QEMU AVR 8 bit cores
PATCH v37 3/17 target/avr: Add instruction decoding
In reply to Michael RolnikThis includes:
- encoding of all 16 bit instructions
- encoding of all 32 bit instructions
--- target/avr/insn.decode | 194 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 194 insertions(+) create mode 100644 target/avr/insn.decode
target/avr/insn.decode new file +194 -0
@@ -0,0 +1,194 @@ +# +# AVR instruction decode definitions. +# +# Copyright (c) 2019 Michael Rolnik <[email protected]> +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see <http://www.gnu.org/licenses/>. +# + +# +# regs_16_31_by_one = [16 .. 31] +# regs_16_23_by_one = [16 .. 23] +# regs_24_30_by_two = [24, 26, 28, 30] +# regs_00_30_by_two = [0, 2, 4, 6, 8, .. 30] + +%rd 4:5 +%rr 9:1 0:4 + +&rd_rr rd rr +&rd_imm rd imm + +@op_rd_rr .... .. . ..... .... &rd_rr rd=%rd rr=%rr +ADD 0000 11 . ..... .... @op_rd_rr +ADC 0001 11 . ..... .... @op_rd_rr +AND 0010 00 . ..... .... @op_rd_rr +CP 0001 01 . ..... .... @op_rd_rr +CPC 0000 01 . ..... .... @op_rd_rr +CPSE 0001 00 . ..... .... @op_rd_rr +EOR 0010 01 . ..... .... @op_rd_rr +MOV 0010 11 . ..... .... @op_rd_rr +MUL 1001 11 . ..... .... @op_rd_rr +OR 0010 10 . ..... .... @op_rd_rr +SBC 0000 10 . ..... .... @op_rd_rr +SUB 0001 10 . ..... .... @op_rd_rr + + +%rd_c 4:2 !function=to_regs_24_30_by_two +%imm6 6:2 0:4 + +@op_rd_imm6 .... .... .. .. .... &rd_imm rd=%rd_c imm=%imm6 +ADIW 1001 0110 .. .. .... @op_rd_imm6 +SBIW 1001 0111 .. .. .... @op_rd_imm6 + + +%rd_a 4:4 !function=to_regs_16_31_by_one +%rr_a 0:4 !function=to_regs_16_31_by_one +%rd_d 4:4 !function=to_regs_00_30_by_two +%rr_d 0:4 !function=to_regs_00_30_by_two +%imm8 8:4 0:4 + +@op_rd_imm8 .... .... .... .... &rd_imm rd=%rd_a imm=%imm8 +ANDI 0111 .... .... .... @op_rd_imm8 +CPI 0011 .... .... .... @op_rd_imm8 +LDI 1110 .... .... .... @op_rd_imm8 +ORI 0110 .... .... .... @op_rd_imm8 +SBCI 0100 .... .... .... @op_rd_imm8 +SUBI 0101 .... .... .... @op_rd_imm8 + + +@op_rd .... ... rd:5 .... +ASR 1001 010 ..... 0101 @op_rd +COM 1001 010 ..... 0000 @op_rd +DEC 1001 010 ..... 1010 @op_rd +ELPM2 1001 000 ..... 0110 @op_rd +ELPMX 1001 000 ..... 0111 @op_rd +INC 1001 010 ..... 0011 @op_rd +LDX1 1001 000 ..... 1100 @op_rd +LDX2 1001 000 ..... 1101 @op_rd +LDX3 1001 000 ..... 1110 @op_rd +LDY2 1001 000 ..... 1001 @op_rd +LDY3 1001 000 ..... 1010 @op_rd +LDZ2 1001 000 ..... 0001 @op_rd +LDZ3 1001 000 ..... 0010 @op_rd +LPM2 1001 000 ..... 0100 @op_rd +LPMX 1001 000 ..... 0101 @op_rd +LSR 1001 010 ..... 0110 @op_rd +NEG 1001 010 ..... 0001 @op_rd +POP 1001 000 ..... 1111 @op_rd +PUSH 1001 001 ..... 1111 @op_rd +ROR 1001 010 ..... 0111 @op_rd +STY2 1001 001 ..... 1001 @op_rd +STY3 1001 001 ..... 1010 @op_rd +STZ2 1001 001 ..... 0001 @op_rd +STZ3 1001 001 ..... 0010 @op_rd +SWAP 1001 010 ..... 0010 @op_rd + + +@op_bit .... .... . bit:3 .... +BCLR 1001 0100 1 ... 1000 @op_bit +BSET 1001 0100 0 ... 1000 @op_bit + + +@op_rd_bit .... ... rd:5 . bit:3 +BLD 1111 100 ..... 0 ... @op_rd_bit +BST 1111 101 ..... 0 ... @op_rd_bit + + +@op_bit_imm .... .. imm:s7 bit:3 +BRBC 1111 01 ....... ... @op_bit_imm +BRBS 1111 00 ....... ... @op_bit_imm + + +BREAK 1001 0101 1001 1000 +EICALL 1001 0101 0001 1001 +EIJMP 1001 0100 0001 1001 +ELPM1 1001 0101 1101 1000 +ICALL 1001 0101 0000 1001 +IJMP 1001 0100 0000 1001 +LPM1 1001 0101 1100 1000 +NOP 0000 0000 0000 0000 +RET 1001 0101 0000 1000 +RETI 1001 0101 0001 1000 +SLEEP 1001 0101 1000 1000 +SPM 1001 0101 1110 1000 +SPMX 1001 0101 1111 1000 +WDR 1001 0101 1010 1000 + + +@op_reg_bit .... .... reg:5 bit:3 +CBI 1001 1000 ..... ... @op_reg_bit +SBI 1001 1010 ..... ... @op_reg_bit +SBIC 1001 1001 ..... ... @op_reg_bit +SBIS 1001 1011 ..... ... @op_reg_bit + + +DES 1001 0100 imm:4 1011 + + +%rd_b 4:3 !function=to_regs_16_23_by_one +%rr_b 0:3 !function=to_regs_16_23_by_one +@fmul .... .... . ... . ... &rd_rr rd=%rd_b rr=%rr_b +FMUL 0000 0011 0 ... 1 ... @fmul +FMULS 0000 0011 1 ... 0 ... @fmul +FMULSU 0000 0011 1 ... 1 ... @fmul +MULSU 0000 0011 0 ... 0 ... @fmul + + +%io_imm 9:2 0:4 +@io_rd_imm .... . .. ..... .... &rd_imm rd=%rd imm=%io_imm +IN 1011 0 .. ..... .... @io_rd_imm +OUT 1011 1 .. ..... .... @io_rd_imm + + +XCH 1001 001 rd:5 0100 +LAC 1001 001 rd:5 0110 +LAS 1001 001 rd:5 0101 +LAT 1001 001 rd:5 0111 +STX1 1001 001 rr:5 1100 +STX2 1001 001 rr:5 1101 +STX3 1001 001 rr:5 1110 + + +%ldst_d_imm 13:1 10:2 0:3 +@ldst_d .. . . .. . rd:5 . ... &rd_imm imm=%ldst_d_imm +LDDY 10 . 0 .. 0 ..... 1 ... @ldst_d +LDDZ 10 . 0 .. 0 ..... 0 ... @ldst_d +STDY 10 . 0 .. 1 ..... 1 ... @ldst_d +STDZ 10 . 0 .. 1 ..... 0 ... @ldst_d + + +MOVW 0000 0001 .... .... &rd_rr rd=%rd_d rr=%rr_d +MULS 0000 0010 .... .... &rd_rr rd=%rd_a rr=%rr_a + +RCALL 1101 imm:s12 +RJMP 1100 imm:s12 + +SBRC 1111 110 rr:5 0 bit:3 +SBRS 1111 111 rr:5 0 bit:3 + +# The 22-bit immediate is partially in the opcode word, +# and partially in the next. Use append_16 to build the +# complete 22-bit value. +%imm_call 4:5 0:1 !function=append_16 +CALL 1001 010 ..... 111 . imm=%imm_call +JMP 1001 010 ..... 110 . imm=%imm_call + + +# The 16-bit immediate is completely in the next word. +# Fields cannot be defined with no bits, so we cannot play +# the same trick and append to a zero-bit value. +# Defer reading the immediate until trans_{LDS,STS}. +@ldst_s .... ... rd:5 .... imm=0 +LDS 1001 000 ..... 0000 @ldst_s +STS 1001 001 ..... 0000 @ldst_s
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 1/17 target/avr: Add outward facing interfaces and core CPU logic
In reply to Michael RolnikThis includes:
- CPU data structures
- object model classes and functions
- migration functions
- GDB hooks
--- target/avr/cpu-param.h | 37 +++ target/avr/cpu-qom.h | 54 ++++ target/avr/cpu.h | 253 ++++++++++++++++++ target/avr/cpu.c | 576 +++++++++++++++++++++++++++++++++++++++++ target/avr/gdbstub.c | 85 ++++++ target/avr/machine.c | 121 +++++++++ gdb-xml/avr-cpu.xml | 49 ++++ 7 files changed, 1175 insertions(+) create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/cpu.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/machine.c create mode 100644 gdb-xml/avr-cpu.xml
target/avr/cpu-param.h new file +37 -0
@@ -0,0 +1,37 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +#ifndef AVR_CPU_PARAM_H +#define AVR_CPU_PARAM_H 1 + +#define TARGET_LONG_BITS 32 +/* +* TARGET_PAGE_BITS cannot be more than 8 bits because +* 1. all IO registers occupy [0x0000 .. 0x00ff] address range, and they +* should be implemented as a device and not memory +* 2. SRAM starts at the address 0x0100 +*/ +#define TARGET_PAGE_BITS 8 +#define TARGET_PHYS_ADDR_SPACE_BITS 24 +#define TARGET_VIRT_ADDR_SPACE_BITS 24 +#define NB_MMU_MODES 2 + + +#endif
target/avr/cpu-qom.h new file +54 -0
54 lines not on this page · raw message
target/avr/cpu.h new file +253 -0
253 lines not on this page · raw message
target/avr/cpu.c new file +576 -0
576 lines not on this page · raw message
target/avr/gdbstub.c new file +85 -0
85 lines not on this page · raw message
target/avr/machine.c new file +121 -0
121 lines not on this page · raw message
gdb-xml/avr-cpu.xml new file +49 -0
49 lines not on this page · raw message
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 5/17 target/avr: Add instruction translation - Arithmetic and Logic Instructions
In reply to Michael RolnikThis includes:
- ADD, ADC, ADIW
- SBIW, SUB, SUBI, SBC, SBCI
- AND, ANDI
- OR, ORI, EOR
- COM, NEG
- INC, DEC
- MUL, MULS, MULSU
- FMUL, FMULS, FMULSU
- DES
--- target/avr/translate.c | 822 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 822 insertions(+)
target/avr/translate.c +822 -0
@@ -143,3 +143,825 @@ static bool avr_have_feature(DisasContext *ctx, int feature)static bool decode_insn (DisasContext *ctx,uint16_t insn);#include "decode_insn.inc.c" + +static void gen_add_CHf (TCGv R ,TCGv Rd,TCGv Rr) +{ +TCGv t1 =tcg_temp_new_i32 (); +TCGv t2 =tcg_temp_new_i32 (); +TCGv t3 =tcg_temp_new_i32 (); + +tcg_gen_and_tl (t1, Rd, Rr);/* t1 = Rd & Rr */ +tcg_gen_andc_tl (t2, Rd,R );/* t2 = Rd & ~R */ +tcg_gen_andc_tl (t3, Rr,R );/* t3 = Rr & ~R */ +tcg_gen_or_tl (t1, t1, t2);/* t1 = t1 | t2 | t3 */ +tcg_gen_or_tl (t1, t1, t3); + +tcg_gen_shri_tl (cpu_Cf, t1,7 );/* Cf = t1(7) */ +tcg_gen_shri_tl (cpu_Hf, t1,3 );/* Hf = t1(3) */ +tcg_gen_andi_tl (cpu_Hf, cpu_Hf,1 ); + +tcg_temp_free_i32 (t3); +tcg_temp_free_i32 (t2); +tcg_temp_free_i32 (t1); +} + + +static void gen_add_Vf (TCGv R ,TCGv Rd,TCGv Rr) +{ +TCGv t1 =tcg_temp_new_i32 (); +TCGv t2 =tcg_temp_new_i32 (); + +/* t1 = Rd & Rr & ~R | ~Rd & ~Rr & R = (Rd ^ R) & ~(Rd ^ Rr) */ +tcg_gen_xor_tl (t1, Rd,R ); +tcg_gen_xor_tl (t2, Rd, Rr); +tcg_gen_andc_tl (t1, t1, t2); + +tcg_gen_shri_tl (cpu_Vf, t1,7 );/* Vf = t1(7) */ + +tcg_temp_free_i32 (t2); +tcg_temp_free_i32 (t1); +} + + +static void gen_sub_CHf (TCGv R ,TCGv Rd,TCGv Rr) +{ +TCGv t1 =tcg_temp_new_i32 (); +TCGv t2 =tcg_temp_new_i32 (); +TCGv t3 =tcg_temp_new_i32 (); + +/* Cf & Hf */ +tcg_gen_not_tl (t1, Rd);/* t1 = ~Rd */ +tcg_gen_and_tl (t2, t1, Rr);/* t2 = ~Rd & Rr */ +tcg_gen_or_tl (t3, t1, Rr);/* t3 = (~Rd | Rr) & R */ +tcg_gen_and_tl (t3, t3,R ); +tcg_gen_or_tl (t2, t2, t3);/* t2 = ~Rd & Rr | ~Rd & R | R & Rr */ +tcg_gen_shri_tl (cpu_Cf, t2,7 );/* Cf = t2(7) */ +tcg_gen_shri_tl (cpu_Hf, t2,3 );/* Hf = t2(3) */ +tcg_gen_andi_tl (cpu_Hf, cpu_Hf,1 ); + +tcg_temp_free_i32 (t3); +tcg_temp_free_i32 (t2); +tcg_temp_free_i32 (t1); +} + + +static void gen_sub_Vf (TCGv R ,TCGv Rd,TCGv Rr) +{ +TCGv t1 =tcg_temp_new_i32 (); +TCGv t2 =tcg_temp_new_i32 (); + +/* Vf */ +/* t1 = Rd & ~Rr & ~R | ~Rd & Rr & R = (Rd ^ R) & (Rd ^ R) */ +tcg_gen_xor_tl (t1, Rd,R ); +tcg_gen_xor_tl (t2, Rd, Rr); +tcg_gen_and_tl (t1, t1, t2); +tcg_gen_shri_tl (cpu_Vf, t1,7 );/* Vf = t1(7) */ + +tcg_temp_free_i32 (t2); +tcg_temp_free_i32 (t1); +} + + +static void gen_NSf (TCGv R ) +{ +tcg_gen_shri_tl (cpu_Nf,R ,7 );/* Nf = R(7) */ +tcg_gen_xor_tl (cpu_Sf, cpu_Nf, cpu_Vf);/* Sf = Nf ^ Vf */ +} + + +static void gen_ZNSf (TCGv R ) +{ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ +tcg_gen_shri_tl (cpu_Nf,R ,7 );/* Nf = R(7) */ +tcg_gen_xor_tl (cpu_Sf, cpu_Nf, cpu_Vf);/* Sf = Nf ^ Vf */ +} + + +/* +* Adds two registers without the C Flag and places the result in the +* destination register Rd. +*/ +static bool trans_ADD (DisasContext *ctx,arg_ADD *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_add_tl (R , Rd, Rr);/* Rd = Rd + Rr */ +tcg_gen_andi_tl (R ,R ,0xff );/* make it 8 bits */ + +gen_add_CHf (R , Rd, Rr); +gen_add_Vf (R , Rd, Rr); +gen_ZNSf (R ); + +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* Adds two registers and the contents of the C Flag and places the result in +* the destination register Rd. +*/ +static bool trans_ADC (DisasContext *ctx,arg_ADC *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_add_tl (R , Rd, Rr);/* R = Rd + Rr + Cf */ +tcg_gen_add_tl (R ,R , cpu_Cf); +tcg_gen_andi_tl (R ,R ,0xff );/* make it 8 bits */ + +gen_add_CHf (R , Rd, Rr); +gen_add_Vf (R , Rd, Rr); +gen_ZNSf (R ); + +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* Subtracts an immediate value (0-63) from a register pair and places the +* result in the register pair. This instruction operates on the upper four +* register pairs, and is well suited for operations on the Pointer Registers. +* This instruction is not available in all devices. Refer to the device +* specific instruction set summary. +*/ +static bool trans_SBIW (DisasContext *ctx,arg_SBIW *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_ADIW_SBIW )) { +return true; + } + +TCGv RdL = cpu_r[a->rd]; +TCGv RdH = cpu_r[a->rd +1 ]; +int Imm = (a->imm); +TCGv R =tcg_temp_new_i32 (); +TCGv Rd =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_deposit_tl (Rd, RdL, RdH,8 ,8 );/* Rd = RdH:RdL */ +tcg_gen_subi_tl (R , Rd, Imm);/* R = Rd - Imm */ +tcg_gen_andi_tl (R ,R ,0xffff );/* make it 16 bits */ + +/* Cf */ +tcg_gen_andc_tl (cpu_Cf,R , Rd); +tcg_gen_shri_tl (cpu_Cf, cpu_Cf,15 );/* Cf = R & ~Rd */ + +/* Vf */ +tcg_gen_andc_tl (cpu_Vf, Rd,R ); +tcg_gen_shri_tl (cpu_Vf, cpu_Vf,15 );/* Vf = Rd & ~R */ + +/* Zf */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +/* Nf */ +tcg_gen_shri_tl (cpu_Nf,R ,15 );/* Nf = R(15) */ + +/* Sf */ +tcg_gen_xor_tl (cpu_Sf, cpu_Nf, cpu_Vf);/* Sf = Nf ^ Vf */ + +/* R */ +tcg_gen_andi_tl (RdL,R ,0xff ); +tcg_gen_shri_tl (RdH,R ,8 ); + +tcg_temp_free_i32 (Rd); +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* Adds an immediate value (0 - 63) to a register pair and places the result +* in the register pair. This instruction operates on the upper four register +* pairs, and is well suited for operations on the pointer registers. This +* instruction is not available in all devices. Refer to the device specific +* instruction set summary. +*/ +static bool trans_ADIW (DisasContext *ctx,arg_ADIW *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_ADIW_SBIW )) { +return true; + } + +TCGv RdL = cpu_r[a->rd]; +TCGv RdH = cpu_r[a->rd +1 ]; +int Imm = (a->imm); +TCGv R =tcg_temp_new_i32 (); +TCGv Rd =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_deposit_tl (Rd, RdL, RdH,8 ,8 );/* Rd = RdH:RdL */ +tcg_gen_addi_tl (R , Rd, Imm);/* R = Rd + Imm */ +tcg_gen_andi_tl (R ,R ,0xffff );/* make it 16 bits */ + +/* Cf */ +tcg_gen_andc_tl (cpu_Cf, Rd,R );/* Cf = Rd & ~R */ +tcg_gen_shri_tl (cpu_Cf, cpu_Cf,15 ); + +/* Vf */ +tcg_gen_andc_tl (cpu_Vf,R , Rd);/* Vf = R & ~Rd */ +tcg_gen_shri_tl (cpu_Vf, cpu_Vf,15 ); + +/* Zf */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +/* Nf */ +tcg_gen_shri_tl (cpu_Nf,R ,15 );/* Nf = R(15) */ + +/* Sf */ +tcg_gen_xor_tl (cpu_Sf, cpu_Nf, cpu_Vf);/* Sf = Nf ^ Vf */ + +/* R */ +tcg_gen_andi_tl (RdL,R ,0xff ); +tcg_gen_shri_tl (RdH,R ,8 ); + +tcg_temp_free_i32 (Rd); +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* Subtracts two registers and places the result in the destination +* register Rd. +*/ +static bool trans_SUB (DisasContext *ctx,arg_SUB *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_sub_tl (R , Rd, Rr);/* R = Rd - Rr */ +tcg_gen_andi_tl (R ,R ,0xff );/* make it 8 bits */ + +gen_sub_CHf (R , Rd, Rr); +gen_sub_Vf (R , Rd, Rr); +gen_ZNSf (R ); + +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + +/* +* Subtracts a register and a constant and places the result in the +* destination register Rd. This instruction is working on Register R16 to R31 +* and is very well suited for operations on the X, Y, and Z-pointers. +*/ +static bool trans_SUBI (DisasContext *ctx,arg_SUBI *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr =tcg_const_i32 (a->imm); +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_sub_tl (R , Rd, Rr);/* R = Rd - Imm */ +tcg_gen_andi_tl (R ,R ,0xff );/* make it 8 bits */ + +gen_sub_CHf (R , Rd, Rr); +gen_sub_Vf (R , Rd, Rr); +gen_ZNSf (R ); + +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); +tcg_temp_free_i32 (Rr); + +return true; +} + + +/* +* Subtracts two registers and subtracts with the C Flag and places