From: Michael Rolnik <mrolnik@gmail.com>
To: qemu-devel@nongnu.org
Cc: thuth@redhat.com, Michael Rolnik <mrolnik@gmail.com>,
richard.henderson@linaro.org, dovgaluk@ispras.ru,
imammedo@redhat.com, philmd@redhat.com,
aleksandar.m.mail@gmail.com
Subject: [PATCH v32 05/13] target/avr: Add instruction translation - Arithmetic and Logic Instructions
Date: Mon, 14 Oct 2019 19:18:35 +0300 [thread overview]
Message-ID: <20191014161843.84845-6-mrolnik@gmail.com> (raw)
In-Reply-To: <20191014161843.84845-1-mrolnik@gmail.com>
This includes:
- ADD, ADC, ADIW
Signed-off-by: Michael Rolnik <mrolnik@gmail.com>
---
target/avr/translate.c | 811 +++++++++++++++++++++++++++++++++++++++++
1 file changed, 811 insertions(+)
diff --git a/target/avr/translate.c b/target/avr/translate.c
index 53c9892a60..3eaa3e5099 100644
--- a/target/avr/translate.c
+++ b/target/avr/translate.c
@@ -130,3 +130,814 @@ static bool avr_have_feature(DisasContext *ctx, int feature)
static bool decode_insn(DisasContext *ctx, uint16_t insn);
#include "decode_insn.inc.c"
+
+static void gen_add_CHf(TCGv R, TCGv Rd, TCGv Rr)
+{
+ TCGv t1 = tcg_temp_new_i32();
+ TCGv t2 = tcg_temp_new_i32();
+ TCGv t3 = tcg_temp_new_i32();
+
+ tcg_gen_and_tl(t1, Rd, Rr); /* t1 = Rd & Rr */
+ tcg_gen_andc_tl(t2, Rd, R); /* t2 = Rd & ~R */
+ tcg_gen_andc_tl(t3, Rr, R); /* t3 = Rr & ~R */
+ tcg_gen_or_tl(t1, t1, t2); /* t1 = t1 | t2 | t3 */
+ tcg_gen_or_tl(t1, t1, t3);
+
+ tcg_gen_shri_tl(cpu_Cf, t1, 7); /* Cf = t1(7) */
+ tcg_gen_shri_tl(cpu_Hf, t1, 3); /* Hf = t1(3) */
+ tcg_gen_andi_tl(cpu_Hf, cpu_Hf, 1);
+
+ tcg_temp_free_i32(t3);
+ tcg_temp_free_i32(t2);
+ tcg_temp_free_i32(t1);
+}
+
+
+static void gen_add_Vf(TCGv R, TCGv Rd, TCGv Rr)
+{
+ TCGv t1 = tcg_temp_new_i32();
+ TCGv t2 = tcg_temp_new_i32();
+
+ /* t1 = Rd & Rr & ~R | ~Rd & ~Rr & R = (Rd ^ R) & ~(Rd ^ Rr) */
+ tcg_gen_xor_tl(t1, Rd, R);
+ tcg_gen_xor_tl(t2, Rd, Rr);
+ tcg_gen_andc_tl(t1, t1, t2);
+
+ tcg_gen_shri_tl(cpu_Vf, t1, 7); /* Vf = t1(7) */
+
+ tcg_temp_free_i32(t2);
+ tcg_temp_free_i32(t1);
+}
+
+
+static void gen_sub_CHf(TCGv R, TCGv Rd, TCGv Rr)
+{
+ TCGv t1 = tcg_temp_new_i32();
+ TCGv t2 = tcg_temp_new_i32();
+ TCGv t3 = tcg_temp_new_i32();
+
+ /* Cf & Hf */
+ tcg_gen_not_tl(t1, Rd); /* t1 = ~Rd */
+ tcg_gen_and_tl(t2, t1, Rr); /* t2 = ~Rd & Rr */
+ tcg_gen_or_tl(t3, t1, Rr); /* t3 = (~Rd | Rr) & R */
+ tcg_gen_and_tl(t3, t3, R);
