[PATCH 09/11] RISC-V: Adding T-Head MemIdx extension

[Christoph Muellner]

From: Christoph Müllner <christoph.muellner@vrull.eu>

This patch adds support for the T-Head MemIdx instructions.
The patch uses the T-Head specific decoder and translation.

Signed-off-by: Christoph Müllner <christoph.muellner@vrull.eu>
---
 target/riscv/cpu.c                         |   1 +
 target/riscv/cpu.h                         |   1 +
 target/riscv/insn_trans/trans_xthead.c.inc | 377 +++++++++++++++++++++
 target/riscv/meson.build                   |   1 +
 target/riscv/translate.c                   |   3 +
 target/riscv/xtheadmemidx.decode           |  73 ++++
 6 files changed, 456 insertions(+)
 create mode 100644 target/riscv/xtheadmemidx.decode

diff --git a/target/riscv/cpu.c b/target/riscv/cpu.c
index 9370722ffa..0af9cc7bec 100644
--- a/target/riscv/cpu.c
+++ b/target/riscv/cpu.c
@@ -926,6 +926,7 @@ static Property riscv_cpu_extensions[] = {
     DEFINE_PROP_BOOL("xtheadcmo", RISCVCPU, cfg.ext_xtheadcmo, false),
     DEFINE_PROP_BOOL("xtheadcondmov", RISCVCPU, cfg.ext_xtheadcondmov, false),
     DEFINE_PROP_BOOL("xtheadmac", RISCVCPU, cfg.ext_xtheadmac, false),
+    DEFINE_PROP_BOOL("xtheadmemidx", RISCVCPU, cfg.ext_xtheadmemidx, false),
     DEFINE_PROP_BOOL("xtheadmempair", RISCVCPU, cfg.ext_xtheadmempair, false),
     DEFINE_PROP_BOOL("xtheadsync", RISCVCPU, cfg.ext_xtheadsync, false),
     DEFINE_PROP_BOOL("xtheadxmae", RISCVCPU, cfg.ext_xtheadxmae, false),
diff --git a/target/riscv/cpu.h b/target/riscv/cpu.h
index 6cc2d19075..590a597f39 100644
--- a/target/riscv/cpu.h
+++ b/target/riscv/cpu.h
@@ -446,6 +446,7 @@ struct RISCVCPUConfig {
     bool ext_xtheadcmo;
     bool ext_xtheadcondmov;
     bool ext_xtheadmac;
+    bool ext_xtheadmemidx;
     bool ext_xtheadmempair;
     bool ext_xtheadsync;
     bool ext_xtheadxmae;
diff --git a/target/riscv/insn_trans/trans_xthead.c.inc 
b/target/riscv/insn_trans/trans_xthead.c.inc
index a2bae249fb..95c6b10d77 100644
--- a/target/riscv/insn_trans/trans_xthead.c.inc
+++ b/target/riscv/insn_trans/trans_xthead.c.inc
@@ -374,3 +374,380 @@ static bool trans_th_swd(DisasContext *ctx, arg_th_pair 
*a)
 {
     return gen_storepair_tl(ctx, a, MO_TESL, 3);
 }
+
+/*
+ * Load with memop from indexed address and add sext(imm5 << imm2) to rs1.
+ * If !preinc, then the address is rs1.
+ * If  preinc, then the address is rs1 + (sext(imm5) << imm2).
+ */
+static bool gen_load_inc(DisasContext *ctx, arg_th_meminc *a, MemOp memop,
+                         bool preinc)
+{
+    TCGv rd = dest_gpr(ctx, a->rd);
+    TCGv base = get_gpr(ctx, a->rs1, EXT_NONE);
+    TCGv addr = tcg_temp_new();
+    TCGv offs = tcg_temp_new();
+
+    tcg_gen_movi_tl(offs, a->imm5);
+    tcg_gen_sextract_tl(offs, offs, 0, 5);
+    tcg_gen_shli_tl(offs, offs, a->imm2);
+
+    if (preinc) {
+        tcg_gen_add_tl(addr, base, offs);
+        if (get_xl(ctx) == MXL_RV32) {
+            tcg_gen_ext32u_tl(addr, addr);
+        }
+    } else {
+        tcg_gen_mov_tl(addr, base);
+    }
+
+    tcg_gen_qemu_ld_tl(rd, addr, ctx->mem_idx, memop);
+
+    if (!preinc) {
+        tcg_gen_add_tl(addr, base, offs);
+        if (get_xl(ctx) == MXL_RV32) {
+            tcg_gen_ext32u_tl(addr, addr);
+        }
+    }
+
+    gen_set_gpr(ctx, a->rd, rd);
+    gen_set_gpr(ctx, a->rs1, addr);
+
+    tcg_temp_free(addr);
+    tcg_temp_free(offs);
+    return true;
+}
+
+/*
+ * Store with memop to indexed address and add sext(imm5 << imm2) to rs1.