the +* result in the destination register Rd. +*/ +static bool trans_SBC (DisasContext *ctx,arg_SBC *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); +TCGv zero =tcg_const_i32 (0 ); + +/* op */ +tcg_gen_sub_tl (R , Rd, Rr);/* R = Rd - Rr - Cf */ +tcg_gen_sub_tl (R ,R , cpu_Cf); +tcg_gen_andi_tl (R ,R ,0xff );/* make it 8 bits */ + +gen_sub_CHf (R , Rd, Rr); +gen_sub_Vf (R , Rd, Rr); +gen_NSf (R ); + +/* +* Previous value remains unchanged when the result is zero; +* cleared otherwise. +*/ +tcg_gen_movcond_tl (TCG_COND_EQ , cpu_Zf,R , zero, cpu_Zf, zero); + +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (zero); +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* SBCI -- Subtract Immediate with Carry +*/ +static bool trans_SBCI (DisasContext *ctx,arg_SBCI *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr =tcg_const_i32 (a->imm); +TCGv R =tcg_temp_new_i32 (); +TCGv zero =tcg_const_i32 (0 ); + +/* op */ +tcg_gen_sub_tl (R , Rd, Rr);/* R = Rd - Rr - Cf */ +tcg_gen_sub_tl (R ,R , cpu_Cf); +tcg_gen_andi_tl (R ,R ,0xff );/* make it 8 bits */ + +gen_sub_CHf (R , Rd, Rr); +gen_sub_Vf (R , Rd, Rr); +gen_NSf (R ); + +/* +* Previous value remains unchanged when the result is zero; +* cleared otherwise. +*/ +tcg_gen_movcond_tl (TCG_COND_EQ , cpu_Zf,R , zero, cpu_Zf, zero); + +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (zero); +tcg_temp_free_i32 (R ); +tcg_temp_free_i32 (Rr); + +return true; +} + + +/* +* Performs the logical AND between the contents of register Rd and register +* Rr and places the result in the destination register Rd. +*/ +static bool trans_AND (DisasContext *ctx,arg_AND *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_and_tl (R , Rd, Rr);/* Rd = Rd and Rr */ + +/* Vf */ +tcg_gen_movi_tl (cpu_Vf,0 );/* Vf = 0 */ + +/* Zf */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +gen_ZNSf (R ); + +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* Performs the logical AND between the contents of register Rd and a constant +* and places the result in the destination register Rd. +*/ +static bool trans_ANDI (DisasContext *ctx,arg_ANDI *a) +{ +TCGv Rd = cpu_r[a->rd]; +int Imm = (a->imm); + +/* op */ +tcg_gen_andi_tl (Rd, Rd, Imm);/* Rd = Rd & Imm */ + +tcg_gen_movi_tl (cpu_Vf,0x00 );/* Vf = 0 */ +gen_ZNSf (Rd); + +return true; +} + + +/* +* Performs the logical OR between the contents of register Rd and register +* Rr and places the result in the destination register Rd. +*/ +static bool trans_OR (DisasContext *ctx,arg_OR *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +tcg_gen_or_tl (R , Rd, Rr); + +tcg_gen_movi_tl (cpu_Vf,0 ); +gen_ZNSf (R ); + +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* Performs the logical OR between the contents of register Rd and a +* constant and places the result in the destination register Rd. +*/ +static bool trans_ORI (DisasContext *ctx,arg_ORI *a) +{ +TCGv Rd = cpu_r[a->rd]; +int Imm = (a->imm); + +tcg_gen_ori_tl (Rd, Rd, Imm);/* Rd = Rd | Imm */ + +tcg_gen_movi_tl (cpu_Vf,0x00 );/* Vf = 0 */ +gen_ZNSf (Rd); + +return true; +} + + +/* +* Performs the logical EOR between the contents of register Rd and +* register Rr and places the result in the destination register Rd. +*/ +static bool trans_EOR (DisasContext *ctx,arg_EOR *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; + +tcg_gen_xor_tl (Rd, Rd, Rr); + +tcg_gen_movi_tl (cpu_Vf,0 ); +gen_ZNSf (Rd); + +return true; +} + + +/* +* Clears the specified bits in register Rd. Performs the logical AND +* between the contents of register Rd and the complement of the constant mask +* K. The result will be placed in register Rd. +*/ +static bool trans_COM (DisasContext *ctx,arg_COM *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv R =tcg_temp_new_i32 (); + +tcg_gen_xori_tl (Rd, Rd,0xff ); + +tcg_gen_movi_tl (cpu_Cf,1 );/* Cf = 1 */ +tcg_gen_movi_tl (cpu_Vf,0 );/* Vf = 0 */ +gen_ZNSf (Rd); + +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* Replaces the contents of register Rd with its two's complement; the +* value $80 is left unchanged. +*/ +static bool trans_NEG (DisasContext *ctx,arg_NEG *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv t0 =tcg_const_i32 (0 ); +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_sub_tl (R , t0, Rd);/* R = 0 - Rd */ +tcg_gen_andi_tl (R ,R ,0xff );/* make it 8 bits */ + +gen_sub_CHf (R , t0, Rd); +gen_sub_Vf (R , t0, Rd); +gen_ZNSf (R ); + +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (t0); +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* Adds one -1- to the contents of register Rd and places the result in the +* destination register Rd. The C Flag in SREG is not affected by the +* operation, thus allowing the INC instruction to be used on a loop counter in +* multiple-precision computations. When operating on unsigned numbers, only +* BREQ and BRNE branches can be expected to perform consistently. When +* operating on two's complement values, all signed branches are available. +*/ +static bool trans_INC (DisasContext *ctx,arg_INC *a) +{ +TCGv Rd = cpu_r[a->rd]; + +tcg_gen_addi_tl (Rd, Rd,1 ); +tcg_gen_andi_tl (Rd, Rd,0xff ); + +/* cpu_Vf = Rd == 0x80 */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Vf, Rd,0x80 ); +gen_ZNSf (Rd); + +return true; +} + + +/* +* Subtracts one -1- from the contents of register Rd and places the result +* in the destination register Rd. The C Flag in SREG is not affected by the +* operation, thus allowing the DEC instruction to be used on a loop counter in +* multiple-precision computations. When operating on unsigned values, only +* BREQ and BRNE branches can be expected to perform consistently. When +* operating on two's complement values, all signed branches are available. +*/ +static bool trans_DEC (DisasContext *ctx,arg_DEC *a) +{ +TCGv Rd = cpu_r[a->rd]; + +tcg_gen_subi_tl (Rd, Rd,1 );/* Rd = Rd - 1 */ +tcg_gen_andi_tl (Rd, Rd,0xff );/* make it 8 bits */ + +/* cpu_Vf = Rd == 0x7f */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Vf, Rd,0x7f ); +gen_ZNSf (Rd); + +return true; +} + + +/* +* This instruction performs 8-bit x 8-bit -> 16-bit unsigned multiplication. +*/ +static bool trans_MUL (DisasContext *ctx,arg_MUL *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_MUL )) { +return true; + } + +TCGv R0 = cpu_r[0 ]; +TCGv R1 = cpu_r[1 ]; +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +tcg_gen_mul_tl (R , Rd, Rr);/* R = Rd * Rr */ + +tcg_gen_andi_tl (R0 ,R ,0xff ); +tcg_gen_shri_tl (R1 ,R ,8 ); + +tcg_gen_shri_tl (cpu_Cf,R ,15 );/* Cf = R(15) */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* This instruction performs 8-bit x 8-bit -> 16-bit signed multiplication. +*/ +static bool trans_MULS (DisasContext *ctx,arg_MULS *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_MUL )) { +return true; + } + +TCGv R0 = cpu_r[0 ]; +TCGv R1 = cpu_r[1 ]; +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); +TCGv t0 =tcg_temp_new_i32 (); +TCGv t1 =tcg_temp_new_i32 (); + +tcg_gen_ext8s_tl (t0, Rd);/* make Rd full 32 bit signed */ +tcg_gen_ext8s_tl (t1, Rr);/* make Rr full 32 bit signed */ +tcg_gen_mul_tl (R , t0, t1);/* R = Rd * Rr */ +tcg_gen_andi_tl (R ,R ,0xffff );/* make it 16 bits */ + +tcg_gen_andi_tl (R0 ,R ,0xff ); +tcg_gen_shri_tl (R1 ,R ,8 ); + +tcg_gen_shri_tl (cpu_Cf,R ,15 );/* Cf = R(15) */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +tcg_temp_free_i32 (t1); +tcg_temp_free_i32 (t0); +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* This instruction performs 8-bit x 8-bit -> 16-bit multiplication of a +* signed and an unsigned number. +*/ +static bool trans_MULSU (DisasContext *ctx,arg_MULSU *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_MUL )) { +return true; + } + +TCGv R0 = cpu_r[0 ]; +TCGv R1 = cpu_r[1 ]; +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); +TCGv t0 =tcg_temp_new_i32 (); + +tcg_gen_ext8s_tl (t0, Rd);/* make Rd full 32 bit signed */ +tcg_gen_mul_tl (R , t0, Rr);/* R = Rd *Rr */ + +tcg_gen_andi_tl (R0 ,R ,0xff ); +tcg_gen_shri_tl (R1 ,R ,8 ); + +tcg_gen_shri_tl (cpu_Cf,R ,16 );/* Cf = R(16) */ +tcg_gen_mov_tl (cpu_Zf,R ); + +tcg_temp_free_i32 (t0); +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* This instruction performs 8-bit x 8-bit -> 16-bit unsigned +* multiplication and shifts the result one bit left. +*/ +static bool trans_FMUL (DisasContext *ctx,arg_FMUL *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_MUL )) { +return true; + } + +TCGv R0 = cpu_r[0 ]; +TCGv R1 = cpu_r[1 ]; +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +tcg_gen_mul_tl (R , Rd, Rr);/* R = Rd * Rr */ + +tcg_gen_shri_tl (cpu_Cf,R ,15 );/* Cf = R(15) */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +tcg_gen_shli_tl (R ,R ,1 ); + +tcg_gen_andi_tl (R0 ,R ,0xff ); +tcg_gen_shri_tl (R1 ,R ,8 ); +tcg_gen_andi_tl (R1 ,R1 ,0xff ); + + +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* This instruction performs 8-bit x 8-bit -> 16-bit signed multiplication +* and shifts the result one bit left. +*/ +static bool trans_FMULS (DisasContext *ctx,arg_FMULS *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_MUL )) { +return true; + } + +TCGv R0 = cpu_r[0 ]; +TCGv R1 = cpu_r[1 ]; +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); +TCGv t0 =tcg_temp_new_i32 (); +TCGv t1 =tcg_temp_new_i32 (); + +tcg_gen_ext8s_tl (t0, Rd);/* make Rd full 32 bit signed */ +tcg_gen_ext8s_tl (t1, Rr);/* make Rr full 32 bit signed */ +tcg_gen_mul_tl (R , t0, t1);/* R = Rd * Rr */ +tcg_gen_andi_tl (R ,R ,0xffff );/* make it 16 bits */ + +tcg_gen_shri_tl (cpu_Cf,R ,15 );/* Cf = R(15) */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +tcg_gen_shli_tl (R ,R ,1 ); + +tcg_gen_andi_tl (R0 ,R ,0xff ); +tcg_gen_shri_tl (R1 ,R ,8 ); +tcg_gen_andi_tl (R1 ,R1 ,0xff ); + +tcg_temp_free_i32 (t1); +tcg_temp_free_i32 (t0); +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* This instruction performs 8-bit x 8-bit -> 16-bit signed multiplication +* and shifts the result one bit left. +*/ +static bool trans_FMULSU (DisasContext *ctx,arg_FMULSU *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_MUL )) { +return true; + } + +TCGv R0 = cpu_r[0 ]; +TCGv R1 = cpu_r[1 ]; +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); +TCGv t0 =tcg_temp_new_i32 (); + +tcg_gen_ext8s_tl (t0, Rd);/* make Rd full 32 bit signed */ +tcg_gen_mul_tl (R , t0, Rr);/* R = Rd * Rr */ +tcg_gen_andi_tl (R ,R ,0xffff );/* make it 16 bits */ + +tcg_gen_shri_tl (cpu_Cf,R ,15 );/* Cf = R(15) */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +tcg_gen_shli_tl (R ,R ,1 ); + +tcg_gen_andi_tl (R0 ,R ,0xff ); +tcg_gen_shri_tl (R1 ,R ,8 ); +tcg_gen_andi_tl (R1 ,R1 ,0xff ); + +tcg_temp_free_i32 (t0); +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* The module is an instruction set extension to the AVR CPU, performing +* DES iterations. The 64-bit data block (plaintext or ciphertext) is placed in +* the CPU register file, registers R0-R7, where LSB of data is placed in LSB +* of R0 and MSB of data is placed in MSB of R7. The full 64-bit key (including +* parity bits) is placed in registers R8- R15, organized in the register file +* with LSB of key in LSB of R8 and MSB of key in MSB of R15. Executing one DES +* instruction performs one round in the DES algorithm. Sixteen rounds must be +* executed in increasing order to form the correct DES ciphertext or +* plaintext. Intermediate results are stored in the register file (R0-R15) +* after each DES instruction. The instruction's operand (K) determines which +* round is executed, and the half carry flag (H) determines whether encryption +* or decryption is performed. The DES algorithm is described in +* "Specifications for the Data Encryption Standard" (Federal Information +* Processing Standards Publication 46). Intermediate results in this +* implementation differ from the standard because the initial permutation and +* the inverse initial permutation are performed each iteration. This does not +* affect the result in the final ciphertext or plaintext, but reduces +* execution time. +*/ +static bool trans_DES (DisasContext *ctx,arg_DES *a) +{ +/* TODO */ +if (!avr_have_feature (ctx,AVR_FEATURE_DES )) { +return true; + } + +return true; +}
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 6/17 target/avr: Add instruction translation - Branch Instructions
In reply to Michael RolnikThis includes:
- RJMP, IJMP, EIJMP, JMP
- RCALL, ICALL, EICALL, CALL
- RET, RETI
- CPSE, CP, CPC, CPI
- SBRC, SBRS, SBIC, SBIS
- BRBC, BRBS
--- target/avr/translate.c | 549 ++++++++++++++++++++++++++++++++++++++++- 1 file changed, 546 insertions(+), 3 deletions(-)
target/avr/translate.c +546 -3
@@ -237,6 +237,86 @@ static void gen_ZNSf(TCGv R) } +static void gen_push_ret (DisasContext *ctx,int ret) +{ +if (avr_feature (ctx->env,AVR_FEATURE_1_BYTE_PC )) { + +TCGv t0 =tcg_const_i32 ((ret &0x0000ff )); + +tcg_gen_qemu_st_tl (t0, cpu_sp,MMU_DATA_IDX ,MO_UB ); +tcg_gen_subi_tl (cpu_sp, cpu_sp,1 ); + +tcg_temp_free_i32 (t0); + }else if (avr_feature (ctx->env,AVR_FEATURE_2_BYTE_PC )) { + +TCGv t0 =tcg_const_i32 ((ret &0x00ffff )); + +tcg_gen_subi_tl (cpu_sp, cpu_sp,1 ); +tcg_gen_qemu_st_tl (t0, cpu_sp,MMU_DATA_IDX ,MO_BEUW ); +tcg_gen_subi_tl (cpu_sp, cpu_sp,1 ); + +tcg_temp_free_i32 (t0); + + }else if (avr_feature (ctx->env,AVR_FEATURE_3_BYTE_PC )) { + +TCGv lo =tcg_const_i32 ((ret &0x0000ff )); +TCGv hi =tcg_const_i32 ((ret &0xffff00 ) >>8 ); + +tcg_gen_qemu_st_tl (lo, cpu_sp,MMU_DATA_IDX ,MO_UB ); +tcg_gen_subi_tl (cpu_sp, cpu_sp,2 ); +tcg_gen_qemu_st_tl (hi, cpu_sp,MMU_DATA_IDX ,MO_BEUW ); +tcg_gen_subi_tl (cpu_sp, cpu_sp,1 ); + +tcg_temp_free_i32 (lo); +tcg_temp_free_i32 (hi); + } +} + + +static void gen_pop_ret (DisasContext *ctx,TCGv ret) +{ +if (avr_feature (ctx->env,AVR_FEATURE_1_BYTE_PC )) { +tcg_gen_addi_tl (cpu_sp, cpu_sp,1 ); +tcg_gen_qemu_ld_tl (ret, cpu_sp,MMU_DATA_IDX ,MO_UB ); + }else if (avr_feature (ctx->env,AVR_FEATURE_2_BYTE_PC )) { +tcg_gen_addi_tl (cpu_sp, cpu_sp,1 ); +tcg_gen_qemu_ld_tl (ret, cpu_sp,MMU_DATA_IDX ,MO_BEUW ); +tcg_gen_addi_tl (cpu_sp, cpu_sp,1 ); + }else if (avr_feature (ctx->env,AVR_FEATURE_3_BYTE_PC )) { +TCGv lo =tcg_temp_new_i32 (); +TCGv hi =tcg_temp_new_i32 (); + +tcg_gen_addi_tl (cpu_sp, cpu_sp,1 ); +tcg_gen_qemu_ld_tl (hi, cpu_sp,MMU_DATA_IDX ,MO_BEUW ); + +tcg_gen_addi_tl (cpu_sp, cpu_sp,2 ); +tcg_gen_qemu_ld_tl (lo, cpu_sp,MMU_DATA_IDX ,MO_UB ); + +tcg_gen_deposit_tl (ret, lo, hi,8 ,16 ); + +tcg_temp_free_i32 (lo); +tcg_temp_free_i32 (hi); + } +} + + +static void gen_goto_tb (DisasContext *ctx,int n,target_ulong dest) +{ +TranslationBlock *tb = ctx->tb; + +if (ctx->singlestep ==0 ) { +tcg_gen_goto_tb (n); +tcg_gen_movi_i32 (cpu_pc, dest); +tcg_gen_exit_tb (tb, n); + }else { +tcg_gen_movi_i32 (cpu_pc, dest); +gen_helper_debug (cpu_env); +tcg_gen_exit_tb (NULL ,0 ); + } + ctx->bstate =DISAS_NORETURN ; +} + +/* * Adds two registers without the C Flag and places the result in the * destination register Rd.