+ tcg_gen_or_tl(t2, t2, t3); /* t2 = ~Rd & Rr | ~Rd & R | R & Rr */
+ tcg_gen_shri_tl(cpu_Cf, t2, 7); /* Cf = t2(7) */
+ tcg_gen_shri_tl(cpu_Hf, t2, 3); /* Hf = t2(3) */
+ tcg_gen_andi_tl(cpu_Hf, cpu_Hf, 1);
+
+ tcg_temp_free_i32(t3);
+ tcg_temp_free_i32(t2);
+ tcg_temp_free_i32(t1);
+}
+
+
+static void gen_sub_Vf(TCGv R, TCGv Rd, TCGv Rr)
+{
+ TCGv t1 = tcg_temp_new_i32();
+ TCGv t2 = tcg_temp_new_i32();
+
+ /* Vf */
+ /* t1 = Rd & ~Rr & ~R | ~Rd & Rr & R = (Rd ^ R) & (Rd ^ R) */
+ tcg_gen_xor_tl(t1, Rd, R);
+ tcg_gen_xor_tl(t2, Rd, Rr);
+ tcg_gen_and_tl(t1, t1, t2);
+ tcg_gen_shri_tl(cpu_Vf, t1, 7); /* Vf = t1(7) */
+
+ tcg_temp_free_i32(t2);
+ tcg_temp_free_i32(t1);
+}
+
+
+static void gen_NSf(TCGv R)
+{
+ tcg_gen_shri_tl(cpu_Nf, R, 7); /* Nf = R(7) */
+ tcg_gen_xor_tl(cpu_Sf, cpu_Nf, cpu_Vf); /* Sf = Nf ^ Vf */
+}
+
+
+static void gen_ZNSf(TCGv R)
+{
+ tcg_gen_mov_tl(cpu_Zf, R); /* Zf = R */
+ tcg_gen_shri_tl(cpu_Nf, R, 7); /* Nf = R(7) */
+ tcg_gen_xor_tl(cpu_Sf, cpu_Nf, cpu_Vf); /* Sf = Nf ^ Vf */
+}
+
+
+/*
+ * Adds two registers without the C Flag and places the result in the
+ * destination register Rd.
+ */
+static bool trans_ADD(DisasContext *ctx, arg_ADD *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+
+ /* op */
+ tcg_gen_add_tl(R, Rd, Rr); /* Rd = Rd + Rr */
+ tcg_gen_andi_tl(R, R, 0xff); /* make it 8 bits */
+
+ gen_add_CHf(R, Rd, Rr);
+ gen_add_Vf(R, Rd, Rr);
+ gen_ZNSf(R);
+
+ /* R */
+ tcg_gen_mov_tl(Rd, R);
+
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * Adds two registers and the contents of the C Flag and places the result in
+ * the destination register Rd.
+ */
+static bool trans_ADC(DisasContext *ctx, arg_ADC *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+
+ /* op */
+ tcg_gen_add_tl(R, Rd, Rr); /* R = Rd + Rr + Cf */
+ tcg_gen_add_tl(R, R, cpu_Cf);
+ tcg_gen_andi_tl(R, R, 0xff); /* make it 8 bits */
+
+ gen_add_CHf(R, Rd, Rr);
+ gen_add_Vf(R, Rd, Rr);
+ gen_ZNSf(R);
+
+ /* R */
+ tcg_gen_mov_tl(Rd, R);
+
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * Subtracts an immediate value (0-63) from a register pair and places the
+ * result in the register pair. This instruction operates on the upper four
+ * register pairs, and is well suited for operations on the Pointer Registers.
+ * This instruction is not available in all devices. Refer to the device
+ * specific instruction set summary.
