[Qemu-riscv] [PULL] target/riscv: Convert to decodetree

[Qemu-riscv] [PULL] target/riscv: Convert to decodetree

[Palmer Dabbelt]

merged tag 'pull-request-2019-03-12'
Primary key fingerprint: 27B8 8847 EEE0 2501 18F3  EAB9 2ED9 D774 FE70 2DB5
The following changes since commit 3f3bbfc7cef4490c5ed5550766a81e7d18f08db1:

  Merge remote-tracking branch 'remotes/huth-gitlab/tags/pull-request-2019-03-12' into staging (2019-03-12 21:06:26 +0000)

are available in the Git repository at:

  git://github.com/palmer-dabbelt/qemu.git tags/riscv-for-master-4.0-sf4

for you to fetch changes up to 25e6ca30c668783cd72ff97080ff44e141b99f9b:

  target/riscv: Remove decode_RV32_64G() (2019-03-13 10:40:50 +0100)

----------------------------------------------------------------
target/riscv: Convert to decodetree

Bastian: this patchset converts the RISC-V decoder to decodetree in four major steps:

1) Convert 32-bit instructions to decodetree [Patch 1-15]:
    Many of the gen_* functions are called by the decode functions for 16-bit
    and 32-bit functions. If we move translation code from the gen_*
    functions to the generated trans_* functions of decode-tree, we get a lot of
    duplication. Therefore, we mostly generate calls to the old gen_* function
    which are properly replaced after step 2).

    Each of the trans_ functions are grouped into files corresponding to their
    ISA extension, e.g. addi which is in RV32I is translated in the file
    'trans_rvi.inc.c'.

2) Convert 16-bit instructions to decodetree [Patch 16-18]:
    All 16 bit instructions have a direct mapping to a 32 bit instruction. Thus,
    we convert the arguments in the 16 bit trans_ function to the arguments of
    the corresponding 32 bit instruction and call the 32 bit trans_ function.

3) Remove old manual decoding in gen_* function [Patch 19-29]:
    this move all manual translation code into the trans_* instructions of
    decode tree, such that we can remove the old decode_* functions.

Palmer: This, with some additional cleanup patches, passed Alistar's
testing on rv32 and rv64 as well as my testing on rv64, so I think it's
good to go.  I've run my standard test against this exact tag.

I still don't have a Mac to try this on, sorry!  If this doesn't work
then I'll go try to find one tomorrow.

----------------------------------------------------------------
Bastian Koppelmann (29):
      target/riscv: Activate decodetree and implemnt LUI & AUIPC
      target/riscv: Convert RVXI branch insns to decodetree
      target/riscv: Convert RV32I load/store insns to decodetree
      target/riscv: Convert RV64I load/store insns to decodetree
      target/riscv: Convert RVXI arithmetic insns to decodetree
      target/riscv: Convert RVXI fence insns to decodetree
      target/riscv: Convert RVXI csr insns to decodetree
      target/riscv: Convert RVXM insns to decodetree
      target/riscv: Convert RV32A insns to decodetree
      target/riscv: Convert RV64A insns to decodetree
      target/riscv: Convert RV32F insns to decodetree
      target/riscv: Convert RV64F insns to decodetree
      target/riscv: Convert RV32D insns to decodetree
      target/riscv: Convert RV64D insns to decodetree
      target/riscv: Convert RV priv insns to decodetree
      target/riscv: Convert quadrant 0 of RVXC insns to decodetree
      target/riscv: Convert quadrant 1 of RVXC insns to decodetree
      target/riscv: Convert quadrant 2 of RVXC insns to decodetree
      target/riscv: Remove gen_jalr()
      target/riscv: Remove manual decoding from gen_branch()
      target/riscv: Remove manual decoding from gen_load()
      target/riscv: Remove manual decoding from gen_store()
      target/riscv: Move gen_arith_imm() decoding into trans_* functions
      target/riscv: make ADD/SUB/OR/XOR/AND insn use arg lists
      target/riscv: Remove shift and slt insn manual decoding
      target/riscv: Remove manual decoding of RV32/64M insn
      target/riscv: Rename trans_arith to gen_arith
      target/riscv: Remove gen_system()
      target/riscv: Remove decode_RV32_64G()

 target/riscv/Makefile.objs                     |   19 +
 target/riscv/insn16.decode                     |  129 ++
 target/riscv/insn32-64.decode                  |   72 +
 target/riscv/insn32.decode                     |  201 +++
 target/riscv/insn_trans/trans_privileged.inc.c |  110 ++
 target/riscv/insn_trans/trans_rva.inc.c        |  218 +++
 target/riscv/insn_trans/trans_rvc.inc.c        |  327 +++++
 target/riscv/insn_trans/trans_rvd.inc.c        |  442 ++++++
 target/riscv/insn_trans/trans_rvf.inc.c        |  439 ++++++
 target/riscv/insn_trans/trans_rvi.inc.c        |  568 ++++++++
 target/riscv/insn_trans/trans_rvm.inc.c        |  120 ++
 target/riscv/translate.c                       | 1847 ++++--------------------
 12 files changed, 2897 insertions(+), 1595 deletions(-)
 create mode 100644 target/riscv/insn16.decode
 create mode 100644 target/riscv/insn32-64.decode
 create mode 100644 target/riscv/insn32.decode
 create mode 100644 target/riscv/insn_trans/trans_privileged.inc.c
 create mode 100644 target/riscv/insn_trans/trans_rva.inc.c
 create mode 100644 target/riscv/insn_trans/trans_rvc.inc.c
 create mode 100644 target/riscv/insn_trans/trans_rvd.inc.c
 create mode 100644 target/riscv/insn_trans/trans_rvf.inc.c
 create mode 100644 target/riscv/insn_trans/trans_rvi.inc.c
 create mode 100644 target/riscv/insn_trans/trans_rvm.inc.c
From Palmer Dabbelt <palmer@sifive.com> # This line is ignored.
From: Palmer Dabbelt <palmer@sifive.com>
Reply-To: 
Subject: 
In-Reply-To: 

[Qemu-riscv] [PULL 01/29] target/riscv: Activate decodetree and implemnt LUI & AUIPC

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

for now only LUI & AUIPC are decoded and translated. If decodetree fails, we
fall back to the old decoder.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Acked-by: Alistair Francis <alistair.francis@wdc.com>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/Makefile.objs              | 10 +++++++
 target/riscv/insn32.decode              | 30 +++++++++++++++++++++
 target/riscv/insn_trans/trans_rvi.inc.c | 35 +++++++++++++++++++++++++
 target/riscv/translate.c                | 31 ++++++++++++----------
 4 files changed, 92 insertions(+), 14 deletions(-)
 create mode 100644 target/riscv/insn32.decode
 create mode 100644 target/riscv/insn_trans/trans_rvi.inc.c

diff --git a/target/riscv/Makefile.objs b/target/riscv/Makefile.objs
index 4072abe3e45c..bf0a268033a0 100644
--- a/target/riscv/Makefile.objs
+++ b/target/riscv/Makefile.objs
@@ -1 +1,11 @@
 obj-y += translate.o op_helper.o cpu_helper.o cpu.o csr.o fpu_helper.o gdbstub.o pmp.o
+
+DECODETREE = $(SRC_PATH)/scripts/decodetree.py
+
+target/riscv/decode_insn32.inc.c: \
+  $(SRC_PATH)/target/riscv/insn32.decode $(DECODETREE)
+	$(call quiet-command, \
+	  $(PYTHON) $(DECODETREE) -o $@ --decode decode_insn32 $<, \
+	  "GEN", $(TARGET_DIR)$@)
+
+target/riscv/translate.o: target/riscv/decode_insn32.inc.c
diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
new file mode 100644
index 000000000000..44d4e922b6fa
--- /dev/null
+++ b/target/riscv/insn32.decode
@@ -0,0 +1,30 @@
+#
+# RISC-V translation routines for the RVXI Base Integer Instruction Set.
+#
+# Copyright (c) 2018 Peer Adelt, peer.adelt@hni.uni-paderborn.de
+#                    Bastian Koppelmann, kbastian@mail.uni-paderborn.de
+#
+# 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/>.
+
+# Fields:
+%rd        7:5
+
+# immediates:
+%imm_u    12:s20                 !function=ex_shift_12
+
+# Formats 32:
+@u       ....................      ..... .......         imm=%imm_u          %rd
+
+# *** RV32I Base Instruction Set ***
+lui      ....................       ..... 0110111 @u
+auipc    ....................       ..... 0010111 @u
diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
new file mode 100644
index 000000000000..9885a8d27551
--- /dev/null
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -0,0 +1,35 @@
+/*
+ * RISC-V translation routines for the RVXI Base Integer Instruction Set.
+ *
+ * Copyright (c) 2016-2017 Sagar Karandikar, sagark@eecs.berkeley.edu
+ * Copyright (c) 2018 Peer Adelt, peer.adelt@hni.uni-paderborn.de
+ *                    Bastian Koppelmann, kbastian@mail.uni-paderborn.de
+ *
+ * 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/>.
+ */
+
+static bool trans_lui(DisasContext *ctx, arg_lui *a)
+{
+    if (a->rd != 0) {
+        tcg_gen_movi_tl(cpu_gpr[a->rd], a->imm);
+    }
+    return true;
+}
+
+static bool trans_auipc(DisasContext *ctx, arg_auipc *a)
+{
+    if (a->rd != 0) {
+        tcg_gen_movi_tl(cpu_gpr[a->rd], a->imm + ctx->base.pc_next);
+    }
+    return true;
+}
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index b7176cbf98e1..a273ac827448 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -1874,6 +1874,19 @@ static void decode_RV32_64C(DisasContext *ctx)
     }
 }
 
+#define EX_SH(amount) \
+    static int ex_shift_##amount(int imm) \
+    {                                         \
+        return imm << amount;                 \
+    }
+EX_SH(12)
+
+bool decode_insn32(DisasContext *ctx, uint32_t insn);
+/* Include the auto-generated decoder for 32 bit insn */
+#include "decode_insn32.inc.c"
+/* Include insn module translation function */
+#include "insn_trans/trans_rvi.inc.c"
+
 static void decode_RV32_64G(DisasContext *ctx)
 {
     int rs1;
@@ -1894,19 +1907,6 @@ static void decode_RV32_64G(DisasContext *ctx)
     imm = GET_IMM(ctx->opcode);
 
     switch (op) {
-    case OPC_RISC_LUI:
-        if (rd == 0) {
-            break; /* NOP */
-        }
-        tcg_gen_movi_tl(cpu_gpr[rd], sextract64(ctx->opcode, 12, 20) << 12);
-        break;
-    case OPC_RISC_AUIPC:
-        if (rd == 0) {
-            break; /* NOP */
-        }
-        tcg_gen_movi_tl(cpu_gpr[rd], (sextract64(ctx->opcode, 12, 20) << 12) +
-               ctx->base.pc_next);
-        break;
     case OPC_RISC_JAL:
         imm = GET_JAL_IMM(ctx->opcode);
         gen_jal(ctx, rd, imm);
@@ -2011,7 +2011,10 @@ static void decode_opc(DisasContext *ctx)
         }
     } else {
         ctx->pc_succ_insn = ctx->base.pc_next + 4;
-        decode_RV32_64G(ctx);
+        if (!decode_insn32(ctx, ctx->opcode)) {
+            /* fallback to old decoder */
+            decode_RV32_64G(ctx);
+        }
     }
 }
 
-- 
2.19.2

[Qemu-riscv] [PULL 02/29] target/riscv: Convert RVXI branch insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Palmer Dabbelt <palmer@sifive.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32.decode              | 19 ++++++++++
 target/riscv/insn_trans/trans_rvi.inc.c | 49 +++++++++++++++++++++++++
 target/riscv/translate.c                | 12 +-----
 3 files changed, 69 insertions(+), 11 deletions(-)

diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index 44d4e922b6fa..81f56c16b45f 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -17,14 +17,33 @@
 # this program.  If not, see <http://www.gnu.org/licenses/>.
 
 # Fields:
+%rs2       20:5
+%rs1       15:5
 %rd        7:5
 
 # immediates:
+%imm_i    20:s12
+%imm_b    31:s1 7:1 25:6 8:4     !function=ex_shift_1
+%imm_j    31:s1 12:8 20:1 21:10  !function=ex_shift_1
 %imm_u    12:s20                 !function=ex_shift_12
 
+# Argument sets:
+&b    imm rs2 rs1
+
 # Formats 32:
+@i       ............    ..... ... ..... .......         imm=%imm_i     %rs1 %rd
+@b       .......   ..... ..... ... ..... ....... &b      imm=%imm_b %rs2 %rs1
 @u       ....................      ..... .......         imm=%imm_u          %rd
+@j       ....................      ..... .......         imm=%imm_j          %rd
 
 # *** RV32I Base Instruction Set ***
 lui      ....................       ..... 0110111 @u
 auipc    ....................       ..... 0010111 @u
+jal      ....................       ..... 1101111 @j
+jalr     ............     ..... 000 ..... 1100111 @i
+beq      ....... .....    ..... 000 ..... 1100011 @b
+bne      ....... .....    ..... 001 ..... 1100011 @b
+blt      ....... .....    ..... 100 ..... 1100011 @b
+bge      ....... .....    ..... 101 ..... 1100011 @b
+bltu     ....... .....    ..... 110 ..... 1100011 @b
+bgeu     ....... .....    ..... 111 ..... 1100011 @b
diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index 9885a8d27551..bcf20def50eb 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -33,3 +33,52 @@ static bool trans_auipc(DisasContext *ctx, arg_auipc *a)
     }
     return true;
 }
+
+static bool trans_jal(DisasContext *ctx, arg_jal *a)
+{
+    gen_jal(ctx, a->rd, a->imm);
+    return true;
+}
+
+static bool trans_jalr(DisasContext *ctx, arg_jalr *a)
+{
+    gen_jalr(ctx, OPC_RISC_JALR, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_beq(DisasContext *ctx, arg_beq *a)
+{
+    gen_branch(ctx, OPC_RISC_BEQ, a->rs1, a->rs2, a->imm);
+    return true;
+}
+
+static bool trans_bne(DisasContext *ctx, arg_bne *a)
+{
+    gen_branch(ctx, OPC_RISC_BNE, a->rs1, a->rs2, a->imm);
+    return true;
+}
+
+static bool trans_blt(DisasContext *ctx, arg_blt *a)
+{
+    gen_branch(ctx, OPC_RISC_BLT, a->rs1, a->rs2, a->imm);
+    return true;
+}
+
+static bool trans_bge(DisasContext *ctx, arg_bge *a)
+{
+    gen_branch(ctx, OPC_RISC_BGE, a->rs1, a->rs2, a->imm);
+    return true;
+}
+
+static bool trans_bltu(DisasContext *ctx, arg_bltu *a)
+{
+    gen_branch(ctx, OPC_RISC_BLTU, a->rs1, a->rs2, a->imm);
+    return true;
+}
+
+static bool trans_bgeu(DisasContext *ctx, arg_bgeu *a)
+{
+
+    gen_branch(ctx, OPC_RISC_BGEU, a->rs1, a->rs2, a->imm);
+    return true;
+}
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index a273ac827448..fb284a5e08d6 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -1879,6 +1879,7 @@ static void decode_RV32_64C(DisasContext *ctx)
     {                                         \
         return imm << amount;                 \
     }
+EX_SH(1)
 EX_SH(12)
 
 bool decode_insn32(DisasContext *ctx, uint32_t insn);
@@ -1907,17 +1908,6 @@ static void decode_RV32_64G(DisasContext *ctx)
     imm = GET_IMM(ctx->opcode);
 
     switch (op) {
-    case OPC_RISC_JAL:
-        imm = GET_JAL_IMM(ctx->opcode);
-        gen_jal(ctx, rd, imm);
-        break;
-    case OPC_RISC_JALR:
-        gen_jalr(ctx, MASK_OP_JALR(ctx->opcode), rd, rs1, imm);
-        break;
-    case OPC_RISC_BRANCH:
-        gen_branch(ctx, MASK_OP_BRANCH(ctx->opcode), rs1, rs2,
-                   GET_B_IMM(ctx->opcode));
-        break;
     case OPC_RISC_LOAD:
         gen_load(ctx, MASK_OP_LOAD(ctx->opcode), rd, rs1, imm);
         break;
-- 
2.19.2

[Qemu-riscv] [PULL 03/29] target/riscv: Convert RV32I load/store insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32.decode              | 10 ++++++
 target/riscv/insn_trans/trans_rvi.inc.c | 48 +++++++++++++++++++++++++
 2 files changed, 58 insertions(+)

diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index 81f56c16b45f..076de873c4f1 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -23,6 +23,7 @@
 
 # immediates:
 %imm_i    20:s12
+%imm_s    25:s7 7:5
 %imm_b    31:s1 7:1 25:6 8:4     !function=ex_shift_1
 %imm_j    31:s1 12:8 20:1 21:10  !function=ex_shift_1
 %imm_u    12:s20                 !function=ex_shift_12
@@ -33,6 +34,7 @@
 # Formats 32:
 @i       ............    ..... ... ..... .......         imm=%imm_i     %rs1 %rd
 @b       .......   ..... ..... ... ..... ....... &b      imm=%imm_b %rs2 %rs1
+@s       .......   ..... ..... ... ..... .......         imm=%imm_s %rs2 %rs1
 @u       ....................      ..... .......         imm=%imm_u          %rd
 @j       ....................      ..... .......         imm=%imm_j          %rd
 
@@ -47,3 +49,11 @@ blt      ....... .....    ..... 100 ..... 1100011 @b
 bge      ....... .....    ..... 101 ..... 1100011 @b
 bltu     ....... .....    ..... 110 ..... 1100011 @b
 bgeu     ....... .....    ..... 111 ..... 1100011 @b
+lb       ............     ..... 000 ..... 0000011 @i
+lh       ............     ..... 001 ..... 0000011 @i
+lw       ............     ..... 010 ..... 0000011 @i
+lbu      ............     ..... 100 ..... 0000011 @i
+lhu      ............     ..... 101 ..... 0000011 @i
+sb       .......  .....   ..... 000 ..... 0100011 @s
+sh       .......  .....   ..... 001 ..... 0100011 @s
+sw       .......  .....   ..... 010 ..... 0100011 @s
diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index bcf20def50eb..d13b7b2b6d8f 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -82,3 +82,51 @@ static bool trans_bgeu(DisasContext *ctx, arg_bgeu *a)
     gen_branch(ctx, OPC_RISC_BGEU, a->rs1, a->rs2, a->imm);
     return true;
 }
+
+static bool trans_lb(DisasContext *ctx, arg_lb *a)
+{
+    gen_load(ctx, OPC_RISC_LB, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_lh(DisasContext *ctx, arg_lh *a)
+{
+    gen_load(ctx, OPC_RISC_LH, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_lw(DisasContext *ctx, arg_lw *a)
+{
+    gen_load(ctx, OPC_RISC_LW, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_lbu(DisasContext *ctx, arg_lbu *a)
+{
+    gen_load(ctx, OPC_RISC_LBU, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_lhu(DisasContext *ctx, arg_lhu *a)
+{
+    gen_load(ctx, OPC_RISC_LHU, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_sb(DisasContext *ctx, arg_sb *a)
+{
+    gen_store(ctx, OPC_RISC_SB, a->rs1, a->rs2, a->imm);
+    return true;
+}
+
+static bool trans_sh(DisasContext *ctx, arg_sh *a)
+{
+    gen_store(ctx, OPC_RISC_SH, a->rs1, a->rs2, a->imm);
+    return true;
+}
+
+static bool trans_sw(DisasContext *ctx, arg_sw *a)
+{
+    gen_store(ctx, OPC_RISC_SW, a->rs1, a->rs2, a->imm);
+    return true;
+}
-- 
2.19.2

[Qemu-riscv] [PULL 04/29] target/riscv: Convert RV64I load/store insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

this splits the 64-bit only instructions into its own decode file such
that we generate the decoder for these instructions only for the RISC-V
64 bit target.

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/Makefile.objs              |  8 +++++---
 target/riscv/insn32-64.decode           | 25 +++++++++++++++++++++++++
 target/riscv/insn_trans/trans_rvi.inc.c | 20 ++++++++++++++++++++
 target/riscv/translate.c                |  7 -------
 4 files changed, 50 insertions(+), 10 deletions(-)
 create mode 100644 target/riscv/insn32-64.decode

diff --git a/target/riscv/Makefile.objs b/target/riscv/Makefile.objs
index bf0a268033a0..05087a91bb85 100644
--- a/target/riscv/Makefile.objs
+++ b/target/riscv/Makefile.objs
@@ -2,10 +2,12 @@ obj-y += translate.o op_helper.o cpu_helper.o cpu.o csr.o fpu_helper.o gdbstub.o
 
 DECODETREE = $(SRC_PATH)/scripts/decodetree.py
 
-target/riscv/decode_insn32.inc.c: \
-  $(SRC_PATH)/target/riscv/insn32.decode $(DECODETREE)
+decode32-y = $(SRC_PATH)/target/riscv/insn32.decode
+decode32-$(TARGET_RISCV64) += $(SRC_PATH)/target/riscv/insn32-64.decode
+
+target/riscv/decode_insn32.inc.c: $(decode32-y) $(DECODETREE)
 	$(call quiet-command, \
-	  $(PYTHON) $(DECODETREE) -o $@ --decode decode_insn32 $<, \
+	  $(PYTHON) $(DECODETREE) -o $@ --decode decode_insn32 $(decode32-y), \
 	  "GEN", $(TARGET_DIR)$@)
 
 target/riscv/translate.o: target/riscv/decode_insn32.inc.c
diff --git a/target/riscv/insn32-64.decode b/target/riscv/insn32-64.decode
new file mode 100644
index 000000000000..439d4e2c587b
--- /dev/null
+++ b/target/riscv/insn32-64.decode
@@ -0,0 +1,25 @@
+#
+# RISC-V translation routines for the RV Instruction Set.
+#
+# Copyright (c) 2018 Peer Adelt, peer.adelt@hni.uni-paderborn.de
+#                    Bastian Koppelmann, kbastian@mail.uni-paderborn.de
+#
+# 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/>.
+
+# This is concatenated with insn32.decode for risc64 targets.
+# Most of the fields and formats are there.
+
+# *** RV64I Base Instruction Set (in addition to RV32I) ***
+lwu      ............   ..... 110 ..... 0000011 @i
+ld       ............   ..... 011 ..... 0000011 @i
+sd       ....... .....  ..... 011 ..... 0100011 @s
diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index d13b7b2b6d8f..61f708dba144 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -130,3 +130,23 @@ static bool trans_sw(DisasContext *ctx, arg_sw *a)
     gen_store(ctx, OPC_RISC_SW, a->rs1, a->rs2, a->imm);
     return true;
 }
+
+#ifdef TARGET_RISCV64
+static bool trans_lwu(DisasContext *ctx, arg_lwu *a)
+{
+    gen_load(ctx, OPC_RISC_LWU, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_ld(DisasContext *ctx, arg_ld *a)
+{
+    gen_load(ctx, OPC_RISC_LD, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_sd(DisasContext *ctx, arg_sd *a)
+{
+    gen_store(ctx, OPC_RISC_SD, a->rs1, a->rs2, a->imm);
+    return true;
+}
+#endif
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index fb284a5e08d6..2e35142ca2a4 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -1908,13 +1908,6 @@ static void decode_RV32_64G(DisasContext *ctx)
     imm = GET_IMM(ctx->opcode);
 
     switch (op) {
-    case OPC_RISC_LOAD:
-        gen_load(ctx, MASK_OP_LOAD(ctx->opcode), rd, rs1, imm);
-        break;
-    case OPC_RISC_STORE:
-        gen_store(ctx, MASK_OP_STORE(ctx->opcode), rs1, rs2,
-                  GET_STORE_IMM(ctx->opcode));
-        break;
     case OPC_RISC_ARITH_IMM:
 #if defined(TARGET_RISCV64)
     case OPC_RISC_ARITH_IMM_W:
-- 
2.19.2

[Qemu-riscv] [PULL 05/29] target/riscv: Convert RVXI arithmetic insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

we cannot remove the call to gen_arith() in decode_RV32_64G() since it
is used to translate multiply instructions.

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32-64.decode           |  13 ++
 target/riscv/insn32.decode              |  25 ++++
 target/riscv/insn_trans/trans_rvi.inc.c | 168 ++++++++++++++++++++++++
 target/riscv/translate.c                |   9 --
 4 files changed, 206 insertions(+), 9 deletions(-)

diff --git a/target/riscv/insn32-64.decode b/target/riscv/insn32-64.decode
index 439d4e2c587b..9a35f2aa1920 100644
--- a/target/riscv/insn32-64.decode
+++ b/target/riscv/insn32-64.decode
@@ -19,7 +19,20 @@
 # This is concatenated with insn32.decode for risc64 targets.
 # Most of the fields and formats are there.
 
+%sh5    20:5
+
+@sh5     .......  ..... .....  ... ..... ....... &shift  shamt=%sh5      %rs1 %rd
+
 # *** RV64I Base Instruction Set (in addition to RV32I) ***
 lwu      ............   ..... 110 ..... 0000011 @i
 ld       ............   ..... 011 ..... 0000011 @i
 sd       ....... .....  ..... 011 ..... 0100011 @s
+addiw    ............   ..... 000 ..... 0011011 @i
+slliw    0000000 .....  ..... 001 ..... 0011011 @sh5
+srliw    0000000 .....  ..... 101 ..... 0011011 @sh5
+sraiw    0100000 .....  ..... 101 ..... 0011011 @sh5
+addw     0000000 .....  ..... 000 ..... 0111011 @r
+subw     0100000 .....  ..... 000 ..... 0111011 @r
+sllw     0000000 .....  ..... 001 ..... 0111011 @r
+srlw     0000000 .....  ..... 101 ..... 0111011 @r
+sraw     0100000 .....  ..... 101 ..... 0111011 @r
diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index 076de873c4f1..1f5bf1f6f97d 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -21,6 +21,8 @@
 %rs1       15:5
 %rd        7:5
 
+%sh10    20:10
+
 # immediates:
 %imm_i    20:s12
 %imm_s    25:s7 7:5
@@ -30,14 +32,18 @@
 
 # Argument sets:
 &b    imm rs2 rs1
+&shift     shamt rs1 rd
 
 # Formats 32:
+@r       .......   ..... ..... ... ..... .......                   %rs2 %rs1 %rd
 @i       ............    ..... ... ..... .......         imm=%imm_i     %rs1 %rd
 @b       .......   ..... ..... ... ..... ....... &b      imm=%imm_b %rs2 %rs1
 @s       .......   ..... ..... ... ..... .......         imm=%imm_s %rs2 %rs1
 @u       ....................      ..... .......         imm=%imm_u          %rd
 @j       ....................      ..... .......         imm=%imm_j          %rd
 
+@sh      ......  ...... .....  ... ..... ....... &shift  shamt=%sh10      %rs1 %rd
+
 # *** RV32I Base Instruction Set ***
 lui      ....................       ..... 0110111 @u
 auipc    ....................       ..... 0010111 @u
@@ -57,3 +63,22 @@ lhu      ............     ..... 101 ..... 0000011 @i
 sb       .......  .....   ..... 000 ..... 0100011 @s
 sh       .......  .....   ..... 001 ..... 0100011 @s
 sw       .......  .....   ..... 010 ..... 0100011 @s
+addi     ............     ..... 000 ..... 0010011 @i
+slti     ............     ..... 010 ..... 0010011 @i
+sltiu    ............     ..... 011 ..... 0010011 @i
+xori     ............     ..... 100 ..... 0010011 @i
+ori      ............     ..... 110 ..... 0010011 @i
+andi     ............     ..... 111 ..... 0010011 @i
+slli     00.... ......    ..... 001 ..... 0010011 @sh
+srli     00.... ......    ..... 101 ..... 0010011 @sh
+srai     01.... ......    ..... 101 ..... 0010011 @sh
+add      0000000 .....    ..... 000 ..... 0110011 @r
+sub      0100000 .....    ..... 000 ..... 0110011 @r
+sll      0000000 .....    ..... 001 ..... 0110011 @r
+slt      0000000 .....    ..... 010 ..... 0110011 @r
+sltu     0000000 .....    ..... 011 ..... 0110011 @r
+xor      0000000 .....    ..... 100 ..... 0110011 @r
+srl      0000000 .....    ..... 101 ..... 0110011 @r
+sra      0100000 .....    ..... 101 ..... 0110011 @r
+or       0000000 .....    ..... 110 ..... 0110011 @r
+and      0000000 .....    ..... 111 ..... 0110011 @r
diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index 61f708dba144..136fa54d0655 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -150,3 +150,171 @@ static bool trans_sd(DisasContext *ctx, arg_sd *a)
     return true;
 }
 #endif
+
+static bool trans_addi(DisasContext *ctx, arg_addi *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_ADDI, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_slti(DisasContext *ctx, arg_slti *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_SLTI, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_sltiu(DisasContext *ctx, arg_sltiu *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_SLTIU, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_xori(DisasContext *ctx, arg_xori *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_XORI, a->rd, a->rs1, a->imm);
+    return true;
+}
+static bool trans_ori(DisasContext *ctx, arg_ori *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_ORI, a->rd, a->rs1, a->imm);
+    return true;
+}
+static bool trans_andi(DisasContext *ctx, arg_andi *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_ANDI, a->rd, a->rs1, a->imm);
+    return true;
+}
+static bool trans_slli(DisasContext *ctx, arg_slli *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_SLLI, a->rd, a->rs1, a->shamt);
+    return true;
+}
+
+static bool trans_srli(DisasContext *ctx, arg_srli *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, a->rd, a->rs1, a->shamt);
+    return true;
+}
+
+static bool trans_srai(DisasContext *ctx, arg_srai *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, a->rd, a->rs1, a->shamt | 0x400);
+    return true;
+}
+
+static bool trans_add(DisasContext *ctx, arg_add *a)
+{
+    gen_arith(ctx, OPC_RISC_ADD, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_sub(DisasContext *ctx, arg_sub *a)
+{
+    gen_arith(ctx, OPC_RISC_SUB, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_sll(DisasContext *ctx, arg_sll *a)
+{
+    gen_arith(ctx, OPC_RISC_SLL, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_slt(DisasContext *ctx, arg_slt *a)
+{
+    gen_arith(ctx, OPC_RISC_SLT, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_sltu(DisasContext *ctx, arg_sltu *a)
+{
+    gen_arith(ctx, OPC_RISC_SLTU, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_xor(DisasContext *ctx, arg_xor *a)
+{
+    gen_arith(ctx, OPC_RISC_XOR, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_srl(DisasContext *ctx, arg_srl *a)
+{
+    gen_arith(ctx, OPC_RISC_SRL, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_sra(DisasContext *ctx, arg_sra *a)
+{
+    gen_arith(ctx, OPC_RISC_SRA, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_or(DisasContext *ctx, arg_or *a)
+{
+    gen_arith(ctx, OPC_RISC_OR, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_and(DisasContext *ctx, arg_and *a)
+{
+    gen_arith(ctx, OPC_RISC_AND, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+#ifdef TARGET_RISCV64
+static bool trans_addiw(DisasContext *ctx, arg_addiw *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_ADDIW, a->rd, a->rs1, a->imm);
+    return true;
+}
+
+static bool trans_slliw(DisasContext *ctx, arg_slliw *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_SLLIW, a->rd, a->rs1, a->shamt);
+    return true;
+}
+
+static bool trans_srliw(DisasContext *ctx, arg_srliw *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_IW, a->rd, a->rs1, a->shamt);
+    return true;
+}
+
+static bool trans_sraiw(DisasContext *ctx, arg_sraiw *a)
+{
+    gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_IW , a->rd, a->rs1,
+                  a->shamt | 0x400);
+    return true;
+}
+
+static bool trans_addw(DisasContext *ctx, arg_addw *a)
+{
+    gen_arith(ctx, OPC_RISC_ADDW, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_subw(DisasContext *ctx, arg_subw *a)
+{
+    gen_arith(ctx, OPC_RISC_SUBW, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_sllw(DisasContext *ctx, arg_sllw *a)
+{
+    gen_arith(ctx, OPC_RISC_SLLW, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_srlw(DisasContext *ctx, arg_srlw *a)
+{
+    gen_arith(ctx, OPC_RISC_SRLW, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_sraw(DisasContext *ctx, arg_sraw *a)
+{
+    gen_arith(ctx, OPC_RISC_SRAW, a->rd, a->rs1, a->rs2);
+    return true;
+}
+#endif
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 2e35142ca2a4..1ae84dcd5992 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -1908,15 +1908,6 @@ static void decode_RV32_64G(DisasContext *ctx)
     imm = GET_IMM(ctx->opcode);
 
     switch (op) {
-    case OPC_RISC_ARITH_IMM:
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_ARITH_IMM_W:
-#endif
-        if (rd == 0) {
-            break; /* NOP */
-        }
-        gen_arith_imm(ctx, MASK_OP_ARITH_IMM(ctx->opcode), rd, rs1, imm);
-        break;
     case OPC_RISC_ARITH:
 #if defined(TARGET_RISCV64)
     case OPC_RISC_ARITH_W:
-- 
2.19.2

[Qemu-riscv] [PULL 06/29] target/riscv: Convert RVXI fence insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32.decode              |  2 ++
 target/riscv/insn_trans/trans_rvi.inc.c | 19 +++++++++++++++++++
 target/riscv/translate.c                | 12 ------------
 3 files changed, 21 insertions(+), 12 deletions(-)

diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index 1f5bf1f6f97d..804b721ca51e 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -82,3 +82,5 @@ srl      0000000 .....    ..... 101 ..... 0110011 @r
 sra      0100000 .....    ..... 101 ..... 0110011 @r
 or       0000000 .....    ..... 110 ..... 0110011 @r
 and      0000000 .....    ..... 111 ..... 0110011 @r
+fence    ---- pred:4 succ:4 ----- 000 ----- 0001111
+fence_i  ---- ----   ----   ----- 001 ----- 0001111
diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index 136fa54d0655..973d6371df85 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -318,3 +318,22 @@ static bool trans_sraw(DisasContext *ctx, arg_sraw *a)
     return true;
 }
 #endif
+
+static bool trans_fence(DisasContext *ctx, arg_fence *a)
+{
+    /* FENCE is a full memory barrier. */
+    tcg_gen_mb(TCG_MO_ALL | TCG_BAR_SC);
+    return true;
+}
+
+static bool trans_fence_i(DisasContext *ctx, arg_fence_i *a)
+{
+    /*
+     * FENCE_I is a no-op in QEMU,
+     * however we need to end the translation block
+     */
+    tcg_gen_movi_tl(cpu_pc, ctx->pc_succ_insn);
+    tcg_gen_exit_tb(NULL, 0);
+    ctx->base.is_jmp = DISAS_NORETURN;
+    return true;
+}
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 1ae84dcd5992..f720746cb791 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -1950,18 +1950,6 @@ static void decode_RV32_64G(DisasContext *ctx)
         gen_fp_arith(ctx, MASK_OP_FP_ARITH(ctx->opcode), rd, rs1, rs2,
                      GET_RM(ctx->opcode));
         break;
-    case OPC_RISC_FENCE:
-        if (ctx->opcode & 0x1000) {
-            /* FENCE_I is a no-op in QEMU,
-             * however we need to end the translation block */
-            tcg_gen_movi_tl(cpu_pc, ctx->pc_succ_insn);
-            tcg_gen_exit_tb(NULL, 0);
-            ctx->base.is_jmp = DISAS_NORETURN;
-        } else {
-            /* FENCE is a full memory barrier. */
-            tcg_gen_mb(TCG_MO_ALL | TCG_BAR_SC);
-        }
-        break;
     case OPC_RISC_SYSTEM:
         gen_system(ctx, MASK_OP_SYSTEM(ctx->opcode), rd, rs1,
                    (ctx->opcode & 0xFFF00000) >> 20);
-- 
2.19.2

[Qemu-riscv] [PULL 07/29] target/riscv: Convert RVXI csr insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32.decode              |  8 +++
 target/riscv/insn_trans/trans_rvi.inc.c | 79 +++++++++++++++++++++++++
 target/riscv/translate.c                | 43 +-------------
 3 files changed, 88 insertions(+), 42 deletions(-)

diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index 804b721ca51e..977b1b10a330 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -22,6 +22,7 @@
 %rd        7:5
 
 %sh10    20:10
+%csr    20:12
 
 # immediates:
 %imm_i    20:s12
@@ -43,6 +44,7 @@
 @j       ....................      ..... .......         imm=%imm_j          %rd
 
 @sh      ......  ...... .....  ... ..... ....... &shift  shamt=%sh10      %rs1 %rd
+@csr     ............   .....  ... ..... .......               %csr     %rs1 %rd
 
 # *** RV32I Base Instruction Set ***
 lui      ....................       ..... 0110111 @u
@@ -84,3 +86,9 @@ or       0000000 .....    ..... 110 ..... 0110011 @r
 and      0000000 .....    ..... 111 ..... 0110011 @r
 fence    ---- pred:4 succ:4 ----- 000 ----- 0001111
 fence_i  ---- ----   ----   ----- 001 ----- 0001111
+csrrw    ............     ..... 001 ..... 1110011 @csr
+csrrs    ............     ..... 010 ..... 1110011 @csr
+csrrc    ............     ..... 011 ..... 1110011 @csr
+csrrwi   ............     ..... 101 ..... 1110011 @csr
+csrrsi   ............     ..... 110 ..... 1110011 @csr
+csrrci   ............     ..... 111 ..... 1110011 @csr
diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index 973d6371df85..4a23372cb823 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -337,3 +337,82 @@ static bool trans_fence_i(DisasContext *ctx, arg_fence_i *a)
     ctx->base.is_jmp = DISAS_NORETURN;
     return true;
 }
+
+#define RISCV_OP_CSR_PRE do {\
+    source1 = tcg_temp_new(); \
+    csr_store = tcg_temp_new(); \
+    dest = tcg_temp_new(); \
+    rs1_pass = tcg_temp_new(); \
+    gen_get_gpr(source1, a->rs1); \
+    tcg_gen_movi_tl(cpu_pc, ctx->base.pc_next); \
+    tcg_gen_movi_tl(rs1_pass, a->rs1); \
+    tcg_gen_movi_tl(csr_store, a->csr); \
+    gen_io_start();\
+} while (0)
+
+#define RISCV_OP_CSR_POST do {\
+    gen_io_end(); \
+    gen_set_gpr(a->rd, dest); \
+    tcg_gen_movi_tl(cpu_pc, ctx->pc_succ_insn); \
+    tcg_gen_exit_tb(NULL, 0); \
+    ctx->base.is_jmp = DISAS_NORETURN; \
+    tcg_temp_free(source1); \
+    tcg_temp_free(csr_store); \
+    tcg_temp_free(dest); \
+    tcg_temp_free(rs1_pass); \
+} while (0)
+
+
+static bool trans_csrrw(DisasContext *ctx, arg_csrrw *a)
+{
+    TCGv source1, csr_store, dest, rs1_pass;
+    RISCV_OP_CSR_PRE;
+    gen_helper_csrrw(dest, cpu_env, source1, csr_store);
+    RISCV_OP_CSR_POST;
+    return true;
+}
+
+static bool trans_csrrs(DisasContext *ctx, arg_csrrs *a)
+{
+    TCGv source1, csr_store, dest, rs1_pass;
+    RISCV_OP_CSR_PRE;
+    gen_helper_csrrs(dest, cpu_env, source1, csr_store, rs1_pass);
+    RISCV_OP_CSR_POST;
+    return true;
+}
+
+static bool trans_csrrc(DisasContext *ctx, arg_csrrc *a)
+{
+    TCGv source1, csr_store, dest, rs1_pass;
+    RISCV_OP_CSR_PRE;
+    gen_helper_csrrc(dest, cpu_env, source1, csr_store, rs1_pass);
+    RISCV_OP_CSR_POST;
+    return true;
+}
+
+static bool trans_csrrwi(DisasContext *ctx, arg_csrrwi *a)
+{
+    TCGv source1, csr_store, dest, rs1_pass;
+    RISCV_OP_CSR_PRE;
+    gen_helper_csrrw(dest, cpu_env, rs1_pass, csr_store);
+    RISCV_OP_CSR_POST;
+    return true;
+}
+
+static bool trans_csrrsi(DisasContext *ctx, arg_csrrsi *a)
+{
+    TCGv source1, csr_store, dest, rs1_pass;
+    RISCV_OP_CSR_PRE;
+    gen_helper_csrrs(dest, cpu_env, rs1_pass, csr_store, rs1_pass);
+    RISCV_OP_CSR_POST;
+    return true;
+}
+
+static bool trans_csrrci(DisasContext *ctx, arg_csrrci *a)
+{
+    TCGv source1, csr_store, dest, rs1_pass;
+    RISCV_OP_CSR_PRE;
+    gen_helper_csrrc(dest, cpu_env, rs1_pass, csr_store, rs1_pass);
+    RISCV_OP_CSR_POST;
+    return true;
+}
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index f720746cb791..18555000af35 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -1476,16 +1476,11 @@ static void gen_fp_arith(DisasContext *ctx, uint32_t opc, int rd,
 static void gen_system(DisasContext *ctx, uint32_t opc, int rd, int rs1,
                        int csr)
 {
-    TCGv source1, csr_store, dest, rs1_pass, imm_rs1;
+    TCGv source1, dest;
     source1 = tcg_temp_new();
-    csr_store = tcg_temp_new();
     dest = tcg_temp_new();
-    rs1_pass = tcg_temp_new();
-    imm_rs1 = tcg_temp_new();
     gen_get_gpr(source1, rs1);
     tcg_gen_movi_tl(cpu_pc, ctx->base.pc_next);
-    tcg_gen_movi_tl(rs1_pass, rs1);
-    tcg_gen_movi_tl(csr_store, csr); /* copy into temp reg to feed to helper */
 