+ * If !preinc, then the address is rs1.
+ * If  preinc, then the address is rs1 + (sext(imm5) << imm2).
+ */
+static bool gen_store_inc(DisasContext *ctx, arg_th_meminc *a, MemOp memop,
+                          bool preinc)
+{
+    TCGv data = get_gpr(ctx, a->rd, EXT_NONE);
+    TCGv base = get_gpr(ctx, a->rs1, EXT_NONE);
+    TCGv addr = tcg_temp_new();
+    TCGv offs = tcg_temp_new();
+
+    tcg_gen_movi_tl(offs, a->imm5);
+    tcg_gen_sextract_tl(offs, offs, 0, 5);
+    tcg_gen_shli_tl(offs, offs, a->imm2);
+
+    if (preinc) {
+        tcg_gen_add_tl(addr, base, offs);
+        if (get_xl(ctx) == MXL_RV32) {
+            tcg_gen_ext32u_tl(addr, addr);
+        }
+    } else {
+        tcg_gen_mov_tl(addr, base);
+    }
+
+    tcg_gen_qemu_st_tl(data, addr, ctx->mem_idx, memop);
+
+    if (!preinc) {
+        tcg_gen_add_tl(addr, base, offs);
+        if (get_xl(ctx) == MXL_RV32) {
+            tcg_gen_ext32u_tl(addr, addr);
+        }
+    }
+
+    gen_set_gpr(ctx, a->rs1, addr);
+
+    tcg_temp_free(addr);
+    tcg_temp_free(offs);
+    return true;
+}
+
+static bool trans_th_ldia(DisasContext *ctx, arg_th_meminc *a)
+{
+    REQUIRE_64BIT(ctx);
+    return gen_load_inc(ctx, a, MO_TESQ, false);
+}
+
+static bool trans_th_ldib(DisasContext *ctx, arg_th_meminc *a)
+{
+    REQUIRE_64BIT(ctx);
+    return gen_load_inc(ctx, a, MO_TESQ, true);
+}
+
+static bool trans_th_lwia(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_load_inc(ctx, a, MO_TESL, false);
+}
+
+static bool trans_th_lwib(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_load_inc(ctx, a, MO_TESL, true);
+}
+
+static bool trans_th_lwuia(DisasContext *ctx, arg_th_meminc *a)
+{
+    REQUIRE_64BIT(ctx);
+    return gen_load_inc(ctx, a, MO_TEUL, false);
+}
+
+static bool trans_th_lwuib(DisasContext *ctx, arg_th_meminc *a)
+{
+    REQUIRE_64BIT(ctx);
+    return gen_load_inc(ctx, a, MO_TEUL, true);
+}
+
+static bool trans_th_lhia(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_load_inc(ctx, a, MO_TESW, false);
+}
+
+static bool trans_th_lhib(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_load_inc(ctx, a, MO_TESW, true);
+}
+
+static bool trans_th_lhuia(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_load_inc(ctx, a, MO_TEUW, false);
+}
+
+static bool trans_th_lhuib(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_load_inc(ctx, a, MO_TEUW, true);
+}
+
+static bool trans_th_lbia(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_load_inc(ctx, a, MO_SB, false);
+}
+
+static bool trans_th_lbib(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_load_inc(ctx, a, MO_SB, true);
+}
+
+static bool trans_th_lbuia(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_load_inc(ctx, a, MO_UB, false);
+}
+
+static bool trans_th_lbuib(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_load_inc(ctx, a, MO_UB, true);
+}
+
+static bool trans_th_sdia(DisasContext *ctx, arg_th_meminc *a)
+{
+    REQUIRE_64BIT(ctx);
+    return gen_store_inc(ctx, a, MO_TESQ, false);
+}
+
+static bool trans_th_sdib(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_store_inc(ctx, a, MO_TESQ, true);
+}
+
+static bool trans_th_swia(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_store_inc(ctx, a, MO_TESL, false);
+}
+
+static bool trans_th_swib(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_store_inc(ctx, a, MO_TESL, true);
+}
+
+static bool trans_th_shia(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_store_inc(ctx, a, MO_TESW, false);
+}
+
+static bool trans_th_shib(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_store_inc(ctx, a, MO_TESW, true);
+}
+
+static bool trans_th_sbia(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_store_inc(ctx, a, MO_SB, false);
+}
+
+static bool trans_th_sbib(DisasContext *ctx, arg_th_meminc *a)
+{
+    return gen_store_inc(ctx, a, MO_SB, true);
+}
+
+/*
+ * Load with memop from indexed address.
