
+    return true;
+}