 #ifndef CONFIG_USER_ONLY
     /* Extract funct7 value and check whether it matches SFENCE.VMA */
@@ -1556,45 +1551,9 @@ static void gen_system(DisasContext *ctx, uint32_t opc, int rd, int rs1,
             break;
         }
         break;
-    default:
-        tcg_gen_movi_tl(imm_rs1, rs1);
-        gen_io_start();
-        switch (opc) {
-        case OPC_RISC_CSRRW:
-            gen_helper_csrrw(dest, cpu_env, source1, csr_store);
-            break;
-        case OPC_RISC_CSRRS:
-            gen_helper_csrrs(dest, cpu_env, source1, csr_store, rs1_pass);
-            break;
-        case OPC_RISC_CSRRC:
-            gen_helper_csrrc(dest, cpu_env, source1, csr_store, rs1_pass);
-            break;
-        case OPC_RISC_CSRRWI:
-            gen_helper_csrrw(dest, cpu_env, imm_rs1, csr_store);
-            break;
-        case OPC_RISC_CSRRSI:
-            gen_helper_csrrs(dest, cpu_env, imm_rs1, csr_store, rs1_pass);
-            break;
-        case OPC_RISC_CSRRCI:
-            gen_helper_csrrc(dest, cpu_env, imm_rs1, csr_store, rs1_pass);
-            break;
-        default:
-            gen_exception_illegal(ctx);
-            return;
-        }
-        gen_io_end();
-        gen_set_gpr(rd, dest);
-        /* end tb since we may be changing priv modes, to get mmu_index right */
-        tcg_gen_movi_tl(cpu_pc, ctx->pc_succ_insn);
-        tcg_gen_exit_tb(NULL, 0); /* no chaining */
-        ctx->base.is_jmp = DISAS_NORETURN;
-        break;
     }
     tcg_temp_free(source1);
-    tcg_temp_free(csr_store);
     tcg_temp_free(dest);
-    tcg_temp_free(rs1_pass);
-    tcg_temp_free(imm_rs1);
 }
 
 static void decode_RV32_64C0(DisasContext *ctx)
-- 
2.19.2

[Qemu-riscv] [PULL 08/29] target/riscv: Convert RVXM insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32-64.decode           |   7 ++
 target/riscv/insn32.decode              |  10 +++
 target/riscv/insn_trans/trans_rvm.inc.c | 113 ++++++++++++++++++++++++
 target/riscv/translate.c                |  16 ++--
 4 files changed, 137 insertions(+), 9 deletions(-)
 create mode 100644 target/riscv/insn_trans/trans_rvm.inc.c

diff --git a/target/riscv/insn32-64.decode b/target/riscv/insn32-64.decode
index 9a35f2aa1920..008f1005469e 100644
--- a/target/riscv/insn32-64.decode
+++ b/target/riscv/insn32-64.decode
@@ -36,3 +36,10 @@ subw     0100000 .....  ..... 000 ..... 0111011 @r
 sllw     0000000 .....  ..... 001 ..... 0111011 @r
 srlw     0000000 .....  ..... 101 ..... 0111011 @r
 sraw     0100000 .....  ..... 101 ..... 0111011 @r
+
+# *** RV64M Standard Extension (in addition to RV32M) ***
+mulw     0000001 .....  ..... 000 ..... 0111011 @r
+divw     0000001 .....  ..... 100 ..... 0111011 @r
+divuw    0000001 .....  ..... 101 ..... 0111011 @r
+remw     0000001 .....  ..... 110 ..... 0111011 @r
+remuw    0000001 .....  ..... 111 ..... 0111011 @r
diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index 977b1b10a330..e53944bf0e40 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -92,3 +92,13 @@ csrrc    ............     ..... 011 ..... 1110011 @csr
 csrrwi   ............     ..... 101 ..... 1110011 @csr
 csrrsi   ............     ..... 110 ..... 1110011 @csr
 csrrci   ............     ..... 111 ..... 1110011 @csr
+
+# *** RV32M Standard Extension ***
+mul      0000001 .....  ..... 000 ..... 0110011 @r
+mulh     0000001 .....  ..... 001 ..... 0110011 @r
+mulhsu   0000001 .....  ..... 010 ..... 0110011 @r
+mulhu    0000001 .....  ..... 011 ..... 0110011 @r
+div      0000001 .....  ..... 100 ..... 0110011 @r
+divu     0000001 .....  ..... 101 ..... 0110011 @r
+rem      0000001 .....  ..... 110 ..... 0110011 @r
+remu     0000001 .....  ..... 111 ..... 0110011 @r
diff --git a/target/riscv/insn_trans/trans_rvm.inc.c b/target/riscv/insn_trans/trans_rvm.inc.c
new file mode 100644
index 000000000000..69631c9e3783
--- /dev/null
+++ b/target/riscv/insn_trans/trans_rvm.inc.c
@@ -0,0 +1,113 @@
+/*
+ * RISC-V translation routines for the RV64M Standard Extension.
+ *
+ * Copyright (c) 2016-2017 Sagar Karandikar, sagark@eecs.berkeley.edu
+ * Copyright (c) 2018 Peer Adelt, peer.adelt@hni.uni-paderborn.de
+ *                    Bastian Koppelmann, kbastian@mail.uni-paderborn.de
+ *
+ * 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/>.
+ */
+
+
+static bool trans_mul(DisasContext *ctx, arg_mul *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_MUL, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_mulh(DisasContext *ctx, arg_mulh *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_MULH, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_mulhsu(DisasContext *ctx, arg_mulhsu *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_MULHSU, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_mulhu(DisasContext *ctx, arg_mulhu *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_MULHU, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_div(DisasContext *ctx, arg_div *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_DIV, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_divu(DisasContext *ctx, arg_divu *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_DIVU, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_rem(DisasContext *ctx, arg_rem *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_REM, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_remu(DisasContext *ctx, arg_remu *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_REMU, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+#ifdef TARGET_RISCV64
+static bool trans_mulw(DisasContext *ctx, arg_mulw *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_MULW, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_divw(DisasContext *ctx, arg_divw *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_DIVW, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_divuw(DisasContext *ctx, arg_divuw *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_DIVUW, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_remw(DisasContext *ctx, arg_remw *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_REMW, a->rd, a->rs1, a->rs2);
+    return true;
+}
+
+static bool trans_remuw(DisasContext *ctx, arg_remuw *a)
+{
+    REQUIRE_EXT(ctx, RVM);
+    gen_arith(ctx, OPC_RISC_REMUW, a->rd, a->rs1, a->rs2);
+    return true;
+}
+#endif
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 18555000af35..783ccade5180 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -1841,11 +1841,18 @@ static void decode_RV32_64C(DisasContext *ctx)
 EX_SH(1)
 EX_SH(12)
 
+#define REQUIRE_EXT(ctx, ext) do { \
+    if (!has_ext(ctx, ext)) {      \
+        return false;              \
+    }                              \
+} while (0)
+
 bool decode_insn32(DisasContext *ctx, uint32_t insn);
 /* Include the auto-generated decoder for 32 bit insn */
 #include "decode_insn32.inc.c"
 /* Include insn module translation function */
 #include "insn_trans/trans_rvi.inc.c"
+#include "insn_trans/trans_rvm.inc.c"
 
 static void decode_RV32_64G(DisasContext *ctx)
 {
@@ -1867,15 +1874,6 @@ static void decode_RV32_64G(DisasContext *ctx)
     imm = GET_IMM(ctx->opcode);
 
     switch (op) {
-    case OPC_RISC_ARITH:
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_ARITH_W:
-#endif
-        if (rd == 0) {
-            break; /* NOP */
-        }
-        gen_arith(ctx, MASK_OP_ARITH(ctx->opcode), rd, rs1, rs2);
-        break;
     case OPC_RISC_FP_LOAD:
         gen_fp_load(ctx, MASK_OP_FP_LOAD(ctx->opcode), rd, rs1, imm);
         break;
-- 
2.19.2

[Qemu-riscv] [PULL 09/29] target/riscv: Convert RV32A insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32.decode              |  17 +++
 target/riscv/insn_trans/trans_rva.inc.c | 160 ++++++++++++++++++++++++
 target/riscv/translate.c                |   1 +
 3 files changed, 178 insertions(+)
 create mode 100644 target/riscv/insn_trans/trans_rva.inc.c

diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index e53944bf0e40..00b9e2d9a508 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -34,6 +34,7 @@
 # Argument sets:
 &b    imm rs2 rs1
 &shift     shamt rs1 rd
+&atomic    aq rl rs2 rs1 rd
 
 # Formats 32:
 @r       .......   ..... ..... ... ..... .......                   %rs2 %rs1 %rd
@@ -46,6 +47,9 @@
 @sh      ......  ...... .....  ... ..... ....... &shift  shamt=%sh10      %rs1 %rd
 @csr     ............   .....  ... ..... .......               %csr     %rs1 %rd
 
+@atom_ld ..... aq:1 rl:1 ..... ........ ..... ....... &atomic rs2=0     %rs1 %rd
+@atom_st ..... aq:1 rl:1 ..... ........ ..... ....... &atomic %rs2      %rs1 %rd
+
 # *** RV32I Base Instruction Set ***
 lui      ....................       ..... 0110111 @u
 auipc    ....................       ..... 0010111 @u
@@ -102,3 +106,16 @@ div      0000001 .....  ..... 100 ..... 0110011 @r
 divu     0000001 .....  ..... 101 ..... 0110011 @r
 rem      0000001 .....  ..... 110 ..... 0110011 @r
 remu     0000001 .....  ..... 111 ..... 0110011 @r
+
+# *** RV32A Standard Extension ***
+lr_w       00010 . . 00000 ..... 010 ..... 0101111 @atom_ld
+sc_w       00011 . . ..... ..... 010 ..... 0101111 @atom_st
+amoswap_w  00001 . . ..... ..... 010 ..... 0101111 @atom_st
+amoadd_w   00000 . . ..... ..... 010 ..... 0101111 @atom_st
+amoxor_w   00100 . . ..... ..... 010 ..... 0101111 @atom_st
+amoand_w   01100 . . ..... ..... 010 ..... 0101111 @atom_st
+amoor_w    01000 . . ..... ..... 010 ..... 0101111 @atom_st
+amomin_w   10000 . . ..... ..... 010 ..... 0101111 @atom_st
+amomax_w   10100 . . ..... ..... 010 ..... 0101111 @atom_st
+amominu_w  11000 . . ..... ..... 010 ..... 0101111 @atom_st
+amomaxu_w  11100 . . ..... ..... 010 ..... 0101111 @atom_st
diff --git a/target/riscv/insn_trans/trans_rva.inc.c b/target/riscv/insn_trans/trans_rva.inc.c
new file mode 100644
index 000000000000..5d3c853ca504
--- /dev/null
+++ b/target/riscv/insn_trans/trans_rva.inc.c
@@ -0,0 +1,160 @@
+/*
+ * RISC-V translation routines for the RV64A Standard Extension.
+ *
+ * Copyright (c) 2016-2017 Sagar Karandikar, sagark@eecs.berkeley.edu
+ * Copyright (c) 2018 Peer Adelt, peer.adelt@hni.uni-paderborn.de
+ *                    Bastian Koppelmann, kbastian@mail.uni-paderborn.de
+ *
+ * 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/>.
+ */
+
+static inline bool gen_lr(DisasContext *ctx, arg_atomic *a, TCGMemOp mop)
+{
+    TCGv src1 = tcg_temp_new();
+    /* Put addr in load_res, data in load_val.  */
+    gen_get_gpr(src1, a->rs1);
+    if (a->rl) {
+        tcg_gen_mb(TCG_MO_ALL | TCG_BAR_STRL);
+    }
+    tcg_gen_qemu_ld_tl(load_val, src1, ctx->mem_idx, mop);
+    if (a->aq) {
+        tcg_gen_mb(TCG_MO_ALL | TCG_BAR_LDAQ);
+    }
+    tcg_gen_mov_tl(load_res, src1);
+    gen_set_gpr(a->rd, load_val);
+
+    tcg_temp_free(src1);
+    return true;
+}
+
+static inline bool gen_sc(DisasContext *ctx, arg_atomic *a, TCGMemOp mop)
+{
+    TCGv src1 = tcg_temp_new();
+    TCGv src2 = tcg_temp_new();
+    TCGv dat = tcg_temp_new();
+    TCGLabel *l1 = gen_new_label();
+    TCGLabel *l2 = gen_new_label();
+
+    gen_get_gpr(src1, a->rs1);
+    tcg_gen_brcond_tl(TCG_COND_NE, load_res, src1, l1);
+
+    gen_get_gpr(src2, a->rs2);
+    /*
+     * Note that the TCG atomic primitives are SC,
+     * so we can ignore AQ/RL along this path.
+     */
+    tcg_gen_atomic_cmpxchg_tl(src1, load_res, load_val, src2,
+                              ctx->mem_idx, mop);
+    tcg_gen_setcond_tl(TCG_COND_NE, dat, src1, load_val);
+    gen_set_gpr(a->rd, dat);
+    tcg_gen_br(l2);
+
+    gen_set_label(l1);
+    /*
+     * Address comparion failure.  However, we still need to
+     * provide the memory barrier implied by AQ/RL.
+     */
+    tcg_gen_mb(TCG_MO_ALL + a->aq * TCG_BAR_LDAQ + a->rl * TCG_BAR_STRL);
+    tcg_gen_movi_tl(dat, 1);
+    gen_set_gpr(a->rd, dat);
+
+    gen_set_label(l2);
+    tcg_temp_free(dat);
+    tcg_temp_free(src1);
+    tcg_temp_free(src2);
+    return true;
+}
+
+static bool gen_amo(DisasContext *ctx, arg_atomic *a,
+                    void(*func)(TCGv, TCGv, TCGv, TCGArg, TCGMemOp),
+                    TCGMemOp mop)
+{
+    TCGv src1 = tcg_temp_new();
+    TCGv src2 = tcg_temp_new();
+
+    gen_get_gpr(src1, a->rs1);
+    gen_get_gpr(src2, a->rs2);
+
+    (*func)(src2, src1, src2, ctx->mem_idx, mop);
+
+    gen_set_gpr(a->rd, src2);
+    tcg_temp_free(src1);
+    tcg_temp_free(src2);
+    return true;
+}
+
+static bool trans_lr_w(DisasContext *ctx, arg_lr_w *a)
+{
+    REQUIRE_EXT(ctx, RVA);
+    return gen_lr(ctx, a, (MO_ALIGN | MO_TESL));
+}
+
+static bool trans_sc_w(DisasContext *ctx, arg_sc_w *a)
+{
+    REQUIRE_EXT(ctx, RVA);
+    return gen_sc(ctx, a, (MO_ALIGN | MO_TESL));
+}
+
+static bool trans_amoswap_w(DisasContext *ctx, arg_amoswap_w *a)
+{
+    REQUIRE_EXT(ctx, RVA);
+    return gen_amo(ctx, a, &tcg_gen_atomic_xchg_tl, (MO_ALIGN | MO_TESL));
+}
+
+static bool trans_amoadd_w(DisasContext *ctx, arg_amoadd_w *a)
+{
+    REQUIRE_EXT(ctx, RVA);
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_add_tl, (MO_ALIGN | MO_TESL));
+}
+
+static bool trans_amoxor_w(DisasContext *ctx, arg_amoxor_w *a)
+{
+    REQUIRE_EXT(ctx, RVA);
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_xor_tl, (MO_ALIGN | MO_TESL));
+}
+
+static bool trans_amoand_w(DisasContext *ctx, arg_amoand_w *a)
+{
+    REQUIRE_EXT(ctx, RVA);
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_and_tl, (MO_ALIGN | MO_TESL));
+}
+
+static bool trans_amoor_w(DisasContext *ctx, arg_amoor_w *a)
+{
+    REQUIRE_EXT(ctx, RVA);
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_or_tl, (MO_ALIGN | MO_TESL));
+}
+
+static bool trans_amomin_w(DisasContext *ctx, arg_amomin_w *a)
+{
+    REQUIRE_EXT(ctx, RVA);
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smin_tl, (MO_ALIGN | MO_TESL));
+}
+
+static bool trans_amomax_w(DisasContext *ctx, arg_amomax_w *a)
+{
+    REQUIRE_EXT(ctx, RVA);
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smax_tl, (MO_ALIGN | MO_TESL));
+}
+
+static bool trans_amominu_w(DisasContext *ctx, arg_amominu_w *a)
+{
+    REQUIRE_EXT(ctx, RVA);
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umin_tl, (MO_ALIGN | MO_TESL));
+}
+
+static bool trans_amomaxu_w(DisasContext *ctx, arg_amomaxu_w *a)
+{
+    REQUIRE_EXT(ctx, RVA);
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umax_tl, (MO_ALIGN | MO_TESL));
+}
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 783ccade5180..b0de062a4fd1 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -1853,6 +1853,7 @@ bool decode_insn32(DisasContext *ctx, uint32_t insn);
 /* Include insn module translation function */
 #include "insn_trans/trans_rvi.inc.c"
 #include "insn_trans/trans_rvm.inc.c"
+#include "insn_trans/trans_rva.inc.c"
 
 static void decode_RV32_64G(DisasContext *ctx)
 {
-- 
2.19.2

[Qemu-riscv] [PULL 10/29] target/riscv: Convert RV64A insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32-64.decode           |  13 +++
 target/riscv/insn_trans/trans_rva.inc.c |  58 ++++++++++
 target/riscv/translate.c                | 144 ------------------------
 3 files changed, 71 insertions(+), 144 deletions(-)

diff --git a/target/riscv/insn32-64.decode b/target/riscv/insn32-64.decode
index 008f1005469e..0bee95c9840d 100644
--- a/target/riscv/insn32-64.decode
+++ b/target/riscv/insn32-64.decode
@@ -43,3 +43,16 @@ divw     0000001 .....  ..... 100 ..... 0111011 @r
 divuw    0000001 .....  ..... 101 ..... 0111011 @r
 remw     0000001 .....  ..... 110 ..... 0111011 @r
 remuw    0000001 .....  ..... 111 ..... 0111011 @r
+
+# *** RV64A Standard Extension (in addition to RV32A) ***
+lr_d       00010 . . 00000 ..... 011 ..... 0101111 @atom_ld
+sc_d       00011 . . ..... ..... 011 ..... 0101111 @atom_st
+amoswap_d  00001 . . ..... ..... 011 ..... 0101111 @atom_st
+amoadd_d   00000 . . ..... ..... 011 ..... 0101111 @atom_st
+amoxor_d   00100 . . ..... ..... 011 ..... 0101111 @atom_st
+amoand_d   01100 . . ..... ..... 011 ..... 0101111 @atom_st
+amoor_d    01000 . . ..... ..... 011 ..... 0101111 @atom_st
+amomin_d   10000 . . ..... ..... 011 ..... 0101111 @atom_st
+amomax_d   10100 . . ..... ..... 011 ..... 0101111 @atom_st
+amominu_d  11000 . . ..... ..... 011 ..... 0101111 @atom_st
+amomaxu_d  11100 . . ..... ..... 011 ..... 0101111 @atom_st
diff --git a/target/riscv/insn_trans/trans_rva.inc.c b/target/riscv/insn_trans/trans_rva.inc.c
index 5d3c853ca504..f6dbbc065e15 100644
--- a/target/riscv/insn_trans/trans_rva.inc.c
+++ b/target/riscv/insn_trans/trans_rva.inc.c
@@ -158,3 +158,61 @@ static bool trans_amomaxu_w(DisasContext *ctx, arg_amomaxu_w *a)
     REQUIRE_EXT(ctx, RVA);
     return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umax_tl, (MO_ALIGN | MO_TESL));
 }
+
+#ifdef TARGET_RISCV64
+
+static bool trans_lr_d(DisasContext *ctx, arg_lr_d *a)
+{
+    return gen_lr(ctx, a, MO_ALIGN | MO_TEQ);
+}
+
+static bool trans_sc_d(DisasContext *ctx, arg_sc_d *a)
+{
+    return gen_sc(ctx, a, (MO_ALIGN | MO_TEQ));
+}
+
+static bool trans_amoswap_d(DisasContext *ctx, arg_amoswap_d *a)
+{
+    return gen_amo(ctx, a, &tcg_gen_atomic_xchg_tl, (MO_ALIGN | MO_TEQ));
+}
+
+static bool trans_amoadd_d(DisasContext *ctx, arg_amoadd_d *a)
+{
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_add_tl, (MO_ALIGN | MO_TEQ));
+}
+
+static bool trans_amoxor_d(DisasContext *ctx, arg_amoxor_d *a)
+{
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_xor_tl, (MO_ALIGN | MO_TEQ));
+}
+
+static bool trans_amoand_d(DisasContext *ctx, arg_amoand_d *a)
+{
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_and_tl, (MO_ALIGN | MO_TEQ));
+}
+
+static bool trans_amoor_d(DisasContext *ctx, arg_amoor_d *a)
+{
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_or_tl, (MO_ALIGN | MO_TEQ));
+}
+
+static bool trans_amomin_d(DisasContext *ctx, arg_amomin_d *a)
+{
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smin_tl, (MO_ALIGN | MO_TEQ));
+}
+
+static bool trans_amomax_d(DisasContext *ctx, arg_amomax_d *a)
+{
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_smax_tl, (MO_ALIGN | MO_TEQ));
+}
+
+static bool trans_amominu_d(DisasContext *ctx, arg_amominu_d *a)
+{
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umin_tl, (MO_ALIGN | MO_TEQ));
+}
+
+static bool trans_amomaxu_d(DisasContext *ctx, arg_amomaxu_d *a)
+{
+    return gen_amo(ctx, a, &tcg_gen_atomic_fetch_umax_tl, (MO_ALIGN | MO_TEQ));
+}
+#endif
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index b0de062a4fd1..c2791459993e 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -794,143 +794,6 @@ static void gen_fp_store(DisasContext *ctx, uint32_t opc, int rs1,
     tcg_temp_free(t0);
 }
 
-static void gen_atomic(DisasContext *ctx, uint32_t opc,
-                      int rd, int rs1, int rs2)
-{
-    TCGv src1, src2, dat;
-    TCGLabel *l1, *l2;
-    TCGMemOp mop;
-    bool aq, rl;
-
-    /* Extract the size of the atomic operation.  */
-    switch (extract32(opc, 12, 3)) {
-    case 2: /* 32-bit */
-        mop = MO_ALIGN | MO_TESL;
-        break;
-#if defined(TARGET_RISCV64)
-    case 3: /* 64-bit */
-        mop = MO_ALIGN | MO_TEQ;
-        break;
-#endif
-    default:
-        gen_exception_illegal(ctx);
-        return;
-    }
-    rl = extract32(opc, 25, 1);
-    aq = extract32(opc, 26, 1);
-
-    src1 = tcg_temp_new();
-    src2 = tcg_temp_new();
-
-    switch (MASK_OP_ATOMIC_NO_AQ_RL_SZ(opc)) {
-    case OPC_RISC_LR:
-        /* Put addr in load_res, data in load_val.  */
-        gen_get_gpr(src1, rs1);
-        if (rl) {
-            tcg_gen_mb(TCG_MO_ALL | TCG_BAR_STRL);
-        }
-        tcg_gen_qemu_ld_tl(load_val, src1, ctx->mem_idx, mop);
-        if (aq) {
-            tcg_gen_mb(TCG_MO_ALL | TCG_BAR_LDAQ);
-        }
-        tcg_gen_mov_tl(load_res, src1);
-        gen_set_gpr(rd, load_val);
-        break;
-
-    case OPC_RISC_SC:
-        l1 = gen_new_label();
-        l2 = gen_new_label();
-        dat = tcg_temp_new();
-
-        gen_get_gpr(src1, rs1);
-        tcg_gen_brcond_tl(TCG_COND_NE, load_res, src1, l1);
-
-        gen_get_gpr(src2, rs2);
-        /* Note that the TCG atomic primitives are SC,
-           so we can ignore AQ/RL along this path.  */
-        tcg_gen_atomic_cmpxchg_tl(src1, load_res, load_val, src2,
-                                  ctx->mem_idx, mop);
-        tcg_gen_setcond_tl(TCG_COND_NE, dat, src1, load_val);
-        gen_set_gpr(rd, dat);
-        tcg_gen_br(l2);
-
-        gen_set_label(l1);
-        /* Address comparion failure.  However, we still need to
-           provide the memory barrier implied by AQ/RL.  */
-        tcg_gen_mb(TCG_MO_ALL + aq * TCG_BAR_LDAQ + rl * TCG_BAR_STRL);
-        tcg_gen_movi_tl(dat, 1);
-        gen_set_gpr(rd, dat);
-
-        gen_set_label(l2);
-        tcg_temp_free(dat);
-        break;
-
-    case OPC_RISC_AMOSWAP:
-        /* Note that the TCG atomic primitives are SC,
-           so we can ignore AQ/RL along this path.  */
-        gen_get_gpr(src1, rs1);
-        gen_get_gpr(src2, rs2);
-        tcg_gen_atomic_xchg_tl(src2, src1, src2, ctx->mem_idx, mop);
-        gen_set_gpr(rd, src2);
-        break;
-    case OPC_RISC_AMOADD:
-        gen_get_gpr(src1, rs1);
-        gen_get_gpr(src2, rs2);
-        tcg_gen_atomic_fetch_add_tl(src2, src1, src2, ctx->mem_idx, mop);
-        gen_set_gpr(rd, src2);
-        break;
-    case OPC_RISC_AMOXOR:
-        gen_get_gpr(src1, rs1);
-        gen_get_gpr(src2, rs2);
-        tcg_gen_atomic_fetch_xor_tl(src2, src1, src2, ctx->mem_idx, mop);
-        gen_set_gpr(rd, src2);
-        break;
-    case OPC_RISC_AMOAND:
-        gen_get_gpr(src1, rs1);
-        gen_get_gpr(src2, rs2);
-        tcg_gen_atomic_fetch_and_tl(src2, src1, src2, ctx->mem_idx, mop);
-        gen_set_gpr(rd, src2);
-        break;
-    case OPC_RISC_AMOOR:
-        gen_get_gpr(src1, rs1);
-        gen_get_gpr(src2, rs2);
-        tcg_gen_atomic_fetch_or_tl(src2, src1, src2, ctx->mem_idx, mop);
-        gen_set_gpr(rd, src2);
-        break;
-    case OPC_RISC_AMOMIN:
-        gen_get_gpr(src1, rs1);
-        gen_get_gpr(src2, rs2);
-        tcg_gen_atomic_fetch_smin_tl(src2, src1, src2, ctx->mem_idx, mop);
-        gen_set_gpr(rd, src2);
-        break;
-    case OPC_RISC_AMOMAX:
-        gen_get_gpr(src1, rs1);
-        gen_get_gpr(src2, rs2);
-        tcg_gen_atomic_fetch_smax_tl(src2, src1, src2, ctx->mem_idx, mop);
-        gen_set_gpr(rd, src2);
-        break;
-    case OPC_RISC_AMOMINU:
-        gen_get_gpr(src1, rs1);
-        gen_get_gpr(src2, rs2);
-        tcg_gen_atomic_fetch_umin_tl(src2, src1, src2, ctx->mem_idx, mop);
-        gen_set_gpr(rd, src2);
-        break;
-    case OPC_RISC_AMOMAXU:
-        gen_get_gpr(src1, rs1);
-        gen_get_gpr(src2, rs2);
-        tcg_gen_atomic_fetch_umax_tl(src2, src1, src2, ctx->mem_idx, mop);
-        gen_set_gpr(rd, src2);
-        break;
-
-    default:
-        gen_exception_illegal(ctx);
-        break;
-    }
-
-    tcg_temp_free(src1);
-    tcg_temp_free(src2);
-}
-
 static void gen_set_rm(DisasContext *ctx, int rm)
 {
     TCGv_i32 t0;
@@ -1882,12 +1745,6 @@ static void decode_RV32_64G(DisasContext *ctx)
         gen_fp_store(ctx, MASK_OP_FP_STORE(ctx->opcode), rs1, rs2,
                      GET_STORE_IMM(ctx->opcode));
         break;
-    case OPC_RISC_ATOMIC:
-        if (!has_ext(ctx, RVA)) {
-            goto do_illegal;
-        }
-        gen_atomic(ctx, MASK_OP_ATOMIC(ctx->opcode), rd, rs1, rs2);
-        break;
     case OPC_RISC_FMADD:
         gen_fp_fmadd(ctx, MASK_OP_FP_FMADD(ctx->opcode), rd, rs1, rs2,
                      GET_RS3(ctx->opcode), GET_RM(ctx->opcode));
@@ -1912,7 +1769,6 @@ static void decode_RV32_64G(DisasContext *ctx)
         gen_system(ctx, MASK_OP_SYSTEM(ctx->opcode), rd, rs1,
                    (ctx->opcode & 0xFFF00000) >> 20);
         break;
-    do_illegal:
     default:
         gen_exception_illegal(ctx);
         break;
-- 
2.19.2

[Qemu-riscv] [PULL 11/29] target/riscv: Convert RV32F insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32.decode              |  35 +++
 target/riscv/insn_trans/trans_rvf.inc.c | 379 ++++++++++++++++++++++++
 target/riscv/translate.c                |   1 +
 3 files changed, 415 insertions(+)
 create mode 100644 target/riscv/insn_trans/trans_rvf.inc.c

diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index 00b9e2d9a508..e40836bf032f 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -17,12 +17,14 @@
 # this program.  If not, see <http://www.gnu.org/licenses/>.
 
 # Fields:
+%rs3       27:5
 %rs2       20:5
 %rs1       15:5
 %rd        7:5
 
 %sh10    20:10
 %csr    20:12
+%rm     12:3
 
 # immediates:
 %imm_i    20:s12
@@ -50,6 +52,11 @@
 @atom_ld ..... aq:1 rl:1 ..... ........ ..... ....... &atomic rs2=0     %rs1 %rd
 @atom_st ..... aq:1 rl:1 ..... ........ ..... ....... &atomic %rs2      %rs1 %rd
 
+@r4_rm   ..... ..  ..... ..... ... ..... ....... %rs3 %rs2 %rs1 %rm %rd
+@r_rm    .......   ..... ..... ... ..... ....... %rs2 %rs1 %rm %rd
+@r2_rm   .......   ..... ..... ... ..... ....... %rs1 %rm %rd
+@r2      .......   ..... ..... ... ..... ....... %rs1 %rd
+
 # *** RV32I Base Instruction Set ***
 lui      ....................       ..... 0110111 @u
 auipc    ....................       ..... 0010111 @u
@@ -119,3 +126,31 @@ amomin_w   10000 . . ..... ..... 010 ..... 0101111 @atom_st
 amomax_w   10100 . . ..... ..... 010 ..... 0101111 @atom_st
 amominu_w  11000 . . ..... ..... 010 ..... 0101111 @atom_st
 amomaxu_w  11100 . . ..... ..... 010 ..... 0101111 @atom_st
+
+# *** RV32F Standard Extension ***
+flw        ............   ..... 010 ..... 0000111 @i
+fsw        .......  ..... ..... 010 ..... 0100111 @s
+fmadd_s    ..... 00 ..... ..... ... ..... 1000011 @r4_rm
+fmsub_s    ..... 00 ..... ..... ... ..... 1000111 @r4_rm
+fnmsub_s   ..... 00 ..... ..... ... ..... 1001011 @r4_rm
+fnmadd_s   ..... 00 ..... ..... ... ..... 1001111 @r4_rm
+fadd_s     0000000  ..... ..... ... ..... 1010011 @r_rm
+fsub_s     0000100  ..... ..... ... ..... 1010011 @r_rm
+fmul_s     0001000  ..... ..... ... ..... 1010011 @r_rm
+fdiv_s     0001100  ..... ..... ... ..... 1010011 @r_rm
+fsqrt_s    0101100  00000 ..... ... ..... 1010011 @r2_rm
+fsgnj_s    0010000  ..... ..... 000 ..... 1010011 @r
+fsgnjn_s   0010000  ..... ..... 001 ..... 1010011 @r
+fsgnjx_s   0010000  ..... ..... 010 ..... 1010011 @r
+fmin_s     0010100  ..... ..... 000 ..... 1010011 @r
+fmax_s     0010100  ..... ..... 001 ..... 1010011 @r
+fcvt_w_s   1100000  00000 ..... ... ..... 1010011 @r2_rm
+fcvt_wu_s  1100000  00001 ..... ... ..... 1010011 @r2_rm
+fmv_x_w    1110000  00000 ..... 000 ..... 1010011 @r2
+feq_s      1010000  ..... ..... 010 ..... 1010011 @r
+flt_s      1010000  ..... ..... 001 ..... 1010011 @r
+fle_s      1010000  ..... ..... 000 ..... 1010011 @r
+fclass_s   1110000  00000 ..... 001 ..... 1010011 @r2
+fcvt_s_w   1101000  00000 ..... ... ..... 1010011 @r2_rm
+fcvt_s_wu  1101000  00001 ..... ... ..... 1010011 @r2_rm
+fmv_w_x    1111000  00000 ..... 000 ..... 1010011 @r2
diff --git a/target/riscv/insn_trans/trans_rvf.inc.c b/target/riscv/insn_trans/trans_rvf.inc.c
new file mode 100644
index 000000000000..0f837903491b
--- /dev/null
+++ b/target/riscv/insn_trans/trans_rvf.inc.c
@@ -0,0 +1,379 @@
+/*
+ * RISC-V translation routines for the RV64F Standard Extension.
+ *
+ * Copyright (c) 2016-2017 Sagar Karandikar, sagark@eecs.berkeley.edu
+ * Copyright (c) 2018 Peer Adelt, peer.adelt@hni.uni-paderborn.de
+ *                    Bastian Koppelmann, kbastian@mail.uni-paderborn.de
+ *
+ * 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_FPU do {\
+    if (ctx->mstatus_fs == 0) \
+        return false;                       \
+} while (0)
+
+static bool trans_flw(DisasContext *ctx, arg_flw *a)
+{
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    tcg_gen_addi_tl(t0, t0, a->imm);
+
+    tcg_gen_qemu_ld_i64(cpu_fpr[a->rd], t0, ctx->mem_idx, MO_TEUL);
+    /* RISC-V requires NaN-boxing of narrower width floating point values */
+    tcg_gen_ori_i64(cpu_fpr[a->rd], cpu_fpr[a->rd], 0xffffffff00000000ULL);
+
+    tcg_temp_free(t0);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fsw(DisasContext *ctx, arg_fsw *a)
+{
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    tcg_gen_addi_tl(t0, t0, a->imm);
+
+    tcg_gen_qemu_st_i64(cpu_fpr[a->rs2], t0, ctx->mem_idx, MO_TEUL);
+
+    tcg_temp_free(t0);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fmadd_s(DisasContext *ctx, arg_fmadd_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fmadd_s(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1],
+                       cpu_fpr[a->rs2], cpu_fpr[a->rs3]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fmsub_s(DisasContext *ctx, arg_fmsub_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fmsub_s(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1],
+                       cpu_fpr[a->rs2], cpu_fpr[a->rs3]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fnmsub_s(DisasContext *ctx, arg_fnmsub_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fnmsub_s(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1],
+                        cpu_fpr[a->rs2], cpu_fpr[a->rs3]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fnmadd_s(DisasContext *ctx, arg_fnmadd_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fnmadd_s(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1],
+                        cpu_fpr[a->rs2], cpu_fpr[a->rs3]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fadd_s(DisasContext *ctx, arg_fadd_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fadd_s(cpu_fpr[a->rd], cpu_env,
+                      cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fsub_s(DisasContext *ctx, arg_fsub_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fsub_s(cpu_fpr[a->rd], cpu_env,
+                      cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fmul_s(DisasContext *ctx, arg_fmul_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fmul_s(cpu_fpr[a->rd], cpu_env,
+                      cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fdiv_s(DisasContext *ctx, arg_fdiv_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fdiv_s(cpu_fpr[a->rd], cpu_env,
+                      cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fsqrt_s(DisasContext *ctx, arg_fsqrt_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fsqrt_s(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fsgnj_s(DisasContext *ctx, arg_fsgnj_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    if (a->rs1 == a->rs2) { /* FMOV */
+        tcg_gen_mov_i64(cpu_fpr[a->rd], cpu_fpr[a->rs1]);
+    } else { /* FSGNJ */
+        tcg_gen_deposit_i64(cpu_fpr[a->rd], cpu_fpr[a->rs2], cpu_fpr[a->rs1],
+                            0, 31);
+    }
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fsgnjn_s(DisasContext *ctx, arg_fsgnjn_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    if (a->rs1 == a->rs2) { /* FNEG */
+        tcg_gen_xori_i64(cpu_fpr[a->rd], cpu_fpr[a->rs1], INT32_MIN);
+    } else {
+        TCGv_i64 t0 = tcg_temp_new_i64();
+        tcg_gen_not_i64(t0, cpu_fpr[a->rs2]);
+        tcg_gen_deposit_i64(cpu_fpr[a->rd], t0, cpu_fpr[a->rs1], 0, 31);
+        tcg_temp_free_i64(t0);
+    }
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fsgnjx_s(DisasContext *ctx, arg_fsgnjx_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    if (a->rs1 == a->rs2) { /* FABS */
+        tcg_gen_andi_i64(cpu_fpr[a->rd], cpu_fpr[a->rs1], ~INT32_MIN);
+    } else {
+        TCGv_i64 t0 = tcg_temp_new_i64();
+        tcg_gen_andi_i64(t0, cpu_fpr[a->rs2], INT32_MIN);
+        tcg_gen_xor_i64(cpu_fpr[a->rd], cpu_fpr[a->rs1], t0);
+        tcg_temp_free_i64(t0);
+    }
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fmin_s(DisasContext *ctx, arg_fmin_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    gen_helper_fmin_s(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1],
+                      cpu_fpr[a->rs2]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fmax_s(DisasContext *ctx, arg_fmax_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    gen_helper_fmax_s(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1],
+                      cpu_fpr[a->rs2]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fcvt_w_s(DisasContext *ctx, arg_fcvt_w_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    TCGv t0 = tcg_temp_new();
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_w_s(t0, cpu_env, cpu_fpr[a->rs1]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+
+    return true;
+}
+
+static bool trans_fcvt_wu_s(DisasContext *ctx, arg_fcvt_wu_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    TCGv t0 = tcg_temp_new();
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_wu_s(t0, cpu_env, cpu_fpr[a->rs1]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+
+    return true;
+}
+
+static bool trans_fmv_x_w(DisasContext *ctx, arg_fmv_x_w *a)
+{
+    /* NOTE: This was FMV.X.S in an earlier version of the ISA spec! */
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    TCGv t0 = tcg_temp_new();
+
+#if defined(TARGET_RISCV64)
+    tcg_gen_ext32s_tl(t0, cpu_fpr[a->rs1]);
+#else
+    tcg_gen_extrl_i64_i32(t0, cpu_fpr[a->rs1]);
+#endif
+
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+
+    return true;
+}
+
+static bool trans_feq_s(DisasContext *ctx, arg_feq_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    TCGv t0 = tcg_temp_new();
+    gen_helper_feq_s(t0, cpu_env, cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+    return true;
+}
+
+static bool trans_flt_s(DisasContext *ctx, arg_flt_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    TCGv t0 = tcg_temp_new();
+    gen_helper_flt_s(t0, cpu_env, cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+    return true;
+}
+
+static bool trans_fle_s(DisasContext *ctx, arg_fle_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+    TCGv t0 = tcg_temp_new();
+    gen_helper_fle_s(t0, cpu_env, cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+    return true;
+}
+
+static bool trans_fclass_s(DisasContext *ctx, arg_fclass_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    TCGv t0 = tcg_temp_new();
+
+    gen_helper_fclass_s(t0, cpu_fpr[a->rs1]);
+
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+
+    return true;
+}
+
+static bool trans_fcvt_s_w(DisasContext *ctx, arg_fcvt_s_w *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_s_w(cpu_fpr[a->rd], cpu_env, t0);
+
+    mark_fs_dirty(ctx);
+    tcg_temp_free(t0);
+
+    return true;
+}
+
+static bool trans_fcvt_s_wu(DisasContext *ctx, arg_fcvt_s_wu *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_s_wu(cpu_fpr[a->rd], cpu_env, t0);
+
+    mark_fs_dirty(ctx);
+    tcg_temp_free(t0);
+
+    return true;
+}
+
+static bool trans_fmv_w_x(DisasContext *ctx, arg_fmv_w_x *a)
+{
+    /* NOTE: This was FMV.S.X in an earlier version of the ISA spec! */
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+
+#if defined(TARGET_RISCV64)
+    tcg_gen_mov_i64(cpu_fpr[a->rd], t0);
+#else
+    tcg_gen_extu_i32_i64(cpu_fpr[a->rd], t0);
+#endif
+
+    mark_fs_dirty(ctx);
+    tcg_temp_free(t0);
+
+    return true;
+}
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index c2791459993e..b9f78f5a1f8f 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -1717,6 +1717,7 @@ bool decode_insn32(DisasContext *ctx, uint32_t insn);
 #include "insn_trans/trans_rvi.inc.c"
 #include "insn_trans/trans_rvm.inc.c"
 #include "insn_trans/trans_rva.inc.c"
+#include "insn_trans/trans_rvf.inc.c"
 
 static void decode_RV32_64G(DisasContext *ctx)
 {
-- 
2.19.2

[Qemu-riscv] [PULL 12/29] target/riscv: Convert RV64F insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32-64.decode           |  6 +++
 target/riscv/insn_trans/trans_rvf.inc.c | 60 +++++++++++++++++++++++++
 2 files changed, 66 insertions(+)

diff --git a/target/riscv/insn32-64.decode b/target/riscv/insn32-64.decode
index 0bee95c9840d..6319f872ac1d 100644
--- a/target/riscv/insn32-64.decode
+++ b/target/riscv/insn32-64.decode
@@ -56,3 +56,9 @@ amomin_d   10000 . . ..... ..... 011 ..... 0101111 @atom_st
 amomax_d   10100 . . ..... ..... 011 ..... 0101111 @atom_st
 amominu_d  11000 . . ..... ..... 011 ..... 0101111 @atom_st
 amomaxu_d  11100 . . ..... ..... 011 ..... 0101111 @atom_st
+
+# *** RV64F Standard Extension (in addition to RV32F) ***
+fcvt_l_s   1100000  00010 ..... ... ..... 1010011 @r2_rm
+fcvt_lu_s  1100000  00011 ..... ... ..... 1010011 @r2_rm
+fcvt_s_l   1101000  00010 ..... ... ..... 1010011 @r2_rm
+fcvt_s_lu  1101000  00011 ..... ... ..... 1010011 @r2_rm
diff --git a/target/riscv/insn_trans/trans_rvf.inc.c b/target/riscv/insn_trans/trans_rvf.inc.c
index 0f837903491b..172dbfa919b6 100644
--- a/target/riscv/insn_trans/trans_rvf.inc.c
+++ b/target/riscv/insn_trans/trans_rvf.inc.c
@@ -377,3 +377,63 @@ static bool trans_fmv_w_x(DisasContext *ctx, arg_fmv_w_x *a)
 
     return true;
 }
+
+#ifdef TARGET_RISCV64
+static bool trans_fcvt_l_s(DisasContext *ctx, arg_fcvt_l_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    TCGv t0 = tcg_temp_new();
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_l_s(t0, cpu_env, cpu_fpr[a->rs1]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+    return true;
+}
+
+static bool trans_fcvt_lu_s(DisasContext *ctx, arg_fcvt_lu_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    TCGv t0 = tcg_temp_new();
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_lu_s(t0, cpu_env, cpu_fpr[a->rs1]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+    return true;
+}
+
+static bool trans_fcvt_s_l(DisasContext *ctx, arg_fcvt_s_l *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_s_l(cpu_fpr[a->rd], cpu_env, t0);
+
+    mark_fs_dirty(ctx);
+    tcg_temp_free(t0);
+    return true;
+}
+
+static bool trans_fcvt_s_lu(DisasContext *ctx, arg_fcvt_s_lu *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVF);
+
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_s_lu(cpu_fpr[a->rd], cpu_env, t0);
+
+    mark_fs_dirty(ctx);
+    tcg_temp_free(t0);
+    return true;
+}
+#endif
-- 
2.19.2