@@ -810,13 +890,14 @@ static bool trans_MULSU(DisasContext *ctx, arg_MULSU *a)TCGv t0 =tcg_temp_new_i32 ();tcg_gen_ext8s_tl (t0, Rd);/* make Rd full 32 bit signed */ -tcg_gen_mul_tl (R , t0, Rr);/* R = Rd *Rr */ +tcg_gen_mul_tl (R , t0, Rr);/* R = Rd * Rr */ +tcg_gen_andi_tl (R ,R ,0xffff );/* make R 16 bits */ tcg_gen_andi_tl (R0 ,R ,0xff );tcg_gen_shri_tl (R1 ,R ,8 ); -tcg_gen_shri_tl (cpu_Cf,R ,16 );/* Cf = R(16) */ -tcg_gen_mov_tl (cpu_Zf,R ); +tcg_gen_shri_tl (cpu_Cf,R ,15 );/* Cf = R(15) */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ tcg_temp_free_i32 (t0);tcg_temp_free_i32 (R );
@@ -965,3 +1046,465 @@ static bool trans_DES(DisasContext *ctx, arg_DES *a)return true; } + + +/* +* Relative jump to an address within PC - 2K +1 and PC + 2K (words). For +* AVR microcontrollers with Program memory not exceeding 4K words (8KB) this +* instruction can address the entire memory from every address location. See +* also JMP. +*/ +static bool trans_RJMP (DisasContext *ctx,arg_RJMP *a) +{ +int dst = ctx->npc + a->imm; + +gen_goto_tb (ctx,0 , dst); + +return true; +} + + +/* +* Indirect jump to the address pointed to by the Z (16 bits) Pointer +* Register in the Register File. The Z-pointer Register is 16 bits wide and +* allows jump within the lowest 64K words (128KB) section of Program memory. +* This instruction is not available in all devices. Refer to the device +* specific instruction set summary. +*/ +static bool trans_IJMP (DisasContext *ctx,arg_IJMP *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_IJMP_ICALL )) { +return true; + } + +gen_jmp_z (ctx); + +return true; +} + + +/* +* Indirect jump to the address pointed to by the Z (16 bits) Pointer +* Register in the Register File and the EIND Register in the I/O space. This +* instruction allows for indirect jumps to the entire 4M (words) Program +* memory space. See also IJMP. This instruction is not available in all +* devices. Refer to the device specific instruction set summary. +*/ +static bool trans_EIJMP (DisasContext *ctx,arg_EIJMP *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_EIJMP_EICALL )) { +return true; + } + +gen_jmp_ez (ctx); +return true; +} + + +/* +* Jump to an address within the entire 4M (words) Program memory. See also +* RJMP. This instruction is not available in all devices. Refer to the device +* specific instruction set summary.0 +*/ +static bool trans_JMP (DisasContext *ctx,arg_JMP *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_JMP_CALL )) { +return true; + } + +gen_goto_tb (ctx,0 , a->imm); + +return true; +} + + +/* +* Relative call to an address within PC - 2K + 1 and PC + 2K (words). The +* return address (the instruction after the RCALL) is stored onto the Stack. +* See also CALL. For AVR microcontrollers with Program memory not exceeding 4K +* words (8KB) this instruction can address the entire memory from every +* address location. The Stack Pointer uses a post-decrement scheme during +* RCALL. +*/ +static bool trans_RCALL (DisasContext *ctx,arg_RCALL *a) +{ +int ret = ctx->npc; +int dst = ctx->npc + a->imm; + +gen_push_ret (ctx, ret); +gen_goto_tb (ctx,0 , dst); + +return true; +} + + +/* +* Calls to a subroutine within the entire 4M (words) Program memory. The +* return address (to the instruction after the CALL) will be stored onto the +* Stack. See also RCALL. The Stack Pointer uses a post-decrement scheme during +* CALL. This instruction is not available in all devices. Refer to the device +* specific instruction set summary. +*/ +static bool trans_ICALL (DisasContext *ctx,arg_ICALL *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_IJMP_ICALL )) { +return true; + } + +int ret = ctx->npc; + +gen_push_ret (ctx, ret); +gen_jmp_z (ctx); + +return true; +} + + +/* +* Indirect call of a subroutine pointed to by the Z (16 bits) Pointer +* Register in the Register File and the EIND Register in the I/O space. This +* instruction allows for indirect calls to the entire 4M (words) Program +* memory space. See also ICALL. The Stack Pointer uses a post-decrement scheme +* during EICALL. This instruction is not available in all devices. Refer to +* the device specific instruction set summary. +*/ +static bool trans_EICALL (DisasContext *ctx,arg_EICALL *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_EIJMP_EICALL )) { +return true; + } + +int ret = ctx->npc; + +gen_push_ret (ctx, ret); +gen_jmp_ez (ctx); +return true; +} + + +/* +* Calls to a subroutine within the entire Program memory. The return +* address (to the instruction after the CALL) will be stored onto the Stack. +* (See also RCALL). The Stack Pointer uses a post-decrement scheme during +* CALL. This instruction is not available in all devices. Refer to the device +* specific instruction set summary. +*/ +static bool trans_CALL (DisasContext *ctx,arg_CALL *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_JMP_CALL )) { +return true; + } + +int Imm = a->imm; +int ret = ctx->npc; + +gen_push_ret (ctx, ret); +gen_goto_tb (ctx,0 , Imm); + +return true; +} + + +/* +* Returns from subroutine. The return address is loaded from the STACK. +* The Stack Pointer uses a preincrement scheme during RET. +*/ +static bool trans_RET (DisasContext *ctx,arg_RET *a) +{ +gen_pop_ret (ctx, cpu_pc); + + ctx->bstate =DISAS_LOOKUP ; +return true; +} + + +/* +* Returns from interrupt. The return address is loaded from the STACK and +* the Global Interrupt Flag is set. Note that the Status Register is not +* automatically stored when entering an interrupt routine, and it is not +* restored when returning from an interrupt routine. This must be handled by +* the application program. The Stack Pointer uses a pre-increment scheme +* during RETI. +*/ +static bool trans_RETI (DisasContext *ctx,arg_RETI *a) +{ +gen_pop_ret (ctx, cpu_pc); +tcg_gen_movi_tl (cpu_If,1 ); + +/* Need to return to main loop to re-evaluate interrupts. */ + ctx->bstate =DISAS_EXIT ; +return true; +} + + +/* +* This instruction performs a compare between two registers Rd and Rr, and +* skips the next instruction if Rd = Rr. +*/ +static bool trans_CPSE (DisasContext *ctx,arg_CPSE *a) +{ + ctx->skip_cond =TCG_COND_EQ ; + ctx->skip_var0 = cpu_r[a->rd]; + ctx->skip_var1 = cpu_r[a->rr]; +return true; +} + + +/* +* This instruction performs a compare between two registers Rd and Rr. +* None of the registers are changed. All conditional branches can be used +* after this instruction. +*/ +static bool trans_CP (DisasContext *ctx,arg_CP *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_sub_tl (R , Rd, Rr);/* R = Rd - Rr */ +tcg_gen_andi_tl (R ,R ,0xff );/* make it 8 bits */ + +gen_sub_CHf (R , Rd, Rr); +gen_sub_Vf (R , Rd, Rr); +gen_ZNSf (R ); + +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* This instruction performs a compare between two registers Rd and Rr and +* also takes into account the previous carry. None of the registers are +* changed. All conditional branches can be used after this instruction. +*/ +static bool trans_CPC (DisasContext *ctx,arg_CPC *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); +TCGv zero =tcg_const_i32 (0 ); + +/* op */ +tcg_gen_sub_tl (R , Rd, Rr);/* R = Rd - Rr - Cf */ +tcg_gen_sub_tl (R ,R , cpu_Cf); +tcg_gen_andi_tl (R ,R ,0xff );/* make it 8 bits */ + +gen_sub_CHf (R , Rd, Rr); +gen_sub_Vf (R , Rd, Rr); +gen_NSf (R ); + +/* +* Previous value remains unchanged when the result is zero; +* cleared otherwise. +*/ +tcg_gen_movcond_tl (TCG_COND_EQ , cpu_Zf,R , zero, cpu_Zf, zero); + +tcg_temp_free_i32 (zero); +tcg_temp_free_i32 (R ); + +return true; +} + + +/* +* This instruction performs a compare between register Rd and a constant. +* The register is not changed. All conditional branches can be used after this +* instruction. +*/ +static bool trans_CPI (DisasContext *ctx,arg_CPI *a) +{ +TCGv Rd = cpu_r[a->rd]; +int Imm = a->imm; +TCGv Rr =tcg_const_i32 (Imm); +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_sub_tl (R , Rd, Rr);/* R = Rd - Rr */ +tcg_gen_andi_tl (R ,R ,0xff );/* make it 8 bits */ + +gen_sub_CHf (R , Rd, Rr); +gen_sub_Vf (R , Rd, Rr); +gen_ZNSf (R ); + +tcg_temp_free_i32 (R ); +tcg_temp_free_i32 (Rr); + +return true; +} + + +/* +* This instruction tests a single bit in a register and skips the next +* instruction if the bit is cleared. +*/ +static bool trans_SBRC (DisasContext *ctx,arg_SBRC *a) +{ +TCGv Rr = cpu_r[a->rr]; + + ctx->skip_cond =TCG_COND_EQ ; + ctx->skip_var0 =tcg_temp_new (); + ctx->free_skip_var0 = true; + +tcg_gen_andi_tl (ctx->skip_var0, Rr,1 << a->bit); +return true; +} + + +/* +* This instruction tests a single bit in a register and skips the next +* instruction if the bit is set. +*/ +static bool trans_SBRS (DisasContext *ctx,arg_SBRS *a) +{ +TCGv Rr = cpu_r[a->rr]; + + ctx->skip_cond =TCG_COND_NE ; + ctx->skip_var0 =tcg_temp_new (); + ctx->free_skip_var0 = true; + +tcg_gen_andi_tl (ctx->skip_var0, Rr,1 << a->bit); +return true; +} + + +/* +* This instruction tests a single bit in an I/O Register and skips the +* next instruction if the bit is cleared. This instruction operates on the +* lower 32 I/O Registers -- addresses 0-31. +*/ +static bool trans_SBIC (DisasContext *ctx,arg_SBIC *a) +{ +TCGv temp =tcg_const_i32 (a->reg); + +gen_helper_inb (temp, cpu_env, temp); +tcg_gen_andi_tl (temp, temp,1 << a->bit); + ctx->skip_cond =TCG_COND_EQ ; + ctx->skip_var0 = temp; + ctx->free_skip_var0 = true; + +return true; +} + + +/* +* This instruction tests a single bit in an I/O Register and skips the +* next instruction if the bit is set. This instruction operates on the lower +* 32 I/O Registers -- addresses 0-31. +*/ +static bool trans_SBIS (DisasContext *ctx,arg_SBIS *a) +{ +TCGv temp =tcg_const_i32 (a->reg); + +gen_helper_inb (temp, cpu_env, temp); +tcg_gen_andi_tl (temp, temp,1 << a->bit); + ctx->skip_cond =TCG_COND_NE ; + ctx->skip_var0 = temp; + ctx->free_skip_var0 = true; + +return true; +} + + +/* +* Conditional relative branch. Tests a single bit in SREG and branches +* relatively to PC if the bit is cleared. This instruction branches relatively +* to PC in either direction (PC - 63 < = destination <= PC + 64). The +* parameter k is the offset from PC and is represented in two's complement +* form. +*/ +static bool trans_BRBC (DisasContext *ctx,arg_BRBC *a) +{ +TCGLabel *not_taken =gen_new_label (); + +TCGv var; + +switch (a->bit) { +case 0x00 : + var = cpu_Cf; +break ; +case 0x01 : + var = cpu_Zf; +break ; +case 0x02 : + var = cpu_Nf; +break ; +case 0x03 : + var = cpu_Vf; +break ; +case 0x04 : + var = cpu_Sf; +break ; +case 0x05 : + var = cpu_Hf; +break ; +case 0x06 : + var = cpu_Tf; +break ; +case 0x07 : + var = cpu_If; +break ; +default : +g_assert_not_reached (); + } + +tcg_gen_brcondi_i32 (TCG_COND_NE , var,0 , not_taken); +gen_goto_tb (ctx,0 , ctx->npc + a->imm); +gen_set_label (not_taken); + + ctx->bstate =DISAS_CHAIN ; +return true; +} + + +/* +* Conditional relative branch. Tests a single bit in SREG and branches +* relatively to PC if the bit is set. This instruction branches relatively to +* PC in either direction (PC - 63 < = destination <= PC + 64). The parameter k +* is the offset from PC and is represented in two's complement form. +*/ +static bool trans_BRBS (DisasContext *ctx,arg_BRBS *a) +{ +TCGLabel *not_taken =gen_new_label (); + +TCGv var; + +switch (a->bit) { +case 0x00 : + var = cpu_Cf; +break ; +case 0x01 : + var = cpu_Zf; +break ; +case 0x02 : + var = cpu_Nf; +break ; +case 0x03 : + var = cpu_Vf; +break ; +case 0x04 : + var = cpu_Sf; +break ; +case 0x05 : + var = cpu_Hf; +break ; +case 0x06 : + var = cpu_Tf; +break ; +case 0x07 : + var = cpu_If; +break ; +default : +g_assert_not_reached (); + } + +tcg_gen_brcondi_i32 (TCG_COND_EQ , var,0 , not_taken); +gen_goto_tb (ctx,0 , ctx->npc + a->imm); +gen_set_label (not_taken); + + ctx->bstate =DISAS_CHAIN ; +return true; +} +
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 2/17 target/avr: Add instruction helpers
In reply to Michael RolnikStubs for unimplemented instructions and helpers for instructions that need to interact with QEMU. SPM and WDR are unimplemented because they require emulation of complex peripherals.
The implementation of SLEEP is very limited due to the lack of peripherals to generate wake interrupts. Memory access instructions are implemented here because some address ranges actually refer to CPU registers.
--- target/avr/helper.h | 29 ++++ target/avr/helper.c | 354 ++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 383 insertions(+) create mode 100644 target/avr/helper.h create mode 100644 target/avr/helper.c
target/avr/helper.h new file +29 -0
@@ -0,0 +1,29 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +DEF_HELPER_1(wdr, void, env) +DEF_HELPER_1(debug, void, env) +DEF_HELPER_1(break, void, env) +DEF_HELPER_1(sleep, void, env) +DEF_HELPER_1(unsupported, void, env) +DEF_HELPER_3(outb, void, env, i32, i32) +DEF_HELPER_2(inb, tl, env, i32) +DEF_HELPER_3(fullwr, void, env, i32, i32) +DEF_HELPER_2 (fullrd, tl, env, i32)
target/avr/helper.c new file +354 -0
354 lines not on this page · raw message
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 13/17 target/avr: Register AVR support with the rest of QEMU
In reply to Michael RolnikAdd AVR related definitions into QEMU
--- qapi/machine.json | 3 ++- include/disas/dis-asm.h | 6 ++++++ include/sysemu/arch_init.h | 1 + arch_init.c | 2 ++ tests/machine-none-test.c | 1 + 5 files changed, 12 insertions(+), 1 deletion(-)
qapi/machine.json +2 -1
@@ -21,11 +21,12 @@ # is true even for "qemu-system-x86_64". # # ppcemb: dropped in 3.1 +# avr: since 5.0 # # Since: 3.0 ## { 'enum' : 'SysEmuTarget', - 'data' : [ 'aarch64', 'alpha', 'arm', 'cris', 'hppa', 'i386', 'lm32', + 'data' : [ 'aarch64', 'alpha', 'arm', 'avr', 'cris', 'hppa', 'i386', 'lm32', 'm68k', 'microblaze', 'microblazeel', 'mips', 'mips64', 'mips64el', 'mipsel', 'moxie', 'nios2', 'or1k', 'ppc', 'ppc64', 'riscv32', 'riscv64', 's390x', 'sh4',
include/disas/dis-asm.h +6 -0
@@ -211,6 +211,12 @@ enum bfd_architecture#define bfd_mach_m32r 0/* backwards compatibility */ bfd_arch_mn10200, /* Matsushita MN10200 */ bfd_arch_mn10300, /* Matsushita MN10300 */ + bfd_arch_avr, /* Atmel AVR microcontrollers. */ +#define bfd_mach_avr1 1 +#define bfd_mach_avr2 2 +#define bfd_mach_avr3 3 +#define bfd_mach_avr4 4 +#define bfd_mach_avr5 5 bfd_arch_cris, /* Axis CRIS */#define bfd_mach_cris_v0_v10 255#define bfd_mach_cris_v32 32
include/sysemu/arch_init.h +1 -0
@@ -24,6 +24,7 @@ enum {QEMU_ARCH_NIOS2 = (1 <<17 ),QEMU_ARCH_HPPA = (1 <<18 ),QEMU_ARCH_RISCV = (1 <<19 ), + QEMU_ARCH_AVR = (1 << 20), };extern const uint32_t arch_type;
arch_init.c +2 -0
@@ -89,6 +89,8 @@ int graphic_depth = 32;#define QEMU_ARCH QEMU_ARCH_UNICORE32#elif defined(TARGET_XTENSA )#define QEMU_ARCH QEMU_ARCH_XTENSA +#elif defined(TARGET_AVR ) +#define QEMU_ARCH QEMU_ARCH_AVR#endif const uint32_t arch_type =QEMU_ARCH ;
tests/machine-none-test.c +1 -0
@@ -27,6 +27,7 @@ static struct arch2cpu cpus_map[] = {/* tested targets list */ { "arm", "cortex-a15" }, { "aarch64", "cortex-a57" }, + { "avr", "avr6-avr-cpu" }, { "x86_64", "qemu64,apic-id=0" }, { "i386", "qemu32,apic-id=0" }, { "alpha", "ev67" },
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 7/17 target/avr: Add instruction translation - Bit and Bit-test Instructions
In reply to Michael RolnikThis includes:
- LSR, ROR
- ASR
- SWAP
- SBI, CBI
- BST, BLD
- BSET, BCLR
--- target/avr/translate.c | 1123 ++++++++++++++++++++++++++++++++++++++++ 1 file changed, 1123 insertions(+)
target/avr/translate.c +1123 -0
@@ -317,6 +317,15 @@ static void gen_goto_tb(DisasContext *ctx, int n, target_ulong dest) } +static void gen_rshift_ZNVSf (TCGv R ) +{ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ +tcg_gen_shri_tl (cpu_Nf,R ,7 );/* Nf = R(7) */ +tcg_gen_xor_tl (cpu_Vf, cpu_Nf, cpu_Cf); +tcg_gen_xor_tl (cpu_Sf, cpu_Nf, cpu_Vf);/* Sf = Nf ^ Vf */ +} + +/* * Adds two registers without the C Flag and places the result in the * destination register Rd.