+ */
+static bool trans_SBIW(DisasContext *ctx, arg_SBIW *a)
+{
+ if (!avr_have_feature(ctx, AVR_FEATURE_ADIW_SBIW)) {
+ return true;
+ }
+
+ TCGv RdL = cpu_r[a->rd];
+ TCGv RdH = cpu_r[a->rd + 1];
+ int Imm = (a->imm);
+ TCGv R = tcg_temp_new_i32();
+ TCGv Rd = tcg_temp_new_i32();
+
+ /* op */
+ tcg_gen_deposit_tl(Rd, RdL, RdH, 8, 8); /* Rd = RdH:RdL */
+ tcg_gen_subi_tl(R, Rd, Imm); /* R = Rd - Imm */
+ tcg_gen_andi_tl(R, R, 0xffff); /* make it 16 bits */
+
+ /* Cf */
+ tcg_gen_andc_tl(cpu_Cf, R, Rd);
+ tcg_gen_shri_tl(cpu_Cf, cpu_Cf, 15); /* Cf = R & ~Rd */
+
+ /* Vf */
+ tcg_gen_andc_tl(cpu_Vf, Rd, R);
+ tcg_gen_shri_tl(cpu_Vf, cpu_Vf, 15); /* Vf = Rd & ~R */
+
+ /* Zf */
+ tcg_gen_mov_tl(cpu_Zf, R); /* Zf = R */
+
+ /* Nf */
+ tcg_gen_shri_tl(cpu_Nf, R, 15); /* Nf = R(15) */
+
+ /* Sf */
+ tcg_gen_xor_tl(cpu_Sf, cpu_Nf, cpu_Vf); /* Sf = Nf ^ Vf */
+
+ /* R */
+ tcg_gen_andi_tl(RdL, R, 0xff);
+ tcg_gen_shri_tl(RdH, R, 8);
+
+ tcg_temp_free_i32(Rd);
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * Adds an immediate value (0 - 63) to a register pair and places the result
+ * in the register pair. This instruction operates on the upper four register
+ * pairs, and is well suited for operations on the pointer registers. This
+ * instruction is not available in all devices. Refer to the device specific
+ * instruction set summary.
+ */
+static bool trans_ADIW(DisasContext *ctx, arg_ADIW *a)
+{
+ if (!avr_have_feature(ctx, AVR_FEATURE_ADIW_SBIW)) {
+ return true;
+ }
+
+ TCGv RdL = cpu_r[a->rd];
+ TCGv RdH = cpu_r[a->rd + 1];
+ int Imm = (a->imm);
+ TCGv R = tcg_temp_new_i32();
+ TCGv Rd = tcg_temp_new_i32();
+
+ /* op */
+ tcg_gen_deposit_tl(Rd, RdL, RdH, 8, 8); /* Rd = RdH:RdL */
+ tcg_gen_addi_tl(R, Rd, Imm); /* R = Rd + Imm */
+ tcg_gen_andi_tl(R, R, 0xffff); /* make it 16 bits */
+
+ /* Cf */
+ tcg_gen_andc_tl(cpu_Cf, Rd, R); /* Cf = Rd & ~R */
+ tcg_gen_shri_tl(cpu_Cf, cpu_Cf, 15);
+
+ /* Vf */
+ tcg_gen_andc_tl(cpu_Vf, R, Rd); /* Vf = R & ~Rd */
+ tcg_gen_shri_tl(cpu_Vf, cpu_Vf, 15);
+
+ /* Zf */
+ tcg_gen_mov_tl(cpu_Zf, R); /* Zf = R */
+
+ /* Nf */
+ tcg_gen_shri_tl(cpu_Nf, R, 15); /* Nf = R(15) */
+
+ /* Sf */
+ tcg_gen_xor_tl(cpu_Sf, cpu_Nf, cpu_Vf);/* Sf = Nf ^ Vf */
+
+ /* R */
+ tcg_gen_andi_tl(RdL, R, 0xff);
+ tcg_gen_shri_tl(RdH, R, 8);
+
+ tcg_temp_free_i32(Rd);
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * Subtracts two registers and places the result in the destination
+ * register Rd.