[Qemu-riscv] [PULL 13/29] target/riscv: Convert RV32D insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32.decode              |  28 ++
 target/riscv/insn_trans/trans_rvd.inc.c | 360 ++++++++++++++++++++++++
 target/riscv/translate.c                |   1 +
 3 files changed, 389 insertions(+)
 create mode 100644 target/riscv/insn_trans/trans_rvd.inc.c

diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index e40836bf032f..e64b2b5e3458 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -154,3 +154,31 @@ fclass_s   1110000  00000 ..... 001 ..... 1010011 @r2
 fcvt_s_w   1101000  00000 ..... ... ..... 1010011 @r2_rm
 fcvt_s_wu  1101000  00001 ..... ... ..... 1010011 @r2_rm
 fmv_w_x    1111000  00000 ..... 000 ..... 1010011 @r2
+
+# *** RV32D Standard Extension ***
+fld        ............   ..... 011 ..... 0000111 @i
+fsd        ....... .....  ..... 011 ..... 0100111 @s
+fmadd_d    ..... 01 ..... ..... ... ..... 1000011 @r4_rm
+fmsub_d    ..... 01 ..... ..... ... ..... 1000111 @r4_rm
+fnmsub_d   ..... 01 ..... ..... ... ..... 1001011 @r4_rm
+fnmadd_d   ..... 01 ..... ..... ... ..... 1001111 @r4_rm
+fadd_d     0000001  ..... ..... ... ..... 1010011 @r_rm
+fsub_d     0000101  ..... ..... ... ..... 1010011 @r_rm
+fmul_d     0001001  ..... ..... ... ..... 1010011 @r_rm
+fdiv_d     0001101  ..... ..... ... ..... 1010011 @r_rm
+fsqrt_d    0101101  00000 ..... ... ..... 1010011 @r2_rm
+fsgnj_d    0010001  ..... ..... 000 ..... 1010011 @r
+fsgnjn_d   0010001  ..... ..... 001 ..... 1010011 @r
+fsgnjx_d   0010001  ..... ..... 010 ..... 1010011 @r
+fmin_d     0010101  ..... ..... 000 ..... 1010011 @r
+fmax_d     0010101  ..... ..... 001 ..... 1010011 @r
+fcvt_s_d   0100000  00001 ..... ... ..... 1010011 @r2_rm
+fcvt_d_s   0100001  00000 ..... ... ..... 1010011 @r2_rm
+feq_d      1010001  ..... ..... 010 ..... 1010011 @r
+flt_d      1010001  ..... ..... 001 ..... 1010011 @r
+fle_d      1010001  ..... ..... 000 ..... 1010011 @r
+fclass_d   1110001  00000 ..... 001 ..... 1010011 @r2
+fcvt_w_d   1100001  00000 ..... ... ..... 1010011 @r2_rm
+fcvt_wu_d  1100001  00001 ..... ... ..... 1010011 @r2_rm
+fcvt_d_w   1101001  00000 ..... ... ..... 1010011 @r2_rm
+fcvt_d_wu  1101001  00001 ..... ... ..... 1010011 @r2_rm
diff --git a/target/riscv/insn_trans/trans_rvd.inc.c b/target/riscv/insn_trans/trans_rvd.inc.c
new file mode 100644
index 000000000000..98fc1cdc5a20
--- /dev/null
+++ b/target/riscv/insn_trans/trans_rvd.inc.c
@@ -0,0 +1,360 @@
+/*
+ * RISC-V translation routines for the RV64D Standard Extension.
+ *
+ * Copyright (c) 2016-2017 Sagar Karandikar, sagark@eecs.berkeley.edu
+ * Copyright (c) 2018 Peer Adelt, peer.adelt@hni.uni-paderborn.de
+ *                    Bastian Koppelmann, kbastian@mail.uni-paderborn.de
+ *
+ * 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/>.
+ */
+
+static bool trans_fld(DisasContext *ctx, arg_fld *a)
+{
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+    tcg_gen_addi_tl(t0, t0, a->imm);
+
+    tcg_gen_qemu_ld_i64(cpu_fpr[a->rd], t0, ctx->mem_idx, MO_TEQ);
+
+    mark_fs_dirty(ctx);
+    tcg_temp_free(t0);
+    return true;
+}
+
+static bool trans_fsd(DisasContext *ctx, arg_fsd *a)
+{
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+    tcg_gen_addi_tl(t0, t0, a->imm);
+
+    tcg_gen_qemu_st_i64(cpu_fpr[a->rs2], t0, ctx->mem_idx, MO_TEQ);
+
+    mark_fs_dirty(ctx);
+    tcg_temp_free(t0);
+    return true;
+}
+
+static bool trans_fmadd_d(DisasContext *ctx, arg_fmadd_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fmadd_d(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1],
+                       cpu_fpr[a->rs2], cpu_fpr[a->rs3]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fmsub_d(DisasContext *ctx, arg_fmsub_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fmsub_d(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1],
+                       cpu_fpr[a->rs2], cpu_fpr[a->rs3]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fnmsub_d(DisasContext *ctx, arg_fnmsub_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fnmsub_d(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1],
+                        cpu_fpr[a->rs2], cpu_fpr[a->rs3]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fnmadd_d(DisasContext *ctx, arg_fnmadd_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fnmadd_d(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1],
+                        cpu_fpr[a->rs2], cpu_fpr[a->rs3]);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fadd_d(DisasContext *ctx, arg_fadd_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fadd_d(cpu_fpr[a->rd], cpu_env,
+                      cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fsub_d(DisasContext *ctx, arg_fsub_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fsub_d(cpu_fpr[a->rd], cpu_env,
+                      cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fmul_d(DisasContext *ctx, arg_fmul_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fmul_d(cpu_fpr[a->rd], cpu_env,
+                      cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fdiv_d(DisasContext *ctx, arg_fdiv_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fdiv_d(cpu_fpr[a->rd], cpu_env,
+                      cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fsqrt_d(DisasContext *ctx, arg_fsqrt_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fsqrt_d(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1]);
+
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fsgnj_d(DisasContext *ctx, arg_fsgnj_d *a)
+{
+    if (a->rs1 == a->rs2) { /* FMOV */
+        tcg_gen_mov_i64(cpu_fpr[a->rd], cpu_fpr[a->rs1]);
+    } else {
+        tcg_gen_deposit_i64(cpu_fpr[a->rd], cpu_fpr[a->rs2],
+                            cpu_fpr[a->rs1], 0, 63);
+    }
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fsgnjn_d(DisasContext *ctx, arg_fsgnjn_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+    if (a->rs1 == a->rs2) { /* FNEG */
+        tcg_gen_xori_i64(cpu_fpr[a->rd], cpu_fpr[a->rs1], INT64_MIN);
+    } else {
+        TCGv_i64 t0 = tcg_temp_new_i64();
+        tcg_gen_not_i64(t0, cpu_fpr[a->rs2]);
+        tcg_gen_deposit_i64(cpu_fpr[a->rd], t0, cpu_fpr[a->rs1], 0, 63);
+        tcg_temp_free_i64(t0);
+    }
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fsgnjx_d(DisasContext *ctx, arg_fsgnjx_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+    if (a->rs1 == a->rs2) { /* FABS */
+        tcg_gen_andi_i64(cpu_fpr[a->rd], cpu_fpr[a->rs1], ~INT64_MIN);
+    } else {
+        TCGv_i64 t0 = tcg_temp_new_i64();
+        tcg_gen_andi_i64(t0, cpu_fpr[a->rs2], INT64_MIN);
+        tcg_gen_xor_i64(cpu_fpr[a->rd], cpu_fpr[a->rs1], t0);
+        tcg_temp_free_i64(t0);
+    }
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fmin_d(DisasContext *ctx, arg_fmin_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    gen_helper_fmin_d(cpu_fpr[a->rd], cpu_env,
+                      cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fmax_d(DisasContext *ctx, arg_fmax_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    gen_helper_fmax_d(cpu_fpr[a->rd], cpu_env,
+                      cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fcvt_s_d(DisasContext *ctx, arg_fcvt_s_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_s_d(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1]);
+
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fcvt_d_s(DisasContext *ctx, arg_fcvt_d_s *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_d_s(cpu_fpr[a->rd], cpu_env, cpu_fpr[a->rs1]);
+
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_feq_d(DisasContext *ctx, arg_feq_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_helper_feq_d(t0, cpu_env, cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+
+    return true;
+}
+
+static bool trans_flt_d(DisasContext *ctx, arg_flt_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_helper_flt_d(t0, cpu_env, cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+
+    return true;
+}
+
+static bool trans_fle_d(DisasContext *ctx, arg_fle_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_helper_fle_d(t0, cpu_env, cpu_fpr[a->rs1], cpu_fpr[a->rs2]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+
+    return true;
+}
+
+static bool trans_fclass_d(DisasContext *ctx, arg_fclass_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_helper_fclass_d(t0, cpu_fpr[a->rs1]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+    return true;
+}
+
+static bool trans_fcvt_w_d(DisasContext *ctx, arg_fcvt_w_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_w_d(t0, cpu_env, cpu_fpr[a->rs1]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+
+    return true;
+}
+
+static bool trans_fcvt_wu_d(DisasContext *ctx, arg_fcvt_wu_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_wu_d(t0, cpu_env, cpu_fpr[a->rs1]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+
+    return true;
+}
+
+static bool trans_fcvt_d_w(DisasContext *ctx, arg_fcvt_d_w *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_d_w(cpu_fpr[a->rd], cpu_env, t0);
+    tcg_temp_free(t0);
+
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fcvt_d_wu(DisasContext *ctx, arg_fcvt_d_wu *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_d_wu(cpu_fpr[a->rd], cpu_env, t0);
+    tcg_temp_free(t0);
+
+    mark_fs_dirty(ctx);
+    return true;
+}
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index b9f78f5a1f8f..c201985ef37b 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -1718,6 +1718,7 @@ bool decode_insn32(DisasContext *ctx, uint32_t insn);
 #include "insn_trans/trans_rvm.inc.c"
 #include "insn_trans/trans_rva.inc.c"
 #include "insn_trans/trans_rvf.inc.c"
+#include "insn_trans/trans_rvd.inc.c"
 
 static void decode_RV32_64G(DisasContext *ctx)
 {
-- 
2.19.2

[Qemu-riscv] [PULL 14/29] target/riscv: Convert RV64D insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32-64.decode           |   8 +
 target/riscv/insn_trans/trans_rvd.inc.c |  82 ++++
 target/riscv/translate.c                | 601 +-----------------------
 3 files changed, 91 insertions(+), 600 deletions(-)

diff --git a/target/riscv/insn32-64.decode b/target/riscv/insn32-64.decode
index 6319f872ac1d..380bf791bcdc 100644
--- a/target/riscv/insn32-64.decode
+++ b/target/riscv/insn32-64.decode
@@ -62,3 +62,11 @@ fcvt_l_s   1100000  00010 ..... ... ..... 1010011 @r2_rm
 fcvt_lu_s  1100000  00011 ..... ... ..... 1010011 @r2_rm
 fcvt_s_l   1101000  00010 ..... ... ..... 1010011 @r2_rm
 fcvt_s_lu  1101000  00011 ..... ... ..... 1010011 @r2_rm
+
+# *** RV64D Standard Extension (in addition to RV32D) ***
+fcvt_l_d   1100001  00010 ..... ... ..... 1010011 @r2_rm
+fcvt_lu_d  1100001  00011 ..... ... ..... 1010011 @r2_rm
+fmv_x_d    1110001  00000 ..... 000 ..... 1010011 @r2
+fcvt_d_l   1101001  00010 ..... ... ..... 1010011 @r2_rm
+fcvt_d_lu  1101001  00011 ..... ... ..... 1010011 @r2_rm
+fmv_d_x    1111001  00000 ..... 000 ..... 1010011 @r2
diff --git a/target/riscv/insn_trans/trans_rvd.inc.c b/target/riscv/insn_trans/trans_rvd.inc.c
index 98fc1cdc5a20..393fa0248ce9 100644
--- a/target/riscv/insn_trans/trans_rvd.inc.c
+++ b/target/riscv/insn_trans/trans_rvd.inc.c
@@ -358,3 +358,85 @@ static bool trans_fcvt_d_wu(DisasContext *ctx, arg_fcvt_d_wu *a)
     mark_fs_dirty(ctx);
     return true;
 }
+
+#ifdef TARGET_RISCV64
+
+static bool trans_fcvt_l_d(DisasContext *ctx, arg_fcvt_l_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_l_d(t0, cpu_env, cpu_fpr[a->rs1]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+    return true;
+}
+
+static bool trans_fcvt_lu_d(DisasContext *ctx, arg_fcvt_lu_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_lu_d(t0, cpu_env, cpu_fpr[a->rs1]);
+    gen_set_gpr(a->rd, t0);
+    tcg_temp_free(t0);
+    return true;
+}
+
+static bool trans_fmv_x_d(DisasContext *ctx, arg_fmv_x_d *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    gen_set_gpr(a->rd, cpu_fpr[a->rs1]);
+    return true;
+}
+
+static bool trans_fcvt_d_l(DisasContext *ctx, arg_fcvt_d_l *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_d_l(cpu_fpr[a->rd], cpu_env, t0);
+    tcg_temp_free(t0);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fcvt_d_lu(DisasContext *ctx, arg_fcvt_d_lu *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+
+    gen_set_rm(ctx, a->rm);
+    gen_helper_fcvt_d_lu(cpu_fpr[a->rd], cpu_env, t0);
+    tcg_temp_free(t0);
+    mark_fs_dirty(ctx);
+    return true;
+}
+
+static bool trans_fmv_d_x(DisasContext *ctx, arg_fmv_d_x *a)
+{
+    REQUIRE_FPU;
+    REQUIRE_EXT(ctx, RVD);
+
+    TCGv t0 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+
+    tcg_gen_mov_tl(cpu_fpr[a->rd], t0);
+    tcg_temp_free(t0);
+    mark_fs_dirty(ctx);
+    return true;
+}
+#endif
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index c201985ef37b..2e36deee82f4 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -186,44 +186,6 @@ static void gen_mulhsu(TCGv ret, TCGv arg1, TCGv arg2)
     tcg_temp_free(rh);
 }
 
-static void gen_fsgnj(DisasContext *ctx, uint32_t rd, uint32_t rs1,
-    uint32_t rs2, int rm, uint64_t min)
-{
-    switch (rm) {
-    case 0: /* fsgnj */
-        if (rs1 == rs2) { /* FMOV */
-            tcg_gen_mov_i64(cpu_fpr[rd], cpu_fpr[rs1]);
-        } else {
-            tcg_gen_deposit_i64(cpu_fpr[rd], cpu_fpr[rs2], cpu_fpr[rs1],
-                                0, min == INT32_MIN ? 31 : 63);
-        }
-        break;
-    case 1: /* fsgnjn */
-        if (rs1 == rs2) { /* FNEG */
-            tcg_gen_xori_i64(cpu_fpr[rd], cpu_fpr[rs1], min);
-        } else {
-            TCGv_i64 t0 = tcg_temp_new_i64();
-            tcg_gen_not_i64(t0, cpu_fpr[rs2]);
-            tcg_gen_deposit_i64(cpu_fpr[rd], t0, cpu_fpr[rs1],
-                                0, min == INT32_MIN ? 31 : 63);
-            tcg_temp_free_i64(t0);
-        }
-        break;
-    case 2: /* fsgnjx */
-        if (rs1 == rs2) { /* FABS */
-            tcg_gen_andi_i64(cpu_fpr[rd], cpu_fpr[rs1], ~min);
-        } else {
-            TCGv_i64 t0 = tcg_temp_new_i64();
-            tcg_gen_andi_i64(t0, cpu_fpr[rs2], min);
-            tcg_gen_xor_i64(cpu_fpr[rd], cpu_fpr[rs1], t0);
-            tcg_temp_free_i64(t0);
-        }
-        break;
-    default:
-        gen_exception_illegal(ctx);
-    }
-}
-
 static void gen_arith(DisasContext *ctx, uint32_t opc, int rd, int rs1,
         int rs2)
 {
@@ -807,535 +769,6 @@ static void gen_set_rm(DisasContext *ctx, int rm)
     tcg_temp_free_i32(t0);
 }
 
-static void gen_fp_fmadd(DisasContext *ctx, uint32_t opc, int rd,
-                         int rs1, int rs2, int rs3, int rm)
-{
-    switch (opc) {
-    case OPC_RISC_FMADD_S:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fmadd_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1],
-                           cpu_fpr[rs2], cpu_fpr[rs3]);
-        break;
-    case OPC_RISC_FMADD_D:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fmadd_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1],
-                           cpu_fpr[rs2], cpu_fpr[rs3]);
-        break;
-    do_illegal:
-    default:
-        gen_exception_illegal(ctx);
-        break;
-    }
-}
-
-static void gen_fp_fmsub(DisasContext *ctx, uint32_t opc, int rd,
-                         int rs1, int rs2, int rs3, int rm)
-{
-    switch (opc) {
-    case OPC_RISC_FMSUB_S:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fmsub_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1],
-                           cpu_fpr[rs2], cpu_fpr[rs3]);
-        break;
-    case OPC_RISC_FMSUB_D:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fmsub_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1],
-                           cpu_fpr[rs2], cpu_fpr[rs3]);
-        break;
-    do_illegal:
-    default:
-        gen_exception_illegal(ctx);
-        break;
-    }
-}
-
-static void gen_fp_fnmsub(DisasContext *ctx, uint32_t opc, int rd,
-                          int rs1, int rs2, int rs3, int rm)
-{
-    switch (opc) {
-    case OPC_RISC_FNMSUB_S:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fnmsub_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1],
-                            cpu_fpr[rs2], cpu_fpr[rs3]);
-        break;
-    case OPC_RISC_FNMSUB_D:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fnmsub_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1],
-                            cpu_fpr[rs2], cpu_fpr[rs3]);
-        break;
-    do_illegal:
-    default:
-        gen_exception_illegal(ctx);
-        break;
-    }
-}
-
-static void gen_fp_fnmadd(DisasContext *ctx, uint32_t opc, int rd,
-                          int rs1, int rs2, int rs3, int rm)
-{
-    switch (opc) {
-    case OPC_RISC_FNMADD_S:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fnmadd_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1],
-                            cpu_fpr[rs2], cpu_fpr[rs3]);
-        break;
-    case OPC_RISC_FNMADD_D:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fnmadd_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1],
-                            cpu_fpr[rs2], cpu_fpr[rs3]);
-        break;
-    do_illegal:
-    default:
-        gen_exception_illegal(ctx);
-        break;
-    }
-}
-
-static void gen_fp_arith(DisasContext *ctx, uint32_t opc, int rd,
-                         int rs1, int rs2, int rm)
-{
-    TCGv t0 = NULL;
-    bool fp_output = true;
-
-    if (ctx->mstatus_fs == 0) {
-        goto do_illegal;
-    }
-
-    switch (opc) {
-    case OPC_RISC_FADD_S:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fadd_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-        break;
-    case OPC_RISC_FSUB_S:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fsub_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-        break;
-    case OPC_RISC_FMUL_S:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fmul_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-        break;
-    case OPC_RISC_FDIV_S:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fdiv_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-        break;
-    case OPC_RISC_FSQRT_S:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fsqrt_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1]);
-        break;
-    case OPC_RISC_FSGNJ_S:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        gen_fsgnj(ctx, rd, rs1, rs2, rm, INT32_MIN);
-        break;
-
-    case OPC_RISC_FMIN_S:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        /* also handles: OPC_RISC_FMAX_S */
-        switch (rm) {
-        case 0x0:
-            gen_helper_fmin_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-            break;
-        case 0x1:
-            gen_helper_fmax_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-            break;
-        default:
-            goto do_illegal;
-        }
-        break;
-
-    case OPC_RISC_FEQ_S:
-        /* also handles: OPC_RISC_FLT_S, OPC_RISC_FLE_S */
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        t0 = tcg_temp_new();
-        switch (rm) {
-        case 0x0:
-            gen_helper_fle_s(t0, cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-            break;
-        case 0x1:
-            gen_helper_flt_s(t0, cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-            break;
-        case 0x2:
-            gen_helper_feq_s(t0, cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-            break;
-        default:
-            goto do_illegal;
-        }
-        gen_set_gpr(rd, t0);
-        tcg_temp_free(t0);
-        fp_output = false;
-        break;
-
-    case OPC_RISC_FCVT_W_S:
-        /* also OPC_RISC_FCVT_WU_S, OPC_RISC_FCVT_L_S, OPC_RISC_FCVT_LU_S */
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        t0 = tcg_temp_new();
-        switch (rs2) {
-        case 0: /* FCVT_W_S */
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_w_s(t0, cpu_env, cpu_fpr[rs1]);
-            break;
-        case 1: /* FCVT_WU_S */
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_wu_s(t0, cpu_env, cpu_fpr[rs1]);
-            break;
-#if defined(TARGET_RISCV64)
-        case 2: /* FCVT_L_S */
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_l_s(t0, cpu_env, cpu_fpr[rs1]);
-            break;
-        case 3: /* FCVT_LU_S */
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_lu_s(t0, cpu_env, cpu_fpr[rs1]);
-            break;
-#endif
-        default:
-            goto do_illegal;
-        }
-        gen_set_gpr(rd, t0);
-        tcg_temp_free(t0);
-        fp_output = false;
-        break;
-
-    case OPC_RISC_FCVT_S_W:
-        /* also OPC_RISC_FCVT_S_WU, OPC_RISC_FCVT_S_L, OPC_RISC_FCVT_S_LU */
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        t0 = tcg_temp_new();
-        gen_get_gpr(t0, rs1);
-        switch (rs2) {
-        case 0: /* FCVT_S_W */
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_s_w(cpu_fpr[rd], cpu_env, t0);
-            break;
-        case 1: /* FCVT_S_WU */
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_s_wu(cpu_fpr[rd], cpu_env, t0);
-            break;
-#if defined(TARGET_RISCV64)
-        case 2: /* FCVT_S_L */
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_s_l(cpu_fpr[rd], cpu_env, t0);
-            break;
-        case 3: /* FCVT_S_LU */
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_s_lu(cpu_fpr[rd], cpu_env, t0);
-            break;
-#endif
-        default:
-            goto do_illegal;
-        }
-        tcg_temp_free(t0);
-        break;
-
-    case OPC_RISC_FMV_X_S:
-        /* also OPC_RISC_FCLASS_S */
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        t0 = tcg_temp_new();
-        switch (rm) {
-        case 0: /* FMV */
-#if defined(TARGET_RISCV64)
-            tcg_gen_ext32s_tl(t0, cpu_fpr[rs1]);
-#else
-            tcg_gen_extrl_i64_i32(t0, cpu_fpr[rs1]);
-#endif
-            break;
-        case 1:
-            gen_helper_fclass_s(t0, cpu_fpr[rs1]);
-            break;
-        default:
-            goto do_illegal;
-        }
-        gen_set_gpr(rd, t0);
-        tcg_temp_free(t0);
-        fp_output = false;
-        break;
-
-    case OPC_RISC_FMV_S_X:
-        if (!has_ext(ctx, RVF)) {
-            goto do_illegal;
-        }
-        t0 = tcg_temp_new();
-        gen_get_gpr(t0, rs1);
-#if defined(TARGET_RISCV64)
-        tcg_gen_mov_i64(cpu_fpr[rd], t0);
-#else
-        tcg_gen_extu_i32_i64(cpu_fpr[rd], t0);
-#endif
-        tcg_temp_free(t0);
-        break;
-
-    /* double */
-    case OPC_RISC_FADD_D:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fadd_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-        break;
-    case OPC_RISC_FSUB_D:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fsub_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-        break;
-    case OPC_RISC_FMUL_D:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fmul_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-        break;
-    case OPC_RISC_FDIV_D:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fdiv_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-        break;
-    case OPC_RISC_FSQRT_D:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        gen_set_rm(ctx, rm);
-        gen_helper_fsqrt_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1]);
-        break;
-    case OPC_RISC_FSGNJ_D:
-        gen_fsgnj(ctx, rd, rs1, rs2, rm, INT64_MIN);
-        break;
-
-    case OPC_RISC_FMIN_D:
-        /* also OPC_RISC_FMAX_D */
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        switch (rm) {
-        case 0:
-            gen_helper_fmin_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-            break;
-        case 1:
-            gen_helper_fmax_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-            break;
-        default:
-            goto do_illegal;
-        }
-        break;
-
-    case OPC_RISC_FCVT_S_D:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        switch (rs2) {
-        case 1:
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_s_d(cpu_fpr[rd], cpu_env, cpu_fpr[rs1]);
-            break;
-        default:
-            goto do_illegal;
-        }
-        break;
-
-    case OPC_RISC_FCVT_D_S:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        switch (rs2) {
-        case 0:
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_d_s(cpu_fpr[rd], cpu_env, cpu_fpr[rs1]);
-            break;
-        default:
-            goto do_illegal;
-        }
-        break;
-
-    case OPC_RISC_FEQ_D:
-        /* also OPC_RISC_FLT_D, OPC_RISC_FLE_D */
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        t0 = tcg_temp_new();
-        switch (rm) {
-        case 0:
-            gen_helper_fle_d(t0, cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-            break;
-        case 1:
-            gen_helper_flt_d(t0, cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-            break;
-        case 2:
-            gen_helper_feq_d(t0, cpu_env, cpu_fpr[rs1], cpu_fpr[rs2]);
-            break;
-        default:
-            goto do_illegal;
-        }
-        gen_set_gpr(rd, t0);
-        tcg_temp_free(t0);
-        fp_output = false;
-        break;
-
-    case OPC_RISC_FCVT_W_D:
-        /* also OPC_RISC_FCVT_WU_D, OPC_RISC_FCVT_L_D, OPC_RISC_FCVT_LU_D */
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        t0 = tcg_temp_new();
-        switch (rs2) {
-        case 0:
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_w_d(t0, cpu_env, cpu_fpr[rs1]);
-            break;
-        case 1:
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_wu_d(t0, cpu_env, cpu_fpr[rs1]);
-            break;
-#if defined(TARGET_RISCV64)
-        case 2:
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_l_d(t0, cpu_env, cpu_fpr[rs1]);
-            break;
-        case 3:
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_lu_d(t0, cpu_env, cpu_fpr[rs1]);
-            break;
-#endif
-        default:
-            goto do_illegal;
-        }
-        gen_set_gpr(rd, t0);
-        tcg_temp_free(t0);
-        fp_output = false;
-        break;
-
-    case OPC_RISC_FCVT_D_W:
-        /* also OPC_RISC_FCVT_D_WU, OPC_RISC_FCVT_D_L, OPC_RISC_FCVT_D_LU */
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        t0 = tcg_temp_new();
-        gen_get_gpr(t0, rs1);
-        switch (rs2) {
-        case 0:
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_d_w(cpu_fpr[rd], cpu_env, t0);
-            break;
-        case 1:
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_d_wu(cpu_fpr[rd], cpu_env, t0);
-            break;
-#if defined(TARGET_RISCV64)
-        case 2:
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_d_l(cpu_fpr[rd], cpu_env, t0);
-            break;
-        case 3:
-            gen_set_rm(ctx, rm);
-            gen_helper_fcvt_d_lu(cpu_fpr[rd], cpu_env, t0);
-            break;
-#endif
-        default:
-            goto do_illegal;
-        }
-        tcg_temp_free(t0);
-        break;
-
-    case OPC_RISC_FMV_X_D:
-        /* also OPC_RISC_FCLASS_D */
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        switch (rm) {
-#if defined(TARGET_RISCV64)
-        case 0: /* FMV */
-            gen_set_gpr(rd, cpu_fpr[rs1]);
-            break;
-#endif
-        case 1:
-            t0 = tcg_temp_new();
-            gen_helper_fclass_d(t0, cpu_fpr[rs1]);
-            gen_set_gpr(rd, t0);
-            tcg_temp_free(t0);
-            break;
-        default:
-            goto do_illegal;
-        }
-        fp_output = false;
-        break;
-
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_FMV_D_X:
-        if (!has_ext(ctx, RVD)) {
-            goto do_illegal;
-        }
-        t0 = tcg_temp_new();
-        gen_get_gpr(t0, rs1);
-        tcg_gen_mov_tl(cpu_fpr[rd], t0);
-        tcg_temp_free(t0);
-        break;
-#endif
-
-    default:
-    do_illegal:
-        if (t0) {
-            tcg_temp_free(t0);
-        }
-        gen_exception_illegal(ctx);
-        return;
-    }
-
-    if (fp_output) {
-        mark_fs_dirty(ctx);
-    }
-}
-
 static void gen_system(DisasContext *ctx, uint32_t opc, int rd, int rs1,
                        int csr)
 {
@@ -1722,11 +1155,8 @@ bool decode_insn32(DisasContext *ctx, uint32_t insn);
 
 static void decode_RV32_64G(DisasContext *ctx)
 {
-    int rs1;
-    int rs2;
-    int rd;
+    int rs1, rd;
     uint32_t op;
-    target_long imm;
 
     /* We do not do misaligned address check here: the address should never be
      * misaligned at this point. Instructions that set PC must do the check,
@@ -1735,38 +1165,9 @@ static void decode_RV32_64G(DisasContext *ctx)
 
     op = MASK_OP_MAJOR(ctx->opcode);
     rs1 = GET_RS1(ctx->opcode);
-    rs2 = GET_RS2(ctx->opcode);
     rd = GET_RD(ctx->opcode);
-    imm = GET_IMM(ctx->opcode);
 
     switch (op) {
-    case OPC_RISC_FP_LOAD:
-        gen_fp_load(ctx, MASK_OP_FP_LOAD(ctx->opcode), rd, rs1, imm);
-        break;
-    case OPC_RISC_FP_STORE:
-        gen_fp_store(ctx, MASK_OP_FP_STORE(ctx->opcode), rs1, rs2,
-                     GET_STORE_IMM(ctx->opcode));
-        break;
-    case OPC_RISC_FMADD:
-        gen_fp_fmadd(ctx, MASK_OP_FP_FMADD(ctx->opcode), rd, rs1, rs2,
-                     GET_RS3(ctx->opcode), GET_RM(ctx->opcode));
-        break;
-    case OPC_RISC_FMSUB:
-        gen_fp_fmsub(ctx, MASK_OP_FP_FMSUB(ctx->opcode), rd, rs1, rs2,
-                     GET_RS3(ctx->opcode), GET_RM(ctx->opcode));
-        break;
-    case OPC_RISC_FNMSUB:
-        gen_fp_fnmsub(ctx, MASK_OP_FP_FNMSUB(ctx->opcode), rd, rs1, rs2,
-                      GET_RS3(ctx->opcode), GET_RM(ctx->opcode));
-        break;
-    case OPC_RISC_FNMADD:
-        gen_fp_fnmadd(ctx, MASK_OP_FP_FNMADD(ctx->opcode), rd, rs1, rs2,
-                      GET_RS3(ctx->opcode), GET_RM(ctx->opcode));
-        break;
-    case OPC_RISC_FP_ARITH:
-        gen_fp_arith(ctx, MASK_OP_FP_ARITH(ctx->opcode), rd, rs1, rs2,
-                     GET_RM(ctx->opcode));
-        break;
     case OPC_RISC_SYSTEM:
         gen_system(ctx, MASK_OP_SYSTEM(ctx->opcode), rd, rs1,
                    (ctx->opcode & 0xFFF00000) >> 20);
-- 
2.19.2

[Qemu-riscv] [PULL 15/29] target/riscv: Convert RV priv insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32.decode                    |  15 +++
 .../riscv/insn_trans/trans_privileged.inc.c   | 110 ++++++++++++++++++
 target/riscv/translate.c                      |  57 +--------
 3 files changed, 126 insertions(+), 56 deletions(-)
 create mode 100644 target/riscv/insn_trans/trans_privileged.inc.c

diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index e64b2b5e3458..ecc46a50cc27 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -57,6 +57,21 @@
 @r2_rm   .......   ..... ..... ... ..... ....... %rs1 %rm %rd
 @r2      .......   ..... ..... ... ..... ....... %rs1 %rd
 
+@sfence_vma ....... ..... .....   ... ..... ....... %rs2 %rs1
+@sfence_vm  ....... ..... .....   ... ..... ....... %rs1
+
+
+# *** Privileged Instructions ***
+ecall      000000000000     00000 000 00000 1110011
+ebreak     000000000001     00000 000 00000 1110011
+uret       0000000    00010 00000 000 00000 1110011
+sret       0001000    00010 00000 000 00000 1110011
+hret       0010000    00010 00000 000 00000 1110011
+mret       0011000    00010 00000 000 00000 1110011
+wfi        0001000    00101 00000 000 00000 1110011
+sfence_vma 0001001    ..... ..... 000 00000 1110011 @sfence_vma
+sfence_vm  0001000    00100 ..... 000 00000 1110011 @sfence_vm
+
 # *** RV32I Base Instruction Set ***
 lui      ....................       ..... 0110111 @u
 auipc    ....................       ..... 0010111 @u
diff --git a/target/riscv/insn_trans/trans_privileged.inc.c b/target/riscv/insn_trans/trans_privileged.inc.c
new file mode 100644
index 000000000000..acb605923e68
--- /dev/null
+++ b/target/riscv/insn_trans/trans_privileged.inc.c
@@ -0,0 +1,110 @@
+/*
+ * RISC-V translation routines for the RISC-V privileged instructions.
+ *
+ * Copyright (c) 2016-2017 Sagar Karandikar, sagark@eecs.berkeley.edu
+ * Copyright (c) 2018 Peer Adelt, peer.adelt@hni.uni-paderborn.de
+ *                    Bastian Koppelmann, kbastian@mail.uni-paderborn.de
+ *
+ * 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/>.
+ */
+
+static bool trans_ecall(DisasContext *ctx, arg_ecall *a)
+{
+    /* always generates U-level ECALL, fixed in do_interrupt handler */
+    generate_exception(ctx, RISCV_EXCP_U_ECALL);
+    tcg_gen_exit_tb(NULL, 0); /* no chaining */
+    ctx->base.is_jmp = DISAS_NORETURN;
+    return true;
+}
+
+static bool trans_ebreak(DisasContext *ctx, arg_ebreak *a)
+{
+    generate_exception(ctx, RISCV_EXCP_BREAKPOINT);
+    tcg_gen_exit_tb(NULL, 0); /* no chaining */
+    ctx->base.is_jmp = DISAS_NORETURN;
+    return true;
+}
+
+static bool trans_uret(DisasContext *ctx, arg_uret *a)
+{
+    return false;
+}
+
+static bool trans_sret(DisasContext *ctx, arg_sret *a)
+{
+#ifndef CONFIG_USER_ONLY
+    tcg_gen_movi_tl(cpu_pc, ctx->base.pc_next);
+
+    if (has_ext(ctx, RVS)) {
+        gen_helper_sret(cpu_pc, cpu_env, cpu_pc);
+        tcg_gen_exit_tb(NULL, 0); /* no chaining */
+        ctx->base.is_jmp = DISAS_NORETURN;
+    } else {
+        return false;
+    }
+    return true;
+#else
+    return false;
+#endif
+}
+
+static bool trans_hret(DisasContext *ctx, arg_hret *a)
+{
+    return false;
+}
+
+static bool trans_mret(DisasContext *ctx, arg_mret *a)
+{
+#ifndef CONFIG_USER_ONLY
+    tcg_gen_movi_tl(cpu_pc, ctx->base.pc_next);
+    gen_helper_mret(cpu_pc, cpu_env, cpu_pc);
+    tcg_gen_exit_tb(NULL, 0); /* no chaining */
+    ctx->base.is_jmp = DISAS_NORETURN;
+    return true;
+#else
+    return false;
+#endif
+}
+
+static bool trans_wfi(DisasContext *ctx, arg_wfi *a)
+{
+#ifndef CONFIG_USER_ONLY
+    tcg_gen_movi_tl(cpu_pc, ctx->pc_succ_insn);
+    gen_helper_wfi(cpu_env);
+    return true;
+#else
+    return false;
+#endif
+}
+
+static bool trans_sfence_vma(DisasContext *ctx, arg_sfence_vma *a)
+{
+#ifndef CONFIG_USER_ONLY
+    if (ctx->priv_ver == PRIV_VERSION_1_10_0) {
+        gen_helper_tlb_flush(cpu_env);
+        return true;
+    }
+#endif
+    return false;
+}
+
+static bool trans_sfence_vm(DisasContext *ctx, arg_sfence_vm *a)
+{
+#ifndef CONFIG_USER_ONLY
+    if (ctx->priv_ver <= PRIV_VERSION_1_09_1) {
+        gen_helper_tlb_flush(cpu_env);
+        return true;
+    }
+#endif
+    return false;
+}
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 2e36deee82f4..02f64ed8a734 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -772,26 +772,8 @@ static void gen_set_rm(DisasContext *ctx, int rm)
 static void gen_system(DisasContext *ctx, uint32_t opc, int rd, int rs1,
                        int csr)
 {
-    TCGv source1, dest;
-    source1 = tcg_temp_new();
-    dest = tcg_temp_new();
-    gen_get_gpr(source1, rs1);
     tcg_gen_movi_tl(cpu_pc, ctx->base.pc_next);
 
-#ifndef CONFIG_USER_ONLY
-    /* Extract funct7 value and check whether it matches SFENCE.VMA */
-    if ((opc == OPC_RISC_ECALL) && ((csr >> 5) == 9)) {
-        if (ctx->priv_ver == PRIV_VERSION_1_10_0) {
-            /* sfence.vma */
-            /* TODO: handle ASID specific fences */
-            gen_helper_tlb_flush(cpu_env);
-            return;
-        } else {
-            gen_exception_illegal(ctx);
-        }
-    }
-#endif
-
     switch (opc) {
     case OPC_RISC_ECALL:
         switch (csr) {
@@ -806,50 +788,12 @@ static void gen_system(DisasContext *ctx, uint32_t opc, int rd, int rs1,
             tcg_gen_exit_tb(NULL, 0); /* no chaining */
             ctx->base.is_jmp = DISAS_NORETURN;
             break;
-#ifndef CONFIG_USER_ONLY
-        case 0x002: /* URET */
-            gen_exception_illegal(ctx);
-            break;
-        case 0x102: /* SRET */
-            if (has_ext(ctx, RVS)) {
-                gen_helper_sret(cpu_pc, cpu_env, cpu_pc);
-                tcg_gen_exit_tb(NULL, 0); /* no chaining */
-                ctx->base.is_jmp = DISAS_NORETURN;
-            } else {
-                gen_exception_illegal(ctx);
-            }
-            break;
-        case 0x202: /* HRET */
-            gen_exception_illegal(ctx);
-            break;
-        case 0x302: /* MRET */
-            gen_helper_mret(cpu_pc, cpu_env, cpu_pc);
-            tcg_gen_exit_tb(NULL, 0); /* no chaining */
-            ctx->base.is_jmp = DISAS_NORETURN;
-            break;
-        case 0x7b2: /* DRET */
-            gen_exception_illegal(ctx);
-            break;
-        case 0x105: /* WFI */
-            tcg_gen_movi_tl(cpu_pc, ctx->pc_succ_insn);
-            gen_helper_wfi(cpu_env);
-            break;
-        case 0x104: /* SFENCE.VM */
-            if (ctx->priv_ver <= PRIV_VERSION_1_09_1) {
-                gen_helper_tlb_flush(cpu_env);
-            } else {
-                gen_exception_illegal(ctx);
-            }
-            break;
-#endif
         default:
             gen_exception_illegal(ctx);
             break;
         }
         break;
     }
-    tcg_temp_free(source1);
-    tcg_temp_free(dest);
 }
 
 static void decode_RV32_64C0(DisasContext *ctx)
@@ -1152,6 +1096,7 @@ bool decode_insn32(DisasContext *ctx, uint32_t insn);
 #include "insn_trans/trans_rva.inc.c"
 #include "insn_trans/trans_rvf.inc.c"
 #include "insn_trans/trans_rvd.inc.c"
+#include "insn_trans/trans_privileged.inc.c"
 
 static void decode_RV32_64G(DisasContext *ctx)
 {
-- 
2.19.2

[Qemu-riscv] [PULL 16/29] target/riscv: Convert quadrant 0 of RVXC insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/Makefile.objs              |  9 ++-
 target/riscv/insn16.decode              | 55 ++++++++++++++++++
 target/riscv/insn_trans/trans_rvc.inc.c | 75 +++++++++++++++++++++++++
 target/riscv/translate.c                | 53 ++++++-----------
 4 files changed, 154 insertions(+), 38 deletions(-)
 create mode 100644 target/riscv/insn16.decode
 create mode 100644 target/riscv/insn_trans/trans_rvc.inc.c

diff --git a/target/riscv/Makefile.objs b/target/riscv/Makefile.objs
index 05087a91bb85..9c6c1093271e 100644
--- a/target/riscv/Makefile.objs
+++ b/target/riscv/Makefile.objs
@@ -10,4 +10,11 @@ target/riscv/decode_insn32.inc.c: $(decode32-y) $(DECODETREE)
 	  $(PYTHON) $(DECODETREE) -o $@ --decode decode_insn32 $(decode32-y), \
 	  "GEN", $(TARGET_DIR)$@)
 