@@ -1508,3 +1517,1117 @@ static bool trans_BRBS(DisasContext *ctx, arg_BRBS *a)return true; } + +/* +* This instruction makes a copy of one register into another. The source +* register Rr is left unchanged, while the destination register Rd is loaded +* with a copy of Rr. +*/ +static bool trans_MOV (DisasContext *ctx,arg_MOV *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; + +tcg_gen_mov_tl (Rd, Rr); + +return true; +} + + +/* +* This instruction makes a copy of one register pair into another register +* pair. The source register pair Rr+1:Rr is left unchanged, while the +* destination register pair Rd+1:Rd is loaded with a copy of Rr + 1:Rr. This +* instruction is not available in all devices. Refer to the device specific +* instruction set summary. +*/ +static bool trans_MOVW (DisasContext *ctx,arg_MOVW *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_MOVW )) { +return true; + } + +TCGv RdL = cpu_r[a->rd]; +TCGv RdH = cpu_r[a->rd +1 ]; +TCGv RrL = cpu_r[a->rr]; +TCGv RrH = cpu_r[a->rr +1 ]; + +tcg_gen_mov_tl (RdH, RrH); +tcg_gen_mov_tl (RdL, RrL); + +return true; +} + + +/* +* Loads an 8 bit constant directly to register 16 to 31. +*/ +static bool trans_LDI (DisasContext *ctx,arg_LDI *a) +{ +TCGv Rd = cpu_r[a->rd]; +int imm = a->imm; + +tcg_gen_movi_tl (Rd, imm); + +return true; +} + + +/* +* Loads one byte from the data space to a register. For parts with SRAM, +* the data space consists of the Register File, I/O memory and internal SRAM +* (and external SRAM if applicable). For parts without SRAM, the data space +* consists of the register file only. The EEPROM has a separate address space. +* A 16-bit address must be supplied. Memory access is limited to the current +* data segment of 64KB. The LDS instruction uses the RAMPD Register to access +* memory above 64KB. To access another data segment in devices with more than +* 64KB data space, the RAMPD in register in the I/O area has to be changed. +* This instruction is not available in all devices. Refer to the device +* specific instruction set summary. +*/ +static bool trans_LDS (DisasContext *ctx,arg_LDS *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =tcg_temp_new_i32 (); +TCGv H = cpu_rampD; + a->imm =next_word (ctx); + +tcg_gen_mov_tl (addr,H );/* addr = H:M:L */ +tcg_gen_shli_tl (addr, addr,16 ); +tcg_gen_ori_tl (addr, addr, a->imm); + +gen_data_load (ctx, Rd, addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +/* +* Loads one byte indirect from the data space to a register. For parts +* with SRAM, the data space consists of the Register File, I/O memory and +* internal SRAM (and external SRAM if applicable). For parts without SRAM, the +* data space consists of the Register File only. In some parts the Flash +* Memory has been mapped to the data space and can be read using this command. +* The EEPROM has a separate address space. The data location is pointed to by +* the X (16 bits) Pointer Register in the Register File. Memory access is +* limited to the current data segment of 64KB. To access another data segment +* in devices with more than 64KB data space, the RAMPX in register in the I/O +* area has to be changed. The X-pointer Register can either be left unchanged +* by the operation, or it can be post-incremented or predecremented. These +* features are especially suited for accessing arrays, tables, and Stack +* Pointer usage of the X-pointer Register. Note that only the low byte of the +* X-pointer is updated in devices with no more than 256 bytes data space. For +* such devices, the high byte of the pointer is not used by this instruction +* and can be used for other purposes. The RAMPX Register in the I/O area is +* updated in parts with more than 64KB data space or more than 64KB Program +* memory, and the increment/decrement is added to the entire 24-bit address on +* such devices. Not all variants of this instruction is available in all +* devices. Refer to the device specific instruction set summary. In the +* Reduced Core tinyAVR the LD instruction can be used to achieve the same +* operation as LPM since the program memory is mapped to the data memory +* space. +*/ +static bool trans_LDX1 (DisasContext *ctx,arg_LDX1 *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_xaddr (); + +gen_data_load (ctx, Rd, addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_LDX2 (DisasContext *ctx,arg_LDX2 *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_xaddr (); + +gen_data_load (ctx, Rd, addr); +tcg_gen_addi_tl (addr, addr,1 );/* addr = addr + 1 */ + +gen_set_xaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_LDX3 (DisasContext *ctx,arg_LDX3 *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_xaddr (); + +tcg_gen_subi_tl (addr, addr,1 );/* addr = addr - 1 */ +gen_data_load (ctx, Rd, addr); +gen_set_xaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_STX1 (DisasContext *ctx,arg_STX1 *a) +{ +TCGv Rd = cpu_r[a->rr]; +TCGv addr =gen_get_xaddr (); + +gen_data_store (ctx, Rd, addr); + +tcg_temp_free_i32 (addr); + +return true; +} + +static bool trans_STX2 (DisasContext *ctx,arg_STX2 *a) +{ +TCGv Rd = cpu_r[a->rr]; +TCGv addr =gen_get_xaddr (); + +gen_data_store (ctx, Rd, addr); +tcg_gen_addi_tl (addr, addr,1 );/* addr = addr + 1 */ +gen_set_xaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + +static bool trans_STX3 (DisasContext *ctx,arg_STX3 *a) +{ +TCGv Rd = cpu_r[a->rr]; +TCGv addr =gen_get_xaddr (); + +tcg_gen_subi_tl (addr, addr,1 );/* addr = addr - 1 */ +gen_data_store (ctx, Rd, addr); +gen_set_xaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +/* +* Loads one byte indirect with or without displacement from the data space +* to a register. For parts with SRAM, the data space consists of the Register +* File, I/O memory and internal SRAM (and external SRAM if applicable). For +* parts without SRAM, the data space consists of the Register File only. In +* some parts the Flash Memory has been mapped to the data space and can be +* read using this command. The EEPROM has a separate address space. The data +* location is pointed to by the Y (16 bits) Pointer Register in the Register +* File. Memory access is limited to the current data segment of 64KB. To +* access another data segment in devices with more than 64KB data space, the +* RAMPY in register in the I/O area has to be changed. The Y-pointer Register +* can either be left unchanged by the operation, or it can be post-incremented +* or predecremented. These features are especially suited for accessing +* arrays, tables, and Stack Pointer usage of the Y-pointer Register. Note that +* only the low byte of the Y-pointer is updated in devices with no more than +* 256 bytes data space. For such devices, the high byte of the pointer is not +* used by this instruction and can be used for other purposes. The RAMPY +* Register in the I/O area is updated in parts with more than 64KB data space +* or more than 64KB Program memory, and the increment/decrement/displacement +* is added to the entire 24-bit address on such devices. Not all variants of +* this instruction is available in all devices. Refer to the device specific +* instruction set summary. In the Reduced Core tinyAVR the LD instruction can +* be used to achieve the same operation as LPM since the program memory is +* mapped to the data memory space. +*/ +static bool trans_LDY2 (DisasContext *ctx,arg_LDY2 *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_yaddr (); + +gen_data_load (ctx, Rd, addr); +tcg_gen_addi_tl (addr, addr,1 );/* addr = addr + 1 */ + +gen_set_yaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_LDY3 (DisasContext *ctx,arg_LDY3 *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_yaddr (); + +tcg_gen_subi_tl (addr, addr,1 );/* addr = addr - 1 */ +gen_data_load (ctx, Rd, addr); +gen_set_yaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_LDDY (DisasContext *ctx,arg_LDDY *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_yaddr (); + +tcg_gen_addi_tl (addr, addr, a->imm);/* addr = addr + q */ +gen_data_load (ctx, Rd, addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_STY2 (DisasContext *ctx,arg_STY2 *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_yaddr (); + +gen_data_store (ctx, Rd, addr); +tcg_gen_addi_tl (addr, addr,1 );/* addr = addr + 1 */ +gen_set_yaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_STY3 (DisasContext *ctx,arg_STY3 *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_yaddr (); + +tcg_gen_subi_tl (addr, addr,1 );/* addr = addr - 1 */ +gen_data_store (ctx, Rd, addr); +gen_set_yaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_STDY (DisasContext *ctx,arg_STDY *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_yaddr (); + +tcg_gen_addi_tl (addr, addr, a->imm);/* addr = addr + q */ +gen_data_store (ctx, Rd, addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +/* +* Loads one byte indirect with or without displacement from the data space +* to a register. For parts with SRAM, the data space consists of the Register +* File, I/O memory and internal SRAM (and external SRAM if applicable). For +* parts without SRAM, the data space consists of the Register File only. In +* some parts the Flash Memory has been mapped to the data space and can be +* read using this command. The EEPROM has a separate address space. The data +* location is pointed to by the Z (16 bits) Pointer Register in the Register +* File. Memory access is limited to the current data segment of 64KB. To +* access another data segment in devices with more than 64KB data space, the +* RAMPZ in register in the I/O area has to be changed. The Z-pointer Register +* can either be left unchanged by the operation, or it can be post-incremented +* or predecremented. These features are especially suited for Stack Pointer +* usage of the Z-pointer Register, however because the Z-pointer Register can +* be used for indirect subroutine calls, indirect jumps and table lookup, it +* is often more convenient to use the X or Y-pointer as a dedicated Stack +* Pointer. Note that only the low byte of the Z-pointer is updated in devices +* with no more than 256 bytes data space. For such devices, the high byte of +* the pointer is not used by this instruction and can be used for other +* purposes. The RAMPZ Register in the I/O area is updated in parts with more +* than 64KB data space or more than 64KB Program memory, and the +* increment/decrement/displacement is added to the entire 24-bit address on +* such devices. Not all variants of this instruction is available in all +* devices. Refer to the device specific instruction set summary. In the +* Reduced Core tinyAVR the LD instruction can be used to achieve the same +* operation as LPM since the program memory is mapped to the data memory +* space. For using the Z-pointer for table lookup in Program memory see the +* LPM and ELPM instructions. +*/ +static bool trans_LDZ2 (DisasContext *ctx,arg_LDZ2 *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_zaddr (); + +gen_data_load (ctx, Rd, addr); +tcg_gen_addi_tl (addr, addr,1 );/* addr = addr + 1 */ + +gen_set_zaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_LDZ3 (DisasContext *ctx,arg_LDZ3 *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_zaddr (); + +tcg_gen_subi_tl (addr, addr,1 );/* addr = addr - 1 */ +gen_data_load (ctx, Rd, addr); + +gen_set_zaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_LDDZ (DisasContext *ctx,arg_LDDZ *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_zaddr (); + +tcg_gen_addi_tl (addr, addr, a->imm);/* addr = addr + q */ +gen_data_load (ctx, Rd, addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_STZ2 (DisasContext *ctx,arg_STZ2 *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_zaddr (); + +gen_data_store (ctx, Rd, addr); +tcg_gen_addi_tl (addr, addr,1 );/* addr = addr + 1 */ + +gen_set_zaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_STZ3 (DisasContext *ctx,arg_STZ3 *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_zaddr (); + +tcg_gen_subi_tl (addr, addr,1 );/* addr = addr - 1 */ +gen_data_store (ctx, Rd, addr); + +gen_set_zaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_STDZ (DisasContext *ctx,arg_STDZ *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_zaddr (); + +tcg_gen_addi_tl (addr, addr, a->imm);/* addr = addr + q */ +gen_data_store (ctx, Rd, addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + + +/* +* Stores one byte from a Register to the data space. For parts with SRAM, +* the data space consists of the Register File, I/O memory and internal SRAM +* (and external SRAM if applicable). For parts without SRAM, the data space +* consists of the Register File only. The EEPROM has a separate address space. +* A 16-bit address must be supplied. Memory access is limited to the current +* data segment of 64KB. The STS instruction uses the RAMPD Register to access +* memory above 64KB. To access another data segment in devices with more than +* 64KB data space, the RAMPD in register in the I/O area has to be changed. +* This instruction is not available in all devices. Refer to the device +* specific instruction set summary. +*/ +static bool trans_STS (DisasContext *ctx,arg_STS *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv addr =tcg_temp_new_i32 (); +TCGv H = cpu_rampD; + a->imm =next_word (ctx); + +tcg_gen_mov_tl (addr,H );/* addr = H:M:L */ +tcg_gen_shli_tl (addr, addr,16 ); +tcg_gen_ori_tl (addr, addr, a->imm); + +gen_data_store (ctx, Rd, addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +/* +* Loads one byte pointed to by the Z-register into the destination +* register Rd. This instruction features a 100% space effective constant +* initialization or constant data fetch. The Program memory is organized in +* 16-bit words while the Z-pointer is a byte address. Thus, the least +* significant bit of the Z-pointer selects either low byte (ZLSB = 0) or high +* byte (ZLSB = 1). This instruction can address the first 64KB (32K words) of +* Program memory. The Zpointer Register can either be left unchanged by the +* operation, or it can be incremented. The incrementation does not apply to +* the RAMPZ Register. Devices with Self-Programming capability can use the +* LPM instruction to read the Fuse and Lock bit values. Refer to the device +* documentation for a detailed description. The LPM instruction is not +* available in all devices. Refer to the device specific instruction set +* summary +*/ +static bool trans_LPM1 (DisasContext *ctx,arg_LPM1 *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_LPM )) { +return true; + } + +TCGv Rd = cpu_r[0 ]; +TCGv addr =tcg_temp_new_i32 (); +TCGv H = cpu_r[31 ]; +TCGv L = cpu_r[30 ]; + +tcg_gen_shli_tl (addr,H ,8 );/* addr = H:L */ +tcg_gen_or_tl (addr, addr,L ); + +tcg_gen_qemu_ld8u (Rd, addr,MMU_CODE_IDX );/* Rd = mem[addr] */ + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_LPM2 (DisasContext *ctx,arg_LPM2 *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_LPM )) { +return true; + } + +TCGv Rd = cpu_r[a->rd]; +TCGv addr =tcg_temp_new_i32 (); +TCGv H = cpu_r[31 ]; +TCGv L = cpu_r[30 ]; + +tcg_gen_shli_tl (addr,H ,8 );/* addr = H:L */ +tcg_gen_or_tl (addr, addr,L ); + +tcg_gen_qemu_ld8u (Rd, addr,MMU_CODE_IDX );/* Rd = mem[addr] */ + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_LPMX (DisasContext *ctx,arg_LPMX *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_LPMX )) { +return true; + } + +TCGv Rd = cpu_r[a->rd]; +TCGv addr =tcg_temp_new_i32 (); +TCGv H = cpu_r[31 ]; +TCGv L = cpu_r[30 ]; + +tcg_gen_shli_tl (addr,H ,8 );/* addr = H:L */ +tcg_gen_or_tl (addr, addr,L ); + +tcg_gen_qemu_ld8u (Rd, addr,MMU_CODE_IDX );/* Rd = mem[addr] */ + +tcg_gen_addi_tl (addr, addr,1 );/* addr = addr + 1 */ + +tcg_gen_andi_tl (L , addr,0xff ); + +tcg_gen_shri_tl (addr, addr,8 ); +tcg_gen_andi_tl (H , addr,0xff ); + +tcg_temp_free_i32 (addr); + +return true; +} + + +/* +* Loads one byte pointed to by the Z-register and the RAMPZ Register in +* the I/O space, and places this byte in the destination register Rd. This +* instruction features a 100% space effective constant initialization or +* constant data fetch. The Program memory is organized in 16-bit words while +* the Z-pointer is a byte address. Thus, the least significant bit of the +* Z-pointer selects either low byte (ZLSB = 0) or high byte (ZLSB = 1). This +* instruction can address the entire Program memory space. The Z-pointer +* Register can either be left unchanged by the operation, or it can be +* incremented. The incrementation applies to the entire 24-bit concatenation +* of the RAMPZ and Z-pointer Registers. Devices with Self-Programming +* capability can use the ELPM instruction to read the Fuse and Lock bit value. +* Refer to the device documentation for a detailed description. This +* instruction is not available in all devices. Refer to the device specific +* instruction set summary. +*/ +static bool trans_ELPM1 (DisasContext *ctx,arg_ELPM1 *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_ELPM )) { +return true; + } + +TCGv Rd = cpu_r[0 ]; +TCGv addr =gen_get_zaddr (); + +tcg_gen_qemu_ld8u (Rd, addr,MMU_CODE_IDX );/* Rd = mem[addr] */ + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_ELPM2 (DisasContext *ctx,arg_ELPM2 *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_ELPM )) { +return true; + } + +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_zaddr (); + +tcg_gen_qemu_ld8u (Rd, addr,MMU_CODE_IDX );/* Rd = mem[addr] */ + +tcg_temp_free_i32 (addr); + +return true; +} + + +static bool trans_ELPMX (DisasContext *ctx,arg_ELPMX *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_ELPMX )) { +return true; + } + +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_zaddr (); + +tcg_gen_qemu_ld8u (Rd, addr,MMU_CODE_IDX );/* Rd = mem[addr] */ + +tcg_gen_addi_tl (addr, addr,1 );/* addr = addr + 1 */ + +gen_set_zaddr (addr); + +tcg_temp_free_i32 (addr); + +return true; +} + + +/* +* SPM can be used to erase a page in the Program memory, to write a page +* in the Program memory (that is already erased), and to set Boot Loader Lock +* bits. In some devices, the Program memory can be written one word at a time, +* in other devices an entire page can be programmed simultaneously after first +* filling a temporary page buffer. In all cases, the Program memory must be +* erased one page at a time. When erasing the Program memory, the RAMPZ and +* Z-register are used as page address. When writing the Program memory, the +* RAMPZ and Z-register are used as page or word address, and the R1:R0 +* register pair is used as data(1). When setting the Boot Loader Lock bits, +* the R1:R0 register pair is used as data. Refer to the device documentation +* for detailed description of SPM usage. This instruction can address the +* entire Program memory. The SPM instruction is not available in all devices. +* Refer to the device specific instruction set summary. Note: 1. R1 +* determines the instruction high byte, and R0 determines the instruction low +* byte. +*/ +static bool trans_SPM (DisasContext *ctx,arg_SPM *a) +{ +/* TODO */ +if (!avr_have_feature (ctx,AVR_FEATURE_SPM )) { +return true; + } + +return true; +} + + +static bool trans_SPMX (DisasContext *ctx,arg_SPMX *a) +{ +/* TODO */ +if (!avr_have_feature (ctx,AVR_FEATURE_SPMX )) { +return true; + } + +return true; +} + + +/* +* Loads data from the I/O Space (Ports, Timers, Configuration Registers, +* etc.) into register Rd in the Register File. +*/ +static bool trans_IN (DisasContext *ctx,arg_IN *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv port =tcg_const_i32 (a->imm); + +gen_helper_inb (Rd, cpu_env, port); + +tcg_temp_free_i32 (port); + +return true; +} + + +/* +* Stores data from register Rr in the Register File to I/O Space (Ports, +* Timers, Configuration Registers, etc.). +*/ +static bool trans_OUT (DisasContext *ctx,arg_OUT *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv port =tcg_const_i32 (a->imm); + +gen_helper_outb (cpu_env, port, Rd); + +tcg_temp_free_i32 (port); + +return true; +} + + +/* +* This instruction stores the contents of register Rr on the STACK. The +* Stack Pointer is post-decremented by 1 after the PUSH. This instruction is +* not available in all devices. Refer to the device specific instruction set +* summary. +*/ +static bool trans_PUSH (DisasContext *ctx,arg_PUSH *a) +{ +TCGv Rd = cpu_r[a->rd]; + +gen_data_store (ctx, Rd, cpu_sp); +tcg_gen_subi_tl (cpu_sp, cpu_sp,1 ); + +return true; +} + + +/* +* This instruction loads register Rd with a byte from the STACK. The Stack +* Pointer is pre-incremented by 1 before the POP. This instruction is not +* available in all devices. Refer to the device specific instruction set +* summary. +*/ +static bool trans_POP (DisasContext *ctx,arg_POP *a) +{ +/* +* Using a temp to work around some strange behaviour: +* tcg_gen_addi_tl(cpu_sp, cpu_sp, 1); +* gen_data_load(ctx, Rd, cpu_sp); +* seems to cause the add to happen twice. +* This doesn't happen if either the add or the load is removed. +*/ +TCGv t1 =tcg_temp_new_i32 (); +TCGv Rd = cpu_r[a->rd]; + +tcg_gen_addi_tl (t1, cpu_sp,1 ); +gen_data_load (ctx, Rd, t1); +tcg_gen_mov_tl (cpu_sp, t1); + +return true; +} + + +/* +* Exchanges one byte indirect between register and data space. The data +* location is pointed to by the Z (16 bits) Pointer Register in the Register +* File. Memory access is limited to the current data segment of 64KB. To +* access another data segment in devices with more than 64KB data space, the +* RAMPZ in register in the I/O area has to be changed. The Z-pointer Register +* is left unchanged by the operation. This instruction is especially suited +* for writing/reading status bits stored in SRAM. +*/ +static bool trans_XCH (DisasContext *ctx,arg_XCH *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_RMW )) { +return true; + } + +TCGv Rd = cpu_r[a->rd]; +TCGv t0 =tcg_temp_new_i32 (); +TCGv addr =gen_get_zaddr (); + +gen_data_load (ctx, t0, addr); +gen_data_store (ctx, Rd, addr); +tcg_gen_mov_tl (Rd, t0); + +tcg_temp_free_i32 (t0); +tcg_temp_free_i32 (addr); + +return true; +} + + +/* +* Load one byte indirect from data space to register and set bits in data +* space specified by the register. The instruction can only be used towards +* internal SRAM. The data location is pointed to by the Z (16 bits) Pointer +* Register in the Register File. Memory access is limited to the current data +* segment of 64KB. To access another data segment in devices with more than +* 64KB data space, the RAMPZ in register in the I/O area has to be changed. +* The Z-pointer Register is left unchanged by the operation. This instruction +* is especially suited for setting status bits stored in SRAM. +*/ +static bool trans_LAS (DisasContext *ctx,arg_LAS *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_RMW )) { +return true; + } + +TCGv Rr = cpu_r[a->rd]; +TCGv addr =gen_get_zaddr (); +TCGv t0 =tcg_temp_new_i32 (); +TCGv t1 =tcg_temp_new_i32 (); + +gen_data_load (ctx, t0, addr);/* t0 = mem[addr] */ +tcg_gen_or_tl (t1, t0, Rr); + +tcg_gen_mov_tl (Rr, t0);/* Rr = t0 */ +gen_data_store (ctx, t1, addr);/* mem[addr] = t1 */ + +tcg_temp_free_i32 (t1); +tcg_temp_free_i32 (t0); +tcg_temp_free_i32 (addr); + +return true; +} + + +/* +* Load one byte indirect from data space to register and stores and clear +* the bits in data space specified by the register. The instruction can +* only be used towards internal SRAM. The data location is pointed to by +* the Z (16 bits) Pointer Register in the Register File. Memory access is +* limited to the current data segment of 64KB. To access another data +* segment in devices with more than 64KB data space, the RAMPZ in register +* in the I/O area has to be changed. The Z-pointer Register is left +* unchanged by the operation. This instruction is especially suited for +* clearing status bits stored in SRAM. +*/ +static bool trans_LAC (DisasContext *ctx,arg_LAC *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_RMW )) { +return true; + } + +TCGv Rr = cpu_r[a->rd]; +TCGv addr =gen_get_zaddr (); +TCGv t0 =tcg_temp_new_i32 (); +TCGv t1 =tcg_temp_new_i32 (); + +gen_data_load (ctx, t0, addr);/* t0 = mem[addr] */ +/* t1 = t0 & (0xff - Rr) = t0 and ~Rr */ +tcg_gen_andc_tl (t1, t0, Rr); + +tcg_gen_mov_tl (Rr, t0);/* Rr = t0 */ +gen_data_store (ctx, t1, addr);/* mem[addr] = t1 */ + +tcg_temp_free_i32 (t1); +tcg_temp_free_i32 (t0); +tcg_temp_free_i32 (addr); + +return true; +} + + +/* +* Load one byte indirect from data space to register and toggles bits in +* the data space specified by the register. The instruction can only be used +* towards SRAM. The data location is pointed to by the Z (16 bits) Pointer +* Register in the Register File. Memory access is limited to the current data +* segment of 64KB. To access another data segment in devices with more than +* 64KB data space, the RAMPZ in register in the I/O area has to be changed. +* The Z-pointer Register is left unchanged by the operation. This instruction +* is especially suited for changing status bits stored in SRAM. +*/ +static bool trans_LAT (DisasContext *ctx,arg_LAT *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_RMW )) { +return true; + } + +TCGv Rd = cpu_r[a->rd]; +TCGv addr =gen_get_zaddr (); +TCGv t0 =tcg_temp_new_i32 (); +TCGv t1 =tcg_temp_new_i32 (); + +gen_data_load (ctx, t0, addr);/* t0 = mem[addr] */ +tcg_gen_xor_tl (t1, t0, Rd); + +tcg_gen_mov_tl (Rd, t0);/* Rd = t0 */ +gen_data_store (ctx, t1, addr);/* mem[addr] = t1 */ + +tcg_temp_free_i32 (t1); +tcg_temp_free_i32 (t0); +tcg_temp_free_i32 (addr); + +return true; +} + + +/* +* Shifts all bits in Rd one place to the right. Bit 7 is cleared. Bit 0 is +* loaded into the C Flag of the SREG. This operation effectively divides an +* unsigned value by two. The C Flag can be used to round the result. +*/ +static bool trans_LSR (DisasContext *ctx,arg_LSR *a) +{ +TCGv Rd = cpu_r[a->rd]; + +tcg_gen_andi_tl (cpu_Cf, Rd,1 ); + +tcg_gen_shri_tl (Rd, Rd,1 ); + +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf, Rd,0 );/* Zf = Rd == 0 */ +tcg_gen_movi_tl (cpu_Nf,0 ); +tcg_gen_mov_tl (cpu_Vf, cpu_Cf); +tcg_gen_mov_tl (cpu_Sf, cpu_Vf); + +return true; +} + + +/* +* Shifts all bits in Rd one place to the right. The C Flag is shifted into +* bit 7 of Rd. Bit 0 is shifted into the C Flag. This operation, combined +* with ASR, effectively divides multi-byte signed values by two. Combined with +* LSR it effectively divides multi-byte unsigned values by two. The Carry Flag +* can be used to round the result. +*/ +static bool trans_ROR (DisasContext *ctx,arg_ROR *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv t0 =tcg_temp_new_i32 (); + +tcg_gen_shli_tl (t0, cpu_Cf,7 ); +tcg_gen_andi_tl (cpu_Cf, Rd,1 ); +tcg_gen_shri_tl (Rd, Rd,1 ); +tcg_gen_or_tl (Rd, Rd, t0); + +gen_rshift_ZNVSf (Rd); + +tcg_temp_free_i32 (t0); + +return true; +} + + +/* +* Shifts all bits in Rd one place to the right. Bit 7 is held constant. Bit 0 +* is loaded into the C Flag of the SREG. This operation effectively divides a +* signed value by two without changing its sign. The Carry Flag can be used to +* round the result. +*/ +static bool trans_ASR (DisasContext *ctx,arg_ASR *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv t0 =tcg_temp_new_i32 (); + +/* Cf */ +tcg_gen_andi_tl (cpu_Cf, Rd,1 );/* Cf = Rd(0) */ + +/* op */ +tcg_gen_andi_tl (t0, Rd,0x80 );/* Rd = (Rd & 0x80) | (Rd >> 1) */ +tcg_gen_shri_tl (Rd, Rd,1 ); +tcg_gen_or_tl (Rd, Rd, t0); + +gen_rshift_ZNVSf (Rd); + +tcg_temp_free_i32 (t0); + +return true; +} + + +/* +* Swaps high and low nibbles in a register. +*/ +static bool trans_SWAP (DisasContext *ctx,arg_SWAP *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv t0 =tcg_temp_new_i32 (); +TCGv t1 =tcg_temp_new_i32 (); + +tcg_gen_andi_tl (t0, Rd,0x0f ); +tcg_gen_shli_tl (t0, t0,4 ); +tcg_gen_andi_tl (t1, Rd,0xf0 ); +tcg_gen_shri_tl (t1, t1,4 ); +tcg_gen_or_tl (Rd, t0, t1); + +tcg_temp_free_i32 (t1); +tcg_temp_free_i32 (t0); + +return true; +} + + +/* +* Sets a specified bit in an I/O Register. This instruction operates on +* the lower 32 I/O Registers -- addresses 0-31. +*/ +static bool trans_SBI (DisasContext *ctx,arg_SBI *a) +{ +TCGv data =tcg_temp_new_i32 (); +TCGv port =tcg_const_i32 (a->reg); + +gen_helper_inb (data, cpu_env, port); +tcg_gen_ori_tl (data, data,1 << a->bit); +gen_helper_outb (cpu_env, port, data); + +tcg_temp_free_i32 (port); +tcg_temp_free_i32 (data); + +return true; +} + + +/* +* Clears a specified bit in an I/O Register. This instruction operates on +* the lower 32 I/O Registers -- addresses 0-31. +*/ +static bool trans_CBI (DisasContext *ctx,arg_CBI *a) +{ +TCGv data =tcg_temp_new_i32 (); +TCGv port =tcg_const_i32 (a->reg); + +gen_helper_inb (data, cpu_env, port); +tcg_gen_andi_tl (data, data, ~(1 << a->bit)); +gen_helper_outb (cpu_env, port, data); + +tcg_temp_free_i32 (data); +tcg_temp_free_i32 (port); + +return true; +} + + +/* +* Stores bit b from Rd to the T Flag in SREG (Status Register). +*/ +static bool trans_BST (DisasContext *ctx,arg_BST *a) +{ +TCGv Rd = cpu_r[a->rd]; + +tcg_gen_andi_tl (cpu_Tf, Rd,1 << a->bit); +tcg_gen_shri_tl (cpu_Tf, cpu_Tf, a->bit); + +return true; +} + + +/* +* Copies the T Flag in the SREG (Status Register) to bit b in register Rd. +*/ +static bool trans_BLD (DisasContext *ctx,arg_BLD *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv t1 =tcg_temp_new_i32 (); + +tcg_gen_andi_tl (Rd, Rd, ~(1u << a->bit));/* clear bit */ +tcg_gen_shli_tl (t1, cpu_Tf, a->bit);/* create mask */ +tcg_gen_or_tl (Rd, Rd, t1); + +tcg_temp_free_i32 (t1); + +return true; +} + + +/* +* Sets a single Flag or bit in SREG. +*/ +static bool trans_BSET (DisasContext *ctx,arg_BSET *a) +{ +switch (a->bit) { +case 0x00 : +tcg_gen_movi_tl (cpu_Cf,0x01 ); +break ; +case 0x01 : +tcg_gen_movi_tl (cpu_Zf,0x01 ); +break ; +case 0x02 : +tcg_gen_movi_tl (cpu_Nf,0x01 ); +break ; +case 0x03 : +tcg_gen_movi_tl (cpu_Vf,0x01 ); +break ; +case 0x04 : +tcg_gen_movi_tl (cpu_Sf,0x01 ); +break ; +case 0x05 : +tcg_gen_movi_tl (cpu_Hf,0x01 ); +break ; +case 0x06 : +tcg_gen_movi_tl (cpu_Tf,0x01 ); +break ; +case 0x07 : +tcg_gen_movi_tl (cpu_If,0x01 ); +break ; + } + +return true; +} + + +/* +* Clears a single Flag in SREG. +*/ +static bool trans_BCLR (DisasContext *ctx,arg_BCLR *a) +{ +switch (a->bit) { +case 0x00 : +tcg_gen_movi_tl (cpu_Cf,0x00 ); +break ; +case 0x01 : +tcg_gen_movi_tl (cpu_Zf,0x00 ); +break ; +case 0x02 : +tcg_gen_movi_tl (cpu_Nf,0x00 ); +break ; +case 0x03 : +tcg_gen_movi_tl (cpu_Vf,0x00 ); +break ; +case 0x04 : +tcg_gen_movi_tl (cpu_Sf,0x00 ); +break ; +case 0x05 : +tcg_gen_movi_tl (cpu_Hf,0x00 ); +break ; +case 0x06 : +tcg_gen_movi_tl (cpu_Tf,0x00 ); +break ; +case 0x07 : +tcg_gen_movi_tl (cpu_If,0x00 ); +break ; + } + +return true; +}
Signature · 1 line
2.17.2 (Apple Git-113)
v37 cover QEMU AVR 8 bit cores
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fully tested yet. However I was able to execute simple code with functions. e.g fibonacci calculation. This series of patches include a non real, sample board.
No fuses support yet. PC is set to 0 at reset.
Following are examples of possible usages, assuming program.elf is compiled for AVR cpu
1. Continious non interrupted execution
run qemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnet window run qemu-system-avr -kernel program.elf -serial tcp::5678,server,nowait -nographic run telent localhost 5678
3. Continious non interrupted execution with serial output into stdout
run qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
run qemu-system-avr -kernel program.elf -s -S
run avr-gdb program.elf and then within GDB shell target remote :1234
5. Print out executed instructions
run qemu-system-avr -kernel program.elf -d in_asm
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding or translation
2. CPU features which allow define the following 8bit AVR cores
avr1
avr2 avr25
avr3 avr31 avr35
avr4
avr5 avr51
avr6
xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows to execute simple code
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features together
changes since v3
1. rampD/X/Y/Z registers are encoded as 0x00ff0000 (instead of 0x000000ff) for faster address manipulaton
2. ffs changed to ctz32
3. duplicate code removed at avr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7. tcg_const_local_i32 -> tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of 0xff - r at COM
11.fixing MULS/MULSU not to modify inputs' content
12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCH
changes since v4
1. target is now defined as big endian in order to optimize push_ret/pop_ret
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifying gen_goto_tb as there is no real paging
5. env->pc -> env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removing avr_set_feature
10. SPL/SPH set bug fix
11. switching stb_phys to cpu_stb_data
12. cleaning up avr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)
changes since v5
1. BLD bug fix
2. decoder generator is added
chages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_cpu_gdb_write_register
2. configure the target as little endian because otherwise GDB does not work
3. fixing and testing gen_push_ret/gen_pop_ret
changes since v7
1. folding back v6
2. logging at helper_outb and helper_inb are done for non supported yet registers only
3. MAINTAINERS updated
changes since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built for all
2. making linux compilable
3. testing on
a. Mac, Apple LLVM version 7.0.0
b. Ubuntu 12.04, gcc 4.9.2
c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memory
changes since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed to any device
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7. memory_region_allocate_system_memory is used to create RAM
8. now there are helper_fullrd & helper_fullwr when LD/ST try to access registers
changes since v10
1. movig back fullwr & fullrd into the commit where outb and inb were introduced
2. changing tlb_fill function signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. using tb->flags to generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objs
changes since v11
1. updating to v2.7.0-rc
2. removing assignment to env->fullacc from gen_intermediate_code
changes since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removing target-avr/machine.h file
4. VMSTATE_SINGLE_TEST -> VMSTATE_SINGLE
5. comment spelling
6. removing hw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixing avr_cpu_gdb_write_register & avr_cpu_gdb_read_register functions 12.any cpu is removed
12.feature bits are not saved into vm state
changes since v13
1. rebasing to v2.7.0-rc1
changes since v14
1. I made self review with git gui tool. (I did not know such a thing exists)
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chris
changes since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI as read(helper_inb), modify, write(helper_outb)
3. implementing CBIC/SBIC as read(helper_inb), check, branch
4. adding missing tcg_temp_free_i32 for tcg_const_i32
changes since v16
1. removing EXT IO registers knoledge from CPU. These registers are accessible by LD/ST only. CPU has no interest in them
changes since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:
1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation of translate.c.
I retained the two separate files that could be regenerated from the included cpugen program, but merged in translate-insn.c. Not that it matters, but the code generated is about 3k smaller.
changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18
changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fill
changes since v20
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for now
changes since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixes a couple of bugs wrt single stepping by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Add gdb-xml/avr-cpu.xml
changes since v22
1. Rebase
2. Split long comment
changes since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list function
changes since v24
1. remove AVR_CPU_TYPE_NAME macro
changes since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1
changes since v26
1. add avocado acceptence test
2. add boot serial test
changes since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/size
changes since v28
1. rebase
2. fix includes & build
changes since v29
1. fixing ownership
2. using 'since' instread of 'added in'
changes since v30
1. rebase
changes since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instruction translation - CPU main translation funcions' commit
changes since v32
1. modify cpu_get_sreg to treat sreg C as other flags, except sreg Z
changes since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logic
changes since v34
1. rebase
changes since v35
1. rebase
2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c
3. split "target/avr: Register AVR support with the rest of QEMU" into three patches
1. "target/avr: Register AVR support with the rest of QEMU"
2. "target/avr: Update build system"
3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"
changes since v36
1. rebase
2. tename
1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS
2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS
3. to_A -> to_regs_16_31_by_one
4. to_B -> to_regs_16_23_by_one
5. to_C -> to_regs_24_30_by_two
6. to_D -> to_regs_00_30_by_two
3. add missing licences
4. add usage example (see above)
5. ass Sarah Harris <[email protected]> as a reviewer to MAINTAINERS
7. use git commit sha1 instead of `master`` in avocado testMichael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: Update MAINTAINERS file
Sarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer peripherals
configure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 8/17 target/avr: Add instruction translation - MCU Control Instructions
This includes:
- BREAK
- NOP
- SLEEP
- WDR
--- target/avr/translate.c | 174 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 174 insertions(+)
target/avr/translate.c +174 -0
@@ -381,6 +381,115 @@ static bool trans_ADC(DisasContext *ctx, arg_ADC *a) } +static void gen_jmp_ez (DisasContext *ctx) +{ +tcg_gen_deposit_tl (cpu_pc, cpu_r[30 ], cpu_r[31 ],8 ,8 ); +tcg_gen_or_tl (cpu_pc, cpu_pc, cpu_eind); + ctx->bstate =DISAS_LOOKUP ; +} + + +static void gen_jmp_z (DisasContext *ctx) +{ +tcg_gen_deposit_tl (cpu_pc, cpu_r[30 ], cpu_r[31 ],8 ,8 ); + ctx->bstate =DISAS_LOOKUP ; +} + + +/* +* in the gen_set_addr & gen_get_addr functions +* H assumed to be in 0x00ff0000 format +* M assumed to be in 0x000000ff format +* L assumed to be in 0x000000ff format +*/ +static void gen_set_addr (TCGv addr,TCGv H ,TCGv M ,TCGv L ) +{ + +tcg_gen_andi_tl (L , addr,0x000000ff ); + +tcg_gen_andi_tl (M , addr,0x0000ff00 ); +tcg_gen_shri_tl (M ,M ,8 ); + +tcg_gen_andi_tl (H , addr,0x00ff0000 ); +} + + +static void gen_set_xaddr (TCGv addr) +{ +gen_set_addr (addr, cpu_rampX, cpu_r[27 ], cpu_r[26 ]); +} + + +static void gen_set_yaddr (TCGv addr) +{ +gen_set_addr (addr, cpu_rampY, cpu_r[29 ], cpu_r[28 ]); +} + + +static void gen_set_zaddr (TCGv addr) +{ +gen_set_addr (addr, cpu_rampZ, cpu_r[31 ], cpu_r[30 ]); +} + + +static TCGv gen_get_addr (TCGv H ,TCGv M ,TCGv L ) +{ +TCGv addr =tcg_temp_new_i32 (); + +tcg_gen_deposit_tl (addr,M ,H ,8 ,8 ); +tcg_gen_deposit_tl (addr,L , addr,8 ,16 ); + +return addr; +} + + +static TCGv gen_get_xaddr (void ) +{ +return gen_get_addr (cpu_rampX, cpu_r[27 ], cpu_r[26 ]); +} + + +static TCGv gen_get_yaddr (void ) +{ +return gen_get_addr (cpu_rampY, cpu_r[29 ], cpu_r[28 ]); +} + + +static TCGv gen_get_zaddr (void ) +{ +return gen_get_addr (cpu_rampZ, cpu_r[31 ], cpu_r[30 ]); +} + + +/* +* Load one byte indirect from data space to register and stores an clear +* the bits in data space specified by the register. The instruction can only +* be used towards internal SRAM. The data location is pointed to by the Z (16 +* bits) Pointer Register in the Register File. Memory access is limited to the +* current data segment of 64KB. To access another data segment in devices with +* more than 64KB data space, the RAMPZ in register in the I/O area has to be +* changed. The Z-pointer Register is left unchanged by the operation. This +* instruction is especially suited for clearing status bits stored in SRAM. +*/ +static void gen_data_store (DisasContext *ctx,TCGv data,TCGv addr) +{ +if (ctx->tb->flags &TB_FLAGS_FULL_ACCESS ) { +gen_helper_fullwr (cpu_env, data, addr); + }else { +tcg_gen_qemu_st8 (data, addr,MMU_DATA_IDX );/* mem[addr] = data */ + } +} + +static void gen_data_load (DisasContext *ctx,TCGv data,TCGv addr) +{ +if (ctx->tb->flags &TB_FLAGS_FULL_ACCESS ) { +gen_helper_fullrd (data, cpu_env, addr); + }else { +tcg_gen_qemu_ld8u (data, addr,MMU_DATA_IDX );/* data = mem[addr] */ + } +} + +/* * Subtracts an immediate value (0-63) from a register pair and places the * result in the register pair. This instruction operates on the upper four
@@ -2631,3 +2740,68 @@ static bool trans_BCLR(DisasContext *ctx, arg_BCLR *a)return true; } + + +/* +* The BREAK instruction is used by the On-chip Debug system, and is +* normally not used in the application software. When the BREAK instruction is +* executed, the AVR CPU is set in the Stopped Mode. This gives the On-chip +* Debugger access to internal resources. If any Lock bits are set, or either +* the JTAGEN or OCDEN Fuses are unprogrammed, the CPU will treat the BREAK +* instruction as a NOP and will not enter the Stopped mode. This instruction +* is not available in all devices. Refer to the device specific instruction +* set summary. +*/ +static bool trans_BREAK (DisasContext *ctx,arg_BREAK *a) +{ +if (!avr_have_feature (ctx,AVR_FEATURE_BREAK )) { +return true; + } + +#ifdef BREAKPOINT_ON_BREAK +tcg_gen_movi_tl (cpu_pc, ctx->npc -1 ); +gen_helper_debug (cpu_env); + ctx->bstate =DISAS_EXIT ; +#else +/* NOP */ +#endif + +return true; +} + + +/* +* This instruction performs a single cycle No Operation. +*/ +static bool trans_NOP (DisasContext *ctx,arg_NOP *a) +{ + +/* NOP */ + +return true; +} + + +/* +* This instruction sets the circuit in sleep mode defined by the MCU +* Control Register. +*/ +static bool trans_SLEEP (DisasContext *ctx,arg_SLEEP *a) +{ +gen_helper_sleep (cpu_env); + ctx->bstate =DISAS_NORETURN ; +return true; +} + + +/* +* This instruction resets the Watchdog Timer. This instruction must be +* executed within a limited time given by the WD prescaler. See the Watchdog +* Timer hardware specification. +*/ +static bool trans_WDR (DisasContext *ctx,arg_WDR *a) +{ +gen_helper_wdr (cpu_env); + +return true; +}
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 15/17 target/avr: Add boot serial test
In reply to Michael RolnikPrint out 'T' through serial port
--- tests/boot-serial-test.c | 10 ++++++++++ tests/Makefile.include | 2 ++ 2 files changed, 12 insertions(+)
tests/boot-serial-test.c +10 -0
@@ -16,6 +16,15 @@#include "qemu/osdep.h" #include "libqtest.h" +static const uint8_t bios_avr[] = { +0x88 ,0xe0 ,/* ldi r24, 0x08 */ +0x80 ,0x93 ,0xc1 ,0x00 ,/* sts 0x00C1, r24 ; Enable tx */ +0x86 ,0xe0 ,/* ldi r24, 0x06 */ +0x80 ,0x93 ,0xc2 ,0x00 ,/* sts 0x00C2, r24 ; Set the data bits to 8 */ +0x84 ,0xe5 ,/* ldi r24, 0x54 */ +0x80 ,0x93 ,0xc6 ,0x00 ,/* sts 0x00C6, r24 ; Output 'T' */ +}; +static const uint8_t kernel_mcf5208[] = {0x41 ,0xf9 ,0xfc ,0x06 ,0x00 ,0x00 ,/* lea 0xfc060000,%a0 */ 0x10 ,0x3c ,0x00 ,0x54 ,/* move.b #'T',%d0 */
@@ -103,6 +112,7 @@ typedef struct testdef {static testdef_t tests[] = { {"alpha" ,"clipper" ,"" ,"PCI:" }, + {"avr" ,"sample" ,"" ,"T" ,sizeof (bios_avr),NULL , bios_avr }, {"ppc" ,"ppce500" ,"" ,"U-Boot" }, {"ppc" ,"40p" ,"-vga none -boot d" ,"Trying cd:," }, {"ppc" ,"g3beige" ,"" ,"PowerPC,750" },
tests/Makefile.include +2 -0
@@ -208,6 +208,8 @@ check-qtest-i386-y += tests/test-x86-cpuid-compat$(EXESUF) check-qtest-i386-y += tests/numa-test$(EXESUF) check-qtest-x86_64-y += $(check-qtest-i386-y) +check-qtest-avr-y += tests/boot-serial-test$(EXESUF) + check-qtest-alpha-y += tests/boot-serial-test$(EXESUF) check-qtest-alpha-$(CONFIG_VGA) += tests/display-vga-test$(EXESUF)
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 10/17 target/avr: Add instruction disassembly function
In reply to Michael RolnikProvide function disassembles executed instruction when -d in_asm is provided
Example:./avr-softmmu/qemu-system-avr -bios free-rtos/Demo/AVR_ATMega2560_GCC/demo.elf -d in_asm will produce something like the following
```
...