+ */
+static bool trans_SUB(DisasContext *ctx, arg_SUB *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+
+ /* op */
+ tcg_gen_sub_tl(R, Rd, Rr); /* R = Rd - Rr */
+ tcg_gen_andi_tl(R, R, 0xff); /* make it 8 bits */
+
+ gen_sub_CHf(R, Rd, Rr);
+ gen_sub_Vf(R, Rd, Rr);
+ gen_ZNSf(R);
+
+ /* R */
+ tcg_gen_mov_tl(Rd, R);
+
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+/*
+ * Subtracts a register and a constant and places the result in the
+ * destination register Rd. This instruction is working on Register R16 to R31
+ * and is very well suited for operations on the X, Y, and Z-pointers.
+ */
+static bool trans_SUBI(DisasContext *ctx, arg_SUBI *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = tcg_const_i32(a->imm);
+ TCGv R = tcg_temp_new_i32();
+
+ /* op */
+ tcg_gen_sub_tl(R, Rd, Rr); /* R = Rd - Imm */
+ tcg_gen_andi_tl(R, R, 0xff); /* make it 8 bits */
+
+ gen_sub_CHf(R, Rd, Rr);
+ gen_sub_Vf(R, Rd, Rr);
+ gen_ZNSf(R);
+
+ /* R */
+ tcg_gen_mov_tl(Rd, R);
+
+ tcg_temp_free_i32(R);
+ tcg_temp_free_i32(Rr);
+
+ return true;
+}
+
+
+/*
+ * Subtracts two registers and subtracts with the C Flag and places the
+ * result in the destination register Rd.
+ */
+static bool trans_SBC(DisasContext *ctx, arg_SBC *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+
+ /* op */
+ 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 */
+
+ 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_or_tl(cpu_Zf, cpu_Zf, R);
+
+ /* R */
+ tcg_gen_mov_tl(Rd, R);
+
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * SBCI -- Subtract Immediate with Carry
+ */
+static bool trans_SBCI(DisasContext *ctx, arg_SBCI *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = tcg_const_i32(a->imm);
+ TCGv R = tcg_temp_new_i32();
+
+ /* op */
+ 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 */
+
+ 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_or_tl(cpu_Zf, cpu_Zf, R);
+
+ /* R */
+ tcg_gen_mov_tl(Rd, R);
+
+ tcg_temp_free_i32(R);
+ tcg_temp_free_i32(Rr);
+
+ return true;
+}
+
+
+/*
+ * Performs the logical AND between the contents of register Rd and register
+ * Rr and places the result in the destination register Rd.
+ */
+static bool trans_AND(DisasContext *ctx, arg_AND *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+
+ /* op */
+ tcg_gen_and_tl(R, Rd, Rr); /* Rd = Rd and Rr */
+
+ /* Vf */
+ tcg_gen_movi_tl(cpu_Vf, 0x00); /* Vf = 0 */
+
+ /* Zf */
+ tcg_gen_mov_tl(cpu_Zf, R); /* Zf = R */
+
+ gen_ZNSf(R);
+
+ /* R */
+ tcg_gen_mov_tl(Rd, R);
+
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * Performs the logical AND between the contents of register Rd and a constant
+ * and places the result in the destination register Rd.
+ */
+static bool trans_ANDI(DisasContext *ctx, arg_ANDI *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ int Imm = (a->imm);
+
+ /* op */
+ tcg_gen_andi_tl(Rd, Rd, Imm); /* Rd = Rd & Imm */
+
+ tcg_gen_movi_tl(cpu_Vf, 0x00); /* Vf = 0 */
+ gen_ZNSf(Rd);
+
+ return true;
+}
+
+
+/*
+ * Performs the logical OR between the contents of register Rd and register
+ * Rr and places the result in the destination register Rd.