-target/riscv/translate.o: target/riscv/decode_insn32.inc.c
+target/riscv/decode_insn16.inc.c: \
+  $(SRC_PATH)/target/riscv/insn16.decode $(DECODETREE)
+	$(call quiet-command, \
+	  $(PYTHON) $(DECODETREE) -o $@ --decode decode_insn16 --insnwidth 16 $<, \
+	  "GEN", $(TARGET_DIR)$@)
+
+target/riscv/translate.o: target/riscv/decode_insn32.inc.c \
+	target/riscv/decode_insn16.inc.c
diff --git a/target/riscv/insn16.decode b/target/riscv/insn16.decode
new file mode 100644
index 000000000000..558c0c41f0b5
--- /dev/null
+++ b/target/riscv/insn16.decode
@@ -0,0 +1,55 @@
+#
+# RISC-V translation routines for the RVXI Base Integer Instruction Set.
+#
+# Copyright (c) 2018 Peer Adelt, peer.adelt@hni.uni-paderborn.de
+#                    Bastian Koppelmann, kbastian@mail.uni-paderborn.de
+#
+# 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/>.
+
+# Fields:
+%rd        7:5
+%rs1_3     7:3                !function=ex_rvc_register
+%rs2_3     2:3                !function=ex_rvc_register
+
+# Immediates:
+%nzuimm_ciw    7:4 11:2 5:1 6:1   !function=ex_shift_2
+%uimm_cl_d     5:2 10:3           !function=ex_shift_3
+%uimm_cl_w     5:1 10:3 6:1       !function=ex_shift_2
+
+
+# Argument sets:
+&cl               rs1 rd
+&cl_dw     uimm   rs1 rd
+&ciw       nzuimm     rd
+&cs               rs1 rs2
+&cs_dw     uimm   rs1 rs2
+
+
+# Formats 16:
+@ciw       ...   ........ ... .. &ciw    nzuimm=%nzuimm_ciw           rd=%rs2_3
+@cl_d      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rd=%rs2_3
+@cl_w      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rd=%rs2_3
+@cl        ... ... ... .. ... .. &cl                      rs1=%rs1_3  rd=%rs2_3
+@cs        ... ... ... .. ... .. &cs                      rs1=%rs1_3  rs2=%rs2_3
+@cs_d      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rs2=%rs2_3
+@cs_w      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rs2=%rs2_3
+
+
+# *** RV64C Standard Extension (Quadrant 0) ***
+c_addi4spn        000    ........ ... 00 @ciw
+c_fld             001  ... ... .. ... 00 @cl_d
+c_lw              010  ... ... .. ... 00 @cl_w
+c_flw_ld          011  --- ... -- ... 00 @cl    #Note: Must parse uimm manually
+c_fsd             101  ... ... .. ... 00 @cs_d
+c_sw              110  ... ... .. ... 00 @cs_w
+c_fsw_sd          111  --- ... -- ... 00 @cs    #Note: Must parse uimm manually
diff --git a/target/riscv/insn_trans/trans_rvc.inc.c b/target/riscv/insn_trans/trans_rvc.inc.c
new file mode 100644
index 000000000000..93ec8aa30b95
--- /dev/null
+++ b/target/riscv/insn_trans/trans_rvc.inc.c
@@ -0,0 +1,75 @@
+/*
+ * RISC-V translation routines for the RVC Compressed Instruction Set.
+ *
+ * Copyright (c) 2016-2017 Sagar Karandikar, sagark@eecs.berkeley.edu
+ * Copyright (c) 2018 Peer Adelt, peer.adelt@hni.uni-paderborn.de
+ *                    Bastian Koppelmann, kbastian@mail.uni-paderborn.de
+ *
+ * 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/>.
+ */
+
+static bool trans_c_addi4spn(DisasContext *ctx, arg_c_addi4spn *a)
+{
+    if (a->nzuimm == 0) {
+        /* Reserved in ISA */
+        return false;
+    }
+    arg_addi arg = { .rd = a->rd, .rs1 = 2, .imm = a->nzuimm };
+    return trans_addi(ctx, &arg);
+}
+
+static bool trans_c_fld(DisasContext *ctx, arg_c_fld *a)
+{
+    arg_fld arg = { .rd = a->rd, .rs1 = a->rs1, .imm = a->uimm };
+    return trans_fld(ctx, &arg);
+}
+
+static bool trans_c_lw(DisasContext *ctx, arg_c_lw *a)
+{
+    arg_lw arg = { .rd = a->rd, .rs1 = a->rs1, .imm = a->uimm };
+    return trans_lw(ctx, &arg);
+}
+
+static bool trans_c_flw_ld(DisasContext *ctx, arg_c_flw_ld *a)
+{
+#ifdef TARGET_RISCV32
+    /* C.FLW ( RV32FC-only ) */
+    return false;
+#else
+    /* C.LD ( RV64C/RV128C-only ) */
+    return false;
+#endif
+}
+
+static bool trans_c_fsd(DisasContext *ctx, arg_c_fsd *a)
+{
+    arg_fsd arg = { .rs1 = a->rs1, .rs2 = a->rs2, .imm = a->uimm };
+    return trans_fsd(ctx, &arg);
+}
+
+static bool trans_c_sw(DisasContext *ctx, arg_c_sw *a)
+{
+    arg_sw arg = { .rs1 = a->rs1, .rs2 = a->rs2, .imm = a->uimm };
+    return trans_sw(ctx, &arg);
+}
+
+static bool trans_c_fsw_sd(DisasContext *ctx, arg_c_fsw_sd *a)
+{
+#ifdef TARGET_RISCV32
+    /* C.FSW ( RV32FC-only ) */
+    return false;
+#else
+    /* C.SD ( RV64C/RV128C-only ) */
+    return false;
+#endif
+}
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 02f64ed8a734..0106fa8d51b3 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -803,27 +803,6 @@ static void decode_RV32_64C0(DisasContext *ctx)
     uint8_t rs1s = GET_C_RS1S(ctx->opcode);
 
     switch (funct3) {
-    case 0:
-        /* illegal */
-        if (ctx->opcode == 0) {
-            gen_exception_illegal(ctx);
-        } else {
-            /* C.ADDI4SPN -> addi rd', x2, zimm[9:2]*/
-            gen_arith_imm(ctx, OPC_RISC_ADDI, rd_rs2, 2,
-                          GET_C_ADDI4SPN_IMM(ctx->opcode));
-        }
-        break;
-    case 1:
-        /* C.FLD -> fld rd', offset[7:3](rs1')*/
-        gen_fp_load(ctx, OPC_RISC_FLD, rd_rs2, rs1s,
-                    GET_C_LD_IMM(ctx->opcode));
-        /* C.LQ(RV128) */
-        break;
-    case 2:
-        /* C.LW -> lw rd', offset[6:2](rs1') */
-        gen_load(ctx, OPC_RISC_LW, rd_rs2, rs1s,
-                 GET_C_LW_IMM(ctx->opcode));
-        break;
     case 3:
 #if defined(TARGET_RISCV64)
         /* C.LD(RV64/128) -> ld rd', offset[7:3](rs1')*/
@@ -835,21 +814,6 @@ static void decode_RV32_64C0(DisasContext *ctx)
                     GET_C_LW_IMM(ctx->opcode));
 #endif
         break;
-    case 4:
-        /* reserved */
-        gen_exception_illegal(ctx);
-        break;
-    case 5:
-        /* C.FSD(RV32/64) -> fsd rs2', offset[7:3](rs1') */
-        gen_fp_store(ctx, OPC_RISC_FSD, rs1s, rd_rs2,
-                     GET_C_LD_IMM(ctx->opcode));
-        /* C.SQ (RV128) */
-        break;
-    case 6:
-        /* C.SW -> sw rs2', offset[6:2](rs1')*/
-        gen_store(ctx, OPC_RISC_SW, rs1s, rd_rs2,
-                  GET_C_LW_IMM(ctx->opcode));
-        break;
     case 7:
 #if defined(TARGET_RISCV64)
         /* C.SD (RV64/128) -> sd rs2', offset[7:3](rs1')*/
@@ -1079,6 +1043,8 @@ static void decode_RV32_64C(DisasContext *ctx)
         return imm << amount;                 \
     }
 EX_SH(1)
+EX_SH(2)
+EX_SH(3)
 EX_SH(12)
 
 #define REQUIRE_EXT(ctx, ext) do { \
@@ -1087,6 +1053,11 @@ EX_SH(12)
     }                              \
 } while (0)
 
+static int ex_rvc_register(int reg)
+{
+    return 8 + reg;
+}
+
 bool decode_insn32(DisasContext *ctx, uint32_t insn);
 /* Include the auto-generated decoder for 32 bit insn */
 #include "decode_insn32.inc.c"
@@ -1098,6 +1069,11 @@ bool decode_insn32(DisasContext *ctx, uint32_t insn);
 #include "insn_trans/trans_rvd.inc.c"
 #include "insn_trans/trans_privileged.inc.c"
 
+bool decode_insn16(DisasContext *ctx, uint16_t insn);
+/* auto-generated decoder*/
+#include "decode_insn16.inc.c"
+#include "insn_trans/trans_rvc.inc.c"
+
 static void decode_RV32_64G(DisasContext *ctx)
 {
     int rs1, rd;
@@ -1131,7 +1107,10 @@ static void decode_opc(DisasContext *ctx)
             gen_exception_illegal(ctx);
         } else {
             ctx->pc_succ_insn = ctx->base.pc_next + 2;
-            decode_RV32_64C(ctx);
+            if (!decode_insn16(ctx, ctx->opcode)) {
+                /* fall back to old decoder */
+                decode_RV32_64C(ctx);
+            }
         }
     } else {
         ctx->pc_succ_insn = ctx->base.pc_next + 4;
-- 
2.19.2

[Qemu-riscv] [PULL 17/29] target/riscv: Convert quadrant 1 of RVXC insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn16.decode              |  43 +++++++
 target/riscv/insn_trans/trans_rvc.inc.c | 151 ++++++++++++++++++++++++
 target/riscv/translate.c                | 118 +-----------------
 3 files changed, 195 insertions(+), 117 deletions(-)

diff --git a/target/riscv/insn16.decode b/target/riscv/insn16.decode
index 558c0c41f0b5..45109301c6d4 100644
--- a/target/riscv/insn16.decode
+++ b/target/riscv/insn16.decode
@@ -22,28 +22,53 @@
 %rs2_3     2:3                !function=ex_rvc_register
 
 # Immediates:
+%imm_ci        12:s1 2:5
 %nzuimm_ciw    7:4 11:2 5:1 6:1   !function=ex_shift_2
 %uimm_cl_d     5:2 10:3           !function=ex_shift_3
 %uimm_cl_w     5:1 10:3 6:1       !function=ex_shift_2
+%imm_cb        12:s1 5:2 2:1 10:2 3:2 !function=ex_shift_1
+%imm_cj        12:s1 8:1 9:2 6:1 7:1 2:1 11:1 3:3 !function=ex_shift_1
+
+%nzuimm_6bit   12:1 2:5
+
+%imm_addi16sp  12:s1 3:2 5:1 2:1 6:1 !function=ex_shift_4
+%imm_lui       12:s1 2:5             !function=ex_shift_12
+
 
 
 # Argument sets:
 &cl               rs1 rd
 &cl_dw     uimm   rs1 rd
+&ci        imm        rd
 &ciw       nzuimm     rd
 &cs               rs1 rs2
 &cs_dw     uimm   rs1 rs2
+&cb        imm    rs1
+&cr               rd  rs2
+&cj       imm
+&c_shift   shamt      rd
+
 
+&caddi16sp_lui  imm_lui imm_addi16sp rd
 
 # Formats 16:
+@ci        ... . ..... .....  .. &ci     imm=%imm_ci                  %rd
 @ciw       ...   ........ ... .. &ciw    nzuimm=%nzuimm_ciw           rd=%rs2_3
 @cl_d      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rd=%rs2_3
 @cl_w      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rd=%rs2_3
 @cl        ... ... ... .. ... .. &cl                      rs1=%rs1_3  rd=%rs2_3
 @cs        ... ... ... .. ... .. &cs                      rs1=%rs1_3  rs2=%rs2_3
+@cs_2      ... ... ... .. ... .. &cr                      rd=%rs1_3   rs2=%rs2_3
 @cs_d      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rs2=%rs2_3
 @cs_w      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rs2=%rs2_3
+@cb        ... ... ... .. ... .. &cb     imm=%imm_cb      rs1=%rs1_3
+@cj        ...    ........... .. &cj     imm=%imm_cj
 
+@c_addi16sp_lui ... .  ..... ..... .. &caddi16sp_lui %imm_lui %imm_addi16sp %rd
+
+@c_shift        ... . .. ... ..... .. &c_shift rd=%rs1_3 shamt=%nzuimm_6bit
+
+@c_andi         ... . .. ... ..... .. &ci imm=%imm_ci rd=%rs1_3
 
 # *** RV64C Standard Extension (Quadrant 0) ***
 c_addi4spn        000    ........ ... 00 @ciw
@@ -53,3 +78,21 @@ c_flw_ld          011  --- ... -- ... 00 @cl    #Note: Must parse uimm manually
 c_fsd             101  ... ... .. ... 00 @cs_d
 c_sw              110  ... ... .. ... 00 @cs_w
 c_fsw_sd          111  --- ... -- ... 00 @cs    #Note: Must parse uimm manually
+
+# *** RV64C Standard Extension (Quadrant 1) ***
+c_addi            000 .  .....  ..... 01 @ci
+c_jal_addiw       001 .  .....  ..... 01 @ci #Note: parse rd and/or imm manually
+c_li              010 .  .....  ..... 01 @ci
+c_addi16sp_lui    011 .  .....  ..... 01 @c_addi16sp_lui # shares opc with C.LUI
+c_srli            100 . 00 ...  ..... 01 @c_shift
+c_srai            100 . 01 ...  ..... 01 @c_shift
+c_andi            100 . 10 ...  ..... 01 @c_andi
+c_sub             100 0 11 ... 00 ... 01 @cs_2
+c_xor             100 0 11 ... 01 ... 01 @cs_2
+c_or              100 0 11 ... 10 ... 01 @cs_2
+c_and             100 0 11 ... 11 ... 01 @cs_2
+c_subw            100 1 11 ... 00 ... 01 @cs_2
+c_addw            100 1 11 ... 01 ... 01 @cs_2
+c_j               101     ........... 01 @cj
+c_beqz            110  ... ...  ..... 01 @cb
+c_bnez            111  ... ...  ..... 01 @cb
diff --git a/target/riscv/insn_trans/trans_rvc.inc.c b/target/riscv/insn_trans/trans_rvc.inc.c
index 93ec8aa30b95..b06c435c9800 100644
--- a/target/riscv/insn_trans/trans_rvc.inc.c
+++ b/target/riscv/insn_trans/trans_rvc.inc.c
@@ -73,3 +73,154 @@ static bool trans_c_fsw_sd(DisasContext *ctx, arg_c_fsw_sd *a)
     return false;
 #endif
 }
+
+static bool trans_c_addi(DisasContext *ctx, arg_c_addi *a)
+{
+    if (a->imm == 0) {
+        /* Hint: insn is valid but does not affect state */
+        return true;
+    }
+    arg_addi arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
+    return trans_addi(ctx, &arg);
+}
+
+static bool trans_c_jal_addiw(DisasContext *ctx, arg_c_jal_addiw *a)
+{
+#ifdef TARGET_RISCV32
+    /* C.JAL */
+    arg_jal arg = { .rd = 1, .imm = a->imm };
+    return trans_jal(ctx, &arg);
+#else
+    /* C.ADDIW */
+    arg_addiw arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
+    return trans_addiw(ctx, &arg);
+#endif
+}
+
+static bool trans_c_li(DisasContext *ctx, arg_c_li *a)
+{
+    if (a->rd == 0) {
+        /* Hint: insn is valid but does not affect state */
+        return true;
+    }
+    arg_addi arg = { .rd = a->rd, .rs1 = 0, .imm = a->imm };
+    return trans_addi(ctx, &arg);
+}
+
+static bool trans_c_addi16sp_lui(DisasContext *ctx, arg_c_addi16sp_lui *a)
+{
+    if (a->rd == 2) {
+        /* C.ADDI16SP */
+        arg_addi arg = { .rd = 2, .rs1 = 2, .imm = a->imm_addi16sp };
+        return trans_addi(ctx, &arg);
+    } else if (a->imm_lui != 0) {
+        /* C.LUI */
+        if (a->rd == 0) {
+            /* Hint: insn is valid but does not affect state */
+            return true;
+        }
+        arg_lui arg = { .rd = a->rd, .imm = a->imm_lui };
+        return trans_lui(ctx, &arg);
+    }
+    return false;
+}
+
+static bool trans_c_srli(DisasContext *ctx, arg_c_srli *a)
+{
+    int shamt = a->shamt;
+    if (shamt == 0) {
+        /* For RV128 a shamt of 0 means a shift by 64 */
+        shamt = 64;
+    }
+    /* Ensure, that shamt[5] is zero for RV32 */
+    if (shamt >= TARGET_LONG_BITS) {
+        return false;
+    }
+
+    arg_srli arg = { .rd = a->rd, .rs1 = a->rd, .shamt = a->shamt };
+    return trans_srli(ctx, &arg);
+}
+
+static bool trans_c_srai(DisasContext *ctx, arg_c_srai *a)
+{
+    int shamt = a->shamt;
+    if (shamt == 0) {
+        /* For RV128 a shamt of 0 means a shift by 64 */
+        shamt = 64;
+    }
+    /* Ensure, that shamt[5] is zero for RV32 */
+    if (shamt >= TARGET_LONG_BITS) {
+        return false;
+    }
+
+    arg_srai arg = { .rd = a->rd, .rs1 = a->rd, .shamt = a->shamt };
+    return trans_srai(ctx, &arg);
+}
+
+static bool trans_c_andi(DisasContext *ctx, arg_c_andi *a)
+{
+    arg_andi arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
+    return trans_andi(ctx, &arg);
+}
+
+static bool trans_c_sub(DisasContext *ctx, arg_c_sub *a)
+{
+    arg_sub arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
+    return trans_sub(ctx, &arg);
+}
+
+static bool trans_c_xor(DisasContext *ctx, arg_c_xor *a)
+{
+    arg_xor arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
+    return trans_xor(ctx, &arg);
+}
+
+static bool trans_c_or(DisasContext *ctx, arg_c_or *a)
+{
+    arg_or arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
+    return trans_or(ctx, &arg);
+}
+
+static bool trans_c_and(DisasContext *ctx, arg_c_and *a)
+{
+    arg_and arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
+    return trans_and(ctx, &arg);
+}
+
+static bool trans_c_subw(DisasContext *ctx, arg_c_subw *a)
+{
+#ifdef TARGET_RISCV64
+    arg_subw arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
+    return trans_subw(ctx, &arg);
+#else
+    return false;
+#endif
+}
+
+static bool trans_c_addw(DisasContext *ctx, arg_c_addw *a)
+{
+#ifdef TARGET_RISCV64
+    arg_addw arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
+    return trans_addw(ctx, &arg);
+#else
+    return false;
+#endif
+}
+
+static bool trans_c_j(DisasContext *ctx, arg_c_j *a)
+{
+    arg_jal arg = { .rd = 0, .imm = a->imm };
+    return trans_jal(ctx, &arg);
+}
+
+static bool trans_c_beqz(DisasContext *ctx, arg_c_beqz *a)
+{
+    arg_beq arg = { .rs1 = a->rs1, .rs2 = 0, .imm = a->imm };
+    return trans_beq(ctx, &arg);
+}
+
+static bool trans_c_bnez(DisasContext *ctx, arg_c_bnez *a)
+{
+    arg_bne arg = { .rs1 = a->rs1, .rs2 = 0, .imm = a->imm };
+    return trans_bne(ctx, &arg);
+}
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 0106fa8d51b3..a584c24fbf6b 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -828,120 +828,6 @@ static void decode_RV32_64C0(DisasContext *ctx)
     }
 }
 
-static void decode_RV32_64C1(DisasContext *ctx)
-{
-    uint8_t funct3 = extract32(ctx->opcode, 13, 3);
-    uint8_t rd_rs1 = GET_C_RS1(ctx->opcode);
-    uint8_t rs1s, rs2s;
-    uint8_t funct2;
-
-    switch (funct3) {
-    case 0:
-        /* C.ADDI -> addi rd, rd, nzimm[5:0] */
-        gen_arith_imm(ctx, OPC_RISC_ADDI, rd_rs1, rd_rs1,
-                      GET_C_IMM(ctx->opcode));
-        break;
-    case 1:
-#if defined(TARGET_RISCV64)
-        /* C.ADDIW (RV64/128) -> addiw rd, rd, imm[5:0]*/
-        gen_arith_imm(ctx, OPC_RISC_ADDIW, rd_rs1, rd_rs1,
-                      GET_C_IMM(ctx->opcode));
-#else
-        /* C.JAL(RV32) -> jal x1, offset[11:1] */
-        gen_jal(ctx, 1, GET_C_J_IMM(ctx->opcode));
-#endif
-        break;
-    case 2:
-        /* C.LI -> addi rd, x0, imm[5:0]*/
-        gen_arith_imm(ctx, OPC_RISC_ADDI, rd_rs1, 0, GET_C_IMM(ctx->opcode));
-        break;
-    case 3:
-        if (rd_rs1 == 2) {
-            /* C.ADDI16SP -> addi x2, x2, nzimm[9:4]*/
-            gen_arith_imm(ctx, OPC_RISC_ADDI, 2, 2,
-                          GET_C_ADDI16SP_IMM(ctx->opcode));
-        } else if (rd_rs1 != 0) {
-            /* C.LUI (rs1/rd =/= {0,2}) -> lui rd, nzimm[17:12]*/
-            tcg_gen_movi_tl(cpu_gpr[rd_rs1],
-                            GET_C_IMM(ctx->opcode) << 12);
-        }
-        break;
-    case 4:
-        funct2 = extract32(ctx->opcode, 10, 2);
-        rs1s = GET_C_RS1S(ctx->opcode);
-        switch (funct2) {
-        case 0: /* C.SRLI(RV32) -> srli rd', rd', shamt[5:0] */
-            gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, rs1s, rs1s,
-                               GET_C_ZIMM(ctx->opcode));
-            /* C.SRLI64(RV128) */
-            break;
-        case 1:
-            /* C.SRAI -> srai rd', rd', shamt[5:0]*/
-            gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, rs1s, rs1s,
-                            GET_C_ZIMM(ctx->opcode) | 0x400);
-            /* C.SRAI64(RV128) */
-            break;
-        case 2:
-            /* C.ANDI -> andi rd', rd', imm[5:0]*/
-            gen_arith_imm(ctx, OPC_RISC_ANDI, rs1s, rs1s,
-                          GET_C_IMM(ctx->opcode));
-            break;
-        case 3:
-            funct2 = extract32(ctx->opcode, 5, 2);
-            rs2s = GET_C_RS2S(ctx->opcode);
-            switch (funct2) {
-            case 0:
-                /* C.SUB -> sub rd', rd', rs2' */
-                if (extract32(ctx->opcode, 12, 1) == 0) {
-                    gen_arith(ctx, OPC_RISC_SUB, rs1s, rs1s, rs2s);
-                }
-#if defined(TARGET_RISCV64)
-                else {
-                    gen_arith(ctx, OPC_RISC_SUBW, rs1s, rs1s, rs2s);
-                }
-#endif
-                break;
-            case 1:
-                /* C.XOR -> xor rs1', rs1', rs2' */
-                if (extract32(ctx->opcode, 12, 1) == 0) {
-                    gen_arith(ctx, OPC_RISC_XOR, rs1s, rs1s, rs2s);
-                }
-#if defined(TARGET_RISCV64)
-                else {
-                    /* C.ADDW (RV64/128) */
-                    gen_arith(ctx, OPC_RISC_ADDW, rs1s, rs1s, rs2s);
-                }
-#endif
-                break;
-            case 2:
-                /* C.OR -> or rs1', rs1', rs2' */
-                gen_arith(ctx, OPC_RISC_OR, rs1s, rs1s, rs2s);
-                break;
-            case 3:
-                /* C.AND -> and rs1', rs1', rs2' */
-                gen_arith(ctx, OPC_RISC_AND, rs1s, rs1s, rs2s);
-                break;
-            }
-            break;
-        }
-        break;
-    case 5:
-        /* C.J -> jal x0, offset[11:1]*/
-        gen_jal(ctx, 0, GET_C_J_IMM(ctx->opcode));
-        break;
-    case 6:
-        /* C.BEQZ -> beq rs1', x0, offset[8:1]*/
-        rs1s = GET_C_RS1S(ctx->opcode);
-        gen_branch(ctx, OPC_RISC_BEQ, rs1s, 0, GET_C_B_IMM(ctx->opcode));
-        break;
-    case 7:
-        /* C.BNEZ -> bne rs1', x0, offset[8:1]*/
-        rs1s = GET_C_RS1S(ctx->opcode);
-        gen_branch(ctx, OPC_RISC_BNE, rs1s, 0, GET_C_B_IMM(ctx->opcode));
-        break;
-    }
-}
-
 static void decode_RV32_64C2(DisasContext *ctx)
 {
     uint8_t rd, rs2;
@@ -1028,9 +914,6 @@ static void decode_RV32_64C(DisasContext *ctx)
     case 0:
         decode_RV32_64C0(ctx);
         break;
-    case 1:
-        decode_RV32_64C1(ctx);
-        break;
     case 2:
         decode_RV32_64C2(ctx);
         break;
@@ -1045,6 +928,7 @@ static void decode_RV32_64C(DisasContext *ctx)
 EX_SH(1)
 EX_SH(2)
 EX_SH(3)
+EX_SH(4)
 EX_SH(12)
 
 #define REQUIRE_EXT(ctx, ext) do { \
-- 
2.19.2

[Qemu-riscv] [PULL 18/29] target/riscv: Convert quadrant 2 of RVXC insns to decodetree

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn16.decode              |  31 ++++++++
 target/riscv/insn_trans/trans_rvc.inc.c | 101 ++++++++++++++++++++++++
 target/riscv/translate.c                |  83 +------------------
 3 files changed, 134 insertions(+), 81 deletions(-)

diff --git a/target/riscv/insn16.decode b/target/riscv/insn16.decode
index 45109301c6d4..17cc52cf2ac2 100644
--- a/target/riscv/insn16.decode
+++ b/target/riscv/insn16.decode
@@ -20,6 +20,7 @@
 %rd        7:5
 %rs1_3     7:3                !function=ex_rvc_register
 %rs2_3     2:3                !function=ex_rvc_register
+%rs2_5     2:5
 
 # Immediates:
 %imm_ci        12:s1 2:5
@@ -30,6 +31,10 @@
 %imm_cj        12:s1 8:1 9:2 6:1 7:1 2:1 11:1 3:3 !function=ex_shift_1
 
 %nzuimm_6bit   12:1 2:5
+%uimm_6bit_ld 2:3 12:1 5:2           !function=ex_shift_3
+%uimm_6bit_lw 2:2 12:1 4:3           !function=ex_shift_2
+%uimm_6bit_sd 7:3 10:3               !function=ex_shift_3
+%uimm_6bit_sw 7:2 9:4                !function=ex_shift_2
 
 %imm_addi16sp  12:s1 3:2 5:1 2:1 6:1 !function=ex_shift_4
 %imm_lui       12:s1 2:5             !function=ex_shift_12
@@ -48,10 +53,15 @@
 &cj       imm
 &c_shift   shamt      rd
 
+&c_ld      uimm  rd
+&c_sd      uimm  rs2
 
 &caddi16sp_lui  imm_lui imm_addi16sp rd
+&cflwsp_ldsp    uimm_flwsp uimm_ldsp rd
+&cfswsp_sdsp    uimm_fswsp uimm_sdsp rs2
 
 # Formats 16:
+@cr        ....  ..... .....  .. &cr                      rs2=%rs2_5  %rd
 @ci        ... . ..... .....  .. &ci     imm=%imm_ci                  %rd
 @ciw       ...   ........ ... .. &ciw    nzuimm=%nzuimm_ciw           rd=%rs2_3
 @cl_d      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rd=%rs2_3
@@ -64,9 +74,19 @@
 @cb        ... ... ... .. ... .. &cb     imm=%imm_cb      rs1=%rs1_3
 @cj        ...    ........... .. &cj     imm=%imm_cj
 
+@c_ld      ... . .....  ..... .. &c_ld     uimm=%uimm_6bit_ld  %rd
+@c_lw      ... . .....  ..... .. &c_ld     uimm=%uimm_6bit_lw  %rd
+@c_sd      ... . .....  ..... .. &c_sd     uimm=%uimm_6bit_sd  rs2=%rs2_5
+@c_sw      ... . .....  ..... .. &c_sd     uimm=%uimm_6bit_sw  rs2=%rs2_5
+
 @c_addi16sp_lui ... .  ..... ..... .. &caddi16sp_lui %imm_lui %imm_addi16sp %rd
+@c_flwsp_ldsp   ... .  ..... ..... .. &cflwsp_ldsp uimm_flwsp=%uimm_6bit_lw \
+    uimm_ldsp=%uimm_6bit_ld %rd
+@c_fswsp_sdsp   ... .  ..... ..... .. &cfswsp_sdsp uimm_fswsp=%uimm_6bit_sw \
+    uimm_sdsp=%uimm_6bit_sd rs2=%rs2_5
 
 @c_shift        ... . .. ... ..... .. &c_shift rd=%rs1_3 shamt=%nzuimm_6bit
+@c_shift2       ... . .. ... ..... .. &c_shift rd=%rd    shamt=%nzuimm_6bit
 
 @c_andi         ... . .. ... ..... .. &ci imm=%imm_ci rd=%rs1_3
 
@@ -96,3 +116,14 @@ c_addw            100 1 11 ... 01 ... 01 @cs_2
 c_j               101     ........... 01 @cj
 c_beqz            110  ... ...  ..... 01 @cb
 c_bnez            111  ... ...  ..... 01 @cb
+
+# *** RV64C Standard Extension (Quadrant 2) ***
+c_slli            000 .  .....  ..... 10 @c_shift2
+c_fldsp           001 .  .....  ..... 10 @c_ld
+c_lwsp            010 .  .....  ..... 10 @c_lw
+c_flwsp_ldsp      011 .  .....  ..... 10 @c_flwsp_ldsp #C.LDSP:RV64;C.FLWSP:RV32
+c_jr_mv           100 0  .....  ..... 10 @cr
+c_ebreak_jalr_add 100 1  .....  ..... 10 @cr
+c_fsdsp           101   ......  ..... 10 @c_sd
+c_swsp            110 .  .....  ..... 10 @c_sw
+c_fswsp_sdsp      111 .  .....  ..... 10 @c_fswsp_sdsp #C.SDSP:RV64;C.FSWSP:RV32
diff --git a/target/riscv/insn_trans/trans_rvc.inc.c b/target/riscv/insn_trans/trans_rvc.inc.c
index b06c435c9800..bcdf64d3b705 100644
--- a/target/riscv/insn_trans/trans_rvc.inc.c
+++ b/target/riscv/insn_trans/trans_rvc.inc.c
@@ -224,3 +224,104 @@ static bool trans_c_bnez(DisasContext *ctx, arg_c_bnez *a)
     arg_bne arg = { .rs1 = a->rs1, .rs2 = 0, .imm = a->imm };
     return trans_bne(ctx, &arg);
 }
+
+static bool trans_c_slli(DisasContext *ctx, arg_c_slli *a)
+{
+    int shamt = a->shamt;
+    if (shamt == 0) {
+        /* For RV128 a shamt of 0 means a shift by 64 */
+        shamt = 64;
+    }
+    /* Ensure, that shamt[5] is zero for RV32 */
+    if (shamt >= TARGET_LONG_BITS) {
+        return false;
+    }
+
+    arg_slli arg = { .rd = a->rd, .rs1 = a->rd, .shamt = a->shamt };
+    return trans_slli(ctx, &arg);
+}
+
+static bool trans_c_fldsp(DisasContext *ctx, arg_c_fldsp *a)
+{
+    arg_fld arg = { .rd = a->rd, .rs1 = 2, .imm = a->uimm };
+    return trans_fld(ctx, &arg);
+}
+
+static bool trans_c_lwsp(DisasContext *ctx, arg_c_lwsp *a)
+{
+    arg_lw arg = { .rd = a->rd, .rs1 = 2, .imm = a->uimm };
+    return trans_lw(ctx, &arg);
+}
+
+static bool trans_c_flwsp_ldsp(DisasContext *ctx, arg_c_flwsp_ldsp *a)
+{
+#ifdef TARGET_RISCV32
+    /* C.FLWSP */
+    arg_flw arg_flw = { .rd = a->rd, .rs1 = 2, .imm = a->uimm_flwsp };
+    return trans_flw(ctx, &arg_flw);
+#else
+    /* C.LDSP */
+    arg_ld arg_ld = { .rd = a->rd, .rs1 = 2, .imm = a->uimm_ldsp };
+    return trans_ld(ctx, &arg_ld);
+#endif
+    return false;
+}
+
+static bool trans_c_jr_mv(DisasContext *ctx, arg_c_jr_mv *a)
+{
+    if (a->rd != 0 && a->rs2 == 0) {
+        /* C.JR */
+        arg_jalr arg = { .rd = 0, .rs1 = a->rd, .imm = 0 };
+        return trans_jalr(ctx, &arg);
+    } else if (a->rd != 0 && a->rs2 != 0) {
+        /* C.MV */
+        arg_add arg = { .rd = a->rd, .rs1 = 0, .rs2 = a->rs2 };
+        return trans_add(ctx, &arg);
+    }
+    return false;
+}
+
+static bool trans_c_ebreak_jalr_add(DisasContext *ctx, arg_c_ebreak_jalr_add *a)
+{
+    if (a->rd == 0 && a->rs2 == 0) {
+        /* C.EBREAK */
+        arg_ebreak arg = { };
+        return trans_ebreak(ctx, &arg);
+    } else if (a->rd != 0) {
+        if (a->rs2 == 0) {
+            /* C.JALR */
+            arg_jalr arg = { .rd = 1, .rs1 = a->rd, .imm = 0 };
+            return trans_jalr(ctx, &arg);
+        } else {
+            /* C.ADD */
+            arg_add arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
+            return trans_add(ctx, &arg);
+        }
+    }
+    return false;
+}
+
+static bool trans_c_fsdsp(DisasContext *ctx, arg_c_fsdsp *a)
+{
+    arg_fsd arg = { .rs1 = 2, .rs2 = a->rs2, .imm = a->uimm };
+    return trans_fsd(ctx, &arg);
+}
+
+static bool trans_c_swsp(DisasContext *ctx, arg_c_swsp *a)
+{
+    arg_sw arg = { .rs1 = 2, .rs2 = a->rs2, .imm = a->uimm };
+    return trans_sw(ctx, &arg);
+}
+
+static bool trans_c_fswsp_sdsp(DisasContext *ctx, arg_c_fswsp_sdsp *a)
+{
+#ifdef TARGET_RISCV32
+    /* C.FSWSP */
+    arg_fsw a_fsw = { .rs1 = a->rs2, .rs2 = 2, .imm = a->uimm_fswsp };
+    return trans_fsw(ctx, &a_fsw);
+#else
+    /* C.SDSP */
+    arg_sd a_sd = { .rs1 = 2, .rs2 = a->rs2, .imm = a->uimm_sdsp };
+    return trans_sd(ctx, &a_sd);
+#endif
+}
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index a584c24fbf6b..80afa2c1e62b 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -681,6 +681,7 @@ static void mark_fs_dirty(DisasContext *ctx)
 static inline void mark_fs_dirty(DisasContext *ctx) { }
 #endif
 
+#if !defined(TARGET_RISCV64)
 static void gen_fp_load(DisasContext *ctx, uint32_t opc, int rd,
         int rs1, target_long imm)
 {
@@ -755,6 +756,7 @@ static void gen_fp_store(DisasContext *ctx, uint32_t opc, int rs1,
 
     tcg_temp_free(t0);
 }
+#endif
 
 static void gen_set_rm(DisasContext *ctx, int rm)
 {
@@ -828,84 +830,6 @@ static void decode_RV32_64C0(DisasContext *ctx)
     }
 }
 
-static void decode_RV32_64C2(DisasContext *ctx)
-{
-    uint8_t rd, rs2;
-    uint8_t funct3 = extract32(ctx->opcode, 13, 3);
-
-
-    rd = GET_RD(ctx->opcode);
-
-    switch (funct3) {
-    case 0: /* C.SLLI -> slli rd, rd, shamt[5:0]
-               C.SLLI64 -> */
-        gen_arith_imm(ctx, OPC_RISC_SLLI, rd, rd, GET_C_ZIMM(ctx->opcode));
-        break;
-    case 1: /* C.FLDSP(RV32/64DC) -> fld rd, offset[8:3](x2) */
-        gen_fp_load(ctx, OPC_RISC_FLD, rd, 2, GET_C_LDSP_IMM(ctx->opcode));
-        break;
-    case 2: /* C.LWSP -> lw rd, offset[7:2](x2) */
-        gen_load(ctx, OPC_RISC_LW, rd, 2, GET_C_LWSP_IMM(ctx->opcode));
-        break;
-    case 3:
-#if defined(TARGET_RISCV64)
-        /* C.LDSP(RVC64) -> ld rd, offset[8:3](x2) */
-        gen_load(ctx, OPC_RISC_LD, rd, 2, GET_C_LDSP_IMM(ctx->opcode));
-#else
-        /* C.FLWSP(RV32FC) -> flw rd, offset[7:2](x2) */
-        gen_fp_load(ctx, OPC_RISC_FLW, rd, 2, GET_C_LWSP_IMM(ctx->opcode));
-#endif
-        break;
-    case 4:
-        rs2 = GET_C_RS2(ctx->opcode);
-
-        if (extract32(ctx->opcode, 12, 1) == 0) {
-            if (rs2 == 0) {
-                /* C.JR -> jalr x0, rs1, 0*/
-                gen_jalr(ctx, OPC_RISC_JALR, 0, rd, 0);
-            } else {
-                /* C.MV -> add rd, x0, rs2 */
-                gen_arith(ctx, OPC_RISC_ADD, rd, 0, rs2);
-            }
-        } else {
-            if (rd == 0) {
-                /* C.EBREAK -> ebreak*/
-                gen_system(ctx, OPC_RISC_ECALL, 0, 0, 0x1);
-            } else {
-                if (rs2 == 0) {
-                    /* C.JALR -> jalr x1, rs1, 0*/
-                    gen_jalr(ctx, OPC_RISC_JALR, 1, rd, 0);
-                } else {
-                    /* C.ADD -> add rd, rd, rs2 */
-                    gen_arith(ctx, OPC_RISC_ADD, rd, rd, rs2);
-                }
-            }
-        }
-        break;
-    case 5:
-        /* C.FSDSP -> fsd rs2, offset[8:3](x2)*/
-        gen_fp_store(ctx, OPC_RISC_FSD, 2, GET_C_RS2(ctx->opcode),
-                     GET_C_SDSP_IMM(ctx->opcode));
-        /* C.SQSP */
-        break;
-    case 6: /* C.SWSP -> sw rs2, offset[7:2](x2)*/
-        gen_store(ctx, OPC_RISC_SW, 2, GET_C_RS2(ctx->opcode),
-                  GET_C_SWSP_IMM(ctx->opcode));
-        break;
-    case 7:
-#if defined(TARGET_RISCV64)
-        /* C.SDSP(Rv64/128) -> sd rs2, offset[8:3](x2)*/
-        gen_store(ctx, OPC_RISC_SD, 2, GET_C_RS2(ctx->opcode),
-                  GET_C_SDSP_IMM(ctx->opcode));
-#else
-        /* C.FSWSP(RV32) -> fsw rs2, offset[7:2](x2) */
-        gen_fp_store(ctx, OPC_RISC_FSW, 2, GET_C_RS2(ctx->opcode),
-                     GET_C_SWSP_IMM(ctx->opcode));
-#endif
-        break;
-    }
-}
-
 static void decode_RV32_64C(DisasContext *ctx)
 {
     uint8_t op = extract32(ctx->opcode, 0, 2);
@@ -914,9 +838,6 @@ static void decode_RV32_64C(DisasContext *ctx)
     case 0:
         decode_RV32_64C0(ctx);
         break;
-    case 2:
-        decode_RV32_64C2(ctx);
-        break;
     }
 }
 
-- 
2.19.2

[Qemu-riscv] [PULL 19/29] target/riscv: Remove gen_jalr()

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

trans_jalr() is the only caller, so move the code into trans_jalr().