+ * If !zero_extend_offset, then address is rs1 + (rs2 << imm2).
+ * If  zero_extend_offset, then address is rs1 + (zext(rs2[31:0]) << imm2).
+ */
+static bool gen_load_idx(DisasContext *ctx, arg_th_memidx *a, MemOp memop,
+                         bool zero_extend_offset)
+{
+    TCGv rd = dest_gpr(ctx, a->rd);
+    TCGv base = get_gpr(ctx, a->rs1, EXT_NONE);
+    TCGv offs = get_gpr(ctx, a->rs2, EXT_NONE);
+    TCGv addr = tcg_temp_new();
+
+    if (zero_extend_offset) {
+        tcg_gen_extract_tl(addr, offs, 0, 32);
+    } else {
+        tcg_gen_mov_tl(addr, offs);
+    }
+    tcg_gen_shli_tl(addr, addr, a->imm2);
+    tcg_gen_add_tl(addr, base, addr);
+
+    if (get_xl(ctx) == MXL_RV32) {
+        tcg_gen_ext32u_tl(addr, addr);
+    }
+
+    tcg_gen_qemu_ld_tl(rd, addr, ctx->mem_idx, memop);
+    gen_set_gpr(ctx, a->rd, rd);
+
+    tcg_temp_free(addr);
+    return true;
+}
+
+/*
+ * Store with memop to indexed address.
+ * If !zero_extend_offset, then address is rs1 + (rs2 << imm2).
+ * If  zero_extend_offset, then address is rs1 + (zext(rs2[31:0]) << imm2).
+ */
+static bool gen_store_idx(DisasContext *ctx, arg_th_memidx *a, MemOp memop,
+                          bool zero_extend_offset)
+{
+    TCGv data = get_gpr(ctx, a->rd, EXT_NONE);
+    TCGv base = get_gpr(ctx, a->rs1, EXT_NONE);
+    TCGv offs = get_gpr(ctx, a->rs2, EXT_NONE);
+    TCGv addr = tcg_temp_new();
+
+    if (zero_extend_offset) {
+        tcg_gen_extract_tl(addr, offs, 0, 32);
+    } else {
+        tcg_gen_mov_tl(addr, offs);
+    }
+    tcg_gen_shli_tl(addr, addr, a->imm2);
+    tcg_gen_add_tl(addr, base, addr);
+
+    if (get_xl(ctx) == MXL_RV32) {
+        tcg_gen_ext32u_tl(addr, addr);
+    }
+
+    tcg_gen_qemu_st_tl(data, addr, ctx->mem_idx, memop);
+
+    tcg_temp_free(addr);
+    return true;
+}
+
+static bool trans_th_lrd(DisasContext *ctx, arg_th_memidx *a)
+{
+    REQUIRE_64BIT(ctx);
+    return gen_load_idx(ctx, a, MO_TESQ, false);
+}
+
+static bool trans_th_lrw(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_load_idx(ctx, a, MO_TESL, false);
+}
+
+static bool trans_th_lrwu(DisasContext *ctx, arg_th_memidx *a)
+{
+    REQUIRE_64BIT(ctx);
+    return gen_load_idx(ctx, a, MO_TEUL, false);
+}
+
+static bool trans_th_lrh(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_load_idx(ctx, a, MO_TESW, false);
+}
+
+static bool trans_th_lrhu(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_load_idx(ctx, a, MO_TEUW, false);
+}
+
+static bool trans_th_lrb(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_load_idx(ctx, a, MO_SB, false);
+}
+
+static bool trans_th_lrbu(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_load_idx(ctx, a, MO_UB, false);
+}
+
+static bool trans_th_srd(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_store_idx(ctx, a, MO_TESQ, false);
+}
+
+static bool trans_th_srw(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_store_idx(ctx, a, MO_TESL, false);
+}
+
+static bool trans_th_srh(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_store_idx(ctx, a, MO_TESW, false);
+}
+
+static bool trans_th_srb(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_store_idx(ctx, a, MO_SB, false);
+}
+static bool trans_th_lurd(DisasContext *ctx, arg_th_memidx *a)
+{
+    REQUIRE_64BIT(ctx);
+    return gen_load_idx(ctx, a, MO_TESQ, true);
+}
+
+static bool trans_th_lurw(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_load_idx(ctx, a, MO_TESL, true);
+}
+
+static bool trans_th_lurwu(DisasContext *ctx, arg_th_memidx *a)
+{
+    REQUIRE_64BIT(ctx);