IN:
0x0000014a: CALL 0x3808
IN: main
0x00003808: CALL 0x4b4
IN: vParTestInitialise
0x000004b4: LDI r24, 255
0x000004b6: STS r24, 0
0x000004b8: MULS r16, r20
0x000004ba: OUT $1, r24
0x000004bc: LDS r24, 0
0x000004be: MULS r16, r20
0x000004c0: OUT $2, r24
0x000004c2: RET
...
```--- target/avr/cpu.h | 1 + target/avr/cpu.c | 2 +- target/avr/disas.c | 228 +++++++++++++++++++++++++++++++++++++++++ target/avr/translate.c | 11 ++ 4 files changed, 241 insertions(+), 1 deletion(-) create mode 100644 target/avr/disas.c
target/avr/cpu.h +1 -0
@@ -157,6 +157,7 @@ bool avr_cpu_exec_interrupt(CPUState *cpu, int int_req);hwaddr avr_cpu_get_phys_page_debug (CPUState *cpu,vaddr addr);int avr_cpu_gdb_read_register (CPUState *cpu,uint8_t *buf,int reg);int avr_cpu_gdb_write_register (CPUState *cpu,uint8_t *buf,int reg); +int avr_print_insn (bfd_vma addr,disassemble_info *info);static inline int avr_feature (CPUAVRState *env,int feature) {
target/avr/cpu.c +1 -1
@@ -83,7 +83,7 @@ static void avr_cpu_reset(CPUState *cs)static void avr_cpu_disas_set_info (CPUState *cpu,disassemble_info *info) { info->mach = bfd_arch_avr; - info->print_insn =NULL ; + info->print_insn = avr_print_insn; }static void avr_cpu_realizefn (DeviceState *dev,Error **errp)
target/avr/disas.c new file +228 -0
228 lines not on this page · raw message
target/avr/translate.c +11 -0
@@ -3032,6 +3032,17 @@ done_generating: tb->size = (ctx.npc - pc_start) *2 ; tb->icount = num_insns; + +#ifdef DEBUG_DISAS +if (qemu_loglevel_mask (CPU_LOG_TB_IN_ASM ) + &&qemu_log_in_addr_range (tb->pc)) { +qemu_log_lock (); +qemu_log ("IN: %s\n" ,lookup_symbol (tb->pc)); +log_target_disas (cs, tb->pc, tb->size); +qemu_log ("\n" ); +qemu_log_unlock (); + } +#endif } void restore_state_to_opc(CPUAVRState *env, TranslationBlock *tb,
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 4/17 target/avr: Add instruction translation - Registers definition
In reply to Michael Rolnik--- target/avr/translate.c | 145 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 145 insertions(+) create mode 100644 target/avr/translate.c
target/avr/translate.c new file +145 -0
@@ -0,0 +1,145 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +#include "qemu/osdep.h" +#include "qemu/qemu-print.h" +#include "tcg/tcg.h" +#include "cpu.h" +#include "exec/exec-all.h" +#include "disas/disas.h" +#include "tcg-op.h" +#include "exec/cpu_ldst.h" +#include "exec/helper-proto.h" +#include "exec/helper-gen.h" +#include "exec/log.h" +#include "exec/gdbstub.h" +#include "exec/translator.h" +#include "exec/gen-icount.h" + +/* +* Define if you want a BREAK instruction translated to a breakpoint +* Active debugging connection is assumed +* This is for +* https://github.com/seharris/qemu-avr-tests/tree/master/instruction-tests +* tests +*/ +#undef BREAKPOINT_ON_BREAK + +static TCGv cpu_pc; + +static TCGv cpu_Cf; +static TCGv cpu_Zf; +static TCGv cpu_Nf; +static TCGv cpu_Vf; +static TCGv cpu_Sf; +static TCGv cpu_Hf; +static TCGv cpu_Tf; +static TCGv cpu_If; + +static TCGv cpu_rampD; +static TCGv cpu_rampX; +static TCGv cpu_rampY; +static TCGv cpu_rampZ; + +static TCGv cpu_r[NUMBER_OF_CPU_REGISTERS ]; +static TCGv cpu_eind; +static TCGv cpu_sp; + +static TCGv cpu_skip; + +static const char reg_names[NUMBER_OF_CPU_REGISTERS ][8 ] = { +"r0" ,"r1" ,"r2" ,"r3" ,"r4" ,"r5" ,"r6" ,"r7" , +"r8" ,"r9" ,"r10" ,"r11" ,"r12" ,"r13" ,"r14" ,"r15" , +"r16" ,"r17" ,"r18" ,"r19" ,"r20" ,"r21" ,"r22" ,"r23" , +"r24" ,"r25" ,"r26" ,"r27" ,"r28" ,"r29" ,"r30" ,"r31" , +}; +#define REG (x) (cpu_r[x]) + +enum { +DISAS_EXIT =DISAS_TARGET_0 ,/* We want return to the cpu main loop. */ +DISAS_LOOKUP =DISAS_TARGET_1 ,/* We have a variable condition exit. */ +DISAS_CHAIN =DISAS_TARGET_2 ,/* We have a single condition exit. */ +}; + +typedef struct DisasContext DisasContext ; + +/* This is the state at translation time. */ +struct DisasContext { +TranslationBlock *tb; + +CPUAVRState *env; +CPUState *cs; + +target_long npc; +uint32_t opcode; + +/* Routine used to access memory */ +int memidx; +int bstate; +int singlestep; + +TCGv skip_var0; +TCGv skip_var1; +TCGCond skip_cond; +bool free_skip_var0; +}; + +static int to_regs_16_31_by_one (DisasContext *ctx,int indx) +{ +return 16 + (indx %16 ); +} + +static int to_regs_16_23_by_one (DisasContext *ctx,int indx) +{ +return 16 + (indx %8 ); +} +static int to_regs_24_30_by_two (DisasContext *ctx,int indx) +{ +return 24 + (indx %4 ) *2 ; +} +static int to_regs_00_30_by_two (DisasContext *ctx,int indx) +{ +return (indx %16 ) *2 ; +} + +static uint16_t next_word (DisasContext *ctx) +{ +return cpu_lduw_code (ctx->env, ctx->npc++ *2 ); +} + +static int append_16 (DisasContext *ctx,int x) +{ +return x <<16 |next_word (ctx); +} + + +static bool avr_have_feature (DisasContext *ctx,int feature) +{ +if (!avr_feature (ctx->env, feature)) { +gen_helper_unsupported (cpu_env); + ctx->bstate =DISAS_NORETURN ; +return false; + } +return true; +} + +static bool decode_insn (DisasContext *ctx,uint16_t insn); +#include "decode_insn.inc.c" +
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 9/17 target/avr: Add instruction translation - CPU main translation function
In reply to Michael Rolnik--- target/avr/translate.c | 234 +++++++++++++++++++++++++++++++++++++++++ 1 file changed, 234 insertions(+)
target/avr/translate.c +234 -0
@@ -2805,3 +2805,237 @@ static bool trans_WDR(DisasContext *ctx, arg_WDR *a)return true; } + + +void avr_cpu_tcg_init (void ) +{ +int i; + +#define AVR_REG_OFFS (x) offsetof(CPUAVRState, x) + cpu_pc =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (pc_w),"pc" ); + cpu_Cf =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (sregC),"Cf" ); + cpu_Zf =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (sregZ),"Zf" ); + cpu_Nf =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (sregN),"Nf" ); + cpu_Vf =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (sregV),"Vf" ); + cpu_Sf =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (sregS),"Sf" ); + cpu_Hf =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (sregH),"Hf" ); + cpu_Tf =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (sregT),"Tf" ); + cpu_If =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (sregI),"If" ); + cpu_rampD =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (rampD),"rampD" ); + cpu_rampX =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (rampX),"rampX" ); + cpu_rampY =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (rampY),"rampY" ); + cpu_rampZ =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (rampZ),"rampZ" ); + cpu_eind =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (eind),"eind" ); + cpu_sp =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (sp),"sp" ); + cpu_skip =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (skip),"skip" ); + +for (i =0 ; i <NUMBER_OF_CPU_REGISTERS ; i++) { + cpu_r[i] =tcg_global_mem_new_i32 (cpu_env,AVR_REG_OFFS (r[i]), + reg_names[i]); + } +#undef AVR_REG_OFFS +} + +static void translate (DisasContext *ctx) +{ +uint32_t opcode =next_word (ctx); + +if (!decode_insn (ctx, opcode)) { +gen_helper_unsupported (cpu_env); + ctx->bstate =DISAS_NORETURN ; + } +} + +/* Standardize the cpu_skip condition to NE. */ +static bool canonicalize_skip (DisasContext *ctx) +{ +switch (ctx->skip_cond) { +case TCG_COND_NEVER : +/* Normal case: cpu_skip is known to be false. */ +return false; + +case TCG_COND_ALWAYS : +/* +* Breakpoint case: cpu_skip is known to be true, via TB_FLAGS_SKIP. +* The breakpoint is on the instruction being skipped, at the start +* of the TranslationBlock. No need to update. +*/ +return false; + +case TCG_COND_NE : +if (ctx->skip_var1 ==NULL ) { +tcg_gen_mov_tl (cpu_skip, ctx->skip_var0); + }else { +tcg_gen_xor_tl (cpu_skip, ctx->skip_var0, ctx->skip_var1); + ctx->skip_var1 =NULL ; + } +break ; + +default : +/* Convert to a NE condition vs 0. */ +if (ctx->skip_var1 ==NULL ) { +tcg_gen_setcondi_tl (ctx->skip_cond, cpu_skip, ctx->skip_var0,0 ); + }else { +tcg_gen_setcond_tl (ctx->skip_cond, cpu_skip, + ctx->skip_var0, ctx->skip_var1); + ctx->skip_var1 =NULL ; + } + ctx->skip_cond =TCG_COND_NE ; +break ; + } +if (ctx->free_skip_var0) { +tcg_temp_free (ctx->skip_var0); + ctx->free_skip_var0 = false; + } + ctx->skip_var0 = cpu_skip; +return true; +} + +void gen_intermediate_code (CPUState *cs,TranslationBlock *tb,int max_insns) +{ +CPUAVRState *env = cs->env_ptr; +DisasContext ctx = { + .tb = tb, + .cs = cs, + .env = env, + .memidx =0 , + .bstate =DISAS_NEXT , + .skip_cond =TCG_COND_NEVER , + .singlestep = cs->singlestep_enabled, + }; +target_ulong pc_start = tb->pc /2 ; +int num_insns =0 ; + +if (tb->flags &TB_FLAGS_FULL_ACCESS ) { +/* +* This flag is set by ST/LD instruction we will regenerate it ONLY +* with mem/cpu memory access instead of mem access +*/ + max_insns =1 ; + } +if (ctx.singlestep) { + max_insns =1 ; + } + +gen_tb_start (tb); + + ctx.npc = pc_start; +if (tb->flags &TB_FLAGS_SKIP ) { + ctx.skip_cond =TCG_COND_ALWAYS ; + ctx.skip_var0 = cpu_skip; + } + +do { +TCGLabel *skip_label =NULL ; + +/* translate current instruction */ +tcg_gen_insn_start (ctx.npc); + num_insns++; + +/* +* this is due to some strange GDB behavior +* let's assume main has address 0x100 +* b main - sets breakpoint at address 0x00000100 (code) +* b *0x100 - sets breakpoint at address 0x00800100 (data) +*/ +if (unlikely (!ctx.singlestep && + (cpu_breakpoint_test (cs,OFFSET_CODE + ctx.npc *2 ,BP_ANY ) || +cpu_breakpoint_test (cs,OFFSET_DATA + ctx.npc *2 ,BP_ANY )))) { +canonicalize_skip (&ctx); +tcg_gen_movi_tl (cpu_pc, ctx.npc); +gen_helper_debug (cpu_env); + goto done_generating; + } + +/* Conditionally skip the next instruction, if indicated. */ +if (ctx.skip_cond !=TCG_COND_NEVER ) { + skip_label =gen_new_label (); +if (ctx.skip_var0 == cpu_skip) { +/* +* Copy cpu_skip so that we may zero it before the branch. +* This ensures that cpu_skip is non-zero after the label +* if and only if the skipped insn itself sets a skip. +*/ + ctx.free_skip_var0 = true; + ctx.skip_var0 =tcg_temp_new (); +tcg_gen_mov_tl (ctx.skip_var0, cpu_skip); +tcg_gen_movi_tl (cpu_skip,0 ); + } +if (ctx.skip_var1 ==NULL ) { +tcg_gen_brcondi_tl (ctx.skip_cond, ctx.skip_var0,0 , skip_label); + }else { +tcg_gen_brcond_tl (ctx.skip_cond, ctx.skip_var0, + ctx.skip_var1, skip_label); + ctx.skip_var1 =NULL ; + } +if (ctx.free_skip_var0) { +tcg_temp_free (ctx.skip_var0); + ctx.free_skip_var0 = false; + } + ctx.skip_cond =TCG_COND_NEVER ; + ctx.skip_var0 =NULL ; + } + +translate (&ctx); + +if (skip_label) { +canonicalize_skip (&ctx); +gen_set_label (skip_label); +if (ctx.bstate ==DISAS_NORETURN ) { + ctx.bstate =DISAS_CHAIN ; + } + } + }while (ctx.bstate ==DISAS_NEXT + && num_insns < max_insns + && (ctx.npc - pc_start) *2 <TARGET_PAGE_SIZE -4 + && !tcg_op_buf_full ()); + +if (tb->cflags &CF_LAST_IO ) { +gen_io_end (); + } + +bool nonconst_skip =canonicalize_skip (&ctx); + +switch (ctx.bstate) { +case DISAS_NORETURN : +assert (!nonconst_skip); +break ; +case DISAS_NEXT : +case DISAS_TOO_MANY : +case DISAS_CHAIN : +if (!nonconst_skip) { +/* Note gen_goto_tb checks singlestep. */ +gen_goto_tb (&ctx,1 , ctx.npc); +break ; + } +tcg_gen_movi_tl (cpu_pc, ctx.npc); +/* fall through */ +case DISAS_LOOKUP : +if (!ctx.singlestep) { +tcg_gen_lookup_and_goto_ptr (); +break ; + } +/* fall through */ +case DISAS_EXIT : +if (ctx.singlestep) { +gen_helper_debug (cpu_env); + }else { +tcg_gen_exit_tb (NULL ,0 ); + } +break ; +default : +g_assert_not_reached (); + } + +done_generating: +gen_tb_end (tb, num_insns); + + tb->size = (ctx.npc - pc_start) *2 ; + tb->icount = num_insns; +} + +void restore_state_to_opc (CPUAVRState *env,TranslationBlock *tb, +target_ulong *data) +{ + env->pc_w = data[0 ]; +}
Signature · 1 line
2.17.2 (Apple Git-113)
Michael Rolnik
NAKPATCH v37 12/17 target/avr: Add example board configuration
In reply to Michael RolnikA simple board setup that configures an AVR CPU to run a given firmware image.
This is all that's useful to implement without peripheral emulation as AVR CPUs include a lot of on-board peripherals.
NOTE: this is not a real board !!!!
NOTE: it's used for CPU testing!!!!