+ */
+static bool trans_OR(DisasContext *ctx, arg_OR *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+
+ tcg_gen_or_tl(R, Rd, Rr);
+
+ tcg_gen_movi_tl(cpu_Vf, 0);
+ gen_ZNSf(R);
+
+ tcg_gen_mov_tl(Rd, R);
+
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * Performs the logical OR between the contents of register Rd and a
+ * constant and places the result in the destination register Rd.
+ */
+static bool trans_ORI(DisasContext *ctx, arg_ORI *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ int Imm = (a->imm);
+
+ tcg_gen_ori_tl(Rd, Rd, Imm); /* Rd = Rd | Imm */
+
+ tcg_gen_movi_tl(cpu_Vf, 0x00); /* Vf = 0 */
+ gen_ZNSf(Rd);
+
+ return true;
+}
+
+
+/*
+ * Performs the logical EOR between the contents of register Rd and
+ * register Rr and places the result in the destination register Rd.
+ */
+static bool trans_EOR(DisasContext *ctx, arg_EOR *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+
+ tcg_gen_xor_tl(Rd, Rd, Rr);
+
+ tcg_gen_movi_tl(cpu_Vf, 0);
+ gen_ZNSf(Rd);
+
+ return true;
+}
+
+
+/*
+ * Clears the specified bits in register Rd. Performs the logical AND
+ * between the contents of register Rd and the complement of the constant mask
+ * K. The result will be placed in register Rd.
+ */
+static bool trans_COM(DisasContext *ctx, arg_COM *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ TCGv R = tcg_temp_new_i32();
+
+ tcg_gen_xori_tl(Rd, Rd, 0xff);
+
+ tcg_gen_movi_tl(cpu_Cf, 1); /* Cf = 1 */
+ tcg_gen_movi_tl(cpu_Vf, 0); /* Vf = 0 */
+ gen_ZNSf(Rd);
+
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * Replaces the contents of register Rd with its two's complement; the
+ * value $80 is left unchanged.
+ */
+static bool trans_NEG(DisasContext *ctx, arg_NEG *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+ TCGv t0 = tcg_const_i32(0);
+ TCGv R = tcg_temp_new_i32();
+
+ /* op */
+ tcg_gen_sub_tl(R, t0, Rd); /* R = 0 - Rd */
+ tcg_gen_andi_tl(R, R, 0xff); /* make it 8 bits */
+
+ gen_sub_CHf(R, t0, Rd);
+ gen_sub_Vf(R, t0, Rd);
+ gen_ZNSf(R);
+
+ /* R */
+ tcg_gen_mov_tl(Rd, R);
+
+ tcg_temp_free_i32(t0);
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * Adds one -1- to the contents of register Rd and places the result in the
+ * destination register Rd. The C Flag in SREG is not affected by the
+ * operation, thus allowing the INC instruction to be used on a loop counter in
+ * multiple-precision computations. When operating on unsigned numbers, only
+ * BREQ and BRNE branches can be expected to perform consistently. When
+ * operating on two's complement values, all signed branches are available.
+ */
+static bool trans_INC(DisasContext *ctx, arg_INC *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+
+ tcg_gen_addi_tl(Rd, Rd, 1);
+ tcg_gen_andi_tl(Rd, Rd, 0xff);
+
+ /* cpu_Vf = Rd == 0x80 */
+ tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_Vf, Rd, 0x80);
+ gen_ZNSf(Rd);
+
+ return true;
+}
+
+
+/*
+ * Subtracts one -1- from the contents of register Rd and places the result
+ * in the destination register Rd. The C Flag in SREG is not affected by the
+ * operation, thus allowing the DEC instruction to be used on a loop counter in
+ * multiple-precision computations. When operating on unsigned values, only
+ * BREQ and BRNE branches can be expected to perform consistently. When
+ * operating on two's complement values, all signed branches are available.