+    return gen_load_idx(ctx, a, MO_TEUL, true);
+}
+
+static bool trans_th_lurh(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_load_idx(ctx, a, MO_TESW, true);
+}
+
+static bool trans_th_lurhu(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_load_idx(ctx, a, MO_TEUW, true);
+}
+
+static bool trans_th_lurb(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_load_idx(ctx, a, MO_SB, true);
+}
+
+static bool trans_th_lurbu(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_load_idx(ctx, a, MO_UB, true);
+}
+
+static bool trans_th_surd(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_store_idx(ctx, a, MO_TESQ, true);
+}
+
+static bool trans_th_surw(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_store_idx(ctx, a, MO_TESL, true);
+}
+
+static bool trans_th_surh(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_store_idx(ctx, a, MO_TESW, true);
+}
+
+static bool trans_th_surb(DisasContext *ctx, arg_th_memidx *a)
+{
+    return gen_store_idx(ctx, a, MO_SB, true);
+}
+
diff --git a/target/riscv/meson.build b/target/riscv/meson.build
index 998f0ba336..30bb4c5bab 100644
--- a/target/riscv/meson.build
+++ b/target/riscv/meson.build
@@ -8,6 +8,7 @@ gen = [
   decodetree.process('xtheadcmo.decode', extra_args: 
'--static-decode=decode_xtheadcmo'),
   decodetree.process('xtheadcondmov.decode', extra_args: 
'--static-decode=decode_xtheadcondmov'),
   decodetree.process('xtheadmac.decode', extra_args: 
'--static-decode=decode_xtheadmac'),
+  decodetree.process('xtheadmemidx.decode', extra_args: 
'--static-decode=decode_xtheadmemidx'),
   decodetree.process('xtheadmempair.decode', extra_args: 
'--static-decode=decode_xtheadmempair'),
   decodetree.process('xtheadsync.decode', extra_args: 
'--static-decode=decode_xtheadsync'),
   decodetree.process('XVentanaCondOps.decode', extra_args: 
'--static-decode=decode_XVentanaCodeOps'),
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 308de419cb..1cb0d885b8 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -138,6 +138,7 @@ MATERIALISE_EXT_PREDICATE(xtheadbs)
 MATERIALISE_EXT_PREDICATE(xtheadcmo)
 MATERIALISE_EXT_PREDICATE(xtheadcondmov);
 MATERIALISE_EXT_PREDICATE(xtheadmac);
+MATERIALISE_EXT_PREDICATE(xtheadmemidx);
 MATERIALISE_EXT_PREDICATE(xtheadmempair);
 MATERIALISE_EXT_PREDICATE(xtheadsync)
 MATERIALISE_EXT_PREDICATE(XVentanaCondOps)
@@ -732,6 +733,7 @@ static int ex_rvc_shifti(DisasContext *ctx, int imm)
 #include "decode-xtheadcmo.c.inc"
 #include "decode-xtheadcondmov.c.inc"
 #include "decode-xtheadmac.c.inc"
+#include "decode-xtheadmemidx.c.inc"
 #include "decode-xtheadmempair.c.inc"
 #include "decode-xtheadsync.c.inc"
 #include "decode-XVentanaCondOps.c.inc"
@@ -1060,6 +1062,7 @@ static void decode_opc(CPURISCVState *env, DisasContext 
*ctx, uint16_t opcode)
         { has_xtheadcmo_p, decode_xtheadcmo },
         { has_xtheadcondmov_p, decode_xtheadcondmov },
         { has_xtheadmac_p, decode_xtheadmac },
+        { has_xtheadmemidx_p, decode_xtheadmemidx },
         { has_xtheadmempair_p, decode_xtheadmempair },
         { has_xtheadsync_p, decode_xtheadsync },
         { has_XVentanaCondOps_p,  decode_XVentanaCodeOps },
diff --git a/target/riscv/xtheadmemidx.decode b/target/riscv/xtheadmemidx.decode
new file mode 100644
index 0000000000..d2e0f2af6f
--- /dev/null
+++ b/target/riscv/xtheadmemidx.decode
@@ -0,0 +1,73 @@
+#
+# RISC-V instruction decode for the XTheadMemIdx extension
+#
+# Copyright (c) 2022 Christoph Muellner, christoph.muellner@vrull.eu
+#
+# SPDX-License-Identifier: LGPL-2.1-or-later