--- hw/avr/sample.c | 282 +++++++++++++++++++++++++++++++++++++++++++ hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + 4 files changed, 290 insertions(+) create mode 100644 hw/avr/sample.c create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs
hw/avr/sample.c new file +282 -0
@@ -0,0 +1,282 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +/* +* NOTE: +* This is not a real AVR board, this is an example! +* The CPU is an approximation of an ATmega2560, but is missing various +* built-in peripherals. +* +* This example board loads provided binary file into flash memory and +* executes it from 0x00000000 address in the code memory space. +* +* Currently used for AVR CPU validation +* +*/ + +#include "qemu/osdep.h" +#include "qapi/error.h" +#include "qemu-common.h" +#include "cpu.h" +#include "hw/hw.h" +#include "sysemu/sysemu.h" +#include "sysemu/qtest.h" +#include "ui/console.h" +#include "hw/boards.h" +#include "hw/loader.h" +#include "qemu/error-report.h" +#include "exec/address-spaces.h" +#include "include/hw/sysbus.h" +#include "include/hw/char/avr_usart.h" +#include "include/hw/timer/avr_timer16.h" +#include "include/hw/misc/avr_mask.h" +#include "elf.h" +#include "hw/misc/unimp.h" + +#define SIZE_FLASH 0x00040000 +#define SIZE_SRAM 0x00002000 +/* +* Size of additional "external" memory, as if the AVR were configured to use +* an external RAM chip. +* Note that the configuration registers that normally enable this feature are +* unimplemented. +*/ +#define SIZE_EXMEM 0x00000000 + +/* Offsets of peripherals in emulated memory space (i.e. not host addresses) */ +#define PRR0_BASE 0x64 +#define PRR1_BASE 0x65 +#define USART_BASE 0xc0 +#define TIMER1_BASE 0x80 +#define TIMER1_IMSK_BASE 0x6f +#define TIMER1_IFR_BASE 0x36 + +/* Interrupt numbers used by peripherals */ +#define USART_RXC_IRQ 24 +#define USART_DRE_IRQ 25 +#define USART_TXC_IRQ 26 + +#define TIMER1_CAPT_IRQ 15 +#define TIMER1_COMPA_IRQ 16 +#define TIMER1_COMPB_IRQ 17 +#define TIMER1_COMPC_IRQ 18 +#define TIMER1_OVF_IRQ 19 + +/* Power reduction */ +#define PRR1_BIT_PRTIM5 0x05/* Timer/Counter5 */ +#define PRR1_BIT_PRTIM4 0x04/* Timer/Counter4 */ +#define PRR1_BIT_PRTIM3 0x03/* Timer/Counter3 */ +#define PRR1_BIT_PRUSART3 0x02/* USART3 */ +#define PRR1_BIT_PRUSART2 0x01/* USART2 */ +#define PRR1_BIT_PRUSART1 0x00/* USART1 */ + +#define PRR0_BIT_PRTWI 0x06/* TWI */ +#define PRR0_BIT_PRTIM2 0x05/* Timer/Counter2 */ +#define PRR0_BIT_PRTIM0 0x04/* Timer/Counter0 */ +#define PRR0_BIT_PRTIM1 0x03/* Timer/Counter1 */ +#define PRR0_BIT_PRSPI 0x02/* Serial Peripheral Interface */ +#define PRR0_BIT_PRUSART0 0x01/* USART0 */ +#define PRR0_BIT_PRADC 0x00/* ADC */ + +typedef struct { +MachineClass parent; +}SampleMachineClass ; + +typedef struct { +MachineState parent; +MemoryRegion *ram; +MemoryRegion *flash; +AVRUsartState *usart0; +AVRTimer16State *timer1; +AVRMaskState *prr[2 ]; +}SampleMachineState ; + +#define TYPE_SAMPLE_MACHINE MACHINE_TYPE_NAME("sample") + +#define SAMPLE_MACHINE (obj) \ + OBJECT_CHECK(SampleMachineState, obj, TYPE_SAMPLE_MACHINE) +#define SAMPLE_MACHINE_GET_CLASS (obj) \ + OBJECT_GET_CLASS(SampleMachineClass, obj, TYPE_SAMPLE_MACHINE) +#define SAMPLE_MACHINE_CLASS (klass) \ + OBJECT_CLASS_CHECK(SampleMachineClass, klass, TYPE_SAMPLE_MACHINE) + +static void sample_init (MachineState *machine) +{ +SampleMachineState *sms =SAMPLE_MACHINE (machine); +MemoryRegion *system_memory =get_system_memory (); +AVRCPU *cpu; +const char *firmware =NULL ; +const char *filename; +int bytes_loaded; +SysBusDevice *busdev; +DeviceState *cpudev; + + system_memory =get_system_memory (); + sms->ram =g_new (MemoryRegion,1 ); + sms->flash =g_new (MemoryRegion,1 ); + + cpu =AVR_CPU (cpu_create (machine->cpu_type)); + cpudev =DEVICE (cpu); + + +memory_region_init_rom (sms->flash,NULL ,"avr.flash" ,SIZE_FLASH , + &error_fatal); +memory_region_add_subregion (system_memory,OFFSET_CODE , sms->flash); + +/* following are atmel2560 device */ +create_unimplemented_device ("usart 3" ,OFFSET_DATA +0x0130 ,0x0007 ); +create_unimplemented_device ("timer-counter-16bit 5" , +OFFSET_DATA +0x0120 ,0x000e ); +create_unimplemented_device ("gpio L" ,OFFSET_DATA +0x0109 ,0x0003 ); +create_unimplemented_device ("gpio K" ,OFFSET_DATA +0x0106 ,0x0003 ); +create_unimplemented_device ("gpio J" ,OFFSET_DATA +0x0103 ,0x0003 ); +create_unimplemented_device ("gpio H" ,OFFSET_DATA +0x0100 ,0x0003 ); +create_unimplemented_device ("usart 2" ,OFFSET_DATA +0x00d0 ,0x0007 ); +create_unimplemented_device ("usart 1" ,OFFSET_DATA +0x00c8 ,0x0007 ); +create_unimplemented_device ("usart 0" ,OFFSET_DATA +0x00c0 ,0x0007 ); +create_unimplemented_device ("twi" ,OFFSET_DATA +0x00b8 ,0x0006 ); +create_unimplemented_device ("timer-counter-async-8bit 2" , +OFFSET_DATA +0x00b0 ,0x0007 ); +create_unimplemented_device ("timer-counter-16bit 4" , +OFFSET_DATA +0x00a0 ,0x000e ); +create_unimplemented_device ("timer-counter-16bit 3" , +OFFSET_DATA +0x0090 ,0x000e ); +create_unimplemented_device ("timer-counter-16bit 1" , +OFFSET_DATA +0x0080 ,0x000e ); +create_unimplemented_device ("ac / adc" , +OFFSET_DATA +0x0078 ,0x0008 ); +create_unimplemented_device ("ext-mem-iface" , +OFFSET_DATA +0x0074 ,0x0002 ); +create_unimplemented_device ("int-controller" , +OFFSET_DATA +0x0068 ,0x000c ); +create_unimplemented_device ("sys" , +OFFSET_DATA +0x0060 ,0x0007 ); +create_unimplemented_device ("spi" , +OFFSET_DATA +0x004c ,0x0003 ); +create_unimplemented_device ("ext-mem-iface" , +OFFSET_DATA +0x004a ,0x0002 ); +create_unimplemented_device ("timer-counter-pwm-8bit 0" , +OFFSET_DATA +0x0043 ,0x0006 ); +create_unimplemented_device ("ext-mem-iface" , +OFFSET_DATA +0x003e ,0x0005 ); +create_unimplemented_device ("int-controller" , +OFFSET_DATA +0x0035 ,0x0009 ); +create_unimplemented_device ("gpio G" ,OFFSET_DATA +0x0032 ,0x0003 ); +create_unimplemented_device ("gpio F" ,OFFSET_DATA +0x002f ,0x0003 ); +create_unimplemented_device ("gpio E" ,OFFSET_DATA +0x002c ,0x0003 ); +create_unimplemented_device ("gpio D" ,OFFSET_DATA +0x0029 ,0x0003 ); +create_unimplemented_device ("gpio C" ,OFFSET_DATA +0x0026 ,0x0003 ); +create_unimplemented_device ("gpio B" ,OFFSET_DATA +0x0023 ,0x0003 ); +create_unimplemented_device ("gpio A" ,OFFSET_DATA +0x0020 ,0x0003 ); + +memory_region_allocate_system_memory ( + sms->ram,NULL ,"avr.ram" ,SIZE_SRAM +SIZE_EXMEM ); +memory_region_add_subregion (system_memory,OFFSET_DATA +0x200 , sms->ram); + +/* Power Reduction built-in peripheral */ + sms->prr[0 ] =AVR_MASK (sysbus_create_simple (TYPE_AVR_MASK , +OFFSET_DATA +PRR0_BASE ,NULL )); + sms->prr[1 ] =AVR_MASK (sysbus_create_simple (TYPE_AVR_MASK , +OFFSET_DATA +PRR1_BASE ,NULL )); + +/* USART 0 built-in peripheral */ + sms->usart0 =AVR_USART (object_new (TYPE_AVR_USART )); + busdev =SYS_BUS_DEVICE (sms->usart0); +qdev_prop_set_chr (DEVICE (sms->usart0),"chardev" ,serial_hd (0 )); +object_property_set_bool (OBJECT (sms->usart0), true,"realized" , + &error_fatal); +sysbus_mmio_map (busdev,0 ,OFFSET_DATA +USART_BASE ); +/* +* These IRQ numbers don't match the datasheet because we're counting from +* zero and not including reset. +*/ +sysbus_connect_irq (busdev,0 ,qdev_get_gpio_in (cpudev,USART_RXC_IRQ )); +sysbus_connect_irq (busdev,1 ,qdev_get_gpio_in (cpudev,USART_DRE_IRQ )); +sysbus_connect_irq (busdev,2 ,qdev_get_gpio_in (cpudev,USART_TXC_IRQ )); +sysbus_connect_irq (SYS_BUS_DEVICE (sms->prr[1 ]),PRR1_BIT_PRUSART1 , +qdev_get_gpio_in (DEVICE (sms->usart0),0 )); + +/* Timer 1 built-in periphal */ + sms->timer1 =AVR_TIMER16 (object_new (TYPE_AVR_TIMER16 )); +object_property_set_bool (OBJECT (sms->timer1), true,"realized" , + &error_fatal); + busdev =SYS_BUS_DEVICE (sms->timer1); +sysbus_mmio_map (busdev,0 ,OFFSET_DATA +TIMER1_BASE ); +sysbus_mmio_map (busdev,1 ,OFFSET_DATA +TIMER1_IMSK_BASE ); +sysbus_mmio_map (busdev,2 ,OFFSET_DATA +TIMER1_IFR_BASE ); +sysbus_connect_irq (busdev,0 ,qdev_get_gpio_in (cpudev,TIMER1_CAPT_IRQ )); +sysbus_connect_irq (busdev,1 ,qdev_get_gpio_in (cpudev,TIMER1_COMPA_IRQ )); +sysbus_connect_irq (busdev,2 ,qdev_get_gpio_in (cpudev,TIMER1_COMPB_IRQ )); +sysbus_connect_irq (busdev,3 ,qdev_get_gpio_in (cpudev,TIMER1_COMPC_IRQ )); +sysbus_connect_irq (busdev,4 ,qdev_get_gpio_in (cpudev,TIMER1_OVF_IRQ )); +sysbus_connect_irq (SYS_BUS_DEVICE (sms->prr[0 ]),PRR0_BIT_PRTIM1 , +qdev_get_gpio_in (DEVICE (sms->timer1),0 )); + +/* Load firmware (contents of flash) trying to auto-detect format */ + firmware = machine->firmware; +if (firmware !=NULL ) { + filename =qemu_find_file (QEMU_FILE_TYPE_BIOS , firmware); +if (filename ==NULL ) { +error_report ("Unable to find %s" , firmware); +exit (1 ); + } + + bytes_loaded =load_elf ( + filename,NULL ,NULL ,NULL ,NULL ,NULL ,NULL ,0 ,EM_NONE ,0 ,0 ); +if (bytes_loaded <0 ) { + bytes_loaded =load_image_targphys ( + filename,OFFSET_CODE ,SIZE_FLASH ); + } +if (bytes_loaded <0 ) { +error_report ( +"Unable to load firmware image %s as ELF or raw binary" , + firmware); +exit (1 ); + } + } +} + +static void sample_class_init (ObjectClass *oc,void *data) +{ +MachineClass *mc =MACHINE_CLASS (oc); + + mc->desc ="AVR sample/example board (ATmega2560)" ; + mc->init = sample_init; + mc->default_cpus =1 ; + mc->min_cpus = mc->default_cpus; + mc->max_cpus = mc->default_cpus; + mc->default_cpu_type ="avr6-avr-cpu" ;/* ATmega2560. */ + mc->is_default =1 ; +} + +static const TypeInfo sample_info = { + .name =TYPE_SAMPLE_MACHINE , + .parent =TYPE_MACHINE , + .instance_size =sizeof (SampleMachineState), + .class_size =sizeof (SampleMachineClass), + .class_init = sample_class_init, +}; + +static void sample_machine_init (void ) +{ +type_register_static (&sample_info); +} + +type_init (sample_machine_init);
hw/Kconfig +1 -0
@@ -44,6 +44,7 @@ source watchdog/Kconfig # arch Kconfig source arm/Kconfig source alpha/Kconfig +source avr/Kconfig source cris/Kconfig source hppa/Kconfig source i386/Kconfig
hw/avr/Kconfig new file +6 -0
@@ -0,0 +1,6 @@ +config AVR_SAMPLE + bool + select AVR_TIMER16 + select AVR_USART + select AVR_MASK + select UNIMP
hw/avr/Makefile.objs new file +1 -0
@@ -0,0 +1 @@ +obj-y += sample.o
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 17/17 target/avr: Update MAINTAINERS file
In reply to Michael RolnikInclude AVR maintaners in MAINTAINERS file
--- MAINTAINERS | 11 +++++++++++ 1 file changed, 11 insertions(+)
MAINTAINERS +11 -0
@@ -163,6 +163,17 @@ S: Maintained F: hw/arm/smmu* F: include/hw/arm/smmu* +AVR TCG CPUs +M: Michael Rolnik <[email protected]> +R: Sarah Harris <[email protected]> +S: Maintained +F: target/avr/ +F: hw/misc/avr_mask.c +F: hw/char/avr_usart.c +F: hw/timer/avr_timer16.c +F: hw/avr/ +F: tests/acceptance/machine_avr6.py + CRIS TCG CPUs M: Edgar E. Iglesias <[email protected]> S: Maintained
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 14/17 target/avr: Update build system
In reply to Michael RolnikMake AVR support buildable
--- configure | 7 +++++++ default-configs/avr-softmmu.mak | 5 +++++ target/avr/Makefile.objs | 34 +++++++++++++++++++++++++++++++++ 3 files changed, 46 insertions(+) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 target/avr/Makefile.objs
configure +7 -0
@@ -7649,6 +7649,10 @@ case "$target_name" in mttcg="yes" gdb_xml_files="aarch64-core.xml aarch64-fpu.xml arm-core.xml arm-vfp.xml arm-vfp3.xml arm-neon.xml" ;; + avr) + gdb_xml_files="avr-cpu.xml" + target_compiler=$cross_cc_avr + ;; cris) ;; hppa)
@@ -7868,6 +7872,9 @@ for i in $ARCH $TARGET_BASE_ARCH ; do disas_config "ARM_A64" fi ;; + avr) + disas_config "AVR" + ;; cris) disas_config "CRIS" ;;
default-configs/avr-softmmu.mak new file +5 -0
@@ -0,0 +1,5 @@ +# Default configuration for avr-softmmu + +# Boards: +# +CONFIG_AVR_SAMPLE=y
target/avr/Makefile.objs new file +34 -0
@@ -0,0 +1,34 @@ +# +# QEMU AVR CPU +# +# Copyright (c) 2019 Michael Rolnik +# +# This library is free software; you can redistribute it and/or +# modify it under the terms of the GNU Lesser General Public +# License as published by the Free Software Foundation; either +# version 2.1 of the License, or (at your option) any later version. +# +# This library is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +# Lesser General Public License for more details. +# +# You should have received a copy of the GNU Lesser General Public +# License along with this library; if not, see +# <http://www.gnu.org/licenses/lgpl-2.1.html> +# + +DECODETREE = $(SRC_PATH)/scripts/decodetree.py +decode-y = $(SRC_PATH)/target/avr/insn.decode + +target/avr/decode_insn.inc.c: $(decode-y) $(DECODETREE) + $(call quiet-command, \ + $(PYTHON) $(DECODETREE) -o $@ --decode decode_insn --insnwidth 16 $<, \ + "GEN", $(TARGET_DIR)$@) + +target/avr/translate.o: target/avr/decode_insn.inc.c + +obj-y += translate.o cpu.o helper.o +obj-y += gdbstub.o +obj-y += disas.o +obj-$(CONFIG_SOFTMMU) += machine.o
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 16/17 target/avr: Add Avocado test
In reply to Michael RolnikThe test is based on
https://github.com/seharris/qemu-avr-tests/tree/master/free-rtos/Demo
demo which. If working correctly, prints 'ABCDEFGHIJKLMNOPQRSTUVWX' out. it also demostrates that timer and IRQ are working
--- tests/acceptance/machine_avr6.py | 56 ++++++++++++++++++++++++++++++++ 1 file changed, 56 insertions(+) create mode 100644 tests/acceptance/machine_avr6.py
tests/acceptance/machine_avr6.py new file +56 -0
@@ -0,0 +1,56 @@ +# +# QEMU AVR +# +# Copyright (c) 2019 Michael Rolnik <[email protected]> +# +# This program is free software: you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation, either version 2 of the License, or +# (at your option) any later version. +# +# This program is distributed in the hope that 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/>. +# + +import logging +import time +import distutils.spawn + +from avocado import skipUnless +from avocado_qemu import Test +from avocado.utils import process + +class AVR6Machine(Test): + timeout = 5 + + def test_freertos(self): + """ + :avocado: tags=arch:avr + :avocado: tags=machine:sample + """ + """ + https://github.com/seharris/qemu-avr-tests/raw/master/free-rtos/Demo/AVR_ATMega2560_GCC/demo.elf + constantly prints out 'ABCDEFGHIJKLMNOPQRSTUVWXABCDEFGHIJKLMNOPQRSTUVWX' + """ + rom_url = 'https://github.com/seharris/qemu-avr-tests' + rom_sha1= '36c3e67b8755dcf37e06af6730ef5d477b8ed16d' + rom_url += '/raw/' + rom_sha1 + '/free-rtos/Demo/AVR_ATMega2560_GCC/demo.elf' + rom_hash = '7eb521f511ca8f2622e0a3c5e8dd686efbb911d4' + rom_path = self.fetch_asset(rom_url, asset_hash=rom_hash) + + self.vm.set_machine('sample') + self.vm.add_args('-bios', rom_path) + self.vm.add_args('-nographic') + self.vm.launch() + + time.sleep(2) + self.vm.shutdown() + + match = 'ABCDEFGHIJKLMNOPQRSTUVWXABCDEFGHIJKLMNOPQRSTUVWX' + + self.assertIn(match, self.vm.get_log())
Signature · 1 line
2.17.2 (Apple Git-113)
PATCH v37 11/17 target/avr: Add limited support for USART and 16 bit timer peripherals
In reply to Michael RolnikFrom: Sarah Harris <[email protected]>
These were designed to facilitate testing but should provide enough function to be useful in other contexts.
Only a subset of the functions of each peripheral is implemented, mainly due to the lack of a standard way to handle electrical connections (like GPIO pins).