Acked-by: Alistair Francis <alistair.francis@wdc.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn_trans/trans_rvi.inc.c | 28 +++++++++++++++++-
 target/riscv/translate.c                | 38 -------------------------
 2 files changed, 27 insertions(+), 39 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index 4a23372cb823..631a88906bce 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -42,7 +42,33 @@ static bool trans_jal(DisasContext *ctx, arg_jal *a)
 
 static bool trans_jalr(DisasContext *ctx, arg_jalr *a)
 {
-    gen_jalr(ctx, OPC_RISC_JALR, a->rd, a->rs1, a->imm);
+    /* no chaining with JALR */
+    TCGLabel *misaligned = NULL;
+    TCGv t0 = tcg_temp_new();
+
+
+    gen_get_gpr(cpu_pc, a->rs1);
+    tcg_gen_addi_tl(cpu_pc, cpu_pc, a->imm);
+    tcg_gen_andi_tl(cpu_pc, cpu_pc, (target_ulong)-2);
+
+    if (!has_ext(ctx, RVC)) {
+        misaligned = gen_new_label();
+        tcg_gen_andi_tl(t0, cpu_pc, 0x2);
+        tcg_gen_brcondi_tl(TCG_COND_NE, t0, 0x0, misaligned);
+    }
+
+    if (a->rd != 0) {
+        tcg_gen_movi_tl(cpu_gpr[a->rd], ctx->pc_succ_insn);
+    }
+    tcg_gen_lookup_and_goto_ptr();
+
+    if (misaligned) {
+        gen_set_label(misaligned);
+        gen_exception_inst_addr_mis(ctx);
+    }
+    ctx->base.is_jmp = DISAS_NORETURN;
+
+    tcg_temp_free(t0);
     return true;
 }
 
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 80afa2c1e62b..9dee2ec24287 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -531,44 +531,6 @@ static void gen_jal(DisasContext *ctx, int rd, target_ulong imm)
     ctx->base.is_jmp = DISAS_NORETURN;
 }
 
-static void gen_jalr(DisasContext *ctx, uint32_t opc, int rd, int rs1,
-                     target_long imm)
-{
-    /* no chaining with JALR */
-    TCGLabel *misaligned = NULL;
-    TCGv t0 = tcg_temp_new();
-
-    switch (opc) {
-    case OPC_RISC_JALR:
-        gen_get_gpr(cpu_pc, rs1);
-        tcg_gen_addi_tl(cpu_pc, cpu_pc, imm);
-        tcg_gen_andi_tl(cpu_pc, cpu_pc, (target_ulong)-2);
-
-        if (!has_ext(ctx, RVC)) {
-            misaligned = gen_new_label();
-            tcg_gen_andi_tl(t0, cpu_pc, 0x2);
-            tcg_gen_brcondi_tl(TCG_COND_NE, t0, 0x0, misaligned);
-        }
-
-        if (rd != 0) {
-            tcg_gen_movi_tl(cpu_gpr[rd], ctx->pc_succ_insn);
-        }
-        tcg_gen_lookup_and_goto_ptr();
-
-        if (misaligned) {
-            gen_set_label(misaligned);
-            gen_exception_inst_addr_mis(ctx);
-        }
-        ctx->base.is_jmp = DISAS_NORETURN;
-        break;
-
-    default:
-        gen_exception_illegal(ctx);
-        break;
-    }
-    tcg_temp_free(t0);
-}
-
 static void gen_branch(DisasContext *ctx, uint32_t opc, int rs1, int rs2,
                        target_long bimm)
 {
-- 
2.19.2

[Qemu-riscv] [PULL 20/29] target/riscv: Remove manual decoding from gen_branch()

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

We now utilizes argument-sets of decodetree such that no manual
decoding is necessary.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn_trans/trans_rvi.inc.c | 46 +++++++++++++++++-------
 target/riscv/translate.c                | 47 -------------------------
 2 files changed, 33 insertions(+), 60 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index 631a88906bce..ae4b0a2bcb78 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -72,41 +72,61 @@ static bool trans_jalr(DisasContext *ctx, arg_jalr *a)
     return true;
 }
 
-static bool trans_beq(DisasContext *ctx, arg_beq *a)
+static bool gen_branch(DisasContext *ctx, arg_b *a, TCGCond cond)
 {
-    gen_branch(ctx, OPC_RISC_BEQ, a->rs1, a->rs2, a->imm);
+    TCGLabel *l = gen_new_label();
+    TCGv source1, source2;
+    source1 = tcg_temp_new();
+    source2 = tcg_temp_new();
+    gen_get_gpr(source1, a->rs1);
+    gen_get_gpr(source2, a->rs2);
+
+    tcg_gen_brcond_tl(cond, source1, source2, l);
+    gen_goto_tb(ctx, 1, ctx->pc_succ_insn);
+    gen_set_label(l); /* branch taken */
+
+    if (!has_ext(ctx, RVC) && ((ctx->base.pc_next + a->imm) & 0x3)) {
+        /* misaligned */
+        gen_exception_inst_addr_mis(ctx);
+    } else {
+        gen_goto_tb(ctx, 0, ctx->base.pc_next + a->imm);
+    }
+    ctx->base.is_jmp = DISAS_NORETURN;
+
+    tcg_temp_free(source1);
+    tcg_temp_free(source2);
+
     return true;
 }
 
+static bool trans_beq(DisasContext *ctx, arg_beq *a)
+{
+    return gen_branch(ctx, a, TCG_COND_EQ);
+}
+
 static bool trans_bne(DisasContext *ctx, arg_bne *a)
 {
-    gen_branch(ctx, OPC_RISC_BNE, a->rs1, a->rs2, a->imm);
-    return true;
+    return gen_branch(ctx, a, TCG_COND_NE);
 }
 
 static bool trans_blt(DisasContext *ctx, arg_blt *a)
 {
-    gen_branch(ctx, OPC_RISC_BLT, a->rs1, a->rs2, a->imm);
-    return true;
+    return gen_branch(ctx, a, TCG_COND_LT);
 }
 
 static bool trans_bge(DisasContext *ctx, arg_bge *a)
 {
-    gen_branch(ctx, OPC_RISC_BGE, a->rs1, a->rs2, a->imm);
-    return true;
+    return gen_branch(ctx, a, TCG_COND_GE);
 }
 
 static bool trans_bltu(DisasContext *ctx, arg_bltu *a)
 {
-    gen_branch(ctx, OPC_RISC_BLTU, a->rs1, a->rs2, a->imm);
-    return true;
+    return gen_branch(ctx, a, TCG_COND_LTU);
 }
 
 static bool trans_bgeu(DisasContext *ctx, arg_bgeu *a)
 {
-
-    gen_branch(ctx, OPC_RISC_BGEU, a->rs1, a->rs2, a->imm);
-    return true;
+    return gen_branch(ctx, a, TCG_COND_GEU);
 }
 
 static bool trans_lb(DisasContext *ctx, arg_lb *a)
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 9dee2ec24287..a3d5cdbad82d 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -531,53 +531,6 @@ static void gen_jal(DisasContext *ctx, int rd, target_ulong imm)
     ctx->base.is_jmp = DISAS_NORETURN;
 }
 
-static void gen_branch(DisasContext *ctx, uint32_t opc, int rs1, int rs2,
-                       target_long bimm)
-{
-    TCGLabel *l = gen_new_label();
-    TCGv source1, source2;
-    source1 = tcg_temp_new();
-    source2 = tcg_temp_new();
-    gen_get_gpr(source1, rs1);
-    gen_get_gpr(source2, rs2);
-
-    switch (opc) {
-    case OPC_RISC_BEQ:
-        tcg_gen_brcond_tl(TCG_COND_EQ, source1, source2, l);
-        break;
-    case OPC_RISC_BNE:
-        tcg_gen_brcond_tl(TCG_COND_NE, source1, source2, l);
-        break;
-    case OPC_RISC_BLT:
-        tcg_gen_brcond_tl(TCG_COND_LT, source1, source2, l);
-        break;
-    case OPC_RISC_BGE:
-        tcg_gen_brcond_tl(TCG_COND_GE, source1, source2, l);
-        break;
-    case OPC_RISC_BLTU:
-        tcg_gen_brcond_tl(TCG_COND_LTU, source1, source2, l);
-        break;
-    case OPC_RISC_BGEU:
-        tcg_gen_brcond_tl(TCG_COND_GEU, source1, source2, l);
-        break;
-    default:
-        gen_exception_illegal(ctx);
-        return;
-    }
-    tcg_temp_free(source1);
-    tcg_temp_free(source2);
-
-    gen_goto_tb(ctx, 1, ctx->pc_succ_insn);
-    gen_set_label(l); /* branch taken */
-    if (!has_ext(ctx, RVC) && ((ctx->base.pc_next + bimm) & 0x3)) {
-        /* misaligned */
-        gen_exception_inst_addr_mis(ctx);
-    } else {
-        gen_goto_tb(ctx, 0, ctx->base.pc_next + bimm);
-    }
-    ctx->base.is_jmp = DISAS_NORETURN;
-}
-
 static void gen_load(DisasContext *ctx, uint32_t opc, int rd, int rs1,
         target_long imm)
 {
-- 
2.19.2

[Qemu-riscv] [PULL 21/29] target/riscv: Remove manual decoding from gen_load()

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

With decodetree we don't need to convert RISC-V opcodes into to MemOps
as the old gen_load() did.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn_trans/trans_rvi.inc.c | 35 +++++++++++++++----------
 target/riscv/translate.c                |  6 +++--
 2 files changed, 25 insertions(+), 16 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index ae4b0a2bcb78..cc361ed4d151 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -129,34 +129,43 @@ static bool trans_bgeu(DisasContext *ctx, arg_bgeu *a)
     return gen_branch(ctx, a, TCG_COND_GEU);
 }
 
-static bool trans_lb(DisasContext *ctx, arg_lb *a)
+static bool gen_load(DisasContext *ctx, arg_lb *a, TCGMemOp memop)
 {
-    gen_load(ctx, OPC_RISC_LB, a->rd, a->rs1, a->imm);
+    TCGv t0 = tcg_temp_new();
+    TCGv t1 = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+    tcg_gen_addi_tl(t0, t0, a->imm);
+
+    tcg_gen_qemu_ld_tl(t1, t0, ctx->mem_idx, memop);
+    gen_set_gpr(a->rd, t1);
+    tcg_temp_free(t0);
+    tcg_temp_free(t1);
     return true;
 }
 
+static bool trans_lb(DisasContext *ctx, arg_lb *a)
+{
+    return gen_load(ctx, a, MO_SB);
+}
+
 static bool trans_lh(DisasContext *ctx, arg_lh *a)
 {
-    gen_load(ctx, OPC_RISC_LH, a->rd, a->rs1, a->imm);
-    return true;
+    return gen_load(ctx, a, MO_TESW);
 }
 
 static bool trans_lw(DisasContext *ctx, arg_lw *a)
 {
-    gen_load(ctx, OPC_RISC_LW, a->rd, a->rs1, a->imm);
-    return true;
+    return gen_load(ctx, a, MO_TESL);
 }
 
 static bool trans_lbu(DisasContext *ctx, arg_lbu *a)
 {
-    gen_load(ctx, OPC_RISC_LBU, a->rd, a->rs1, a->imm);
-    return true;
+    return gen_load(ctx, a, MO_UB);
 }
 
 static bool trans_lhu(DisasContext *ctx, arg_lhu *a)
 {
-    gen_load(ctx, OPC_RISC_LHU, a->rd, a->rs1, a->imm);
-    return true;
+    return gen_load(ctx, a, MO_TEUW);
 }
 
 static bool trans_sb(DisasContext *ctx, arg_sb *a)
@@ -180,14 +189,12 @@ static bool trans_sw(DisasContext *ctx, arg_sw *a)
 #ifdef TARGET_RISCV64
 static bool trans_lwu(DisasContext *ctx, arg_lwu *a)
 {
-    gen_load(ctx, OPC_RISC_LWU, a->rd, a->rs1, a->imm);
-    return true;
+    return gen_load(ctx, a, MO_TEUL);
 }
 
 static bool trans_ld(DisasContext *ctx, arg_ld *a)
 {
-    gen_load(ctx, OPC_RISC_LD, a->rd, a->rs1, a->imm);
-    return true;
+    return gen_load(ctx, a, MO_TEQ);
 }
 
 static bool trans_sd(DisasContext *ctx, arg_sd *a)
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index a3d5cdbad82d..99d6d3b4ae91 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -531,7 +531,8 @@ static void gen_jal(DisasContext *ctx, int rd, target_ulong imm)
     ctx->base.is_jmp = DISAS_NORETURN;
 }
 
-static void gen_load(DisasContext *ctx, uint32_t opc, int rd, int rs1,
+#ifdef TARGET_RISCV64
+static void gen_load_c(DisasContext *ctx, uint32_t opc, int rd, int rs1,
         target_long imm)
 {
     TCGv t0 = tcg_temp_new();
@@ -550,6 +551,7 @@ static void gen_load(DisasContext *ctx, uint32_t opc, int rd, int rs1,
     tcg_temp_free(t0);
     tcg_temp_free(t1);
 }
+#endif
 
 static void gen_store(DisasContext *ctx, uint32_t opc, int rs1, int rs2,
         target_long imm)
@@ -723,7 +725,7 @@ static void decode_RV32_64C0(DisasContext *ctx)
     case 3:
 #if defined(TARGET_RISCV64)
         /* C.LD(RV64/128) -> ld rd', offset[7:3](rs1')*/
-        gen_load(ctx, OPC_RISC_LD, rd_rs2, rs1s,
+        gen_load_c(ctx, OPC_RISC_LD, rd_rs2, rs1s,
                  GET_C_LD_IMM(ctx->opcode));
 #else
         /* C.FLW (RV32) -> flw rd', offset[6:2](rs1')*/
-- 
2.19.2

[Qemu-riscv] [PULL 22/29] target/riscv: Remove manual decoding from gen_store()

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

With decodetree we don't need to convert RISC-V opcodes into to MemOps
as the old gen_store() did.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn_trans/trans_rvi.inc.c | 27 +++++++++++++++++--------
 target/riscv/translate.c                |  8 +++++---
 2 files changed, 24 insertions(+), 11 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index cc361ed4d151..5a09c6335ae0 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -168,22 +168,34 @@ static bool trans_lhu(DisasContext *ctx, arg_lhu *a)
     return gen_load(ctx, a, MO_TEUW);
 }
 
-static bool trans_sb(DisasContext *ctx, arg_sb *a)
+static bool gen_store(DisasContext *ctx, arg_sb *a, TCGMemOp memop)
 {
-    gen_store(ctx, OPC_RISC_SB, a->rs1, a->rs2, a->imm);
+    TCGv t0 = tcg_temp_new();
+    TCGv dat = tcg_temp_new();
+    gen_get_gpr(t0, a->rs1);
+    tcg_gen_addi_tl(t0, t0, a->imm);
+    gen_get_gpr(dat, a->rs2);
+
+    tcg_gen_qemu_st_tl(dat, t0, ctx->mem_idx, memop);
+    tcg_temp_free(t0);
+    tcg_temp_free(dat);
     return true;
 }
 
+
+static bool trans_sb(DisasContext *ctx, arg_sb *a)
+{
+    return gen_store(ctx, a, MO_SB);
+}
+
 static bool trans_sh(DisasContext *ctx, arg_sh *a)
 {
-    gen_store(ctx, OPC_RISC_SH, a->rs1, a->rs2, a->imm);
-    return true;
+    return gen_store(ctx, a, MO_TESW);
 }
 
 static bool trans_sw(DisasContext *ctx, arg_sw *a)
 {
-    gen_store(ctx, OPC_RISC_SW, a->rs1, a->rs2, a->imm);
-    return true;
+    return gen_store(ctx, a, MO_TESL);
 }
 
 #ifdef TARGET_RISCV64
@@ -199,8 +211,7 @@ static bool trans_ld(DisasContext *ctx, arg_ld *a)
 
 static bool trans_sd(DisasContext *ctx, arg_sd *a)
 {
-    gen_store(ctx, OPC_RISC_SD, a->rs1, a->rs2, a->imm);
-    return true;
+    return gen_store(ctx, a, MO_TEQ);
 }
 #endif
 
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 99d6d3b4ae91..cdc08b1bff20 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -56,6 +56,7 @@ typedef struct DisasContext {
     int frm;
 } DisasContext;
 
+#ifdef TARGET_RISCV64
 /* convert riscv funct3 to qemu memop for load/store */
 static const int tcg_memop_lookup[8] = {
     [0 ... 7] = -1,
@@ -69,6 +70,7 @@ static const int tcg_memop_lookup[8] = {
     [6] = MO_TEUL,
 #endif
 };
+#endif
 
 #ifdef TARGET_RISCV64
 #define CASE_OP_32_64(X) case X: case glue(X, W)
@@ -551,9 +553,8 @@ static void gen_load_c(DisasContext *ctx, uint32_t opc, int rd, int rs1,
     tcg_temp_free(t0);
     tcg_temp_free(t1);
 }
-#endif
 
-static void gen_store(DisasContext *ctx, uint32_t opc, int rs1, int rs2,
+static void gen_store_c(DisasContext *ctx, uint32_t opc, int rs1, int rs2,
         target_long imm)
 {
     TCGv t0 = tcg_temp_new();
@@ -572,6 +573,7 @@ static void gen_store(DisasContext *ctx, uint32_t opc, int rs1, int rs2,
     tcg_temp_free(t0);
     tcg_temp_free(dat);
 }
+#endif
 
 #ifndef CONFIG_USER_ONLY
 /* The states of mstatus_fs are:
@@ -736,7 +738,7 @@ static void decode_RV32_64C0(DisasContext *ctx)
     case 7:
 #if defined(TARGET_RISCV64)
         /* C.SD (RV64/128) -> sd rs2', offset[7:3](rs1')*/
-        gen_store(ctx, OPC_RISC_SD, rs1s, rd_rs2,
+        gen_store_c(ctx, OPC_RISC_SD, rs1s, rd_rs2,
                   GET_C_LD_IMM(ctx->opcode));
 #else
         /* C.FSW (RV32) -> fsw rs2', offset[6:2](rs1')*/
-- 
2.19.2

[Qemu-riscv] [PULL 23/29] target/riscv: Move gen_arith_imm() decoding into trans_* functions

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

gen_arith_imm() does a lot of decoding manually, which was hard to read
in case of the shift instructions and is not necessary anymore with
decodetree.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32.decode              |   3 +-
 target/riscv/insn_trans/trans_rvi.inc.c |  98 +++++++++++++++++-----
 target/riscv/translate.c                | 107 ++++++------------------
 3 files changed, 108 insertions(+), 100 deletions(-)

diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index ecc46a50cc27..d6b4197841f5 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -35,12 +35,13 @@
 
 # Argument sets:
 &b    imm rs2 rs1
+&i    imm rs1 rd
 &shift     shamt rs1 rd
 &atomic    aq rl rs2 rs1 rd
 
 # Formats 32:
 @r       .......   ..... ..... ... ..... .......                   %rs2 %rs1 %rd
-@i       ............    ..... ... ..... .......         imm=%imm_i     %rs1 %rd
+@i       ............    ..... ... ..... ....... &i      imm=%imm_i     %rs1 %rd
 @b       .......   ..... ..... ... ..... ....... &b      imm=%imm_b %rs2 %rs1
 @s       .......   ..... ..... ... ..... .......         imm=%imm_s %rs2 %rs1
 @u       ....................      ..... .......         imm=%imm_u          %rd
diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index 5a09c6335ae0..0265740bdb69 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -217,52 +217,96 @@ static bool trans_sd(DisasContext *ctx, arg_sd *a)
 
 static bool trans_addi(DisasContext *ctx, arg_addi *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_ADDI, a->rd, a->rs1, a->imm);
-    return true;
+    return gen_arith_imm(ctx, a, &tcg_gen_add_tl);
 }
 
 static bool trans_slti(DisasContext *ctx, arg_slti *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_SLTI, a->rd, a->rs1, a->imm);
+    TCGv source1;
+    source1 = tcg_temp_new();
+    gen_get_gpr(source1, a->rs1);
+
+    tcg_gen_setcondi_tl(TCG_COND_LT, source1, source1, a->imm);
+
+    gen_set_gpr(a->rd, source1);
+    tcg_temp_free(source1);
     return true;
 }
 
 static bool trans_sltiu(DisasContext *ctx, arg_sltiu *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_SLTIU, a->rd, a->rs1, a->imm);
+    TCGv source1;
+    source1 = tcg_temp_new();
+    gen_get_gpr(source1, a->rs1);
+
+    tcg_gen_setcondi_tl(TCG_COND_LTU, source1, source1, a->imm);
+
+    gen_set_gpr(a->rd, source1);
+    tcg_temp_free(source1);
     return true;
 }
 
 static bool trans_xori(DisasContext *ctx, arg_xori *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_XORI, a->rd, a->rs1, a->imm);
-    return true;
+    return gen_arith_imm(ctx, a, &tcg_gen_xor_tl);
 }
 static bool trans_ori(DisasContext *ctx, arg_ori *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_ORI, a->rd, a->rs1, a->imm);
-    return true;
+    return gen_arith_imm(ctx, a, &tcg_gen_or_tl);
 }
 static bool trans_andi(DisasContext *ctx, arg_andi *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_ANDI, a->rd, a->rs1, a->imm);
-    return true;
+    return gen_arith_imm(ctx, a, &tcg_gen_and_tl);
 }
 static bool trans_slli(DisasContext *ctx, arg_slli *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_SLLI, a->rd, a->rs1, a->shamt);
+    if (a->shamt >= TARGET_LONG_BITS) {
+        return false;
+    }
+
+    if (a->rd != 0) {
+        TCGv t = tcg_temp_new();
+        gen_get_gpr(t, a->rs1);
+
+        tcg_gen_shli_tl(t, t, a->shamt);
+
+        gen_set_gpr(a->rd, t);
+        tcg_temp_free(t);
+    } /* NOP otherwise */
     return true;
 }
 
 static bool trans_srli(DisasContext *ctx, arg_srli *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, a->rd, a->rs1, a->shamt);
+    if (a->shamt >= TARGET_LONG_BITS) {
+        return false;
+    }
+
+    if (a->rd != 0) {
+        TCGv t = tcg_temp_new();
+        gen_get_gpr(t, a->rs1);
+
+        tcg_gen_shri_tl(t, t, a->shamt);
+        gen_set_gpr(a->rd, t);
+        tcg_temp_free(t);
+    } /* NOP otherwise */
     return true;
 }
 
 static bool trans_srai(DisasContext *ctx, arg_srai *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, a->rd, a->rs1, a->shamt | 0x400);
+    if (a->shamt >= TARGET_LONG_BITS) {
+        return false;
+    }
+
+    if (a->rd != 0) {
+        TCGv t = tcg_temp_new();
+        gen_get_gpr(t, a->rs1);
+
+        tcg_gen_sari_tl(t, t, a->shamt);
+        gen_set_gpr(a->rd, t);
+        tcg_temp_free(t);
+    } /* NOP otherwise */
     return true;
 }
 
@@ -329,26 +373,42 @@ static bool trans_and(DisasContext *ctx, arg_and *a)
 #ifdef TARGET_RISCV64
 static bool trans_addiw(DisasContext *ctx, arg_addiw *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_ADDIW, a->rd, a->rs1, a->imm);
-    return true;
+    return gen_arith_imm(ctx, a, &gen_addw);
 }
 
 static bool trans_slliw(DisasContext *ctx, arg_slliw *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_SLLIW, a->rd, a->rs1, a->shamt);
+    TCGv source1;
+    source1 = tcg_temp_new();
+    gen_get_gpr(source1, a->rs1);
+
+    tcg_gen_shli_tl(source1, source1, a->shamt);
+    tcg_gen_ext32s_tl(source1, source1);
+    gen_set_gpr(a->rd, source1);
+
+    tcg_temp_free(source1);
     return true;
 }
 
 static bool trans_srliw(DisasContext *ctx, arg_srliw *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_IW, a->rd, a->rs1, a->shamt);
+    TCGv t = tcg_temp_new();
+    gen_get_gpr(t, a->rs1);
+    tcg_gen_extract_tl(t, t, a->shamt, 32 - a->shamt);
+    /* sign-extend for W instructions */
+    tcg_gen_ext32s_tl(t, t);
+    gen_set_gpr(a->rd, t);
+    tcg_temp_free(t);
     return true;
 }
 
 static bool trans_sraiw(DisasContext *ctx, arg_sraiw *a)
 {
-    gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_IW , a->rd, a->rs1,
-                  a->shamt | 0x400);
+    TCGv t = tcg_temp_new();
+    gen_get_gpr(t, a->rs1);
+    tcg_gen_sextract_tl(t, t, a->shamt, 32 - a->shamt);
+    gen_set_gpr(a->rd, t);
+    tcg_temp_free(t);
     return true;
 }
 
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index cdc08b1bff20..0157758a160a 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -433,86 +433,6 @@ static void gen_arith(DisasContext *ctx, uint32_t opc, int rd, int rs1,
     tcg_temp_free(source2);
 }
 
-static void gen_arith_imm(DisasContext *ctx, uint32_t opc, int rd,
-        int rs1, target_long imm)
-{
-    TCGv source1 = tcg_temp_new();
-    int shift_len = TARGET_LONG_BITS;
-    int shift_a;
-
-    gen_get_gpr(source1, rs1);
-
-    switch (opc) {
-    case OPC_RISC_ADDI:
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_ADDIW:
-#endif
-        tcg_gen_addi_tl(source1, source1, imm);
-        break;
-    case OPC_RISC_SLTI:
-        tcg_gen_setcondi_tl(TCG_COND_LT, source1, source1, imm);
-        break;
-    case OPC_RISC_SLTIU:
-        tcg_gen_setcondi_tl(TCG_COND_LTU, source1, source1, imm);
-        break;
-    case OPC_RISC_XORI:
-        tcg_gen_xori_tl(source1, source1, imm);
-        break;
-    case OPC_RISC_ORI:
-        tcg_gen_ori_tl(source1, source1, imm);
-        break;
-    case OPC_RISC_ANDI:
-        tcg_gen_andi_tl(source1, source1, imm);
-        break;
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_SLLIW:
-        shift_len = 32;
-        /* FALLTHRU */
-#endif
-    case OPC_RISC_SLLI:
-        if (imm >= shift_len) {
-            goto do_illegal;
-        }
-        tcg_gen_shli_tl(source1, source1, imm);
-        break;
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_SHIFT_RIGHT_IW:
-        shift_len = 32;
-        /* FALLTHRU */
-#endif
-    case OPC_RISC_SHIFT_RIGHT_I:
-        /* differentiate on IMM */
-        shift_a = imm & 0x400;
-        imm &= 0x3ff;
-        if (imm >= shift_len) {
-            goto do_illegal;
-        }
-        if (imm != 0) {
-            if (shift_a) {
-                /* SRAI[W] */
-                tcg_gen_sextract_tl(source1, source1, imm, shift_len - imm);
-            } else {
-                /* SRLI[W] */
-                tcg_gen_extract_tl(source1, source1, imm, shift_len - imm);
-            }
-            /* No further sign-extension needed for W instructions.  */
-            opc &= ~0x8;
-        }
-        break;
-    default:
-    do_illegal:
-        gen_exception_illegal(ctx);
-        return;
-    }
-
-    if (opc & 0x8) { /* sign-extend for W instructions */
-        tcg_gen_ext32s_tl(source1, source1);
-    }
-
-    gen_set_gpr(rd, source1);
-    tcg_temp_free(source1);
-}
-
 static void gen_jal(DisasContext *ctx, int rd, target_ulong imm)
 {
     target_ulong next_pc;
@@ -785,6 +705,33 @@ static int ex_rvc_register(int reg)
 bool decode_insn32(DisasContext *ctx, uint32_t insn);
 /* Include the auto-generated decoder for 32 bit insn */
 #include "decode_insn32.inc.c"
+
+static bool gen_arith_imm(DisasContext *ctx, arg_i *a,
+                          void(*func)(TCGv, TCGv, TCGv))
+{
+    TCGv source1, source2;
+    source1 = tcg_temp_new();
+    source2 = tcg_temp_new();
+
+    gen_get_gpr(source1, a->rs1);
+    tcg_gen_movi_tl(source2, a->imm);
+
+    (*func)(source1, source1, source2);
+
+    gen_set_gpr(a->rd, source1);
+    tcg_temp_free(source1);
+    tcg_temp_free(source2);
+    return true;
+}
+
+#ifdef TARGET_RISCV64
+static void gen_addw(TCGv ret, TCGv arg1, TCGv arg2)
+{
+    tcg_gen_add_tl(ret, arg1, arg2);
+    tcg_gen_ext32s_tl(ret, ret);
+}
+#endif
+
 /* Include insn module translation function */
 #include "insn_trans/trans_rvi.inc.c"
 #include "insn_trans/trans_rvm.inc.c"
-- 
2.19.2

[Qemu-riscv] [PULL 24/29] target/riscv: make ADD/SUB/OR/XOR/AND insn use arg lists

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

manual decoding in gen_arith() is not necessary with decodetree. For now
the function is called trans_arith as the original gen_arith still
exists. The former will be renamed to gen_arith as soon as the old
gen_arith can be removed.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn32.decode              |  3 +-
 target/riscv/insn_trans/trans_rvi.inc.c | 21 +++++--------
 target/riscv/translate.c                | 40 +++++++++++++++----------
 3 files changed, 34 insertions(+), 30 deletions(-)

diff --git a/target/riscv/insn32.decode b/target/riscv/insn32.decode
index d6b4197841f5..6f3ab7aa52d3 100644
--- a/target/riscv/insn32.decode
+++ b/target/riscv/insn32.decode
@@ -36,11 +36,12 @@
 # Argument sets:
 &b    imm rs2 rs1
 &i    imm rs1 rd
+&r    rd rs1 rs2
 &shift     shamt rs1 rd
 &atomic    aq rl rs2 rs1 rd
 
 # Formats 32:
-@r       .......   ..... ..... ... ..... .......                   %rs2 %rs1 %rd
+@r       .......   ..... ..... ... ..... ....... &r                %rs2 %rs1 %rd
 @i       ............    ..... ... ..... ....... &i      imm=%imm_i     %rs1 %rd
 @b       .......   ..... ..... ... ..... ....... &b      imm=%imm_b %rs2 %rs1
 @s       .......   ..... ..... ... ..... .......         imm=%imm_s %rs2 %rs1
diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index 0265740bdb69..8879f2da35c7 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -312,14 +312,12 @@ static bool trans_srai(DisasContext *ctx, arg_srai *a)
 
 static bool trans_add(DisasContext *ctx, arg_add *a)
 {
-    gen_arith(ctx, OPC_RISC_ADD, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &tcg_gen_add_tl);
 }
 
 static bool trans_sub(DisasContext *ctx, arg_sub *a)
 {
-    gen_arith(ctx, OPC_RISC_SUB, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &tcg_gen_sub_tl);
 }
 
 static bool trans_sll(DisasContext *ctx, arg_sll *a)
@@ -342,8 +340,7 @@ static bool trans_sltu(DisasContext *ctx, arg_sltu *a)
 
 static bool trans_xor(DisasContext *ctx, arg_xor *a)
 {
-    gen_arith(ctx, OPC_RISC_XOR, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &tcg_gen_xor_tl);
 }
 
 static bool trans_srl(DisasContext *ctx, arg_srl *a)
@@ -360,14 +357,12 @@ static bool trans_sra(DisasContext *ctx, arg_sra *a)
 
 static bool trans_or(DisasContext *ctx, arg_or *a)
 {
-    gen_arith(ctx, OPC_RISC_OR, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &tcg_gen_or_tl);
 }
 
 static bool trans_and(DisasContext *ctx, arg_and *a)
 {
-    gen_arith(ctx, OPC_RISC_AND, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &tcg_gen_and_tl);
 }
 
 #ifdef TARGET_RISCV64
@@ -414,14 +409,12 @@ static bool trans_sraiw(DisasContext *ctx, arg_sraiw *a)
 
 static bool trans_addw(DisasContext *ctx, arg_addw *a)
 {
-    gen_arith(ctx, OPC_RISC_ADDW, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &gen_addw);
 }
 
 static bool trans_subw(DisasContext *ctx, arg_subw *a)
 {
-    gen_arith(ctx, OPC_RISC_SUBW, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &gen_subw);
 }
 
 static bool trans_sllw(DisasContext *ctx, arg_sllw *a)
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 0157758a160a..8eb883463322 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -198,12 +198,6 @@ static void gen_arith(DisasContext *ctx, uint32_t opc, int rd, int rs1,
     gen_get_gpr(source2, rs2);
 
     switch (opc) {
-    CASE_OP_32_64(OPC_RISC_ADD):
-        tcg_gen_add_tl(source1, source1, source2);
-        break;
-    CASE_OP_32_64(OPC_RISC_SUB):
-        tcg_gen_sub_tl(source1, source1, source2);
-        break;
 #if defined(TARGET_RISCV64)
     case OPC_RISC_SLLW:
         tcg_gen_andi_tl(source2, source2, 0x1F);
@@ -220,9 +214,6 @@ static void gen_arith(DisasContext *ctx, uint32_t opc, int rd, int rs1,
     case OPC_RISC_SLTU:
         tcg_gen_setcond_tl(TCG_COND_LTU, source1, source1, source2);
         break;
-    case OPC_RISC_XOR:
-        tcg_gen_xor_tl(source1, source1, source2);
-        break;
 #if defined(TARGET_RISCV64)
     case OPC_RISC_SRLW:
         /* clear upper 32 */
@@ -248,12 +239,6 @@ static void gen_arith(DisasContext *ctx, uint32_t opc, int rd, int rs1,
         tcg_gen_andi_tl(source2, source2, TARGET_LONG_BITS - 1);
         tcg_gen_sar_tl(source1, source1, source2);
         break;
-    case OPC_RISC_OR:
-        tcg_gen_or_tl(source1, source1, source2);
-        break;
-    case OPC_RISC_AND:
-        tcg_gen_and_tl(source1, source1, source2);
-        break;
     CASE_OP_32_64(OPC_RISC_MUL):
         if (!has_ext(ctx, RVM)) {
             goto do_illegal;
@@ -730,8 +715,33 @@ static void gen_addw(TCGv ret, TCGv arg1, TCGv arg2)
     tcg_gen_add_tl(ret, arg1, arg2);
     tcg_gen_ext32s_tl(ret, ret);
 }
+
+static void gen_subw(TCGv ret, TCGv arg1, TCGv arg2)
+{
+    tcg_gen_sub_tl(ret, arg1, arg2);
+    tcg_gen_ext32s_tl(ret, ret);
+}
+
 #endif
 
+static bool trans_arith(DisasContext *ctx, arg_r *a,
+                        void(*func)(TCGv, TCGv, TCGv))
+{
+    TCGv source1, source2;
+    source1 = tcg_temp_new();
+    source2 = tcg_temp_new();
+
+    gen_get_gpr(source1, a->rs1);
+    gen_get_gpr(source2, a->rs2);
+
+    (*func)(source1, source1, source2);
+
+    gen_set_gpr(a->rd, source1);
+    tcg_temp_free(source1);
+    tcg_temp_free(source2);
+    return true;
+}
+
 /* Include insn module translation function */
 #include "insn_trans/trans_rvi.inc.c"
 #include "insn_trans/trans_rvm.inc.c"
-- 
2.19.2

[Qemu-riscv] [PULL 25/29] target/riscv: Remove shift and slt insn manual decoding

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn_trans/trans_rvi.inc.c | 93 +++++++++++++++++--------
 target/riscv/translate.c                | 59 +++++-----------
 2 files changed, 81 insertions(+), 71 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index 8879f2da35c7..88ef0003ec17 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -220,30 +220,25 @@ static bool trans_addi(DisasContext *ctx, arg_addi *a)
     return gen_arith_imm(ctx, a, &tcg_gen_add_tl);
 }
 
-static bool trans_slti(DisasContext *ctx, arg_slti *a)
+static void gen_slt(TCGv ret, TCGv s1, TCGv s2)
 {
-    TCGv source1;
-    source1 = tcg_temp_new();
-    gen_get_gpr(source1, a->rs1);
+    tcg_gen_setcond_tl(TCG_COND_LT, ret, s1, s2);
+}
+
+static void gen_sltu(TCGv ret, TCGv s1, TCGv s2)
+{
+    tcg_gen_setcond_tl(TCG_COND_LTU, ret, s1, s2);
+}
 
-    tcg_gen_setcondi_tl(TCG_COND_LT, source1, source1, a->imm);
 
-    gen_set_gpr(a->rd, source1);
-    tcg_temp_free(source1);
-    return true;
+static bool trans_slti(DisasContext *ctx, arg_slti *a)
+{
+    return gen_arith_imm(ctx, a, &gen_slt);
 }
 
 static bool trans_sltiu(DisasContext *ctx, arg_sltiu *a)
 {
-    TCGv source1;
-    source1 = tcg_temp_new();
-    gen_get_gpr(source1, a->rs1);
-
-    tcg_gen_setcondi_tl(TCG_COND_LTU, source1, source1, a->imm);
-
-    gen_set_gpr(a->rd, source1);
-    tcg_temp_free(source1);
-    return true;
+    return gen_arith_imm(ctx, a, &gen_sltu);
 }
 
 static bool trans_xori(DisasContext *ctx, arg_xori *a)
@@ -322,20 +317,17 @@ static bool trans_sub(DisasContext *ctx, arg_sub *a)
 
 static bool trans_sll(DisasContext *ctx, arg_sll *a)
 {
-    gen_arith(ctx, OPC_RISC_SLL, a->rd, a->rs1, a->rs2);
-    return true;
+    return gen_shift(ctx, a, &tcg_gen_shl_tl);
 }
 
 static bool trans_slt(DisasContext *ctx, arg_slt *a)
 {
-    gen_arith(ctx, OPC_RISC_SLT, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &gen_slt);
 }
 
 static bool trans_sltu(DisasContext *ctx, arg_sltu *a)
 {
-    gen_arith(ctx, OPC_RISC_SLTU, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &gen_sltu);
 }
 
 static bool trans_xor(DisasContext *ctx, arg_xor *a)
@@ -345,14 +337,12 @@ static bool trans_xor(DisasContext *ctx, arg_xor *a)
 
 static bool trans_srl(DisasContext *ctx, arg_srl *a)
 {
-    gen_arith(ctx, OPC_RISC_SRL, a->rd, a->rs1, a->rs2);
-    return true;
+    return gen_shift(ctx, a, &tcg_gen_shr_tl);
 }
 
 static bool trans_sra(DisasContext *ctx, arg_sra *a)
 {
-    gen_arith(ctx, OPC_RISC_SRA, a->rd, a->rs1, a->rs2);
-    return true;
+    return gen_shift(ctx, a, &tcg_gen_sar_tl);
 }
 
 static bool trans_or(DisasContext *ctx, arg_or *a)
@@ -419,19 +409,62 @@ static bool trans_subw(DisasContext *ctx, arg_subw *a)
 
 static bool trans_sllw(DisasContext *ctx, arg_sllw *a)
 {
-    gen_arith(ctx, OPC_RISC_SLLW, a->rd, a->rs1, a->rs2);
+    TCGv source1 = tcg_temp_new();
+    TCGv source2 = tcg_temp_new();
+
+    gen_get_gpr(source1, a->rs1);
+    gen_get_gpr(source2, a->rs2);
+
+    tcg_gen_andi_tl(source2, source2, 0x1F);
+    tcg_gen_shl_tl(source1, source1, source2);
+
+    tcg_gen_ext32s_tl(source1, source1);
+    gen_set_gpr(a->rd, source1);
+    tcg_temp_free(source1);
+    tcg_temp_free(source2);
     return true;
 }
 
 static bool trans_srlw(DisasContext *ctx, arg_srlw *a)
 {
-    gen_arith(ctx, OPC_RISC_SRLW, a->rd, a->rs1, a->rs2);
+    TCGv source1 = tcg_temp_new();
+    TCGv source2 = tcg_temp_new();
+
+    gen_get_gpr(source1, a->rs1);
+    gen_get_gpr(source2, a->rs2);
+
+    /* clear upper 32 */
+    tcg_gen_ext32u_tl(source1, source1);
+    tcg_gen_andi_tl(source2, source2, 0x1F);
+    tcg_gen_shr_tl(source1, source1, source2);
+
+    tcg_gen_ext32s_tl(source1, source1);
+    gen_set_gpr(a->rd, source1);
+    tcg_temp_free(source1);
+    tcg_temp_free(source2);
     return true;
 }
 
 static bool trans_sraw(DisasContext *ctx, arg_sraw *a)
 {
-    gen_arith(ctx, OPC_RISC_SRAW, a->rd, a->rs1, a->rs2);
+    TCGv source1 = tcg_temp_new();
+    TCGv source2 = tcg_temp_new();
+
+    gen_get_gpr(source1, a->rs1);
+    gen_get_gpr(source2, a->rs2);
+
+    /*
+     * first, trick to get it to act like working on 32 bits (get rid of
+     * upper 32, sign extend to fill space)
+     */
+    tcg_gen_ext32s_tl(source1, source1);
+    tcg_gen_andi_tl(source2, source2, 0x1F);
+    tcg_gen_sar_tl(source1, source1, source2);
+
+    gen_set_gpr(a->rd, source1);
+    tcg_temp_free(source1);
+    tcg_temp_free(source2);
+
     return true;
 }
 #endif
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 8eb883463322..9ae40f65096f 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -198,47 +198,6 @@ static void gen_arith(DisasContext *ctx, uint32_t opc, int rd, int rs1,
     gen_get_gpr(source2, rs2);
 
     switch (opc) {
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_SLLW:
-        tcg_gen_andi_tl(source2, source2, 0x1F);
-        tcg_gen_shl_tl(source1, source1, source2);
-        break;
-#endif
-    case OPC_RISC_SLL:
-        tcg_gen_andi_tl(source2, source2, TARGET_LONG_BITS - 1);
-        tcg_gen_shl_tl(source1, source1, source2);
-        break;
-    case OPC_RISC_SLT:
-        tcg_gen_setcond_tl(TCG_COND_LT, source1, source1, source2);
-        break;
-    case OPC_RISC_SLTU:
-        tcg_gen_setcond_tl(TCG_COND_LTU, source1, source1, source2);
-        break;
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_SRLW:
-        /* clear upper 32 */
-        tcg_gen_ext32u_tl(source1, source1);
-        tcg_gen_andi_tl(source2, source2, 0x1F);
-        tcg_gen_shr_tl(source1, source1, source2);
-        break;
-#endif
-    case OPC_RISC_SRL:
-        tcg_gen_andi_tl(source2, source2, TARGET_LONG_BITS - 1);
-        tcg_gen_shr_tl(source1, source1, source2);
-        break;
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_SRAW:
-        /* first, trick to get it to act like working on 32 bits (get rid of
-        upper 32, sign extend to fill space) */
-        tcg_gen_ext32s_tl(source1, source1);
-        tcg_gen_andi_tl(source2, source2, 0x1F);
-        tcg_gen_sar_tl(source1, source1, source2);
-        break;
-#endif
-    case OPC_RISC_SRA:
-        tcg_gen_andi_tl(source2, source2, TARGET_LONG_BITS - 1);
-        tcg_gen_sar_tl(source1, source1, source2);
-        break;
     CASE_OP_32_64(OPC_RISC_MUL):
         if (!has_ext(ctx, RVM)) {
             goto do_illegal;
@@ -742,6 +701,24 @@ static bool trans_arith(DisasContext *ctx, arg_r *a,
     return true;
 }
 