+#
+# The XTheadMemIdx extension provides GPR memory operations.
+#
+# It is documented in
+# 
https://github.com/T-head-Semi/thead-extension-spec/releases/download/2.0.0/xthead-2022-09-05-2.0.0.pdf
+
+# Fields
+%imm2      25:2
+%imm5      20:5
+%rs2       20:5
+%rs1       15:5
+%rd         7:5
+
+# Argument sets
+&th_meminc      rd rs1 imm5 imm2
+&th_memidx      rd rs1 rs2 imm2
+
+# Formats:
+@th_meminc   ..... .. ..... ..... ... ..... ....... &th_meminc %rd %rs1 %imm5 
%imm2
+@th_memidx   ..... .. ..... ..... ... ..... ....... &th_memidx %rd %rs1 %rs2 
%imm2
+
+th_ldia      01111 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_ldib      01101 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lwia      01011 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lwib      01001 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lwuia     11011 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lwuib     11001 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lhia      00111 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lhib      00101 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lhuia     10111 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lhuib     10101 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lbia      00011 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lbib      00001 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lbuia     10011 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_lbuib     10001 .. ..... ..... 100 ..... 0001011 @th_meminc
+th_sdia      01111 .. ..... ..... 101 ..... 0001011 @th_meminc
+th_sdib      01101 .. ..... ..... 101 ..... 0001011 @th_meminc
+th_swia      01011 .. ..... ..... 101 ..... 0001011 @th_meminc
+th_swib      01001 .. ..... ..... 101 ..... 0001011 @th_meminc
+th_shia      00111 .. ..... ..... 101 ..... 0001011 @th_meminc
+th_shib      00101 .. ..... ..... 101 ..... 0001011 @th_meminc
+th_sbia      00011 .. ..... ..... 101 ..... 0001011 @th_meminc
+th_sbib      00001 .. ..... ..... 101 ..... 0001011 @th_meminc
+
+th_lrd       01100 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lrw       01000 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lrwu      11000 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lrh       00100 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lrhu      10100 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lrb       00000 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lrbu      10000 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_srd       01100 .. ..... ..... 101 ..... 0001011 @th_memidx
+th_srw       01000 .. ..... ..... 101 ..... 0001011 @th_memidx
+th_srh       00100 .. ..... ..... 101 ..... 0001011 @th_memidx
+th_srb       00000 .. ..... ..... 101 ..... 0001011 @th_memidx
+
+th_lurd      01110 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lurw      01010 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lurwu     11010 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lurh      00110 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lurhu     10110 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lurb      00010 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_lurbu     10010 .. ..... ..... 100 ..... 0001011 @th_memidx
+th_surd      01110 .. ..... ..... 101 ..... 0001011 @th_memidx
+th_surw      01010 .. ..... ..... 101 ..... 0001011 @th_memidx
+th_surh      00110 .. ..... ..... 101 ..... 0001011 @th_memidx
+th_surb      00010 .. ..... ..... 101 ..... 0001011 @th_memidx
-- 
2.37.2