target/riscv: Add support for Zacas extension

Add support for amocas.w/d/q instructions which are part of the ratified
Zacas extension: https://github.com/riscv/riscv-zacas

Signed-off-by: Weiwei Li <liweiwei@iscas.ac.cn>
Signed-off-by: Junqiang Wang <wangjunqiang@iscas.ac.cn>
Signed-off-by: Rob Bradford <rbradford@rivosinc.com>
Reviewed-by: Daniel Henrique Barboza <dbarboza@ventanamicro.com>
Message-ID: <20231207153842.32401-2-rbradford@rivosinc.com>
Signed-off-by: Alistair Francis <alistair.francis@wdc.com>

target/riscv/insn_trans/trans_rvzacas.c.inc

@@ -0,0 +1,150 @@
+/*
+ * RISC-V translation routines for the RV64 Zacas Standard Extension.
+ *
+ * Copyright (c) 2020-2023 PLCT Lab
+ *
+ * This program is free software; you can redistribute it and/or modify it
+ * under the terms and conditions of the GNU General Public License,
+ * version 2 or later, as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
+ * more details.
+ *
+ * You should have received a copy of the GNU General Public License along with
+ * this program.  If not, see <http://www.gnu.org/licenses/>.
+ */
+
+#define REQUIRE_ZACAS(ctx) do {          \
+    if (!ctx->cfg_ptr->ext_zacas) {      \
+        return false;                     \
+    }                                     \
+} while (0)
+
+static bool gen_cmpxchg(DisasContext *ctx, arg_atomic *a, MemOp mop)
+{
+    TCGv dest = get_gpr(ctx, a->rd, EXT_NONE);
+    TCGv src1 = get_address(ctx, a->rs1, 0);
+    TCGv src2 = get_gpr(ctx, a->rs2, EXT_NONE);
+
+    decode_save_opc(ctx);
+    tcg_gen_atomic_cmpxchg_tl(dest, src1, dest, src2, ctx->mem_idx, mop);
+
+    gen_set_gpr(ctx, a->rd, dest);
+    return true;
+}
+
+static bool trans_amocas_w(DisasContext *ctx, arg_amocas_w *a)
+{
+    REQUIRE_ZACAS(ctx);
+    return gen_cmpxchg(ctx, a, MO_ALIGN | MO_TESL);
+}
+
+static TCGv_i64 get_gpr_pair(DisasContext *ctx, int reg_num)
+{
+    TCGv_i64 t;
+
+    assert(get_ol(ctx) == MXL_RV32);
+
+    if (reg_num == 0) {
+        return tcg_constant_i64(0);
+    }
+
+    t = tcg_temp_new_i64();
+    tcg_gen_concat_tl_i64(t, cpu_gpr[reg_num], cpu_gpr[reg_num + 1]);
+    return t;
+}
+
+static void gen_set_gpr_pair(DisasContext *ctx, int reg_num, TCGv_i64 t)
+{
+    assert(get_ol(ctx) == MXL_RV32);
+
+    if (reg_num != 0) {
+#ifdef TARGET_RISCV32
+        tcg_gen_extr_i64_i32(cpu_gpr[reg_num], cpu_gpr[reg_num + 1], t);
+#else
+        tcg_gen_ext32s_i64(cpu_gpr[reg_num], t);
+        tcg_gen_sari_i64(cpu_gpr[reg_num + 1], t, 32);
+#endif
+
+        if (get_xl_max(ctx) == MXL_RV128) {
+            tcg_gen_sari_tl(cpu_gprh[reg_num], cpu_gpr[reg_num], 63);
+            tcg_gen_sari_tl(cpu_gprh[reg_num + 1], cpu_gpr[reg_num + 1], 63);
+        }
+    }
+}
+
+static bool gen_cmpxchg64(DisasContext *ctx, arg_atomic *a, MemOp mop)
+{
+    /*
+     * Encodings with odd numbered registers specified in rs2 and rd are
+     * reserved.
+     */
+    if ((a->rs2 | a->rd) & 1) {
+        return false;
+    }
+
+    TCGv_i64 dest = get_gpr_pair(ctx, a->rd);
+    TCGv src1 = get_address(ctx, a->rs1, 0);
+    TCGv_i64 src2 = get_gpr_pair(ctx, a->rs2);
+
+    decode_save_opc(ctx);
+    tcg_gen_atomic_cmpxchg_i64(dest, src1, dest, src2, ctx->mem_idx, mop);
+
+    gen_set_gpr_pair(ctx, a->rd, dest);
+    return true;
+}
+
+static bool trans_amocas_d(DisasContext *ctx, arg_amocas_d *a)
+{
+    REQUIRE_ZACAS(ctx);
+    switch (get_ol(ctx)) {
+    case MXL_RV32:
+        return gen_cmpxchg64(ctx, a, MO_ALIGN | MO_TEUQ);
+    case MXL_RV64:
+    case MXL_RV128:
+        return gen_cmpxchg(ctx, a, MO_ALIGN | MO_TEUQ);
+    default:
+        g_assert_not_reached();
+    }
+}
+
+static bool trans_amocas_q(DisasContext *ctx, arg_amocas_q *a)
+{
+    REQUIRE_ZACAS(ctx);
+    REQUIRE_64BIT(ctx);
+
+    /*
+     * Encodings with odd numbered registers specified in rs2 and rd are
+     * reserved.
+     */
+    if ((a->rs2 | a->rd) & 1) {
+        return false;
+    }
+
+#ifdef TARGET_RISCV64
+    TCGv_i128 dest = tcg_temp_new_i128();
+    TCGv src1 = get_address(ctx, a->rs1, 0);
+    TCGv_i128 src2 = tcg_temp_new_i128();
+    TCGv_i64 src2l = get_gpr(ctx, a->rs2, EXT_NONE);
+    TCGv_i64 src2h = get_gpr(ctx, a->rs2 == 0 ? 0 : a->rs2 + 1, EXT_NONE);
+    TCGv_i64 destl = get_gpr(ctx, a->rd, EXT_NONE);
+    TCGv_i64 desth = get_gpr(ctx, a->rd == 0 ? 0 : a->rd + 1, EXT_NONE);
+
+    tcg_gen_concat_i64_i128(src2, src2l, src2h);
+    tcg_gen_concat_i64_i128(dest, destl, desth);
+    decode_save_opc(ctx);
+    tcg_gen_atomic_cmpxchg_i128(dest, src1, dest, src2, ctx->mem_idx,
+                                (MO_ALIGN | MO_TEUO));
+
+    tcg_gen_extr_i128_i64(destl, desth, dest);
+
+    if (a->rd != 0) {
+        gen_set_gpr(ctx, a->rd, destl);
+        gen_set_gpr(ctx, a->rd + 1, desth);
+    }
+#endif
+
+    return true;
+}