+ */
+static bool trans_DEC(DisasContext *ctx, arg_DEC *a)
+{
+ TCGv Rd = cpu_r[a->rd];
+
+ tcg_gen_subi_tl(Rd, Rd, 1); /* Rd = Rd - 1 */
+ tcg_gen_andi_tl(Rd, Rd, 0xff); /* make it 8 bits */
+
+ /* cpu_Vf = Rd == 0x7f */
+ tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_Vf, Rd, 0x7f);
+ gen_ZNSf(Rd);
+
+ return true;
+}
+
+
+/*
+ * This instruction performs 8-bit x 8-bit -> 16-bit unsigned multiplication.
+ */
+static bool trans_MUL(DisasContext *ctx, arg_MUL *a)
+{
+ if (!avr_have_feature(ctx, AVR_FEATURE_MUL)) {
+ return true;
+ }
+
+ TCGv R0 = cpu_r[0];
+ TCGv R1 = cpu_r[1];
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+
+ tcg_gen_mul_tl(R, Rd, Rr); /* R = Rd *Rr */
+
+ tcg_gen_andi_tl(R0, R, 0xff);
+ tcg_gen_shri_tl(R1, R, 8);
+
+ tcg_gen_shri_tl(cpu_Cf, R, 15); /* Cf = R(16) */
+ tcg_gen_mov_tl(cpu_Zf, R);
+
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * This instruction performs 8-bit x 8-bit -> 16-bit signed multiplication.
+ */
+static bool trans_MULS(DisasContext *ctx, arg_MULS *a)
+{
+ if (!avr_have_feature(ctx, AVR_FEATURE_MUL)) {
+ return true;
+ }
+
+ TCGv R0 = cpu_r[0];
+ TCGv R1 = cpu_r[1];
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+ TCGv t0 = tcg_temp_new_i32();
+ TCGv t1 = tcg_temp_new_i32();
+
+ tcg_gen_ext8s_tl(t0, Rd); /* make Rd full 32 bit signed */
+ tcg_gen_ext8s_tl(t1, Rr); /* make Rr full 32 bit signed */
+ tcg_gen_mul_tl(R, t0, t1); /* R = Rd * Rr */
+
+ tcg_gen_andi_tl(R0, R, 0xff);
+ tcg_gen_shri_tl(R1, R, 8);
+
+ tcg_gen_shri_tl(cpu_Cf, R, 15); /* Cf = R(16) */
+ tcg_gen_mov_tl(cpu_Zf, R);
+
+ tcg_temp_free_i32(t1);
+ tcg_temp_free_i32(t0);
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * This instruction performs 8-bit x 8-bit -> 16-bit multiplication of a
+ * signed and an unsigned number.
+ */
+static bool trans_MULSU(DisasContext *ctx, arg_MULSU *a)
+{
+ if (!avr_have_feature(ctx, AVR_FEATURE_MUL)) {
+ return true;
+ }
+
+ TCGv R0 = cpu_r[0];
+ TCGv R1 = cpu_r[1];
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+ TCGv t0 = tcg_temp_new_i32();
+
+ tcg_gen_ext8s_tl(t0, Rd); /* make Rd full 32 bit signed */
+ tcg_gen_mul_tl(R, t0, Rr); /* R = Rd *Rr */
+
+ tcg_gen_andi_tl(R0, R, 0xff);
+ tcg_gen_shri_tl(R1, R, 8);
+
+ tcg_gen_shri_tl(cpu_Cf, R, 16); /* Cf = R(16) */
+ tcg_gen_mov_tl(cpu_Zf, R);
+
+ tcg_temp_free_i32(t0);
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * This instruction performs 8-bit x 8-bit -> 16-bit unsigned
+ * multiplication and shifts the result one bit left.