+static bool gen_shift(DisasContext *ctx, arg_r *a,
+                        void(*func)(TCGv, TCGv, TCGv))
+{
+    TCGv source1 = tcg_temp_new();
+    TCGv source2 = tcg_temp_new();
+
+    gen_get_gpr(source1, a->rs1);
+    gen_get_gpr(source2, a->rs2);
+
+    tcg_gen_andi_tl(source2, source2, TARGET_LONG_BITS - 1);
+    (*func)(source1, source1, source2);
+
+    gen_set_gpr(a->rd, source1);
+    tcg_temp_free(source1);
+    tcg_temp_free(source2);
+    return true;
+}
+
 /* Include insn module translation function */
 #include "insn_trans/trans_rvi.inc.c"
 #include "insn_trans/trans_rvm.inc.c"
-- 
2.19.2

[Qemu-riscv] [PULL 26/29] target/riscv: Remove manual decoding of RV32/64M insn

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/insn_trans/trans_rvm.inc.c |  55 ++--
 target/riscv/translate.c                | 320 ++++++++++--------------
 2 files changed, 164 insertions(+), 211 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvm.inc.c b/target/riscv/insn_trans/trans_rvm.inc.c
index 69631c9e3783..d2bf2f171904 100644
--- a/target/riscv/insn_trans/trans_rvm.inc.c
+++ b/target/riscv/insn_trans/trans_rvm.inc.c
@@ -22,92 +22,99 @@
 static bool trans_mul(DisasContext *ctx, arg_mul *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_MUL, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &tcg_gen_mul_tl);
 }
 
 static bool trans_mulh(DisasContext *ctx, arg_mulh *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_MULH, a->rd, a->rs1, a->rs2);
+    TCGv source1 = tcg_temp_new();
+    TCGv source2 = tcg_temp_new();
+    gen_get_gpr(source1, a->rs1);
+    gen_get_gpr(source2, a->rs2);
+
+    tcg_gen_muls2_tl(source2, source1, source1, source2);
+
+    gen_set_gpr(a->rd, source1);
+    tcg_temp_free(source1);
+    tcg_temp_free(source2);
     return true;
 }
 
 static bool trans_mulhsu(DisasContext *ctx, arg_mulhsu *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_MULHSU, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &gen_mulhsu);
 }
 
 static bool trans_mulhu(DisasContext *ctx, arg_mulhu *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_MULHU, a->rd, a->rs1, a->rs2);
+    TCGv source1 = tcg_temp_new();
+    TCGv source2 = tcg_temp_new();
+    gen_get_gpr(source1, a->rs1);
+    gen_get_gpr(source2, a->rs2);
+
+    tcg_gen_mulu2_tl(source2, source1, source1, source2);
+
+    gen_set_gpr(a->rd, source1);
+    tcg_temp_free(source1);
+    tcg_temp_free(source2);
     return true;
 }
 
 static bool trans_div(DisasContext *ctx, arg_div *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_DIV, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &gen_div);
 }
 
 static bool trans_divu(DisasContext *ctx, arg_divu *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_DIVU, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &gen_divu);
 }
 
 static bool trans_rem(DisasContext *ctx, arg_rem *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_REM, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &gen_rem);
 }
 
 static bool trans_remu(DisasContext *ctx, arg_remu *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_REMU, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &gen_remu);
 }
 
 #ifdef TARGET_RISCV64
 static bool trans_mulw(DisasContext *ctx, arg_mulw *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_MULW, a->rd, a->rs1, a->rs2);
-    return true;
+    return trans_arith(ctx, a, &gen_mulw);
 }
 
 static bool trans_divw(DisasContext *ctx, arg_divw *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_DIVW, a->rd, a->rs1, a->rs2);
-    return true;
+    return gen_arith_div_w(ctx, a, &gen_div);
 }
 
 static bool trans_divuw(DisasContext *ctx, arg_divuw *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_DIVUW, a->rd, a->rs1, a->rs2);
-    return true;
+    return gen_arith_div_w(ctx, a, &gen_divu);
 }
 
 static bool trans_remw(DisasContext *ctx, arg_remw *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_REMW, a->rd, a->rs1, a->rs2);
-    return true;
+    return gen_arith_div_w(ctx, a, &gen_rem);
 }
 
 static bool trans_remuw(DisasContext *ctx, arg_remuw *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    gen_arith(ctx, OPC_RISC_REMUW, a->rd, a->rs1, a->rs2);
-    return true;
+    return gen_arith_div_w(ctx, a, &gen_remu);
 }
 #endif
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 9ae40f65096f..3cd7e16c63cf 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -188,193 +188,112 @@ static void gen_mulhsu(TCGv ret, TCGv arg1, TCGv arg2)
     tcg_temp_free(rh);
 }
 
-static void gen_arith(DisasContext *ctx, uint32_t opc, int rd, int rs1,
-        int rs2)
-{
-    TCGv source1, source2, cond1, cond2, zeroreg, resultopt1;
-    source1 = tcg_temp_new();
-    source2 = tcg_temp_new();
-    gen_get_gpr(source1, rs1);
-    gen_get_gpr(source2, rs2);
-
-    switch (opc) {
-    CASE_OP_32_64(OPC_RISC_MUL):
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        tcg_gen_mul_tl(source1, source1, source2);
-        break;
-    case OPC_RISC_MULH:
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        tcg_gen_muls2_tl(source2, source1, source1, source2);
-        break;
-    case OPC_RISC_MULHSU:
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        gen_mulhsu(source1, source1, source2);
-        break;
-    case OPC_RISC_MULHU:
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        tcg_gen_mulu2_tl(source2, source1, source1, source2);
-        break;
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_DIVW:
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        tcg_gen_ext32s_tl(source1, source1);
-        tcg_gen_ext32s_tl(source2, source2);
-        /* fall through to DIV */
-#endif
-    case OPC_RISC_DIV:
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        /* Handle by altering args to tcg_gen_div to produce req'd results:
-         * For overflow: want source1 in source1 and 1 in source2
-         * For div by zero: want -1 in source1 and 1 in source2 -> -1 result */
-        cond1 = tcg_temp_new();
-        cond2 = tcg_temp_new();
-        zeroreg = tcg_const_tl(0);
-        resultopt1 = tcg_temp_new();
-
-        tcg_gen_movi_tl(resultopt1, (target_ulong)-1);
-        tcg_gen_setcondi_tl(TCG_COND_EQ, cond2, source2, (target_ulong)(~0L));
-        tcg_gen_setcondi_tl(TCG_COND_EQ, cond1, source1,
-                            ((target_ulong)1) << (TARGET_LONG_BITS - 1));
-        tcg_gen_and_tl(cond1, cond1, cond2); /* cond1 = overflow */
-        tcg_gen_setcondi_tl(TCG_COND_EQ, cond2, source2, 0); /* cond2 = div 0 */
-        /* if div by zero, set source1 to -1, otherwise don't change */
-        tcg_gen_movcond_tl(TCG_COND_EQ, source1, cond2, zeroreg, source1,
-                resultopt1);
-        /* if overflow or div by zero, set source2 to 1, else don't change */
-        tcg_gen_or_tl(cond1, cond1, cond2);
-        tcg_gen_movi_tl(resultopt1, (target_ulong)1);
-        tcg_gen_movcond_tl(TCG_COND_EQ, source2, cond1, zeroreg, source2,
-                resultopt1);
-        tcg_gen_div_tl(source1, source1, source2);
-
-        tcg_temp_free(cond1);
-        tcg_temp_free(cond2);
-        tcg_temp_free(zeroreg);
-        tcg_temp_free(resultopt1);
-        break;
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_DIVUW:
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        tcg_gen_ext32u_tl(source1, source1);
-        tcg_gen_ext32u_tl(source2, source2);
-        /* fall through to DIVU */
-#endif
-    case OPC_RISC_DIVU:
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        cond1 = tcg_temp_new();
-        zeroreg = tcg_const_tl(0);
-        resultopt1 = tcg_temp_new();
-
-        tcg_gen_setcondi_tl(TCG_COND_EQ, cond1, source2, 0);
-        tcg_gen_movi_tl(resultopt1, (target_ulong)-1);
-        tcg_gen_movcond_tl(TCG_COND_EQ, source1, cond1, zeroreg, source1,
-                resultopt1);
-        tcg_gen_movi_tl(resultopt1, (target_ulong)1);
-        tcg_gen_movcond_tl(TCG_COND_EQ, source2, cond1, zeroreg, source2,
-                resultopt1);
-        tcg_gen_divu_tl(source1, source1, source2);
-
-        tcg_temp_free(cond1);
-        tcg_temp_free(zeroreg);
-        tcg_temp_free(resultopt1);
-        break;
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_REMW:
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        tcg_gen_ext32s_tl(source1, source1);
-        tcg_gen_ext32s_tl(source2, source2);
-        /* fall through to REM */
-#endif
-    case OPC_RISC_REM:
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        cond1 = tcg_temp_new();
-        cond2 = tcg_temp_new();
-        zeroreg = tcg_const_tl(0);
-        resultopt1 = tcg_temp_new();
-
-        tcg_gen_movi_tl(resultopt1, 1L);
-        tcg_gen_setcondi_tl(TCG_COND_EQ, cond2, source2, (target_ulong)-1);
-        tcg_gen_setcondi_tl(TCG_COND_EQ, cond1, source1,
-                            (target_ulong)1 << (TARGET_LONG_BITS - 1));
-        tcg_gen_and_tl(cond2, cond1, cond2); /* cond1 = overflow */
-        tcg_gen_setcondi_tl(TCG_COND_EQ, cond1, source2, 0); /* cond2 = div 0 */
-        /* if overflow or div by zero, set source2 to 1, else don't change */
-        tcg_gen_or_tl(cond2, cond1, cond2);
-        tcg_gen_movcond_tl(TCG_COND_EQ, source2, cond2, zeroreg, source2,
-                resultopt1);
-        tcg_gen_rem_tl(resultopt1, source1, source2);
-        /* if div by zero, just return the original dividend */
-        tcg_gen_movcond_tl(TCG_COND_EQ, source1, cond1, zeroreg, resultopt1,
-                source1);
-
-        tcg_temp_free(cond1);
-        tcg_temp_free(cond2);
-        tcg_temp_free(zeroreg);
-        tcg_temp_free(resultopt1);
-        break;
-#if defined(TARGET_RISCV64)
-    case OPC_RISC_REMUW:
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        tcg_gen_ext32u_tl(source1, source1);
-        tcg_gen_ext32u_tl(source2, source2);
-        /* fall through to REMU */
-#endif
-    case OPC_RISC_REMU:
-        if (!has_ext(ctx, RVM)) {
-            goto do_illegal;
-        }
-        cond1 = tcg_temp_new();
-        zeroreg = tcg_const_tl(0);
-        resultopt1 = tcg_temp_new();
-
-        tcg_gen_movi_tl(resultopt1, (target_ulong)1);
-        tcg_gen_setcondi_tl(TCG_COND_EQ, cond1, source2, 0);
-        tcg_gen_movcond_tl(TCG_COND_EQ, source2, cond1, zeroreg, source2,
-                resultopt1);
-        tcg_gen_remu_tl(resultopt1, source1, source2);
-        /* if div by zero, just return the original dividend */
-        tcg_gen_movcond_tl(TCG_COND_EQ, source1, cond1, zeroreg, resultopt1,
-                source1);
-
-        tcg_temp_free(cond1);
-        tcg_temp_free(zeroreg);
-        tcg_temp_free(resultopt1);
-        break;
-    do_illegal:
-    default:
-        gen_exception_illegal(ctx);
-        return;
-    }
-
-    if (opc & 0x8) { /* sign extend for W instructions */
-        tcg_gen_ext32s_tl(source1, source1);
-    }
-
-    gen_set_gpr(rd, source1);
-    tcg_temp_free(source1);
-    tcg_temp_free(source2);
+static void gen_div(TCGv ret, TCGv source1, TCGv source2)
+{
+    TCGv cond1, cond2, zeroreg, resultopt1;
+    /*
+     * Handle by altering args to tcg_gen_div to produce req'd results:
+     * For overflow: want source1 in source1 and 1 in source2
+     * For div by zero: want -1 in source1 and 1 in source2 -> -1 result
+     */
+    cond1 = tcg_temp_new();
+    cond2 = tcg_temp_new();
+    zeroreg = tcg_const_tl(0);
+    resultopt1 = tcg_temp_new();
+
+    tcg_gen_movi_tl(resultopt1, (target_ulong)-1);
+    tcg_gen_setcondi_tl(TCG_COND_EQ, cond2, source2, (target_ulong)(~0L));
+    tcg_gen_setcondi_tl(TCG_COND_EQ, cond1, source1,
+                        ((target_ulong)1) << (TARGET_LONG_BITS - 1));
+    tcg_gen_and_tl(cond1, cond1, cond2); /* cond1 = overflow */
+    tcg_gen_setcondi_tl(TCG_COND_EQ, cond2, source2, 0); /* cond2 = div 0 */
+    /* if div by zero, set source1 to -1, otherwise don't change */
+    tcg_gen_movcond_tl(TCG_COND_EQ, source1, cond2, zeroreg, source1,
+            resultopt1);
+    /* if overflow or div by zero, set source2 to 1, else don't change */
+    tcg_gen_or_tl(cond1, cond1, cond2);
+    tcg_gen_movi_tl(resultopt1, (target_ulong)1);
+    tcg_gen_movcond_tl(TCG_COND_EQ, source2, cond1, zeroreg, source2,
+            resultopt1);
+    tcg_gen_div_tl(ret, source1, source2);
+
+    tcg_temp_free(cond1);
+    tcg_temp_free(cond2);
+    tcg_temp_free(zeroreg);
+    tcg_temp_free(resultopt1);
+}
+
+static void gen_divu(TCGv ret, TCGv source1, TCGv source2)
+{
+    TCGv cond1, zeroreg, resultopt1;
+    cond1 = tcg_temp_new();
+
+    zeroreg = tcg_const_tl(0);
+    resultopt1 = tcg_temp_new();
+
+    tcg_gen_setcondi_tl(TCG_COND_EQ, cond1, source2, 0);
+    tcg_gen_movi_tl(resultopt1, (target_ulong)-1);
+    tcg_gen_movcond_tl(TCG_COND_EQ, source1, cond1, zeroreg, source1,
+            resultopt1);
+    tcg_gen_movi_tl(resultopt1, (target_ulong)1);
+    tcg_gen_movcond_tl(TCG_COND_EQ, source2, cond1, zeroreg, source2,
+            resultopt1);
+    tcg_gen_divu_tl(ret, source1, source2);
+
+    tcg_temp_free(cond1);
+    tcg_temp_free(zeroreg);
+    tcg_temp_free(resultopt1);
+}
+
+static void gen_rem(TCGv ret, TCGv source1, TCGv source2)
+{
+    TCGv cond1, cond2, zeroreg, resultopt1;
+
+    cond1 = tcg_temp_new();
+    cond2 = tcg_temp_new();
+    zeroreg = tcg_const_tl(0);
+    resultopt1 = tcg_temp_new();
+
+    tcg_gen_movi_tl(resultopt1, 1L);
+    tcg_gen_setcondi_tl(TCG_COND_EQ, cond2, source2, (target_ulong)-1);
+    tcg_gen_setcondi_tl(TCG_COND_EQ, cond1, source1,
+                        (target_ulong)1 << (TARGET_LONG_BITS - 1));
+    tcg_gen_and_tl(cond2, cond1, cond2); /* cond1 = overflow */
+    tcg_gen_setcondi_tl(TCG_COND_EQ, cond1, source2, 0); /* cond2 = div 0 */
+    /* if overflow or div by zero, set source2 to 1, else don't change */
+    tcg_gen_or_tl(cond2, cond1, cond2);
+    tcg_gen_movcond_tl(TCG_COND_EQ, source2, cond2, zeroreg, source2,
+            resultopt1);
+    tcg_gen_rem_tl(resultopt1, source1, source2);
+    /* if div by zero, just return the original dividend */
+    tcg_gen_movcond_tl(TCG_COND_EQ, ret, cond1, zeroreg, resultopt1,
+            source1);
+
+    tcg_temp_free(cond1);
+    tcg_temp_free(cond2);
+    tcg_temp_free(zeroreg);
+    tcg_temp_free(resultopt1);
+}
+
+static void gen_remu(TCGv ret, TCGv source1, TCGv source2)
+{
+    TCGv cond1, zeroreg, resultopt1;
+    cond1 = tcg_temp_new();
+    zeroreg = tcg_const_tl(0);
+    resultopt1 = tcg_temp_new();
+
+    tcg_gen_movi_tl(resultopt1, (target_ulong)1);
+    tcg_gen_setcondi_tl(TCG_COND_EQ, cond1, source2, 0);
+    tcg_gen_movcond_tl(TCG_COND_EQ, source2, cond1, zeroreg, source2,
+            resultopt1);
+    tcg_gen_remu_tl(resultopt1, source1, source2);
+    /* if div by zero, just return the original dividend */
+    tcg_gen_movcond_tl(TCG_COND_EQ, ret, cond1, zeroreg, resultopt1,
+            source1);
+
+    tcg_temp_free(cond1);
+    tcg_temp_free(zeroreg);
+    tcg_temp_free(resultopt1);
 }
 
 static void gen_jal(DisasContext *ctx, int rd, target_ulong imm)
@@ -681,6 +600,33 @@ static void gen_subw(TCGv ret, TCGv arg1, TCGv arg2)
     tcg_gen_ext32s_tl(ret, ret);
 }
 
+static void gen_mulw(TCGv ret, TCGv arg1, TCGv arg2)
+{
+    tcg_gen_mul_tl(ret, arg1, arg2);
+    tcg_gen_ext32s_tl(ret, ret);
+}
+
+static bool gen_arith_div_w(DisasContext *ctx, arg_r *a,
+                            void(*func)(TCGv, TCGv, TCGv))
+{
+    TCGv source1, source2;
+    source1 = tcg_temp_new();
+    source2 = tcg_temp_new();
+
+    gen_get_gpr(source1, a->rs1);
+    gen_get_gpr(source2, a->rs2);
+    tcg_gen_ext32s_tl(source1, source1);
+    tcg_gen_ext32s_tl(source2, source2);
+
+    (*func)(source1, source1, source2);
+
+    tcg_gen_ext32s_tl(source1, source1);
+    gen_set_gpr(a->rd, source1);
+    tcg_temp_free(source1);
+    tcg_temp_free(source2);
+    return true;
+}
+
 #endif
 
 static bool trans_arith(DisasContext *ctx, arg_r *a,
-- 
2.19.2

[Qemu-riscv] [PULL 27/29] target/riscv: Rename trans_arith to gen_arith

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
---
 target/riscv/insn_trans/trans_rvi.inc.c | 18 +++++++++---------
 target/riscv/insn_trans/trans_rvm.inc.c | 14 +++++++-------
 target/riscv/translate.c                |  4 ++--
 3 files changed, 18 insertions(+), 18 deletions(-)

diff --git a/target/riscv/insn_trans/trans_rvi.inc.c b/target/riscv/insn_trans/trans_rvi.inc.c
index 88ef0003ec17..d420a4d8b2e9 100644
--- a/target/riscv/insn_trans/trans_rvi.inc.c
+++ b/target/riscv/insn_trans/trans_rvi.inc.c
@@ -307,12 +307,12 @@ static bool trans_srai(DisasContext *ctx, arg_srai *a)
 
 static bool trans_add(DisasContext *ctx, arg_add *a)
 {
-    return trans_arith(ctx, a, &tcg_gen_add_tl);
+    return gen_arith(ctx, a, &tcg_gen_add_tl);
 }
 
 static bool trans_sub(DisasContext *ctx, arg_sub *a)
 {
-    return trans_arith(ctx, a, &tcg_gen_sub_tl);
+    return gen_arith(ctx, a, &tcg_gen_sub_tl);
 }
 
 static bool trans_sll(DisasContext *ctx, arg_sll *a)
@@ -322,17 +322,17 @@ static bool trans_sll(DisasContext *ctx, arg_sll *a)
 
 static bool trans_slt(DisasContext *ctx, arg_slt *a)
 {
-    return trans_arith(ctx, a, &gen_slt);
+    return gen_arith(ctx, a, &gen_slt);
 }
 
 static bool trans_sltu(DisasContext *ctx, arg_sltu *a)
 {
-    return trans_arith(ctx, a, &gen_sltu);
+    return gen_arith(ctx, a, &gen_sltu);
 }
 
 static bool trans_xor(DisasContext *ctx, arg_xor *a)
 {
-    return trans_arith(ctx, a, &tcg_gen_xor_tl);
+    return gen_arith(ctx, a, &tcg_gen_xor_tl);
 }
 
 static bool trans_srl(DisasContext *ctx, arg_srl *a)
@@ -347,12 +347,12 @@ static bool trans_sra(DisasContext *ctx, arg_sra *a)
 
 static bool trans_or(DisasContext *ctx, arg_or *a)
 {
-    return trans_arith(ctx, a, &tcg_gen_or_tl);
+    return gen_arith(ctx, a, &tcg_gen_or_tl);
 }
 
 static bool trans_and(DisasContext *ctx, arg_and *a)
 {
-    return trans_arith(ctx, a, &tcg_gen_and_tl);
+    return gen_arith(ctx, a, &tcg_gen_and_tl);
 }
 
 #ifdef TARGET_RISCV64
@@ -399,12 +399,12 @@ static bool trans_sraiw(DisasContext *ctx, arg_sraiw *a)
 
 static bool trans_addw(DisasContext *ctx, arg_addw *a)
 {
-    return trans_arith(ctx, a, &gen_addw);
+    return gen_arith(ctx, a, &gen_addw);
 }
 
 static bool trans_subw(DisasContext *ctx, arg_subw *a)
 {
-    return trans_arith(ctx, a, &gen_subw);
+    return gen_arith(ctx, a, &gen_subw);
 }
 
 static bool trans_sllw(DisasContext *ctx, arg_sllw *a)
diff --git a/target/riscv/insn_trans/trans_rvm.inc.c b/target/riscv/insn_trans/trans_rvm.inc.c
index d2bf2f171904..204af225f8f3 100644
--- a/target/riscv/insn_trans/trans_rvm.inc.c
+++ b/target/riscv/insn_trans/trans_rvm.inc.c
@@ -22,7 +22,7 @@
 static bool trans_mul(DisasContext *ctx, arg_mul *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    return trans_arith(ctx, a, &tcg_gen_mul_tl);
+    return gen_arith(ctx, a, &tcg_gen_mul_tl);
 }
 
 static bool trans_mulh(DisasContext *ctx, arg_mulh *a)
@@ -44,7 +44,7 @@ static bool trans_mulh(DisasContext *ctx, arg_mulh *a)
 static bool trans_mulhsu(DisasContext *ctx, arg_mulhsu *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    return trans_arith(ctx, a, &gen_mulhsu);
+    return gen_arith(ctx, a, &gen_mulhsu);
 }
 
 static bool trans_mulhu(DisasContext *ctx, arg_mulhu *a)
@@ -66,32 +66,32 @@ static bool trans_mulhu(DisasContext *ctx, arg_mulhu *a)
 static bool trans_div(DisasContext *ctx, arg_div *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    return trans_arith(ctx, a, &gen_div);
+    return gen_arith(ctx, a, &gen_div);
 }
 
 static bool trans_divu(DisasContext *ctx, arg_divu *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    return trans_arith(ctx, a, &gen_divu);
+    return gen_arith(ctx, a, &gen_divu);
 }
 
 static bool trans_rem(DisasContext *ctx, arg_rem *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    return trans_arith(ctx, a, &gen_rem);
+    return gen_arith(ctx, a, &gen_rem);
 }
 
 static bool trans_remu(DisasContext *ctx, arg_remu *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    return trans_arith(ctx, a, &gen_remu);
+    return gen_arith(ctx, a, &gen_remu);
 }
 
 #ifdef TARGET_RISCV64
 static bool trans_mulw(DisasContext *ctx, arg_mulw *a)
 {
     REQUIRE_EXT(ctx, RVM);
-    return trans_arith(ctx, a, &gen_mulw);
+    return gen_arith(ctx, a, &gen_mulw);
 }
 
 static bool trans_divw(DisasContext *ctx, arg_divw *a)
diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 3cd7e16c63cf..dedf4189d5b7 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -629,8 +629,8 @@ static bool gen_arith_div_w(DisasContext *ctx, arg_r *a,
 
 #endif
 
-static bool trans_arith(DisasContext *ctx, arg_r *a,
-                        void(*func)(TCGv, TCGv, TCGv))
+static bool gen_arith(DisasContext *ctx, arg_r *a,
+                      void(*func)(TCGv, TCGv, TCGv))
 {
     TCGv source1, source2;
     source1 = tcg_temp_new();
-- 
2.19.2

[Qemu-riscv] [PULL 28/29] target/riscv: Remove gen_system()

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

with all 16 bit insns moved to decodetree no path is falling back to
gen_system(), so we can remove it.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/translate.c | 34 ----------------------------------
 1 file changed, 34 deletions(-)

diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index dedf4189d5b7..92be090bc7bb 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -473,33 +473,6 @@ static void gen_set_rm(DisasContext *ctx, int rm)
     tcg_temp_free_i32(t0);
 }
 
-static void gen_system(DisasContext *ctx, uint32_t opc, int rd, int rs1,
-                       int csr)
-{
-    tcg_gen_movi_tl(cpu_pc, ctx->base.pc_next);
-
-    switch (opc) {
-    case OPC_RISC_ECALL:
-        switch (csr) {
-        case 0x0: /* ECALL */
-            /* always generates U-level ECALL, fixed in do_interrupt handler */
-            generate_exception(ctx, RISCV_EXCP_U_ECALL);
-            tcg_gen_exit_tb(NULL, 0); /* no chaining */
-            ctx->base.is_jmp = DISAS_NORETURN;
-            break;
-        case 0x1: /* EBREAK */
-            generate_exception(ctx, RISCV_EXCP_BREAKPOINT);
-            tcg_gen_exit_tb(NULL, 0); /* no chaining */
-            ctx->base.is_jmp = DISAS_NORETURN;
-            break;
-        default:
-            gen_exception_illegal(ctx);
-            break;
-        }
-        break;
-    }
-}
-
 static void decode_RV32_64C0(DisasContext *ctx)
 {
     uint8_t funct3 = extract32(ctx->opcode, 13, 3);
@@ -680,7 +653,6 @@ bool decode_insn16(DisasContext *ctx, uint16_t insn);
 
 static void decode_RV32_64G(DisasContext *ctx)
 {
-    int rs1, rd;
     uint32_t op;
 
     /* We do not do misaligned address check here: the address should never be
@@ -689,14 +661,8 @@ static void decode_RV32_64G(DisasContext *ctx)
      * perform the misaligned instruction fetch */
 
     op = MASK_OP_MAJOR(ctx->opcode);
-    rs1 = GET_RS1(ctx->opcode);
-    rd = GET_RD(ctx->opcode);
 
     switch (op) {
-    case OPC_RISC_SYSTEM:
-        gen_system(ctx, MASK_OP_SYSTEM(ctx->opcode), rd, rs1,
-                   (ctx->opcode & 0xFFF00000) >> 20);
-        break;
     default:
         gen_exception_illegal(ctx);
         break;
-- 
2.19.2

[Qemu-riscv] [PULL 29/29] target/riscv: Remove decode_RV32_64G()

[Palmer Dabbelt]

From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>

decodetree handles all instructions now so the fallback is not necessary
anymore.

Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
---
 target/riscv/translate.c | 21 +--------------------
 1 file changed, 1 insertion(+), 20 deletions(-)

diff --git a/target/riscv/translate.c b/target/riscv/translate.c
index 92be090bc7bb..049fa65c6611 100644
--- a/target/riscv/translate.c
+++ b/target/riscv/translate.c
@@ -651,24 +651,6 @@ bool decode_insn16(DisasContext *ctx, uint16_t insn);
 #include "decode_insn16.inc.c"
 #include "insn_trans/trans_rvc.inc.c"
 
-static void decode_RV32_64G(DisasContext *ctx)
-{
-    uint32_t op;
-
-    /* We do not do misaligned address check here: the address should never be
-     * misaligned at this point. Instructions that set PC must do the check,
-     * since epc must be the address of the instruction that caused us to
-     * perform the misaligned instruction fetch */
-
-    op = MASK_OP_MAJOR(ctx->opcode);
-
-    switch (op) {
-    default:
-        gen_exception_illegal(ctx);
-        break;
-    }
-}
-
 static void decode_opc(DisasContext *ctx)
 {
     /* check for compressed insn */
@@ -685,8 +667,7 @@ static void decode_opc(DisasContext *ctx)
     } else {
         ctx->pc_succ_insn = ctx->base.pc_next + 4;
         if (!decode_insn32(ctx, ctx->opcode)) {
-            /* fallback to old decoder */
-            decode_RV32_64G(ctx);
+            gen_exception_illegal(ctx);
         }
     }
 }
-- 
2.19.2

Re: [Qemu-riscv] [PULL] target/riscv: Convert to decodetree

[Peter Maydell]

On Wed, 13 Mar 2019 at 14:37, Palmer Dabbelt <palmer@sifive.com> wrote:
>
> merged tag 'pull-request-2019-03-12'
> Primary key fingerprint: 27B8 8847 EEE0 2501 18F3  EAB9 2ED9 D774 FE70 2DB5
> The following changes since commit 3f3bbfc7cef4490c5ed5550766a81e7d18f08db1:
>
>   Merge remote-tracking branch 'remotes/huth-gitlab/tags/pull-request-2019-03-12' into staging (2019-03-12 21:06:26 +0000)
>
> are available in the Git repository at:
>
>   git://github.com/palmer-dabbelt/qemu.git tags/riscv-for-master-4.0-sf4
>
> for you to fetch changes up to 25e6ca30c668783cd72ff97080ff44e141b99f9b:
>
>   target/riscv: Remove decode_RV32_64G() (2019-03-13 10:40:50 +0100)
Applied, thanks.

Please update the changelog at https://wiki.qemu.org/ChangeLog/4.0
for any user-visible changes.

-- PMM

Re: [Qemu-riscv] [Qemu-devel] [PULL 17/29] target/riscv: Convert quadrant 1 of RVXC insns to decodetree

[Alistair Francis]

On Wed, Mar 13, 2019 at 7:53 AM Palmer Dabbelt <palmer@sifive.com> wrote:
>
> From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>
> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
> Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
> Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>

This commit is the first bad commit in breaking 32-bit boot.

It looks like the jal doesn't jump to the correct address:

----------------
IN:
0x80000022:  00050433          add             s0,a0,zero
0x80000026:  000584b3          add             s1,a1,zero
0x8000002a:  2c79              jal             ra,670          # 0x800002c8

----------------
IN:
0x800002c8:  00000533          add             a0,zero,zero
0x800002cc:  8082              ret


Alistair

> ---
>  target/riscv/insn16.decode              |  43 +++++++
>  target/riscv/insn_trans/trans_rvc.inc.c | 151 ++++++++++++++++++++++++
>  target/riscv/translate.c                | 118 +-----------------
>  3 files changed, 195 insertions(+), 117 deletions(-)
>
> diff --git a/target/riscv/insn16.decode b/target/riscv/insn16.decode
> index 558c0c41f0b5..45109301c6d4 100644
> --- a/target/riscv/insn16.decode
> +++ b/target/riscv/insn16.decode
> @@ -22,28 +22,53 @@
>  %rs2_3     2:3                !function=ex_rvc_register
>
>  # Immediates:
> +%imm_ci        12:s1 2:5
>  %nzuimm_ciw    7:4 11:2 5:1 6:1   !function=ex_shift_2
>  %uimm_cl_d     5:2 10:3           !function=ex_shift_3
>  %uimm_cl_w     5:1 10:3 6:1       !function=ex_shift_2
> +%imm_cb        12:s1 5:2 2:1 10:2 3:2 !function=ex_shift_1
> +%imm_cj        12:s1 8:1 9:2 6:1 7:1 2:1 11:1 3:3 !function=ex_shift_1
> +
> +%nzuimm_6bit   12:1 2:5
> +
> +%imm_addi16sp  12:s1 3:2 5:1 2:1 6:1 !function=ex_shift_4
> +%imm_lui       12:s1 2:5             !function=ex_shift_12
> +
>
>
>  # Argument sets:
>  &cl               rs1 rd
>  &cl_dw     uimm   rs1 rd
> +&ci        imm        rd
>  &ciw       nzuimm     rd
>  &cs               rs1 rs2
>  &cs_dw     uimm   rs1 rs2
> +&cb        imm    rs1
> +&cr               rd  rs2
> +&cj       imm
> +&c_shift   shamt      rd
> +
>
> +&caddi16sp_lui  imm_lui imm_addi16sp rd
>
>  # Formats 16:
> +@ci        ... . ..... .....  .. &ci     imm=%imm_ci                  %rd
>  @ciw       ...   ........ ... .. &ciw    nzuimm=%nzuimm_ciw           rd=%rs2_3
>  @cl_d      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rd=%rs2_3
>  @cl_w      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rd=%rs2_3
>  @cl        ... ... ... .. ... .. &cl                      rs1=%rs1_3  rd=%rs2_3
>  @cs        ... ... ... .. ... .. &cs                      rs1=%rs1_3  rs2=%rs2_3
> +@cs_2      ... ... ... .. ... .. &cr                      rd=%rs1_3   rs2=%rs2_3
>  @cs_d      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rs2=%rs2_3
>  @cs_w      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rs2=%rs2_3
> +@cb        ... ... ... .. ... .. &cb     imm=%imm_cb      rs1=%rs1_3
> +@cj        ...    ........... .. &cj     imm=%imm_cj
>
> +@c_addi16sp_lui ... .  ..... ..... .. &caddi16sp_lui %imm_lui %imm_addi16sp %rd
> +
> +@c_shift        ... . .. ... ..... .. &c_shift rd=%rs1_3 shamt=%nzuimm_6bit
> +
> +@c_andi         ... . .. ... ..... .. &ci imm=%imm_ci rd=%rs1_3
>
>  # *** RV64C Standard Extension (Quadrant 0) ***
>  c_addi4spn        000    ........ ... 00 @ciw
> @@ -53,3 +78,21 @@ c_flw_ld          011  --- ... -- ... 00 @cl    #Note: Must parse uimm manually
>  c_fsd             101  ... ... .. ... 00 @cs_d
>  c_sw              110  ... ... .. ... 00 @cs_w
>  c_fsw_sd          111  --- ... -- ... 00 @cs    #Note: Must parse uimm manually
> +
> +# *** RV64C Standard Extension (Quadrant 1) ***
> +c_addi            000 .  .....  ..... 01 @ci
> +c_jal_addiw       001 .  .....  ..... 01 @ci #Note: parse rd and/or imm manually
> +c_li              010 .  .....  ..... 01 @ci
> +c_addi16sp_lui    011 .  .....  ..... 01 @c_addi16sp_lui # shares opc with C.LUI
> +c_srli            100 . 00 ...  ..... 01 @c_shift
> +c_srai            100 . 01 ...  ..... 01 @c_shift
> +c_andi            100 . 10 ...  ..... 01 @c_andi
> +c_sub             100 0 11 ... 00 ... 01 @cs_2
> +c_xor             100 0 11 ... 01 ... 01 @cs_2
> +c_or              100 0 11 ... 10 ... 01 @cs_2
> +c_and             100 0 11 ... 11 ... 01 @cs_2
> +c_subw            100 1 11 ... 00 ... 01 @cs_2
> +c_addw            100 1 11 ... 01 ... 01 @cs_2
> +c_j               101     ........... 01 @cj
> +c_beqz            110  ... ...  ..... 01 @cb
> +c_bnez            111  ... ...  ..... 01 @cb
> diff --git a/target/riscv/insn_trans/trans_rvc.inc.c b/target/riscv/insn_trans/trans_rvc.inc.c
> index 93ec8aa30b95..b06c435c9800 100644
> --- a/target/riscv/insn_trans/trans_rvc.inc.c
> +++ b/target/riscv/insn_trans/trans_rvc.inc.c
> @@ -73,3 +73,154 @@ static bool trans_c_fsw_sd(DisasContext *ctx, arg_c_fsw_sd *a)
>      return false;
>  #endif
>  }
> +
> +static bool trans_c_addi(DisasContext *ctx, arg_c_addi *a)
> +{
> +    if (a->imm == 0) {
> +        /* Hint: insn is valid but does not affect state */
> +        return true;
> +    }
> +    arg_addi arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
> +    return trans_addi(ctx, &arg);
> +}
> +
> +static bool trans_c_jal_addiw(DisasContext *ctx, arg_c_jal_addiw *a)
> +{
> +#ifdef TARGET_RISCV32
> +    /* C.JAL */
> +    arg_jal arg = { .rd = 1, .imm = a->imm };
> +    return trans_jal(ctx, &arg);
> +#else
> +    /* C.ADDIW */
> +    arg_addiw arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
> +    return trans_addiw(ctx, &arg);
> +#endif
> +}
> +
> +static bool trans_c_li(DisasContext *ctx, arg_c_li *a)
> +{
> +    if (a->rd == 0) {
> +        /* Hint: insn is valid but does not affect state */
> +        return true;
> +    }
> +    arg_addi arg = { .rd = a->rd, .rs1 = 0, .imm = a->imm };
> +    return trans_addi(ctx, &arg);
> +}
> +
> +static bool trans_c_addi16sp_lui(DisasContext *ctx, arg_c_addi16sp_lui *a)
> +{
> +    if (a->rd == 2) {
> +        /* C.ADDI16SP */
> +        arg_addi arg = { .rd = 2, .rs1 = 2, .imm = a->imm_addi16sp };
> +        return trans_addi(ctx, &arg);
> +    } else if (a->imm_lui != 0) {
> +        /* C.LUI */
> +        if (a->rd == 0) {
> +            /* Hint: insn is valid but does not affect state */
> +            return true;
> +        }
> +        arg_lui arg = { .rd = a->rd, .imm = a->imm_lui };
> +        return trans_lui(ctx, &arg);
> +    }
> +    return false;
> +}
> +
> +static bool trans_c_srli(DisasContext *ctx, arg_c_srli *a)
> +{
> +    int shamt = a->shamt;
> +    if (shamt == 0) {
> +        /* For RV128 a shamt of 0 means a shift by 64 */
> +        shamt = 64;
> +    }
> +    /* Ensure, that shamt[5] is zero for RV32 */
> +    if (shamt >= TARGET_LONG_BITS) {
> +        return false;
> +    }
> +
> +    arg_srli arg = { .rd = a->rd, .rs1 = a->rd, .shamt = a->shamt };
> +    return trans_srli(ctx, &arg);
> +}
> +
> +static bool trans_c_srai(DisasContext *ctx, arg_c_srai *a)
> +{
> +    int shamt = a->shamt;
> +    if (shamt == 0) {
> +        /* For RV128 a shamt of 0 means a shift by 64 */
> +        shamt = 64;
> +    }
> +    /* Ensure, that shamt[5] is zero for RV32 */
> +    if (shamt >= TARGET_LONG_BITS) {
> +        return false;
> +    }
> +
> +    arg_srai arg = { .rd = a->rd, .rs1 = a->rd, .shamt = a->shamt };
> +    return trans_srai(ctx, &arg);
> +}
> +
> +static bool trans_c_andi(DisasContext *ctx, arg_c_andi *a)
> +{
> +    arg_andi arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
> +    return trans_andi(ctx, &arg);
> +}
> +
> +static bool trans_c_sub(DisasContext *ctx, arg_c_sub *a)
> +{
> +    arg_sub arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> +    return trans_sub(ctx, &arg);
> +}
> +
> +static bool trans_c_xor(DisasContext *ctx, arg_c_xor *a)
> +{
> +    arg_xor arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> +    return trans_xor(ctx, &arg);
> +}
> +
> +static bool trans_c_or(DisasContext *ctx, arg_c_or *a)
> +{
> +    arg_or arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> +    return trans_or(ctx, &arg);
> +}
> +
> +static bool trans_c_and(DisasContext *ctx, arg_c_and *a)
> +{
> +    arg_and arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> +    return trans_and(ctx, &arg);
> +}
> +
> +static bool trans_c_subw(DisasContext *ctx, arg_c_subw *a)
> +{
> +#ifdef TARGET_RISCV64
> +    arg_subw arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> +    return trans_subw(ctx, &arg);
> +#else
> +    return false;
> +#endif
> +}
> +
> +static bool trans_c_addw(DisasContext *ctx, arg_c_addw *a)
> +{
> +#ifdef TARGET_RISCV64
> +    arg_addw arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> +    return trans_addw(ctx, &arg);
> +#else
> +    return false;
> +#endif
> +}
> +
> +static bool trans_c_j(DisasContext *ctx, arg_c_j *a)
> +{
> +    arg_jal arg = { .rd = 0, .imm = a->imm };
> +    return trans_jal(ctx, &arg);
> +}
> +
> +static bool trans_c_beqz(DisasContext *ctx, arg_c_beqz *a)
> +{
> +    arg_beq arg = { .rs1 = a->rs1, .rs2 = 0, .imm = a->imm };
> +    return trans_beq(ctx, &arg);
> +}
> +
> +static bool trans_c_bnez(DisasContext *ctx, arg_c_bnez *a)
> +{
> +    arg_bne arg = { .rs1 = a->rs1, .rs2 = 0, .imm = a->imm };
> +    return trans_bne(ctx, &arg);
> +}
> diff --git a/target/riscv/translate.c b/target/riscv/translate.c
> index 0106fa8d51b3..a584c24fbf6b 100644
> --- a/target/riscv/translate.c
> +++ b/target/riscv/translate.c
> @@ -828,120 +828,6 @@ static void decode_RV32_64C0(DisasContext *ctx)
>      }
>  }
>
> -static void decode_RV32_64C1(DisasContext *ctx)
> -{
> -    uint8_t funct3 = extract32(ctx->opcode, 13, 3);
> -    uint8_t rd_rs1 = GET_C_RS1(ctx->opcode);
> -    uint8_t rs1s, rs2s;
> -    uint8_t funct2;
> -
> -    switch (funct3) {
> -    case 0:
> -        /* C.ADDI -> addi rd, rd, nzimm[5:0] */
> -        gen_arith_imm(ctx, OPC_RISC_ADDI, rd_rs1, rd_rs1,
> -                      GET_C_IMM(ctx->opcode));
> -        break;
> -    case 1:
> -#if defined(TARGET_RISCV64)
> -        /* C.ADDIW (RV64/128) -> addiw rd, rd, imm[5:0]*/
> -        gen_arith_imm(ctx, OPC_RISC_ADDIW, rd_rs1, rd_rs1,
> -                      GET_C_IMM(ctx->opcode));
> -#else
> -        /* C.JAL(RV32) -> jal x1, offset[11:1] */
> -        gen_jal(ctx, 1, GET_C_J_IMM(ctx->opcode));
> -#endif
> -        break;
> -    case 2:
> -        /* C.LI -> addi rd, x0, imm[5:0]*/
> -        gen_arith_imm(ctx, OPC_RISC_ADDI, rd_rs1, 0, GET_C_IMM(ctx->opcode));
> -        break;
> -    case 3:
> -        if (rd_rs1 == 2) {
> -            /* C.ADDI16SP -> addi x2, x2, nzimm[9:4]*/
> -            gen_arith_imm(ctx, OPC_RISC_ADDI, 2, 2,
> -                          GET_C_ADDI16SP_IMM(ctx->opcode));
> -        } else if (rd_rs1 != 0) {
> -            /* C.LUI (rs1/rd =/= {0,2}) -> lui rd, nzimm[17:12]*/
> -            tcg_gen_movi_tl(cpu_gpr[rd_rs1],
> -                            GET_C_IMM(ctx->opcode) << 12);
> -        }
> -        break;
> -    case 4:
> -        funct2 = extract32(ctx->opcode, 10, 2);
> -        rs1s = GET_C_RS1S(ctx->opcode);
> -        switch (funct2) {
> -        case 0: /* C.SRLI(RV32) -> srli rd', rd', shamt[5:0] */
> -            gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, rs1s, rs1s,
> -                               GET_C_ZIMM(ctx->opcode));
> -            /* C.SRLI64(RV128) */
> -            break;
> -        case 1:
> -            /* C.SRAI -> srai rd', rd', shamt[5:0]*/
> -            gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, rs1s, rs1s,
> -                            GET_C_ZIMM(ctx->opcode) | 0x400);
> -            /* C.SRAI64(RV128) */
> -            break;
> -        case 2:
> -            /* C.ANDI -> andi rd', rd', imm[5:0]*/
> -            gen_arith_imm(ctx, OPC_RISC_ANDI, rs1s, rs1s,
> -                          GET_C_IMM(ctx->opcode));
> -            break;
> -        case 3:
> -            funct2 = extract32(ctx->opcode, 5, 2);
> -            rs2s = GET_C_RS2S(ctx->opcode);
> -            switch (funct2) {
> -            case 0:
> -                /* C.SUB -> sub rd', rd', rs2' */
> -                if (extract32(ctx->opcode, 12, 1) == 0) {
> -                    gen_arith(ctx, OPC_RISC_SUB, rs1s, rs1s, rs2s);
> -                }
> -#if defined(TARGET_RISCV64)
> -                else {
> -                    gen_arith(ctx, OPC_RISC_SUBW, rs1s, rs1s, rs2s);
> -                }
> -#endif
> -                break;
> -            case 1:
> -                /* C.XOR -> xor rs1', rs1', rs2' */
> -                if (extract32(ctx->opcode, 12, 1) == 0) {
> -                    gen_arith(ctx, OPC_RISC_XOR, rs1s, rs1s, rs2s);
> -                }
> -#if defined(TARGET_RISCV64)
> -                else {
> -                    /* C.ADDW (RV64/128) */
> -                    gen_arith(ctx, OPC_RISC_ADDW, rs1s, rs1s, rs2s);
> -                }
> -#endif
> -                break;
> -            case 2:
> -                /* C.OR -> or rs1', rs1', rs2' */
> -                gen_arith(ctx, OPC_RISC_OR, rs1s, rs1s, rs2s);
> -                break;
> -            case 3:
> -                /* C.AND -> and rs1', rs1', rs2' */
> -                gen_arith(ctx, OPC_RISC_AND, rs1s, rs1s, rs2s);
> -                break;
> -            }
> -            break;
> -        }
> -        break;
> -    case 5:
> -        /* C.J -> jal x0, offset[11:1]*/
> -        gen_jal(ctx, 0, GET_C_J_IMM(ctx->opcode));
> -        break;
> -    case 6:
> -        /* C.BEQZ -> beq rs1', x0, offset[8:1]*/
> -        rs1s = GET_C_RS1S(ctx->opcode);
> -        gen_branch(ctx, OPC_RISC_BEQ, rs1s, 0, GET_C_B_IMM(ctx->opcode));
> -        break;
> -    case 7:
> -        /* C.BNEZ -> bne rs1', x0, offset[8:1]*/
> -        rs1s = GET_C_RS1S(ctx->opcode);
> -        gen_branch(ctx, OPC_RISC_BNE, rs1s, 0, GET_C_B_IMM(ctx->opcode));
> -        break;
> -    }
> -}
> -
>  static void decode_RV32_64C2(DisasContext *ctx)
>  {
>      uint8_t rd, rs2;
> @@ -1028,9 +914,6 @@ static void decode_RV32_64C(DisasContext *ctx)
>      case 0:
>          decode_RV32_64C0(ctx);
>          break;
> -    case 1:
> -        decode_RV32_64C1(ctx);
> -        break;
>      case 2:
>          decode_RV32_64C2(ctx);
>          break;
> @@ -1045,6 +928,7 @@ static void decode_RV32_64C(DisasContext *ctx)
>  EX_SH(1)
>  EX_SH(2)
>  EX_SH(3)
> +EX_SH(4)
>  EX_SH(12)
>
>  #define REQUIRE_EXT(ctx, ext) do { \
> --
> 2.19.2
>
>