+ */
+static bool trans_FMUL(DisasContext *ctx, arg_FMUL *a)
+{
+ if (!avr_have_feature(ctx, AVR_FEATURE_MUL)) {
+ return true;
+ }
+
+ TCGv R0 = cpu_r[0];
+ TCGv R1 = cpu_r[1];
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+
+ tcg_gen_mul_tl(R, Rd, Rr); /* R = Rd *Rr */
+ tcg_gen_shli_tl(R, R, 1);
+
+ tcg_gen_andi_tl(R0, R, 0xff);
+ tcg_gen_shri_tl(R1, R, 8);
+ tcg_gen_andi_tl(R1, R1, 0xff);
+
+ tcg_gen_shri_tl(cpu_Cf, R, 16); /* Cf = R(16) */
+ tcg_gen_andi_tl(cpu_Zf, R, 0x0000ffff);
+
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * This instruction performs 8-bit x 8-bit -> 16-bit signed multiplication
+ * and shifts the result one bit left.
+ */
+static bool trans_FMULS(DisasContext *ctx, arg_FMULS *a)
+{
+ if (!avr_have_feature(ctx, AVR_FEATURE_MUL)) {
+ return true;
+ }
+
+ TCGv R0 = cpu_r[0];
+ TCGv R1 = cpu_r[1];
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+ TCGv t0 = tcg_temp_new_i32();
+ TCGv t1 = tcg_temp_new_i32();
+
+ tcg_gen_ext8s_tl(t0, Rd); /* make Rd full 32 bit signed */
+ tcg_gen_ext8s_tl(t1, Rr); /* make Rr full 32 bit signed */
+ tcg_gen_mul_tl(R, t0, t1); /* R = Rd *Rr */
+ tcg_gen_shli_tl(R, R, 1);
+
+ tcg_gen_andi_tl(R0, R, 0xff);
+ tcg_gen_shri_tl(R1, R, 8);
+ tcg_gen_andi_tl(R1, R1, 0xff);
+
+ tcg_gen_shri_tl(cpu_Cf, R, 16); /* Cf = R(16) */
+ tcg_gen_andi_tl(cpu_Zf, R, 0x0000ffff);
+
+ tcg_temp_free_i32(t1);
+ tcg_temp_free_i32(t0);
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * This instruction performs 8-bit x 8-bit -> 16-bit signed multiplication
+ * and shifts the result one bit left.
+ */
+static bool trans_FMULSU(DisasContext *ctx, arg_FMULSU *a)
+{
+ if (!avr_have_feature(ctx, AVR_FEATURE_MUL)) {
+ return true;
+ }
+
+ TCGv R0 = cpu_r[0];
+ TCGv R1 = cpu_r[1];
+ TCGv Rd = cpu_r[a->rd];
+ TCGv Rr = cpu_r[a->rr];
+ TCGv R = tcg_temp_new_i32();
+ TCGv t0 = tcg_temp_new_i32();
+
+ tcg_gen_ext8s_tl(t0, Rd); /* make Rd full 32 bit signed */
+ tcg_gen_mul_tl(R, t0, Rr); /* R = Rd *Rr */
+ tcg_gen_shli_tl(R, R, 1);
+
+ tcg_gen_andi_tl(R0, R, 0xff);
+ tcg_gen_shri_tl(R1, R, 8);
+ tcg_gen_andi_tl(R1, R1, 0xff);
+
+ tcg_gen_shri_tl(cpu_Cf, R, 16); /* Cf = R(16) */
+ tcg_gen_andi_tl(cpu_Zf, R, 0x0000ffff);
+
+ tcg_temp_free_i32(t0);
+ tcg_temp_free_i32(R);
+
+ return true;
+}
+
+
+/*
+ * The module is an instruction set extension to the AVR CPU, performing
+ * DES iterations. The 64-bit data block (plaintext or ciphertext) is placed in
+ * the CPU register file, registers R0-R7, where LSB of data is placed in LSB
+ * of R0 and MSB of data is placed in MSB of R7. The full 64-bit key (including
+ * parity bits) is placed in registers R8- R15, organized in the register file
+ * with LSB of key in LSB of R8 and MSB of key in MSB of R15. Executing one DES
+ * instruction performs one round in the DES algorithm. Sixteen rounds must be
+ * executed in increasing order to form the correct DES ciphertext or
+ * plaintext. Intermediate results are stored in the register file (R0-R15)
+ * after each DES instruction. The instruction's operand (K) determines which
+ * round is executed, and the half carry flag (H) determines whether encryption
+ * or decryption is performed. The DES algorithm is described in
+ * "Specifications for the Data Encryption Standard" (Federal Information
+ * Processing Standards Publication 46). Intermediate results in this
+ * implementation differ from the standard because the initial permutation and
+ * the inverse initial permutation are performed each iteration. This does not
+ * affect the result in the final ciphertext or plaintext, but reduces
+ * execution time.