Re: [Qemu-riscv] [Qemu-devel] [PULL 17/29] target/riscv: Convert quadrant 1 of RVXC insns to decodetree

[Palmer Dabbelt]

On Thu, 14 Mar 2019 13:28:37 PDT (-0700), alistair23@gmail.com wrote:
> On Wed, Mar 13, 2019 at 7:53 AM Palmer Dabbelt <palmer@sifive.com> wrote:
>>
>> From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>>
>> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
>> Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>> Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
>
> This commit is the first bad commit in breaking 32-bit boot.
>
> It looks like the jal doesn't jump to the correct address:
>
> ----------------
> IN:
> 0x80000022:  00050433          add             s0,a0,zero
> 0x80000026:  000584b3          add             s1,a1,zero
> 0x8000002a:  2c79              jal             ra,670          # 0x800002c8
>
> ----------------
> IN:
> 0x800002c8:  00000533          add             a0,zero,zero
> 0x800002cc:  8082              ret


Sorry, for some reason I thought you'd tested this on 32-bit?  Do you have a 
workflow that I can use to reproduce the bug?

>
>
> Alistair
>
>> ---
>>  target/riscv/insn16.decode              |  43 +++++++
>>  target/riscv/insn_trans/trans_rvc.inc.c | 151 ++++++++++++++++++++++++
>>  target/riscv/translate.c                | 118 +-----------------
>>  3 files changed, 195 insertions(+), 117 deletions(-)
>>
>> diff --git a/target/riscv/insn16.decode b/target/riscv/insn16.decode
>> index 558c0c41f0b5..45109301c6d4 100644
>> --- a/target/riscv/insn16.decode
>> +++ b/target/riscv/insn16.decode
>> @@ -22,28 +22,53 @@
>>  %rs2_3     2:3                !function=ex_rvc_register
>>
>>  # Immediates:
>> +%imm_ci        12:s1 2:5
>>  %nzuimm_ciw    7:4 11:2 5:1 6:1   !function=ex_shift_2
>>  %uimm_cl_d     5:2 10:3           !function=ex_shift_3
>>  %uimm_cl_w     5:1 10:3 6:1       !function=ex_shift_2
>> +%imm_cb        12:s1 5:2 2:1 10:2 3:2 !function=ex_shift_1
>> +%imm_cj        12:s1 8:1 9:2 6:1 7:1 2:1 11:1 3:3 !function=ex_shift_1
>> +
>> +%nzuimm_6bit   12:1 2:5
>> +
>> +%imm_addi16sp  12:s1 3:2 5:1 2:1 6:1 !function=ex_shift_4
>> +%imm_lui       12:s1 2:5             !function=ex_shift_12
>> +
>>
>>
>>  # Argument sets:
>>  &cl               rs1 rd
>>  &cl_dw     uimm   rs1 rd
>> +&ci        imm        rd
>>  &ciw       nzuimm     rd
>>  &cs               rs1 rs2
>>  &cs_dw     uimm   rs1 rs2
>> +&cb        imm    rs1
>> +&cr               rd  rs2
>> +&cj       imm
>> +&c_shift   shamt      rd
>> +
>>
>> +&caddi16sp_lui  imm_lui imm_addi16sp rd
>>
>>  # Formats 16:
>> +@ci        ... . ..... .....  .. &ci     imm=%imm_ci                  %rd
>>  @ciw       ...   ........ ... .. &ciw    nzuimm=%nzuimm_ciw           rd=%rs2_3
>>  @cl_d      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rd=%rs2_3
>>  @cl_w      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rd=%rs2_3
>>  @cl        ... ... ... .. ... .. &cl                      rs1=%rs1_3  rd=%rs2_3
>>  @cs        ... ... ... .. ... .. &cs                      rs1=%rs1_3  rs2=%rs2_3
>> +@cs_2      ... ... ... .. ... .. &cr                      rd=%rs1_3   rs2=%rs2_3
>>  @cs_d      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rs2=%rs2_3
>>  @cs_w      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rs2=%rs2_3
>> +@cb        ... ... ... .. ... .. &cb     imm=%imm_cb      rs1=%rs1_3
>> +@cj        ...    ........... .. &cj     imm=%imm_cj
>>
>> +@c_addi16sp_lui ... .  ..... ..... .. &caddi16sp_lui %imm_lui %imm_addi16sp %rd
>> +
>> +@c_shift        ... . .. ... ..... .. &c_shift rd=%rs1_3 shamt=%nzuimm_6bit
>> +
>> +@c_andi         ... . .. ... ..... .. &ci imm=%imm_ci rd=%rs1_3
>>
>>  # *** RV64C Standard Extension (Quadrant 0) ***
>>  c_addi4spn        000    ........ ... 00 @ciw
>> @@ -53,3 +78,21 @@ c_flw_ld          011  --- ... -- ... 00 @cl    #Note: Must parse uimm manually
>>  c_fsd             101  ... ... .. ... 00 @cs_d
>>  c_sw              110  ... ... .. ... 00 @cs_w
>>  c_fsw_sd          111  --- ... -- ... 00 @cs    #Note: Must parse uimm manually
>> +
>> +# *** RV64C Standard Extension (Quadrant 1) ***
>> +c_addi            000 .  .....  ..... 01 @ci
>> +c_jal_addiw       001 .  .....  ..... 01 @ci #Note: parse rd and/or imm manually
>> +c_li              010 .  .....  ..... 01 @ci
>> +c_addi16sp_lui    011 .  .....  ..... 01 @c_addi16sp_lui # shares opc with C.LUI
>> +c_srli            100 . 00 ...  ..... 01 @c_shift
>> +c_srai            100 . 01 ...  ..... 01 @c_shift
>> +c_andi            100 . 10 ...  ..... 01 @c_andi
>> +c_sub             100 0 11 ... 00 ... 01 @cs_2
>> +c_xor             100 0 11 ... 01 ... 01 @cs_2
>> +c_or              100 0 11 ... 10 ... 01 @cs_2
>> +c_and             100 0 11 ... 11 ... 01 @cs_2
>> +c_subw            100 1 11 ... 00 ... 01 @cs_2
>> +c_addw            100 1 11 ... 01 ... 01 @cs_2
>> +c_j               101     ........... 01 @cj
>> +c_beqz            110  ... ...  ..... 01 @cb
>> +c_bnez            111  ... ...  ..... 01 @cb
>> diff --git a/target/riscv/insn_trans/trans_rvc.inc.c b/target/riscv/insn_trans/trans_rvc.inc.c
>> index 93ec8aa30b95..b06c435c9800 100644
>> --- a/target/riscv/insn_trans/trans_rvc.inc.c
>> +++ b/target/riscv/insn_trans/trans_rvc.inc.c
>> @@ -73,3 +73,154 @@ static bool trans_c_fsw_sd(DisasContext *ctx, arg_c_fsw_sd *a)
>>      return false;
>>  #endif
>>  }
>> +
>> +static bool trans_c_addi(DisasContext *ctx, arg_c_addi *a)
>> +{
>> +    if (a->imm == 0) {
>> +        /* Hint: insn is valid but does not affect state */
>> +        return true;
>> +    }
>> +    arg_addi arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
>> +    return trans_addi(ctx, &arg);
>> +}
>> +
>> +static bool trans_c_jal_addiw(DisasContext *ctx, arg_c_jal_addiw *a)
>> +{
>> +#ifdef TARGET_RISCV32
>> +    /* C.JAL */
>> +    arg_jal arg = { .rd = 1, .imm = a->imm };
>> +    return trans_jal(ctx, &arg);
>> +#else
>> +    /* C.ADDIW */
>> +    arg_addiw arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
>> +    return trans_addiw(ctx, &arg);
>> +#endif
>> +}
>> +
>> +static bool trans_c_li(DisasContext *ctx, arg_c_li *a)
>> +{
>> +    if (a->rd == 0) {
>> +        /* Hint: insn is valid but does not affect state */
>> +        return true;
>> +    }
>> +    arg_addi arg = { .rd = a->rd, .rs1 = 0, .imm = a->imm };
>> +    return trans_addi(ctx, &arg);
>> +}
>> +
>> +static bool trans_c_addi16sp_lui(DisasContext *ctx, arg_c_addi16sp_lui *a)
>> +{
>> +    if (a->rd == 2) {
>> +        /* C.ADDI16SP */
>> +        arg_addi arg = { .rd = 2, .rs1 = 2, .imm = a->imm_addi16sp };
>> +        return trans_addi(ctx, &arg);
>> +    } else if (a->imm_lui != 0) {
>> +        /* C.LUI */
>> +        if (a->rd == 0) {
>> +            /* Hint: insn is valid but does not affect state */
>> +            return true;
>> +        }
>> +        arg_lui arg = { .rd = a->rd, .imm = a->imm_lui };
>> +        return trans_lui(ctx, &arg);
>> +    }
>> +    return false;
>> +}
>> +
>> +static bool trans_c_srli(DisasContext *ctx, arg_c_srli *a)
>> +{
>> +    int shamt = a->shamt;
>> +    if (shamt == 0) {
>> +        /* For RV128 a shamt of 0 means a shift by 64 */
>> +        shamt = 64;
>> +    }
>> +    /* Ensure, that shamt[5] is zero for RV32 */
>> +    if (shamt >= TARGET_LONG_BITS) {
>> +        return false;
>> +    }
>> +
>> +    arg_srli arg = { .rd = a->rd, .rs1 = a->rd, .shamt = a->shamt };
>> +    return trans_srli(ctx, &arg);
>> +}
>> +
>> +static bool trans_c_srai(DisasContext *ctx, arg_c_srai *a)
>> +{
>> +    int shamt = a->shamt;
>> +    if (shamt == 0) {
>> +        /* For RV128 a shamt of 0 means a shift by 64 */
>> +        shamt = 64;
>> +    }
>> +    /* Ensure, that shamt[5] is zero for RV32 */
>> +    if (shamt >= TARGET_LONG_BITS) {
>> +        return false;
>> +    }
>> +
>> +    arg_srai arg = { .rd = a->rd, .rs1 = a->rd, .shamt = a->shamt };
>> +    return trans_srai(ctx, &arg);
>> +}
>> +
>> +static bool trans_c_andi(DisasContext *ctx, arg_c_andi *a)
>> +{
>> +    arg_andi arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
>> +    return trans_andi(ctx, &arg);
>> +}
>> +
>> +static bool trans_c_sub(DisasContext *ctx, arg_c_sub *a)
>> +{
>> +    arg_sub arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> +    return trans_sub(ctx, &arg);
>> +}
>> +
>> +static bool trans_c_xor(DisasContext *ctx, arg_c_xor *a)
>> +{
>> +    arg_xor arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> +    return trans_xor(ctx, &arg);
>> +}
>> +
>> +static bool trans_c_or(DisasContext *ctx, arg_c_or *a)
>> +{
>> +    arg_or arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> +    return trans_or(ctx, &arg);
>> +}
>> +
>> +static bool trans_c_and(DisasContext *ctx, arg_c_and *a)
>> +{
>> +    arg_and arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> +    return trans_and(ctx, &arg);
>> +}
>> +
>> +static bool trans_c_subw(DisasContext *ctx, arg_c_subw *a)
>> +{
>> +#ifdef TARGET_RISCV64
>> +    arg_subw arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> +    return trans_subw(ctx, &arg);
>> +#else
>> +    return false;
>> +#endif
>> +}
>> +
>> +static bool trans_c_addw(DisasContext *ctx, arg_c_addw *a)
>> +{
>> +#ifdef TARGET_RISCV64
>> +    arg_addw arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> +    return trans_addw(ctx, &arg);
>> +#else
>> +    return false;
>> +#endif
>> +}
>> +
>> +static bool trans_c_j(DisasContext *ctx, arg_c_j *a)
>> +{
>> +    arg_jal arg = { .rd = 0, .imm = a->imm };
>> +    return trans_jal(ctx, &arg);
>> +}
>> +
>> +static bool trans_c_beqz(DisasContext *ctx, arg_c_beqz *a)
>> +{
>> +    arg_beq arg = { .rs1 = a->rs1, .rs2 = 0, .imm = a->imm };
>> +    return trans_beq(ctx, &arg);
>> +}
>> +
>> +static bool trans_c_bnez(DisasContext *ctx, arg_c_bnez *a)
>> +{
>> +    arg_bne arg = { .rs1 = a->rs1, .rs2 = 0, .imm = a->imm };
>> +    return trans_bne(ctx, &arg);
>> +}
>> diff --git a/target/riscv/translate.c b/target/riscv/translate.c
>> index 0106fa8d51b3..a584c24fbf6b 100644
>> --- a/target/riscv/translate.c
>> +++ b/target/riscv/translate.c
>> @@ -828,120 +828,6 @@ static void decode_RV32_64C0(DisasContext *ctx)
>>      }
>>  }
>>
>> -static void decode_RV32_64C1(DisasContext *ctx)
>> -{
>> -    uint8_t funct3 = extract32(ctx->opcode, 13, 3);
>> -    uint8_t rd_rs1 = GET_C_RS1(ctx->opcode);
>> -    uint8_t rs1s, rs2s;
>> -    uint8_t funct2;
>> -
>> -    switch (funct3) {
>> -    case 0:
>> -        /* C.ADDI -> addi rd, rd, nzimm[5:0] */
>> -        gen_arith_imm(ctx, OPC_RISC_ADDI, rd_rs1, rd_rs1,
>> -                      GET_C_IMM(ctx->opcode));
>> -        break;
>> -    case 1:
>> -#if defined(TARGET_RISCV64)
>> -        /* C.ADDIW (RV64/128) -> addiw rd, rd, imm[5:0]*/
>> -        gen_arith_imm(ctx, OPC_RISC_ADDIW, rd_rs1, rd_rs1,
>> -                      GET_C_IMM(ctx->opcode));
>> -#else
>> -        /* C.JAL(RV32) -> jal x1, offset[11:1] */
>> -        gen_jal(ctx, 1, GET_C_J_IMM(ctx->opcode));
>> -#endif
>> -        break;
>> -    case 2:
>> -        /* C.LI -> addi rd, x0, imm[5:0]*/
>> -        gen_arith_imm(ctx, OPC_RISC_ADDI, rd_rs1, 0, GET_C_IMM(ctx->opcode));
>> -        break;
>> -    case 3:
>> -        if (rd_rs1 == 2) {
>> -            /* C.ADDI16SP -> addi x2, x2, nzimm[9:4]*/
>> -            gen_arith_imm(ctx, OPC_RISC_ADDI, 2, 2,
>> -                          GET_C_ADDI16SP_IMM(ctx->opcode));
>> -        } else if (rd_rs1 != 0) {
>> -            /* C.LUI (rs1/rd =/= {0,2}) -> lui rd, nzimm[17:12]*/
>> -            tcg_gen_movi_tl(cpu_gpr[rd_rs1],
>> -                            GET_C_IMM(ctx->opcode) << 12);
>> -        }
>> -        break;
>> -    case 4:
>> -        funct2 = extract32(ctx->opcode, 10, 2);
>> -        rs1s = GET_C_RS1S(ctx->opcode);
>> -        switch (funct2) {
>> -        case 0: /* C.SRLI(RV32) -> srli rd', rd', shamt[5:0] */
>> -            gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, rs1s, rs1s,
>> -                               GET_C_ZIMM(ctx->opcode));
>> -            /* C.SRLI64(RV128) */
>> -            break;
>> -        case 1:
>> -            /* C.SRAI -> srai rd', rd', shamt[5:0]*/
>> -            gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, rs1s, rs1s,
>> -                            GET_C_ZIMM(ctx->opcode) | 0x400);
>> -            /* C.SRAI64(RV128) */
>> -            break;
>> -        case 2:
>> -            /* C.ANDI -> andi rd', rd', imm[5:0]*/
>> -            gen_arith_imm(ctx, OPC_RISC_ANDI, rs1s, rs1s,
>> -                          GET_C_IMM(ctx->opcode));
>> -            break;
>> -        case 3:
>> -            funct2 = extract32(ctx->opcode, 5, 2);
>> -            rs2s = GET_C_RS2S(ctx->opcode);
>> -            switch (funct2) {
>> -            case 0:
>> -                /* C.SUB -> sub rd', rd', rs2' */
>> -                if (extract32(ctx->opcode, 12, 1) == 0) {
>> -                    gen_arith(ctx, OPC_RISC_SUB, rs1s, rs1s, rs2s);
>> -                }
>> -#if defined(TARGET_RISCV64)
>> -                else {
>> -                    gen_arith(ctx, OPC_RISC_SUBW, rs1s, rs1s, rs2s);
>> -                }
>> -#endif
>> -                break;
>> -            case 1:
>> -                /* C.XOR -> xor rs1', rs1', rs2' */
>> -                if (extract32(ctx->opcode, 12, 1) == 0) {
>> -                    gen_arith(ctx, OPC_RISC_XOR, rs1s, rs1s, rs2s);
>> -                }
>> -#if defined(TARGET_RISCV64)
>> -                else {
>> -                    /* C.ADDW (RV64/128) */
>> -                    gen_arith(ctx, OPC_RISC_ADDW, rs1s, rs1s, rs2s);
>> -                }
>> -#endif
>> -                break;
>> -            case 2:
>> -                /* C.OR -> or rs1', rs1', rs2' */
>> -                gen_arith(ctx, OPC_RISC_OR, rs1s, rs1s, rs2s);
>> -                break;
>> -            case 3:
>> -                /* C.AND -> and rs1', rs1', rs2' */
>> -                gen_arith(ctx, OPC_RISC_AND, rs1s, rs1s, rs2s);
>> -                break;
>> -            }
>> -            break;
>> -        }
>> -        break;
>> -    case 5:
>> -        /* C.J -> jal x0, offset[11:1]*/
>> -        gen_jal(ctx, 0, GET_C_J_IMM(ctx->opcode));
>> -        break;
>> -    case 6:
>> -        /* C.BEQZ -> beq rs1', x0, offset[8:1]*/
>> -        rs1s = GET_C_RS1S(ctx->opcode);
>> -        gen_branch(ctx, OPC_RISC_BEQ, rs1s, 0, GET_C_B_IMM(ctx->opcode));
>> -        break;
>> -    case 7:
>> -        /* C.BNEZ -> bne rs1', x0, offset[8:1]*/
>> -        rs1s = GET_C_RS1S(ctx->opcode);
>> -        gen_branch(ctx, OPC_RISC_BNE, rs1s, 0, GET_C_B_IMM(ctx->opcode));
>> -        break;
>> -    }
>> -}
>> -
>>  static void decode_RV32_64C2(DisasContext *ctx)
>>  {
>>      uint8_t rd, rs2;
>> @@ -1028,9 +914,6 @@ static void decode_RV32_64C(DisasContext *ctx)
>>      case 0:
>>          decode_RV32_64C0(ctx);
>>          break;
>> -    case 1:
>> -        decode_RV32_64C1(ctx);
>> -        break;
>>      case 2:
>>          decode_RV32_64C2(ctx);
>>          break;
>> @@ -1045,6 +928,7 @@ static void decode_RV32_64C(DisasContext *ctx)
>>  EX_SH(1)
>>  EX_SH(2)
>>  EX_SH(3)
>> +EX_SH(4)
>>  EX_SH(12)
>>
>>  #define REQUIRE_EXT(ctx, ext) do { \
>> --
>> 2.19.2
>>
>>

Re: [Qemu-riscv] [Qemu-devel] [PULL 17/29] target/riscv: Convert quadrant 1 of RVXC insns to decodetree

[Alistair Francis]

On Thu, Mar 14, 2019 at 8:59 PM Palmer Dabbelt <palmer@sifive.com> wrote:
>
> On Thu, 14 Mar 2019 13:28:37 PDT (-0700), alistair23@gmail.com wrote:
> > On Wed, Mar 13, 2019 at 7:53 AM Palmer Dabbelt <palmer@sifive.com> wrote:
> >>
> >> From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
> >>
> >> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
> >> Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
> >> Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
> >
> > This commit is the first bad commit in breaking 32-bit boot.
> >
> > It looks like the jal doesn't jump to the correct address:
> >
> > ----------------
> > IN:
> > 0x80000022:  00050433          add             s0,a0,zero
> > 0x80000026:  000584b3          add             s1,a1,zero
> > 0x8000002a:  2c79              jal             ra,670          # 0x800002c8
> >
> > ----------------
> > IN:
> > 0x800002c8:  00000533          add             a0,zero,zero
> > 0x800002cc:  8082              ret
>
>
> Sorry, for some reason I thought you'd tested this on 32-bit?  Do you have a
> workflow that I can use to reproduce the bug?

I did! Not sure what happened there. Maybe something changed between
what I tested and what was merged or somehow I tested the wrong thing.

Just booting 32-bit OpenSBI fails. It seems that the imm for jal is
wrong. I'll keep digging into it.

Alistair

>
> >
> >
> > Alistair
> >
> >> ---
> >>  target/riscv/insn16.decode              |  43 +++++++
> >>  target/riscv/insn_trans/trans_rvc.inc.c | 151 ++++++++++++++++++++++++
> >>  target/riscv/translate.c                | 118 +-----------------
> >>  3 files changed, 195 insertions(+), 117 deletions(-)
> >>
> >> diff --git a/target/riscv/insn16.decode b/target/riscv/insn16.decode
> >> index 558c0c41f0b5..45109301c6d4 100644
> >> --- a/target/riscv/insn16.decode
> >> +++ b/target/riscv/insn16.decode
> >> @@ -22,28 +22,53 @@
> >>  %rs2_3     2:3                !function=ex_rvc_register
> >>
> >>  # Immediates:
> >> +%imm_ci        12:s1 2:5
> >>  %nzuimm_ciw    7:4 11:2 5:1 6:1   !function=ex_shift_2
> >>  %uimm_cl_d     5:2 10:3           !function=ex_shift_3
> >>  %uimm_cl_w     5:1 10:3 6:1       !function=ex_shift_2
> >> +%imm_cb        12:s1 5:2 2:1 10:2 3:2 !function=ex_shift_1
> >> +%imm_cj        12:s1 8:1 9:2 6:1 7:1 2:1 11:1 3:3 !function=ex_shift_1
> >> +
> >> +%nzuimm_6bit   12:1 2:5
> >> +
> >> +%imm_addi16sp  12:s1 3:2 5:1 2:1 6:1 !function=ex_shift_4
> >> +%imm_lui       12:s1 2:5             !function=ex_shift_12
> >> +
> >>
> >>
> >>  # Argument sets:
> >>  &cl               rs1 rd
> >>  &cl_dw     uimm   rs1 rd
> >> +&ci        imm        rd
> >>  &ciw       nzuimm     rd
> >>  &cs               rs1 rs2
> >>  &cs_dw     uimm   rs1 rs2
> >> +&cb        imm    rs1
> >> +&cr               rd  rs2
> >> +&cj       imm
> >> +&c_shift   shamt      rd
> >> +
> >>
> >> +&caddi16sp_lui  imm_lui imm_addi16sp rd
> >>
> >>  # Formats 16:
> >> +@ci        ... . ..... .....  .. &ci     imm=%imm_ci                  %rd
> >>  @ciw       ...   ........ ... .. &ciw    nzuimm=%nzuimm_ciw           rd=%rs2_3
> >>  @cl_d      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rd=%rs2_3
> >>  @cl_w      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rd=%rs2_3
> >>  @cl        ... ... ... .. ... .. &cl                      rs1=%rs1_3  rd=%rs2_3
> >>  @cs        ... ... ... .. ... .. &cs                      rs1=%rs1_3  rs2=%rs2_3
> >> +@cs_2      ... ... ... .. ... .. &cr                      rd=%rs1_3   rs2=%rs2_3
> >>  @cs_d      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rs2=%rs2_3
> >>  @cs_w      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rs2=%rs2_3
> >> +@cb        ... ... ... .. ... .. &cb     imm=%imm_cb      rs1=%rs1_3
> >> +@cj        ...    ........... .. &cj     imm=%imm_cj
> >>
> >> +@c_addi16sp_lui ... .  ..... ..... .. &caddi16sp_lui %imm_lui %imm_addi16sp %rd
> >> +
> >> +@c_shift        ... . .. ... ..... .. &c_shift rd=%rs1_3 shamt=%nzuimm_6bit
> >> +
> >> +@c_andi         ... . .. ... ..... .. &ci imm=%imm_ci rd=%rs1_3
> >>
> >>  # *** RV64C Standard Extension (Quadrant 0) ***
> >>  c_addi4spn        000    ........ ... 00 @ciw
> >> @@ -53,3 +78,21 @@ c_flw_ld          011  --- ... -- ... 00 @cl    #Note: Must parse uimm manually
> >>  c_fsd             101  ... ... .. ... 00 @cs_d
> >>  c_sw              110  ... ... .. ... 00 @cs_w
> >>  c_fsw_sd          111  --- ... -- ... 00 @cs    #Note: Must parse uimm manually
> >> +
> >> +# *** RV64C Standard Extension (Quadrant 1) ***
> >> +c_addi            000 .  .....  ..... 01 @ci
> >> +c_jal_addiw       001 .  .....  ..... 01 @ci #Note: parse rd and/or imm manually
> >> +c_li              010 .  .....  ..... 01 @ci
> >> +c_addi16sp_lui    011 .  .....  ..... 01 @c_addi16sp_lui # shares opc with C.LUI
> >> +c_srli            100 . 00 ...  ..... 01 @c_shift
> >> +c_srai            100 . 01 ...  ..... 01 @c_shift
> >> +c_andi            100 . 10 ...  ..... 01 @c_andi
> >> +c_sub             100 0 11 ... 00 ... 01 @cs_2
> >> +c_xor             100 0 11 ... 01 ... 01 @cs_2
> >> +c_or              100 0 11 ... 10 ... 01 @cs_2
> >> +c_and             100 0 11 ... 11 ... 01 @cs_2
> >> +c_subw            100 1 11 ... 00 ... 01 @cs_2
> >> +c_addw            100 1 11 ... 01 ... 01 @cs_2
> >> +c_j               101     ........... 01 @cj
> >> +c_beqz            110  ... ...  ..... 01 @cb
> >> +c_bnez            111  ... ...  ..... 01 @cb
> >> diff --git a/target/riscv/insn_trans/trans_rvc.inc.c b/target/riscv/insn_trans/trans_rvc.inc.c
> >> index 93ec8aa30b95..b06c435c9800 100644
> >> --- a/target/riscv/insn_trans/trans_rvc.inc.c
> >> +++ b/target/riscv/insn_trans/trans_rvc.inc.c
> >> @@ -73,3 +73,154 @@ static bool trans_c_fsw_sd(DisasContext *ctx, arg_c_fsw_sd *a)
> >>      return false;
> >>  #endif
> >>  }
> >> +
> >> +static bool trans_c_addi(DisasContext *ctx, arg_c_addi *a)
> >> +{
> >> +    if (a->imm == 0) {
> >> +        /* Hint: insn is valid but does not affect state */
> >> +        return true;
> >> +    }
> >> +    arg_addi arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
> >> +    return trans_addi(ctx, &arg);
> >> +}
> >> +
> >> +static bool trans_c_jal_addiw(DisasContext *ctx, arg_c_jal_addiw *a)
> >> +{
> >> +#ifdef TARGET_RISCV32
> >> +    /* C.JAL */
> >> +    arg_jal arg = { .rd = 1, .imm = a->imm };
> >> +    return trans_jal(ctx, &arg);
> >> +#else
> >> +    /* C.ADDIW */
> >> +    arg_addiw arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
> >> +    return trans_addiw(ctx, &arg);
> >> +#endif
> >> +}
> >> +
> >> +static bool trans_c_li(DisasContext *ctx, arg_c_li *a)
> >> +{
> >> +    if (a->rd == 0) {
> >> +        /* Hint: insn is valid but does not affect state */
> >> +        return true;
> >> +    }
> >> +    arg_addi arg = { .rd = a->rd, .rs1 = 0, .imm = a->imm };
> >> +    return trans_addi(ctx, &arg);
> >> +}
> >> +
> >> +static bool trans_c_addi16sp_lui(DisasContext *ctx, arg_c_addi16sp_lui *a)
> >> +{
> >> +    if (a->rd == 2) {
> >> +        /* C.ADDI16SP */
> >> +        arg_addi arg = { .rd = 2, .rs1 = 2, .imm = a->imm_addi16sp };
> >> +        return trans_addi(ctx, &arg);
> >> +    } else if (a->imm_lui != 0) {
> >> +        /* C.LUI */
> >> +        if (a->rd == 0) {
> >> +            /* Hint: insn is valid but does not affect state */
> >> +            return true;
> >> +        }
> >> +        arg_lui arg = { .rd = a->rd, .imm = a->imm_lui };
> >> +        return trans_lui(ctx, &arg);
> >> +    }
> >> +    return false;
> >> +}
> >> +
> >> +static bool trans_c_srli(DisasContext *ctx, arg_c_srli *a)
> >> +{
> >> +    int shamt = a->shamt;
> >> +    if (shamt == 0) {
> >> +        /* For RV128 a shamt of 0 means a shift by 64 */
> >> +        shamt = 64;
> >> +    }
> >> +    /* Ensure, that shamt[5] is zero for RV32 */
> >> +    if (shamt >= TARGET_LONG_BITS) {
> >> +        return false;
> >> +    }
> >> +
> >> +    arg_srli arg = { .rd = a->rd, .rs1 = a->rd, .shamt = a->shamt };
> >> +    return trans_srli(ctx, &arg);
> >> +}
> >> +
> >> +static bool trans_c_srai(DisasContext *ctx, arg_c_srai *a)
> >> +{
> >> +    int shamt = a->shamt;
> >> +    if (shamt == 0) {
> >> +        /* For RV128 a shamt of 0 means a shift by 64 */
> >> +        shamt = 64;
> >> +    }
> >> +    /* Ensure, that shamt[5] is zero for RV32 */
> >> +    if (shamt >= TARGET_LONG_BITS) {
> >> +        return false;
> >> +    }
> >> +
> >> +    arg_srai arg = { .rd = a->rd, .rs1 = a->rd, .shamt = a->shamt };
> >> +    return trans_srai(ctx, &arg);
> >> +}
> >> +
> >> +static bool trans_c_andi(DisasContext *ctx, arg_c_andi *a)
> >> +{
> >> +    arg_andi arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
> >> +    return trans_andi(ctx, &arg);
> >> +}
> >> +
> >> +static bool trans_c_sub(DisasContext *ctx, arg_c_sub *a)
> >> +{
> >> +    arg_sub arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> >> +    return trans_sub(ctx, &arg);
> >> +}
> >> +
> >> +static bool trans_c_xor(DisasContext *ctx, arg_c_xor *a)
> >> +{
> >> +    arg_xor arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> >> +    return trans_xor(ctx, &arg);
> >> +}
> >> +
> >> +static bool trans_c_or(DisasContext *ctx, arg_c_or *a)
> >> +{
> >> +    arg_or arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> >> +    return trans_or(ctx, &arg);
> >> +}
> >> +
> >> +static bool trans_c_and(DisasContext *ctx, arg_c_and *a)
> >> +{
> >> +    arg_and arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> >> +    return trans_and(ctx, &arg);
> >> +}
> >> +
> >> +static bool trans_c_subw(DisasContext *ctx, arg_c_subw *a)
> >> +{
> >> +#ifdef TARGET_RISCV64
> >> +    arg_subw arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> >> +    return trans_subw(ctx, &arg);
> >> +#else
> >> +    return false;
> >> +#endif
> >> +}
> >> +
> >> +static bool trans_c_addw(DisasContext *ctx, arg_c_addw *a)
> >> +{
> >> +#ifdef TARGET_RISCV64
> >> +    arg_addw arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
> >> +    return trans_addw(ctx, &arg);
> >> +#else
> >> +    return false;
> >> +#endif
> >> +}
> >> +
> >> +static bool trans_c_j(DisasContext *ctx, arg_c_j *a)
> >> +{
> >> +    arg_jal arg = { .rd = 0, .imm = a->imm };
> >> +    return trans_jal(ctx, &arg);
> >> +}
> >> +
> >> +static bool trans_c_beqz(DisasContext *ctx, arg_c_beqz *a)
> >> +{
> >> +    arg_beq arg = { .rs1 = a->rs1, .rs2 = 0, .imm = a->imm };
> >> +    return trans_beq(ctx, &arg);
> >> +}
> >> +
> >> +static bool trans_c_bnez(DisasContext *ctx, arg_c_bnez *a)
> >> +{
> >> +    arg_bne arg = { .rs1 = a->rs1, .rs2 = 0, .imm = a->imm };
> >> +    return trans_bne(ctx, &arg);
> >> +}
> >> diff --git a/target/riscv/translate.c b/target/riscv/translate.c
> >> index 0106fa8d51b3..a584c24fbf6b 100644
> >> --- a/target/riscv/translate.c
> >> +++ b/target/riscv/translate.c
> >> @@ -828,120 +828,6 @@ static void decode_RV32_64C0(DisasContext *ctx)
> >>      }
> >>  }
> >>
> >> -static void decode_RV32_64C1(DisasContext *ctx)
> >> -{
> >> -    uint8_t funct3 = extract32(ctx->opcode, 13, 3);
> >> -    uint8_t rd_rs1 = GET_C_RS1(ctx->opcode);
> >> -    uint8_t rs1s, rs2s;
> >> -    uint8_t funct2;
> >> -
> >> -    switch (funct3) {
> >> -    case 0:
> >> -        /* C.ADDI -> addi rd, rd, nzimm[5:0] */
> >> -        gen_arith_imm(ctx, OPC_RISC_ADDI, rd_rs1, rd_rs1,
> >> -                      GET_C_IMM(ctx->opcode));
> >> -        break;
> >> -    case 1:
> >> -#if defined(TARGET_RISCV64)
> >> -        /* C.ADDIW (RV64/128) -> addiw rd, rd, imm[5:0]*/
> >> -        gen_arith_imm(ctx, OPC_RISC_ADDIW, rd_rs1, rd_rs1,
> >> -                      GET_C_IMM(ctx->opcode));
> >> -#else
> >> -        /* C.JAL(RV32) -> jal x1, offset[11:1] */
> >> -        gen_jal(ctx, 1, GET_C_J_IMM(ctx->opcode));
> >> -#endif
> >> -        break;
> >> -    case 2:
> >> -        /* C.LI -> addi rd, x0, imm[5:0]*/
> >> -        gen_arith_imm(ctx, OPC_RISC_ADDI, rd_rs1, 0, GET_C_IMM(ctx->opcode));
> >> -        break;
> >> -    case 3:
> >> -        if (rd_rs1 == 2) {
> >> -            /* C.ADDI16SP -> addi x2, x2, nzimm[9:4]*/
> >> -            gen_arith_imm(ctx, OPC_RISC_ADDI, 2, 2,
> >> -                          GET_C_ADDI16SP_IMM(ctx->opcode));
> >> -        } else if (rd_rs1 != 0) {
> >> -            /* C.LUI (rs1/rd =/= {0,2}) -> lui rd, nzimm[17:12]*/
> >> -            tcg_gen_movi_tl(cpu_gpr[rd_rs1],
> >> -                            GET_C_IMM(ctx->opcode) << 12);
> >> -        }
> >> -        break;
> >> -    case 4:
> >> -        funct2 = extract32(ctx->opcode, 10, 2);
> >> -        rs1s = GET_C_RS1S(ctx->opcode);
> >> -        switch (funct2) {
> >> -        case 0: /* C.SRLI(RV32) -> srli rd', rd', shamt[5:0] */
> >> -            gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, rs1s, rs1s,
> >> -                               GET_C_ZIMM(ctx->opcode));
> >> -            /* C.SRLI64(RV128) */
> >> -            break;
> >> -        case 1:
> >> -            /* C.SRAI -> srai rd', rd', shamt[5:0]*/
> >> -            gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, rs1s, rs1s,
> >> -                            GET_C_ZIMM(ctx->opcode) | 0x400);
> >> -            /* C.SRAI64(RV128) */
> >> -            break;
> >> -        case 2:
> >> -            /* C.ANDI -> andi rd', rd', imm[5:0]*/
> >> -            gen_arith_imm(ctx, OPC_RISC_ANDI, rs1s, rs1s,
> >> -                          GET_C_IMM(ctx->opcode));
> >> -            break;
> >> -        case 3:
> >> -            funct2 = extract32(ctx->opcode, 5, 2);
> >> -            rs2s = GET_C_RS2S(ctx->opcode);
> >> -            switch (funct2) {
> >> -            case 0:
> >> -                /* C.SUB -> sub rd', rd', rs2' */
> >> -                if (extract32(ctx->opcode, 12, 1) == 0) {
> >> -                    gen_arith(ctx, OPC_RISC_SUB, rs1s, rs1s, rs2s);
> >> -                }
> >> -#if defined(TARGET_RISCV64)
> >> -                else {
> >> -                    gen_arith(ctx, OPC_RISC_SUBW, rs1s, rs1s, rs2s);
> >> -                }
> >> -#endif
> >> -                break;
> >> -            case 1:
> >> -                /* C.XOR -> xor rs1', rs1', rs2' */
> >> -                if (extract32(ctx->opcode, 12, 1) == 0) {
> >> -                    gen_arith(ctx, OPC_RISC_XOR, rs1s, rs1s, rs2s);
> >> -                }
> >> -#if defined(TARGET_RISCV64)
> >> -                else {
> >> -                    /* C.ADDW (RV64/128) */
> >> -                    gen_arith(ctx, OPC_RISC_ADDW, rs1s, rs1s, rs2s);
> >> -                }
> >> -#endif
> >> -                break;
> >> -            case 2:
> >> -                /* C.OR -> or rs1', rs1', rs2' */
> >> -                gen_arith(ctx, OPC_RISC_OR, rs1s, rs1s, rs2s);
> >> -                break;
> >> -            case 3:
> >> -                /* C.AND -> and rs1', rs1', rs2' */
> >> -                gen_arith(ctx, OPC_RISC_AND, rs1s, rs1s, rs2s);
> >> -                break;
> >> -            }
> >> -            break;
> >> -        }
> >> -        break;
> >> -    case 5:
> >> -        /* C.J -> jal x0, offset[11:1]*/
> >> -        gen_jal(ctx, 0, GET_C_J_IMM(ctx->opcode));
> >> -        break;
> >> -    case 6:
> >> -        /* C.BEQZ -> beq rs1', x0, offset[8:1]*/
> >> -        rs1s = GET_C_RS1S(ctx->opcode);
> >> -        gen_branch(ctx, OPC_RISC_BEQ, rs1s, 0, GET_C_B_IMM(ctx->opcode));
> >> -        break;
> >> -    case 7:
> >> -        /* C.BNEZ -> bne rs1', x0, offset[8:1]*/
> >> -        rs1s = GET_C_RS1S(ctx->opcode);
> >> -        gen_branch(ctx, OPC_RISC_BNE, rs1s, 0, GET_C_B_IMM(ctx->opcode));
> >> -        break;
> >> -    }
> >> -}
> >> -
> >>  static void decode_RV32_64C2(DisasContext *ctx)
> >>  {
> >>      uint8_t rd, rs2;
> >> @@ -1028,9 +914,6 @@ static void decode_RV32_64C(DisasContext *ctx)
> >>      case 0:
> >>          decode_RV32_64C0(ctx);
> >>          break;
> >> -    case 1:
> >> -        decode_RV32_64C1(ctx);
> >> -        break;
> >>      case 2:
> >>          decode_RV32_64C2(ctx);
> >>          break;
> >> @@ -1045,6 +928,7 @@ static void decode_RV32_64C(DisasContext *ctx)
> >>  EX_SH(1)
> >>  EX_SH(2)
> >>  EX_SH(3)
> >> +EX_SH(4)
> >>  EX_SH(12)
> >>
> >>  #define REQUIRE_EXT(ctx, ext) do { \
> >> --
> >> 2.19.2
> >>
> >>

Re: [Qemu-riscv] [Qemu-devel] [PULL 17/29] target/riscv: Convert quadrant 1 of RVXC insns to decodetree

[Palmer Dabbelt]

On Thu, 14 Mar 2019 21:57:37 PDT (-0700), alistair23@gmail.com wrote:
> On Thu, Mar 14, 2019 at 8:59 PM Palmer Dabbelt <palmer@sifive.com> wrote:
>>
>> On Thu, 14 Mar 2019 13:28:37 PDT (-0700), alistair23@gmail.com wrote:
>> > On Wed, Mar 13, 2019 at 7:53 AM Palmer Dabbelt <palmer@sifive.com> wrote:
>> >>
>> >> From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>> >>
>> >> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
>> >> Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>> >> Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
>> >
>> > This commit is the first bad commit in breaking 32-bit boot.
>> >
>> > It looks like the jal doesn't jump to the correct address:
>> >
>> > ----------------
>> > IN:
>> > 0x80000022:  00050433          add             s0,a0,zero
>> > 0x80000026:  000584b3          add             s1,a1,zero
>> > 0x8000002a:  2c79              jal             ra,670          # 0x800002c8
>> >
>> > ----------------
>> > IN:
>> > 0x800002c8:  00000533          add             a0,zero,zero
>> > 0x800002cc:  8082              ret
>>
>>
>> Sorry, for some reason I thought you'd tested this on 32-bit?  Do you have a
>> workflow that I can use to reproduce the bug?
>
> I did! Not sure what happened there. Maybe something changed between
> what I tested and what was merged or somehow I tested the wrong thing.
>
> Just booting 32-bit OpenSBI fails. It seems that the imm for jal is
> wrong. I'll keep digging into it.