+ */
+static bool trans_DES(DisasContext *ctx, arg_DES *a)
+{
+ /* TODO */
+ if (!avr_have_feature(ctx, AVR_FEATURE_DES)) {
+ return true;
+ }
+
+ return true;
+}
--
2.17.2 (Apple Git-113)
next prev parent reply other threads:[~2019-10-14 17:41 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-10-14 16:18 [PATCH v32 00/13] QEMU AVR 8 bit cores Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 01/13] target/avr: Add outward facing interfaces and core CPU logic Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 02/13] target/avr: Add instruction helpers Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 03/13] target/avr: Add instruction decoding Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 04/13] target/avr: Add instruction translation - Registers definition Michael Rolnik
2019-10-17 17:25 ` Aleksandar Markovic
2019-10-17 19:15 ` Michael Rolnik
2019-10-17 20:16 ` Aleksandar Markovic
2019-10-17 20:43 ` Michael Rolnik
2019-10-18 8:52 ` Aleksandar Markovic
2019-10-18 11:27 ` Michael Rolnik
2019-10-18 12:10 ` Michael Rolnik
2019-10-18 13:23 ` Aleksandar Markovic
2019-10-18 13:56 ` Michael Rolnik
2019-10-18 16:51 ` Aleksandar Markovic
2019-10-18 18:08 ` Aleksandar Markovic
2019-10-19 20:18 ` Michael Rolnik
2019-10-14 16:18 ` Michael Rolnik [this message]
2019-10-14 16:18 ` [PATCH v32 06/13] target/avr: Add instruction translation - Branch Instructions Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 07/13] target/avr: Add instruction translation - Bit and Bit-test Instructions Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 08/13] target/avr: Add instruction translation - MCU Control Instructions Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 09/13] target/avr: Add instruction translation - CPU main translation function Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 10/13] target/avr: Add limited support for USART and 16 bit timer peripherals Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 11/13] target/avr: Add example board configuration Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 12/13] target/avr: Register AVR support with the rest of QEMU, the build system, and the WMAINTAINERS file Michael Rolnik
2019-10-14 16:18 ` [PATCH v32 13/13] target/avr: Add tests Michael Rolnik
-- strict thread matches above, loose matches on Subject: below --
2019-10-13 7:47 [PATCH v32 00/13] QEMU AVR 8 bit cores Michael Rolnik
2019-10-13 7:47 ` [PATCH v32 05/13] target/avr: Add instruction translation - Arithmetic and Logic Instructions Michael Rolnik
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20191014161843.84845-6-mrolnik@gmail.com \
--to=mrolnik@gmail.com \
--cc=aleksandar.m.mail@gmail.com \
--cc=dovgaluk@ispras.ru \
--cc=imammedo@redhat.com \
--cc=philmd@redhat.com \
--cc=qemu-devel@nongnu.org \
--cc=richard.henderson@linaro.org \
--cc=thuth@redhat.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).