OK, thanks.  There was some jiggling around of the patch due to OSX build 
issues, maybe that broke something in 32-bit land?

>
> Alistair
>
>>
>> >
>> >
>> > Alistair
>> >
>> >> ---
>> >>  target/riscv/insn16.decode              |  43 +++++++
>> >>  target/riscv/insn_trans/trans_rvc.inc.c | 151 ++++++++++++++++++++++++
>> >>  target/riscv/translate.c                | 118 +-----------------
>> >>  3 files changed, 195 insertions(+), 117 deletions(-)
>> >>
>> >> diff --git a/target/riscv/insn16.decode b/target/riscv/insn16.decode
>> >> index 558c0c41f0b5..45109301c6d4 100644
>> >> --- a/target/riscv/insn16.decode
>> >> +++ b/target/riscv/insn16.decode
>> >> @@ -22,28 +22,53 @@
>> >>  %rs2_3     2:3                !function=ex_rvc_register
>> >>
>> >>  # Immediates:
>> >> +%imm_ci        12:s1 2:5
>> >>  %nzuimm_ciw    7:4 11:2 5:1 6:1   !function=ex_shift_2
>> >>  %uimm_cl_d     5:2 10:3           !function=ex_shift_3
>> >>  %uimm_cl_w     5:1 10:3 6:1       !function=ex_shift_2
>> >> +%imm_cb        12:s1 5:2 2:1 10:2 3:2 !function=ex_shift_1
>> >> +%imm_cj        12:s1 8:1 9:2 6:1 7:1 2:1 11:1 3:3 !function=ex_shift_1
>> >> +
>> >> +%nzuimm_6bit   12:1 2:5
>> >> +
>> >> +%imm_addi16sp  12:s1 3:2 5:1 2:1 6:1 !function=ex_shift_4
>> >> +%imm_lui       12:s1 2:5             !function=ex_shift_12
>> >> +
>> >>
>> >>
>> >>  # Argument sets:
>> >>  &cl               rs1 rd
>> >>  &cl_dw     uimm   rs1 rd
>> >> +&ci        imm        rd
>> >>  &ciw       nzuimm     rd
>> >>  &cs               rs1 rs2
>> >>  &cs_dw     uimm   rs1 rs2
>> >> +&cb        imm    rs1
>> >> +&cr               rd  rs2
>> >> +&cj       imm
>> >> +&c_shift   shamt      rd
>> >> +
>> >>
>> >> +&caddi16sp_lui  imm_lui imm_addi16sp rd
>> >>
>> >>  # Formats 16:
>> >> +@ci        ... . ..... .....  .. &ci     imm=%imm_ci                  %rd
>> >>  @ciw       ...   ........ ... .. &ciw    nzuimm=%nzuimm_ciw           rd=%rs2_3
>> >>  @cl_d      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rd=%rs2_3
>> >>  @cl_w      ... ... ... .. ... .. &cl_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rd=%rs2_3
>> >>  @cl        ... ... ... .. ... .. &cl                      rs1=%rs1_3  rd=%rs2_3
>> >>  @cs        ... ... ... .. ... .. &cs                      rs1=%rs1_3  rs2=%rs2_3
>> >> +@cs_2      ... ... ... .. ... .. &cr                      rd=%rs1_3   rs2=%rs2_3
>> >>  @cs_d      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_d  rs1=%rs1_3  rs2=%rs2_3
>> >>  @cs_w      ... ... ... .. ... .. &cs_dw  uimm=%uimm_cl_w  rs1=%rs1_3  rs2=%rs2_3
>> >> +@cb        ... ... ... .. ... .. &cb     imm=%imm_cb      rs1=%rs1_3
>> >> +@cj        ...    ........... .. &cj     imm=%imm_cj
>> >>
>> >> +@c_addi16sp_lui ... .  ..... ..... .. &caddi16sp_lui %imm_lui %imm_addi16sp %rd
>> >> +
>> >> +@c_shift        ... . .. ... ..... .. &c_shift rd=%rs1_3 shamt=%nzuimm_6bit
>> >> +
>> >> +@c_andi         ... . .. ... ..... .. &ci imm=%imm_ci rd=%rs1_3
>> >>
>> >>  # *** RV64C Standard Extension (Quadrant 0) ***
>> >>  c_addi4spn        000    ........ ... 00 @ciw
>> >> @@ -53,3 +78,21 @@ c_flw_ld          011  --- ... -- ... 00 @cl    #Note: Must parse uimm manually
>> >>  c_fsd             101  ... ... .. ... 00 @cs_d
>> >>  c_sw              110  ... ... .. ... 00 @cs_w
>> >>  c_fsw_sd          111  --- ... -- ... 00 @cs    #Note: Must parse uimm manually
>> >> +
>> >> +# *** RV64C Standard Extension (Quadrant 1) ***
>> >> +c_addi            000 .  .....  ..... 01 @ci
>> >> +c_jal_addiw       001 .  .....  ..... 01 @ci #Note: parse rd and/or imm manually
>> >> +c_li              010 .  .....  ..... 01 @ci
>> >> +c_addi16sp_lui    011 .  .....  ..... 01 @c_addi16sp_lui # shares opc with C.LUI
>> >> +c_srli            100 . 00 ...  ..... 01 @c_shift
>> >> +c_srai            100 . 01 ...  ..... 01 @c_shift
>> >> +c_andi            100 . 10 ...  ..... 01 @c_andi
>> >> +c_sub             100 0 11 ... 00 ... 01 @cs_2
>> >> +c_xor             100 0 11 ... 01 ... 01 @cs_2
>> >> +c_or              100 0 11 ... 10 ... 01 @cs_2
>> >> +c_and             100 0 11 ... 11 ... 01 @cs_2
>> >> +c_subw            100 1 11 ... 00 ... 01 @cs_2
>> >> +c_addw            100 1 11 ... 01 ... 01 @cs_2
>> >> +c_j               101     ........... 01 @cj
>> >> +c_beqz            110  ... ...  ..... 01 @cb
>> >> +c_bnez            111  ... ...  ..... 01 @cb
>> >> diff --git a/target/riscv/insn_trans/trans_rvc.inc.c b/target/riscv/insn_trans/trans_rvc.inc.c
>> >> index 93ec8aa30b95..b06c435c9800 100644
>> >> --- a/target/riscv/insn_trans/trans_rvc.inc.c
>> >> +++ b/target/riscv/insn_trans/trans_rvc.inc.c
>> >> @@ -73,3 +73,154 @@ static bool trans_c_fsw_sd(DisasContext *ctx, arg_c_fsw_sd *a)
>> >>      return false;
>> >>  #endif
>> >>  }
>> >> +
>> >> +static bool trans_c_addi(DisasContext *ctx, arg_c_addi *a)
>> >> +{
>> >> +    if (a->imm == 0) {
>> >> +        /* Hint: insn is valid but does not affect state */
>> >> +        return true;
>> >> +    }
>> >> +    arg_addi arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
>> >> +    return trans_addi(ctx, &arg);
>> >> +}
>> >> +
>> >> +static bool trans_c_jal_addiw(DisasContext *ctx, arg_c_jal_addiw *a)
>> >> +{
>> >> +#ifdef TARGET_RISCV32
>> >> +    /* C.JAL */
>> >> +    arg_jal arg = { .rd = 1, .imm = a->imm };
>> >> +    return trans_jal(ctx, &arg);
>> >> +#else
>> >> +    /* C.ADDIW */
>> >> +    arg_addiw arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
>> >> +    return trans_addiw(ctx, &arg);
>> >> +#endif
>> >> +}
>> >> +
>> >> +static bool trans_c_li(DisasContext *ctx, arg_c_li *a)
>> >> +{
>> >> +    if (a->rd == 0) {
>> >> +        /* Hint: insn is valid but does not affect state */
>> >> +        return true;
>> >> +    }
>> >> +    arg_addi arg = { .rd = a->rd, .rs1 = 0, .imm = a->imm };
>> >> +    return trans_addi(ctx, &arg);
>> >> +}
>> >> +
>> >> +static bool trans_c_addi16sp_lui(DisasContext *ctx, arg_c_addi16sp_lui *a)
>> >> +{
>> >> +    if (a->rd == 2) {
>> >> +        /* C.ADDI16SP */
>> >> +        arg_addi arg = { .rd = 2, .rs1 = 2, .imm = a->imm_addi16sp };
>> >> +        return trans_addi(ctx, &arg);
>> >> +    } else if (a->imm_lui != 0) {
>> >> +        /* C.LUI */
>> >> +        if (a->rd == 0) {
>> >> +            /* Hint: insn is valid but does not affect state */
>> >> +            return true;
>> >> +        }
>> >> +        arg_lui arg = { .rd = a->rd, .imm = a->imm_lui };
>> >> +        return trans_lui(ctx, &arg);
>> >> +    }
>> >> +    return false;
>> >> +}
>> >> +
>> >> +static bool trans_c_srli(DisasContext *ctx, arg_c_srli *a)
>> >> +{
>> >> +    int shamt = a->shamt;
>> >> +    if (shamt == 0) {
>> >> +        /* For RV128 a shamt of 0 means a shift by 64 */
>> >> +        shamt = 64;
>> >> +    }
>> >> +    /* Ensure, that shamt[5] is zero for RV32 */
>> >> +    if (shamt >= TARGET_LONG_BITS) {
>> >> +        return false;
>> >> +    }
>> >> +
>> >> +    arg_srli arg = { .rd = a->rd, .rs1 = a->rd, .shamt = a->shamt };
>> >> +    return trans_srli(ctx, &arg);
>> >> +}
>> >> +
>> >> +static bool trans_c_srai(DisasContext *ctx, arg_c_srai *a)
>> >> +{
>> >> +    int shamt = a->shamt;
>> >> +    if (shamt == 0) {
>> >> +        /* For RV128 a shamt of 0 means a shift by 64 */
>> >> +        shamt = 64;
>> >> +    }
>> >> +    /* Ensure, that shamt[5] is zero for RV32 */
>> >> +    if (shamt >= TARGET_LONG_BITS) {
>> >> +        return false;
>> >> +    }
>> >> +
>> >> +    arg_srai arg = { .rd = a->rd, .rs1 = a->rd, .shamt = a->shamt };
>> >> +    return trans_srai(ctx, &arg);
>> >> +}
>> >> +
>> >> +static bool trans_c_andi(DisasContext *ctx, arg_c_andi *a)
>> >> +{
>> >> +    arg_andi arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
>> >> +    return trans_andi(ctx, &arg);
>> >> +}
>> >> +
>> >> +static bool trans_c_sub(DisasContext *ctx, arg_c_sub *a)
>> >> +{
>> >> +    arg_sub arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> >> +    return trans_sub(ctx, &arg);
>> >> +}
>> >> +
>> >> +static bool trans_c_xor(DisasContext *ctx, arg_c_xor *a)
>> >> +{
>> >> +    arg_xor arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> >> +    return trans_xor(ctx, &arg);
>> >> +}
>> >> +
>> >> +static bool trans_c_or(DisasContext *ctx, arg_c_or *a)
>> >> +{
>> >> +    arg_or arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> >> +    return trans_or(ctx, &arg);
>> >> +}
>> >> +
>> >> +static bool trans_c_and(DisasContext *ctx, arg_c_and *a)
>> >> +{
>> >> +    arg_and arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> >> +    return trans_and(ctx, &arg);
>> >> +}
>> >> +
>> >> +static bool trans_c_subw(DisasContext *ctx, arg_c_subw *a)
>> >> +{
>> >> +#ifdef TARGET_RISCV64
>> >> +    arg_subw arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> >> +    return trans_subw(ctx, &arg);
>> >> +#else
>> >> +    return false;
>> >> +#endif
>> >> +}
>> >> +
>> >> +static bool trans_c_addw(DisasContext *ctx, arg_c_addw *a)
>> >> +{
>> >> +#ifdef TARGET_RISCV64
>> >> +    arg_addw arg = { .rd = a->rd, .rs1 = a->rd, .rs2 = a->rs2 };
>> >> +    return trans_addw(ctx, &arg);
>> >> +#else
>> >> +    return false;
>> >> +#endif
>> >> +}
>> >> +
>> >> +static bool trans_c_j(DisasContext *ctx, arg_c_j *a)
>> >> +{
>> >> +    arg_jal arg = { .rd = 0, .imm = a->imm };
>> >> +    return trans_jal(ctx, &arg);
>> >> +}
>> >> +
>> >> +static bool trans_c_beqz(DisasContext *ctx, arg_c_beqz *a)
>> >> +{
>> >> +    arg_beq arg = { .rs1 = a->rs1, .rs2 = 0, .imm = a->imm };
>> >> +    return trans_beq(ctx, &arg);
>> >> +}
>> >> +
>> >> +static bool trans_c_bnez(DisasContext *ctx, arg_c_bnez *a)
>> >> +{
>> >> +    arg_bne arg = { .rs1 = a->rs1, .rs2 = 0, .imm = a->imm };
>> >> +    return trans_bne(ctx, &arg);
>> >> +}
>> >> diff --git a/target/riscv/translate.c b/target/riscv/translate.c
>> >> index 0106fa8d51b3..a584c24fbf6b 100644
>> >> --- a/target/riscv/translate.c
>> >> +++ b/target/riscv/translate.c
>> >> @@ -828,120 +828,6 @@ static void decode_RV32_64C0(DisasContext *ctx)
>> >>      }
>> >>  }
>> >>
>> >> -static void decode_RV32_64C1(DisasContext *ctx)
>> >> -{
>> >> -    uint8_t funct3 = extract32(ctx->opcode, 13, 3);
>> >> -    uint8_t rd_rs1 = GET_C_RS1(ctx->opcode);
>> >> -    uint8_t rs1s, rs2s;
>> >> -    uint8_t funct2;
>> >> -
>> >> -    switch (funct3) {
>> >> -    case 0:
>> >> -        /* C.ADDI -> addi rd, rd, nzimm[5:0] */
>> >> -        gen_arith_imm(ctx, OPC_RISC_ADDI, rd_rs1, rd_rs1,
>> >> -                      GET_C_IMM(ctx->opcode));
>> >> -        break;
>> >> -    case 1:
>> >> -#if defined(TARGET_RISCV64)
>> >> -        /* C.ADDIW (RV64/128) -> addiw rd, rd, imm[5:0]*/
>> >> -        gen_arith_imm(ctx, OPC_RISC_ADDIW, rd_rs1, rd_rs1,
>> >> -                      GET_C_IMM(ctx->opcode));
>> >> -#else
>> >> -        /* C.JAL(RV32) -> jal x1, offset[11:1] */
>> >> -        gen_jal(ctx, 1, GET_C_J_IMM(ctx->opcode));
>> >> -#endif
>> >> -        break;
>> >> -    case 2:
>> >> -        /* C.LI -> addi rd, x0, imm[5:0]*/
>> >> -        gen_arith_imm(ctx, OPC_RISC_ADDI, rd_rs1, 0, GET_C_IMM(ctx->opcode));
>> >> -        break;
>> >> -    case 3:
>> >> -        if (rd_rs1 == 2) {
>> >> -            /* C.ADDI16SP -> addi x2, x2, nzimm[9:4]*/
>> >> -            gen_arith_imm(ctx, OPC_RISC_ADDI, 2, 2,
>> >> -                          GET_C_ADDI16SP_IMM(ctx->opcode));
>> >> -        } else if (rd_rs1 != 0) {
>> >> -            /* C.LUI (rs1/rd =/= {0,2}) -> lui rd, nzimm[17:12]*/
>> >> -            tcg_gen_movi_tl(cpu_gpr[rd_rs1],
>> >> -                            GET_C_IMM(ctx->opcode) << 12);
>> >> -        }
>> >> -        break;
>> >> -    case 4:
>> >> -        funct2 = extract32(ctx->opcode, 10, 2);
>> >> -        rs1s = GET_C_RS1S(ctx->opcode);
>> >> -        switch (funct2) {
>> >> -        case 0: /* C.SRLI(RV32) -> srli rd', rd', shamt[5:0] */
>> >> -            gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, rs1s, rs1s,
>> >> -                               GET_C_ZIMM(ctx->opcode));
>> >> -            /* C.SRLI64(RV128) */
>> >> -            break;
>> >> -        case 1:
>> >> -            /* C.SRAI -> srai rd', rd', shamt[5:0]*/
>> >> -            gen_arith_imm(ctx, OPC_RISC_SHIFT_RIGHT_I, rs1s, rs1s,
>> >> -                            GET_C_ZIMM(ctx->opcode) | 0x400);
>> >> -            /* C.SRAI64(RV128) */
>> >> -            break;
>> >> -        case 2:
>> >> -            /* C.ANDI -> andi rd', rd', imm[5:0]*/
>> >> -            gen_arith_imm(ctx, OPC_RISC_ANDI, rs1s, rs1s,
>> >> -                          GET_C_IMM(ctx->opcode));
>> >> -            break;
>> >> -        case 3:
>> >> -            funct2 = extract32(ctx->opcode, 5, 2);
>> >> -            rs2s = GET_C_RS2S(ctx->opcode);
>> >> -            switch (funct2) {
>> >> -            case 0:
>> >> -                /* C.SUB -> sub rd', rd', rs2' */
>> >> -                if (extract32(ctx->opcode, 12, 1) == 0) {
>> >> -                    gen_arith(ctx, OPC_RISC_SUB, rs1s, rs1s, rs2s);
>> >> -                }
>> >> -#if defined(TARGET_RISCV64)
>> >> -                else {
>> >> -                    gen_arith(ctx, OPC_RISC_SUBW, rs1s, rs1s, rs2s);
>> >> -                }
>> >> -#endif
>> >> -                break;
>> >> -            case 1:
>> >> -                /* C.XOR -> xor rs1', rs1', rs2' */
>> >> -                if (extract32(ctx->opcode, 12, 1) == 0) {
>> >> -                    gen_arith(ctx, OPC_RISC_XOR, rs1s, rs1s, rs2s);
>> >> -                }
>> >> -#if defined(TARGET_RISCV64)
>> >> -                else {
>> >> -                    /* C.ADDW (RV64/128) */
>> >> -                    gen_arith(ctx, OPC_RISC_ADDW, rs1s, rs1s, rs2s);
>> >> -                }
>> >> -#endif
>> >> -                break;
>> >> -            case 2:
>> >> -                /* C.OR -> or rs1', rs1', rs2' */
>> >> -                gen_arith(ctx, OPC_RISC_OR, rs1s, rs1s, rs2s);
>> >> -                break;
>> >> -            case 3:
>> >> -                /* C.AND -> and rs1', rs1', rs2' */
>> >> -                gen_arith(ctx, OPC_RISC_AND, rs1s, rs1s, rs2s);
>> >> -                break;
>> >> -            }
>> >> -            break;
>> >> -        }
>> >> -        break;
>> >> -    case 5:
>> >> -        /* C.J -> jal x0, offset[11:1]*/
>> >> -        gen_jal(ctx, 0, GET_C_J_IMM(ctx->opcode));
>> >> -        break;
>> >> -    case 6:
>> >> -        /* C.BEQZ -> beq rs1', x0, offset[8:1]*/
>> >> -        rs1s = GET_C_RS1S(ctx->opcode);
>> >> -        gen_branch(ctx, OPC_RISC_BEQ, rs1s, 0, GET_C_B_IMM(ctx->opcode));
>> >> -        break;
>> >> -    case 7:
>> >> -        /* C.BNEZ -> bne rs1', x0, offset[8:1]*/
>> >> -        rs1s = GET_C_RS1S(ctx->opcode);
>> >> -        gen_branch(ctx, OPC_RISC_BNE, rs1s, 0, GET_C_B_IMM(ctx->opcode));
>> >> -        break;
>> >> -    }
>> >> -}
>> >> -
>> >>  static void decode_RV32_64C2(DisasContext *ctx)
>> >>  {
>> >>      uint8_t rd, rs2;
>> >> @@ -1028,9 +914,6 @@ static void decode_RV32_64C(DisasContext *ctx)
>> >>      case 0:
>> >>          decode_RV32_64C0(ctx);
>> >>          break;
>> >> -    case 1:
>> >> -        decode_RV32_64C1(ctx);
>> >> -        break;
>> >>      case 2:
>> >>          decode_RV32_64C2(ctx);
>> >>          break;
>> >> @@ -1045,6 +928,7 @@ static void decode_RV32_64C(DisasContext *ctx)
>> >>  EX_SH(1)
>> >>  EX_SH(2)
>> >>  EX_SH(3)
>> >> +EX_SH(4)
>> >>  EX_SH(12)
>> >>
>> >>  #define REQUIRE_EXT(ctx, ext) do { \
>> >> --
>> >> 2.19.2
>> >>
>> >>

Re: [Qemu-riscv] [Qemu-devel] [PULL 17/29] target/riscv: Convert quadrant 1 of RVXC insns to decodetree

[Bastian Koppelmann]

Hi Alistair

On 3/14/19 9:28 PM, Alistair Francis wrote:
> On Wed, Mar 13, 2019 at 7:53 AM Palmer Dabbelt <palmer@sifive.com> wrote:
>> From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>>
>> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
>> Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>> Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
> This commit is the first bad commit in breaking 32-bit boot.
>
> It looks like the jal doesn't jump to the correct address:
>
> ----------------
> IN:
> 0x80000022:  00050433          add             s0,a0,zero
> 0x80000026:  000584b3          add             s1,a1,zero
> 0x8000002a:  2c79              jal             ra,670          # 0x800002c8
>
> ----------------
> IN:
> 0x800002c8:  00000533          add             a0,zero,zero
> 0x800002cc:  8082              ret


Oops! Can you point me to the binary to reproduce this?

Cheers,

Bastian

Re: [Qemu-riscv] [Qemu-devel] [PULL 17/29] target/riscv: Convert quadrant 1 of RVXC insns to decodetree

[Palmer Dabbelt]

On Fri, 15 Mar 2019 02:06:07 PDT (-0700), Bastian Koppelmann wrote:
> Hi Alistair
>
> On 3/14/19 9:28 PM, Alistair Francis wrote:
>> On Wed, Mar 13, 2019 at 7:53 AM Palmer Dabbelt <palmer@sifive.com> wrote:
>>> From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>>>
>>> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
>>> Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>>> Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
>> This commit is the first bad commit in breaking 32-bit boot.
>>
>> It looks like the jal doesn't jump to the correct address:
>>
>> ----------------
>> IN:
>> 0x80000022:  00050433          add             s0,a0,zero
>> 0x80000026:  000584b3          add             s1,a1,zero
>> 0x8000002a:  2c79              jal             ra,670          # 0x800002c8
>>
>> ----------------
>> IN:
>> 0x800002c8:  00000533          add             a0,zero,zero
>> 0x800002cc:  8082              ret
>
>
> Oops! Can you point me to the binary to reproduce this?

I think I've traced it down to something simple: in my hello world binary I see

    20401a8c:       2a45                    jal     20401c3c <atexit>

in the objdump, and I see

    IN: _start
    0x20401a8c:  2a45              jal             ra,432          # 0x20401c3c

but then QEMU jumps to 0x20401a9d.  I have a feeling it's something wrong with 
gen_jal() that disappeared during the cleanups that we dropped in order to fix 
the build issues.

I'm running

    ./riscv32-softmmu/qemu-system-riscv32 -machine sifive_e -kernel ~/work/sifive/freedom-e-sdk/software/hello/hello -nographic -d in_asm,out_asm,exec,cpu -singlestep |& tee out.log

on the "hifive1" branch of github.com/palmer-dabbelt/qemu, which just has a 
PRCI fixup that I forgot about and haven't sent upstream yet (I'll do that 
after this issue).  The binary should be at

    http://www.dabbelt.com/~palmer/hello.elf

and the debug log at 

    http://www.dabbelt.com/~palmer/out.log

You can build the binary from github.com/sifive/freedom-e-sdk via

   make software PROGRAM=hello TARGET=sifive-hifive1

using the riscv64-unknown-elf-gcc-20181127-x86_64-linux-ubuntu14 toolchain 
binaries from our website (newer ones should work, but probably won't produce 
exactly the same output).

I'll poke around after grabbing some dinner...

Re: [Qemu-riscv] [Qemu-devel] [PULL 17/29] target/riscv: Convert quadrant 1 of RVXC insns to decodetree

[Palmer Dabbelt]

[-- Attachment #1: Type: text/plain, Size: 3053 bytes --]

On Fri, Mar 15, 2019 at 4:19 AM Palmer Dabbelt <palmer@sifive.com> wrote:

> On Fri, 15 Mar 2019 02:06:07 PDT (-0700), Bastian Koppelmann wrote:
> > Hi Alistair
> >
> > On 3/14/19 9:28 PM, Alistair Francis wrote:
> >> On Wed, Mar 13, 2019 at 7:53 AM Palmer Dabbelt <palmer@sifive.com>
> wrote:
> >>> From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
> >>>
> >>> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
> >>> Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
> >>> Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
> >> This commit is the first bad commit in breaking 32-bit boot.
> >>
> >> It looks like the jal doesn't jump to the correct address:
> >>
> >> ----------------
> >> IN:
> >> 0x80000022:  00050433          add             s0,a0,zero
> >> 0x80000026:  000584b3          add             s1,a1,zero
> >> 0x8000002a:  2c79              jal             ra,670          #
> 0x800002c8
> >>
> >> ----------------
> >> IN:
> >> 0x800002c8:  00000533          add             a0,zero,zero
> >> 0x800002cc:  8082              ret
> >
> >
> > Oops! Can you point me to the binary to reproduce this?
>
> I think I've traced it down to something simple: in my hello world binary
> I see
>
>     20401a8c:       2a45                    jal     20401c3c <atexit>
>
> in the objdump, and I see
>
>     IN: _start
>     0x20401a8c:  2a45              jal             ra,432          #
> 0x20401c3c
>
> but then QEMU jumps to 0x20401a9d.  I have a feeling it's something wrong
> with
> gen_jal() that disappeared during the cleanups that we dropped in order to
> fix
> the build issues.
>
> I'm running
>
>     ./riscv32-softmmu/qemu-system-riscv32 -machine sifive_e -kernel
> ~/work/sifive/freedom-e-sdk/software/hello/hello -nographic -d
> in_asm,out_asm,exec,cpu -singlestep |& tee out.log
>
> on the "hifive1" branch of github.com/palmer-dabbelt/qemu, which just has
> a
> PRCI fixup that I forgot about and haven't sent upstream yet (I'll do that
> after this issue).  The binary should be at
>
>     http://www.dabbelt.com/~palmer/hello.elf
>
> and the debug log at
>
>     http://www.dabbelt.com/~palmer/out.log
>
> You can build the binary from github.com/sifive/freedom-e-sdk via
>
>    make software PROGRAM=hello TARGET=sifive-hifive1
>
> using the riscv64-unknown-elf-gcc-20181127-x86_64-linux-ubuntu14 toolchain
> binaries from our website (newer ones should work, but probably won't
> produce
> exactly the same output).
>
> I'll poke around after grabbing some dinner...
>

OK, I'm pretty sure this is it:

c_jal_addiw       001 .  .....  ..... 01 @ci #Note: parse rd and/or imm
manually

static bool trans_c_jal_addiw(DisasContext *ctx, arg_c_jal_addiw *a)
{
#ifdef TARGET_RISCV32
    /* C.JAL */
    arg_jal arg = { .rd = 1, .imm = a->imm };
    return trans_jal(ctx, &arg);
#else
    /* C.ADDIW */
    arg_addiw arg = { .rd = a->rd, .rs1 = a->rd, .imm = a->imm };
    return trans_addiw(ctx, &arg);
#endif
}

It parses RD manually, but not IMM :)  I'm actually going to eat now...

[-- Attachment #2: Type: text/html, Size: 4846 bytes --]

Re: [Qemu-riscv] [Qemu-devel] [PULL 17/29] target/riscv: Convert quadrant 1 of RVXC insns to decodetree

[Bastian Koppelmann]

On 3/15/19 1:07 PM, Palmer Dabbelt wrote:
> On Fri, Mar 15, 2019 at 4:19 AM Palmer Dabbelt <palmer@sifive.com> wrote:
>
>> On Fri, 15 Mar 2019 02:06:07 PDT (-0700), Bastian Koppelmann wrote:
>>> Hi Alistair
>>>
>>> On 3/14/19 9:28 PM, Alistair Francis wrote:
>>>> On Wed, Mar 13, 2019 at 7:53 AM Palmer Dabbelt <palmer@sifive.com>
>> wrote:
>>>>> From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>>>>>
>>>>> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
>>>>> Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>>>>> Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
>>>> This commit is the first bad commit in breaking 32-bit boot.
>>>>
>>>> It looks like the jal doesn't jump to the correct address:
>>>>
>>>> ----------------
>>>> IN:
>>>> 0x80000022:  00050433          add             s0,a0,zero
>>>> 0x80000026:  000584b3          add             s1,a1,zero
>>>> 0x8000002a:  2c79              jal             ra,670          #
>> 0x800002c8
>>>> ----------------
>>>> IN:
>>>> 0x800002c8:  00000533          add             a0,zero,zero
>>>> 0x800002cc:  8082              ret
>>>
>>> Oops! Can you point me to the binary to reproduce this?
>> I think I've traced it down to something simple: in my hello world binary
>> I see
>>
>>      20401a8c:       2a45                    jal     20401c3c <atexit>
>>
>> in the objdump, and I see
>>
>>      IN: _start
>>      0x20401a8c:  2a45              jal             ra,432          #
>> 0x20401c3c
>>
>> but then QEMU jumps to 0x20401a9d.  I have a feeling it's something wrong
>> with
>> gen_jal() that disappeared during the cleanups that we dropped in order to
>> fix
>> the build issues.
>>
>> I'm running
>>
>>      ./riscv32-softmmu/qemu-system-riscv32 -machine sifive_e -kernel
>> ~/work/sifive/freedom-e-sdk/software/hello/hello -nographic -d
>> in_asm,out_asm,exec,cpu -singlestep |& tee out.log
>>
>> on the "hifive1" branch of github.com/palmer-dabbelt/qemu, which just has
>> a
>> PRCI fixup that I forgot about and haven't sent upstream yet (I'll do that
>> after this issue).  The binary should be at
>>
>>      http://www.dabbelt.com/~palmer/hello.elf
>>
>> and the debug log at
>>
>>      http://www.dabbelt.com/~palmer/out.log
>>
>> You can build the binary from github.com/sifive/freedom-e-sdk via
>>
>>     make software PROGRAM=hello TARGET=sifive-hifive1
>>
>> using the riscv64-unknown-elf-gcc-20181127-x86_64-linux-ubuntu14 toolchain
>> binaries from our website (newer ones should work, but probably won't
>> produce
>> exactly the same output).
>>
>> I'll poke around after grabbing some dinner...
>>
> OK, I'm pretty sure this is it:
>
> c_jal_addiw       001 .  .....  ..... 01 @ci #Note: parse rd and/or imm
> manually

Thanks for the digging. Yes this bug was fixed in the patches we dropped 
:/. Heres the fix:

diff --git a/target/riscv/insn_trans/trans_rvc.inc.c 
b/target/riscv/insn_trans/trans_rvc.inc.c
index bcdf64d3b7..5241a8414e 100644
--- a/target/riscv/insn_trans/trans_rvc.inc.c
+++ b/target/riscv/insn_trans/trans_rvc.inc.c
@@ -88,7 +88,9 @@ static bool trans_c_jal_addiw(DisasContext *ctx, 
arg_c_jal_addiw *a)
  {
  #ifdef TARGET_RISCV32
      /* C.JAL */
-    arg_jal arg = { .rd = 1, .imm = a->imm };
+    arg_c_j tmp;
+    extract_cj(&tmp, ctx->opcode);
+    arg_jal arg = { .rd = 1, .imm = tmp.imm };
      return trans_jal(ctx, &arg);
  #else
      /* C.ADDIW */

I'll send a patch as soon as I have checked the other cases that require 
manual parsing.

Cheers,

Bastian

Re: [Qemu-riscv] [Qemu-devel] [PULL 17/29] target/riscv: Convert quadrant 1 of RVXC insns to decodetree

[Bastian Koppelmann]

On 3/15/19 1:44 PM, Bastian Koppelmann wrote:
>
> On 3/15/19 1:07 PM, Palmer Dabbelt wrote:
>> On Fri, Mar 15, 2019 at 4:19 AM Palmer Dabbelt <palmer@sifive.com> 
>> wrote:
>>
>>> On Fri, 15 Mar 2019 02:06:07 PDT (-0700), Bastian Koppelmann wrote:
>>>> Hi Alistair
>>>>
>>>> On 3/14/19 9:28 PM, Alistair Francis wrote:
>>>>> On Wed, Mar 13, 2019 at 7:53 AM Palmer Dabbelt <palmer@sifive.com>
>>> wrote:
>>>>>> From: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>>>>>>
>>>>>> Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
>>>>>> Signed-off-by: Bastian Koppelmann <kbastian@mail.uni-paderborn.de>
>>>>>> Signed-off-by: Peer Adelt <peer.adelt@hni.uni-paderborn.de>
>>>>> This commit is the first bad commit in breaking 32-bit boot.
>>>>>
>>>>> It looks like the jal doesn't jump to the correct address:
>>>>>
>>>>> ----------------
>>>>> IN:
>>>>> 0x80000022:  00050433          add             s0,a0,zero
>>>>> 0x80000026:  000584b3          add             s1,a1,zero
>>>>> 0x8000002a:  2c79              jal ra,670          #
>>> 0x800002c8
>>>>> ----------------
>>>>> IN:
>>>>> 0x800002c8:  00000533          add a0,zero,zero
>>>>> 0x800002cc:  8082              ret
>>>>
>>>> Oops! Can you point me to the binary to reproduce this?
>>> I think I've traced it down to something simple: in my hello world 
>>> binary
>>> I see
>>>
>>>      20401a8c:       2a45                    jal     20401c3c <atexit>
>>>
>>> in the objdump, and I see
>>>
>>>      IN: _start
>>>      0x20401a8c:  2a45              jal ra,432          #
>>> 0x20401c3c
>>>
>>> but then QEMU jumps to 0x20401a9d.  I have a feeling it's something 
>>> wrong
>>> with
>>> gen_jal() that disappeared during the cleanups that we dropped in 
>>> order to
>>> fix
>>> the build issues.
>>>
>>> I'm running
>>>
>>>      ./riscv32-softmmu/qemu-system-riscv32 -machine sifive_e -kernel
>>> ~/work/sifive/freedom-e-sdk/software/hello/hello -nographic -d
>>> in_asm,out_asm,exec,cpu -singlestep |& tee out.log
>>>
>>> on the "hifive1" branch of github.com/palmer-dabbelt/qemu, which 
>>> just has
>>> a
>>> PRCI fixup that I forgot about and haven't sent upstream yet (I'll 
>>> do that
>>> after this issue).  The binary should be at
>>>
>>>      http://www.dabbelt.com/~palmer/hello.elf
>>>
>>> and the debug log at
>>>
>>>      http://www.dabbelt.com/~palmer/out.log
>>>
>>> You can build the binary from github.com/sifive/freedom-e-sdk via
>>>
>>>     make software PROGRAM=hello TARGET=sifive-hifive1
>>>
>>> using the riscv64-unknown-elf-gcc-20181127-x86_64-linux-ubuntu14 
>>> toolchain
>>> binaries from our website (newer ones should work, but probably won't
>>> produce
>>> exactly the same output).
>>>
>>> I'll poke around after grabbing some dinner...
>>>
>> OK, I'm pretty sure this is it:
>>
>> c_jal_addiw       001 .  .....  ..... 01 @ci #Note: parse rd and/or imm
>> manually
>
> Thanks for the digging. Yes this bug was fixed in the patches we 
> dropped :/. Heres the fix:
>
> diff --git a/target/riscv/insn_trans/trans_rvc.inc.c 
> b/target/riscv/insn_trans/trans_rvc.inc.c
> index bcdf64d3b7..5241a8414e 100644
> --- a/target/riscv/insn_trans/trans_rvc.inc.c
> +++ b/target/riscv/insn_trans/trans_rvc.inc.c
> @@ -88,7 +88,9 @@ static bool trans_c_jal_addiw(DisasContext *ctx, 
> arg_c_jal_addiw *a)
>  {
>  #ifdef TARGET_RISCV32
>      /* C.JAL */
> -    arg_jal arg = { .rd = 1, .imm = a->imm };
> +    arg_c_j tmp;
> +    extract_cj(&tmp, ctx->opcode);
Sorry, it's decode_insn16_extract_cj()