--- include/hw/char/avr_usart.h | 97 ++++++ include/hw/misc/avr_mask.h | 47 +++ include/hw/timer/avr_timer16.h | 97 ++++++ hw/char/avr_usart.c | 324 ++++++++++++++++++ hw/misc/avr_mask.c | 112 ++++++ hw/timer/avr_timer16.c | 605 +++++++++++++++++++++++++++++++++ hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + 12 files changed, 1296 insertions(+) create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c
include/hw/char/avr_usart.h new file +97 -0
@@ -0,0 +1,97 @@ +/* +* AVR USART +* +* Copyright (c) 2018 University of Kent +* Author: Sarah Harris +* +* Permission is hereby granted, free of charge, to any person obtaining a copy +* of this software and associated documentation files (the "Software"), to deal +* in the Software without restriction, including without limitation the rights +* to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +* copies of the Software, and to permit persons to whom the Software is +* furnished to do so, subject to the following conditions: +* +* The above copyright notice and this permission notice shall be included in +* all copies or substantial portions of the Software. +* +* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL +* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +* OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN +* THE SOFTWARE. +*/ + +#ifndef HW_AVR_USART_H +#define HW_AVR_USART_H + +#include "hw/sysbus.h" +#include "chardev/char-fe.h" +#include "hw/hw.h" + +/* Offsets of registers. */ +#define USART_DR 0x06 +#define USART_CSRA 0x00 +#define USART_CSRB 0x01 +#define USART_CSRC 0x02 +#define USART_BRRH 0x05 +#define USART_BRRL 0x04 + +/* Relevant bits in regiters. */ +#define USART_CSRA_RXC (1 << 7) +#define USART_CSRA_TXC (1 << 6) +#define USART_CSRA_DRE (1 << 5) +#define USART_CSRA_MPCM (1 << 0) + +#define USART_CSRB_RXCIE (1 << 7) +#define USART_CSRB_TXCIE (1 << 6) +#define USART_CSRB_DREIE (1 << 5) +#define USART_CSRB_RXEN (1 << 4) +#define USART_CSRB_TXEN (1 << 3) +#define USART_CSRB_CSZ2 (1 << 2) +#define USART_CSRB_RXB8 (1 << 1) +#define USART_CSRB_TXB8 (1 << 0) + +#define USART_CSRC_MSEL1 (1 << 7) +#define USART_CSRC_MSEL0 (1 << 6) +#define USART_CSRC_PM1 (1 << 5) +#define USART_CSRC_PM0 (1 << 4) +#define USART_CSRC_CSZ1 (1 << 2) +#define USART_CSRC_CSZ0 (1 << 1) + +#define TYPE_AVR_USART "avr-usart" +#define AVR_USART (obj) \ + OBJECT_CHECK(AVRUsartState, (obj), TYPE_AVR_USART) + +typedef struct { +/* <private> */ +SysBusDevice parent_obj; + +/* <public> */ +MemoryRegion mmio; + +CharBackend chr; + +bool enabled; + +uint8_t data; +bool data_valid; +uint8_t char_mask; +/* Control and Status Registers */ +uint8_t csra; +uint8_t csrb; +uint8_t csrc; +/* Baud Rate Registers (low/high byte) */ +uint8_t brrh; +uint8_t brrl; + +/* Receive Complete */ +qemu_irq rxc_irq; +/* Transmit Complete */ +qemu_irq txc_irq; +/* Data Register Empty */ +qemu_irq dre_irq; +}AVRUsartState ; + +#endif /* HW_AVR_USART_H */
include/hw/misc/avr_mask.h new file +47 -0
47 lines not on this page · raw message
include/hw/timer/avr_timer16.h new file +97 -0
97 lines not on this page · raw message
hw/char/avr_usart.c new file +324 -0
324 lines not on this page · raw message
hw/misc/avr_mask.c new file +112 -0
112 lines not on this page · raw message
hw/timer/avr_timer16.c new file +605 -0
605 lines not on this page · raw message
hw/char/Kconfig +3 -0
@@ -46,3 +46,6 @@ config SCLPCONSOLE config TERMINAL3270 bool + +config AVR_USART + bool
hw/char/Makefile.objs +1 -0
@@ -21,6 +21,7 @@ obj-$(CONFIG_PSERIES) += spapr_vty.o obj-$(CONFIG_DIGIC) += digic-uart.o obj-$(CONFIG_STM32F2XX_USART) += stm32f2xx_usart.o obj-$(CONFIG_RASPI) += bcm2835_aux.o +common-obj-$(CONFIG_AVR_USART) += avr_usart.o common-obj-$(CONFIG_CMSDK_APB_UART) += cmsdk-apb-uart.o common-obj-$(CONFIG_ETRAXFS) += etraxfs_ser.o
hw/misc/Kconfig +3 -0
@@ -125,4 +125,7 @@ config MAC_VIA select MOS6522 select ADB +config AVR_MASK + bool + source macio/Kconfig
hw/misc/Makefile.objs +2 -0
@@ -82,3 +82,5 @@ common-obj-$(CONFIG_NRF51_SOC) += nrf51_rng.o obj-$(CONFIG_MAC_VIA) += mac_via.o common-obj-$(CONFIG_GRLIB) += grlib_ahb_apb_pnp.o + +obj-$(CONFIG_AVR_MASK) += avr_mask.o
hw/timer/Kconfig +3 -0
@@ -34,3 +34,6 @@ config CMSDK_APB_TIMER config CMSDK_APB_DUALTIMER bool select PTIMER + +config AVR_TIMER16 + bool
hw/timer/Makefile.objs +2 -0
@@ -35,3 +35,5 @@ common-obj-$(CONFIG_CMSDK_APB_TIMER) += cmsdk-apb-timer.o common-obj-$(CONFIG_CMSDK_APB_DUALTIMER) += cmsdk-apb-dualtimer.o common-obj-$(CONFIG_MSF2) += mss-timer.o common-obj-$(CONFIG_RASPI) += bcm2835_systmr.o + +obj-$(CONFIG_AVR_TIMER16) += avr_timer16.o
Signature · 1 line
2.17.2 (Apple Git-113)
Quoted 7 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fully tested yet. However I was able to execute simple code with functions. e.g fibonacci calculation. This series of patches include a non real, sample board.
No fuses support yet. PC is set to 0 at reset.
I have a couple of general remarks, so I am responding to the cover letter, not individual patches.
1) The licenses for Sarah devices differ than the rest - shouldn't all licenses be harmonized?
2) There is an architectural problem with peripherals. It is possible that they evolve over time, so, for example, USART could not be the same for older and newer CPUs (in principle, newer peripheral is expected to be o sort of "superset" of the older). How do you solve that problem? Right now, it may not looks serious to you, but if you don;t think about that right now, from the outset, soon the code will become so entangled, ti woudl be almost very difficult to fix it. Please think about that, how would you solve it, is there a way to pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?
Quoted 5 lines
Following are examples of possible usages, assuming program.elf is compiled for AVR cpu
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnet window runqemu-system-avr -kernel program.elf -serial tcp::5678,server,nowait -nographic
Quoted 9 lines
run
telent localhost 5678
3. Continious non interrupted execution with serial output into stdout runqemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -S
runavr-gdb program.elfand then within GDB shelltarget remote :1234
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asm
Thank you so much for these examples!
Aleksandar
Quoted 349 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding or translation
2. CPU features which allow define the following 8bit AVR coresavr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows to execute simple code
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of0x000000ff) for faster address manipulaton
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM
11.fixing MULS/MULSU not to modify inputs' content
12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHchanges since v4
1. target is now defined as big endian in order to optimize push_ret/pop_ret
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_cpu_gdb_write_register
2. configure the target as little endian because otherwise GDB does not work
3. fixing and testing gen_push_ret/gen_pop_retchanges since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supported yet registers only
3.MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built for all
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memorychanges since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed to any device
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try to access registerschanges since v10
1. movig back fullwr & fullrd into the commit where outb and inb were introduced
2. changingtlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codechanges since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions 12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thing exists)
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrischanges since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers are accessible by LD/ST only. CPU has no interest in themchanges since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixes a couple of bugs wrt single stepping by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instruction translation - CPU main translation funcions' commitchanges since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicchanges since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU" into three patches 1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tename1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to MAINTAINERS 7. use git commit sha1 instead of `master`` in avocado testMichael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileSarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer peripheralsconfigure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)
Re: [PATCH v37 01/17] target/avr: Add outward facing interfaces and core CPU logic
In reply to Michael RolnikHi Michael,
Quoted 157 lines · Michael Rolnik
This includes:
- CPU data structures
- object model classes and functions
- migration functions
- GDB hooksCo-developed-by Michael Rolnik <[email protected]>Co-developed-by Sarah Harris <[email protected]>Signed-off-by Michael Rolnik <[email protected]>Signed-off-by Sarah Harris <[email protected]>Signed-off-by Michael Rolnik <[email protected]>Acked-by Igor Mammedov <[email protected]>--- target/avr/cpu-param.h | 37 +++ target/avr/cpu-qom.h | 54 ++++ target/avr/cpu.h | 253 ++++++++++++++++++ target/avr/cpu.c | 576 +++++++++++++++++++++++++++++++++++++++++ target/avr/gdbstub.c | 85 ++++++ target/avr/machine.c | 121 +++++++++ gdb-xml/avr-cpu.xml | 49 ++++ 7 files changed, 1175 insertions(+)create mode 100644
target/avr/cpu-param.h
create mode 100644target/avr/cpu-qom.h
create mode 100644target/avr/cpu.h
create mode 100644target/avr/cpu.c
create mode 100644target/avr/gdbstub.c
create mode 100644target/avr/machine.c
create mode 100644gdb-xml/avr-cpu.xmltarget/avr/cpu-param.h new file +37
-0@@ -0,0 +1,37 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +#ifndef AVR_CPU_PARAM_H +#define AVR_CPU_PARAM_H 1 + +#define TARGET_LONG_BITS 32 +/* +* TARGET_PAGE_BITS cannot be more than 8 bits because +* 1. all IO registers occupy [0x0000 .. 0x00ff] address range, and they +* should be implemented as a device and not memory +* 2. SRAM starts at the address 0x0100 +*/ +#define TARGET_PAGE_BITS 8 +#define TARGET_PHYS_ADDR_SPACE_BITS 24 +#define TARGET_VIRT_ADDR_SPACE_BITS 24 +#define NB_MMU_MODES 2 + + +#endif target/avr/cpu-qom.h new file +54
-0@@ -0,0 +1,54 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +#ifndef QEMU_AVR_QOM_H +#define QEMU_AVR_QOM_H + +#include "hw/core/cpu.h" + +#define TYPE_AVR_CPU "avr-cpu" + +#define AVR_CPU_CLASS (klass) \ + OBJECT_CLASS_CHECK(AVRCPUClass, (klass), TYPE_AVR_CPU) +#define AVR_CPU (obj) \ + OBJECT_CHECK(AVRCPU, (obj), TYPE_AVR_CPU) +#define AVR_CPU_GET_CLASS (obj) \ + OBJECT_GET_CLASS(AVRCPUClass, (obj), TYPE_AVR_CPU) + +/** +* AVRCPUClass: +* @parent_realize: The parent class' realize handler. +* @parent_reset: The parent class' reset handler. +* @vr: Version Register value. +* +* A AVR CPU model. +*/ +typedef struct AVRCPUClass { +/*< private >*/ +CPUClass parent_class; +/*< public >*/ +DeviceRealize parent_realize; +void (*parent_reset)(CPUState *cpu); +}AVRCPUClass ; + +typedef struct AVRCPU AVRCPU ; + + +#endif /* !defined (QEMU_AVR_CPU_QOM_H) */ target/avr/cpu.h new file +18
-0@@ -0,0 +1,253 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html>
target/avr/cpu.h line 28
Quoted 10 lines
target/avr/cpu.h new file +10
-0+*/ + +#ifndef QEMU_AVR_CPU_H +#define QEMU_AVR_CPU_H + +#include "cpu-qom.h" +#include "exec/cpu-defs.h" + +#define TCG_GUEST_DEFAULT_MO 0 +
Please add:
#define AVR_CPU_TYPE_SUFFIX "-" TYPE_AVR_CPU #define AVR_CPU_TYPE_NAME(name) (name AVR_CPU_TYPE_SUFFIX)
Quoted 9 lines
target/avr/cpu.h new file +9
-0+#define CPU_RESOLVING_TYPE TYPE_AVR_CPU + +/* +* AVR has two memory spaces, data & code. +* e.g. both have 0 address +* ST/LD instructions access data space +* LPM/SPM and instruction fetching access code memory space +*/ +#define MMU_CODE_IDX 0
target/avr/cpu.h line 47
Quoted 10 lines
target/avr/cpu.h new file +10
-0+#define MMU_DATA_IDX 1 + +#define EXCP_RESET 1 +#define EXCP_INT (n) (EXCP_RESET + (n) + 1) + +/* Number of CPU registers */ +#define NUMBER_OF_CPU_REGISTERS 32 +/* Number of IO registers accessible by ld/st/in/out */ +#define NUMBER_OF_IO_REGISTERS 64 +
Is the following block <...
Quoted 10 lines
target/avr/cpu.h new file +10
-0+/* +* Offsets of AVR memory regions in host memory space. +* +* This is needed because the AVR has separate code and data address +* spaces that both have start from zero but have to go somewhere in +* host memory. +* +* It's also useful to know where some things are, like the IO registers. +*/ +/* Flash program memory */
target/avr/cpu.h line 67
Quoted 10 lines
target/avr/cpu.h new file +10
-0+#define OFFSET_CODE 0x00000000 +/* CPU registers, IO registers, and SRAM */ +#define OFFSET_DATA 0x00800000 +/* CPU registers specifically, these are mapped at the start of data */ +#define OFFSET_CPU_REGISTERS OFFSET_DATA +/* +* IO registers, including status register, stack pointer, and memory +* mapped peripherals, mapped just after CPU registers +*/ +#define OFFSET_IO_REGISTERS (OFFSET_DATA + NUMBER_OF_CPU_REGISTERS)
...> really CPU specific? This doesn't seem used by the CPU (core) code, but by the MCU devices. Maybe we can extract them into "target/avr/addrspace.h".
Can be done later.
[...]
Quoted 34 lines
target/avr/cpu.h new file +34
-0+static void avr_cpu_realizefn (DeviceState *dev,Error **errp) +{ +CPUState *cs =CPU (dev); +AVRCPUClass *mcc =AVR_CPU_GET_CLASS (dev); +Error *local_err =NULL ; + +cpu_exec_realizefn (cs, &local_err); +if (local_err !=NULL ) { +error_propagate (errp, local_err); +return ; + } +qemu_init_vcpu (cs); +cpu_reset (cs); + + mcc->parent_realize (dev, errp); +} + +static void avr_cpu_set_int (void *opaque,int irq,int level) +{ +AVRCPU *cpu = opaque; +CPUAVRState *env = &cpu->env; +CPUState *cs =CPU (cpu); + +uint64_t mask = (1ull << irq); +if (level) { + env->intsrc |= mask; +cpu_interrupt (cs,CPU_INTERRUPT_HARD ); + }else { + env->intsrc &= ~mask; +if (env->intsrc ==0 ) { +cpu_reset_interrupt (cs,CPU_INTERRUPT_HARD ); + } + } +}
target/avr/cpu.h line 111
Quoted 10 lines
target/avr/cpu.h new file +10
-0+ +static void avr_cpu_initfn (Object *obj) +{ +AVRCPU *cpu =AVR_CPU (obj); + +cpu_set_cpustate_pointers (cpu); + +#ifndef CONFIG_USER_ONLY +/* Set the number of interrupts supported by the CPU. */ +qdev_init_gpio_in (DEVICE (cpu), avr_cpu_set_int,57 );
Deeply buried magic number...
I'm trying to understand the following comment in sample.c:
/*
* These IRQ numbers don't match the datasheet because we're
counting from
* zero and not including reset.
*/
RESET is IRQ#1. IIUC qdev_get_gpio_in(cpu, 0) will return IRQ#2? Then qdev_get_gpio_in(cpu, 55) will return IRQ#57?
This is related to:
#define EXCP_RESET 1
#define EXCP_INT(n) (EXCP_RESET + (n) + 1)
But then you use the hardware numbering:
/* Interrupt numbers used by peripherals */
#define USART_RXC_IRQ 24
#define USART_DRE_IRQ 25
#define USART_TXC_IRQ 26
So for USART3_TX=57, have we qdev_get_gpio_in(cpu, 57) out of bound?
Trying so trigger an abort:
#1 0x00007f2ed58e1895 in abort () at /lib64/libc.so.6 #2 0x00007f2ed58e1769 in _nl_load_domain.cold () at /lib64/libc.so.6 #3 0x00007f2ed58ef566 in annobin_assert.c_end () at /lib64/libc.so.6 #4 0x0000561fbbb37417 in qdev_get_gpio_in_named (dev=0x561fbd8b19d0, name=0x0, n=57) at hw/core/qdev.c:478 #5 0x0000561fbbb37454 in qdev_get_gpio_in (dev=0x561fbd8b19d0, n=57) at hw/core/qdev.c:484
[This has to be fixed before merging]
Also, what about other MCUs with more than 57 lines? Can we use some maximum value instead? Maybe 64?
Else we need to add a property and move that into realize(), because we won't know the value at init() time.
Please use a #define for this value.
target/avr/cpu.h line 113
Quoted 2 lines
target/avr/cpu.h new file +2
-0+#endif +}
[...]
Thanks,
Phil.
Re: [PATCH v37 02/17] target/avr: Add instruction helpers
In reply to Michael RolnikQuoted 69 lines · Michael Rolnik
Stubs for unimplemented instructions and helpers for instructions that need to interact with QEMU. SPM and WDR are unimplemented because they require emulation of complex peripherals.
The implementation of SLEEP is very limited due to the lack of peripherals to generate wake interrupts. Memory access instructions are implemented here because some address ranges actually refer to CPU registers.Signed-off-by Michael Rolnik <[email protected]>--- target/avr/helper.h | 29 ++++ target/avr/helper.c | 354 ++++++++++++++++++++++++++++++++++++++++++++ 2 files changed, 383 insertions(+)create mode 100644
target/avr/helper.h
create mode 100644target/avr/helper.ctarget/avr/helper.h new file +29
-0@@ -0,0 +1,29 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +DEF_HELPER_1(wdr, void, env) +DEF_HELPER_1(debug, void, env) +DEF_HELPER_1(break, void, env) +DEF_HELPER_1(sleep, void, env) +DEF_HELPER_1(unsupported, void, env) +DEF_HELPER_3(outb, void, env, i32, i32) +DEF_HELPER_2(inb, tl, env, i32) +DEF_HELPER_3(fullwr, void, env, i32, i32) +DEF_HELPER_2 (fullrd, tl, env, i32)target/avr/helper.c new file +14
-0@@ -0,0 +1,354 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details.
target/avr/helper.c line 24
Quoted 10 lines
target/avr/helper.c new file +10
-0+* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +#include "qemu/osdep.h" + +#include "cpu.h" +#include "hw/irq.h"
Nitpicking again, the previous include doesn't seem used.
Quoted 73 lines
target/avr/helper.c new file +73
-0+#include "hw/sysbus.h" +#include "sysemu/sysemu.h" +#include "exec/exec-all.h" +#include "exec/cpu_ldst.h" +#include "exec/helper-proto.h" +#include "exec/ioport.h" +#include "qemu/host-utils.h" +#include "qemu/error-report.h" + +bool avr_cpu_exec_interrupt (CPUState *cs,int interrupt_request) +{ +bool ret = false; +CPUClass *cc =CPU_GET_CLASS (cs); +AVRCPU *cpu =AVR_CPU (cs); +CPUAVRState *env = &cpu->env; + +if (interrupt_request &CPU_INTERRUPT_RESET ) { +if (cpu_interrupts_enabled (env)) { + cs->exception_index =EXCP_RESET ; + cc->do_interrupt (cs); + + cs->interrupt_request &= ~CPU_INTERRUPT_RESET ; + + ret = true; + } + } +if (interrupt_request &CPU_INTERRUPT_HARD ) { +if (cpu_interrupts_enabled (env) && env->intsrc !=0 ) { +int index =ctz32 (env->intsrc); + cs->exception_index =EXCP_INT (index); + cc->do_interrupt (cs); + + env->intsrc &= env->intsrc -1 ;/* clear the interrupt */ + cs->interrupt_request &= ~CPU_INTERRUPT_HARD ; + + ret = true; + } + } +return ret; +} + +void avr_cpu_do_interrupt (CPUState *cs) +{ +AVRCPU *cpu =AVR_CPU (cs); +CPUAVRState *env = &cpu->env; + +uint32_t ret = env->pc_w; +int vector =0 ; +int size =avr_feature (env,AVR_FEATURE_JMP_CALL ) ?2 :1 ; +int base =0 ; + +if (cs->exception_index ==EXCP_RESET ) { + vector =0 ; + }else if (env->intsrc !=0 ) { + vector =ctz32 (env->intsrc) +1 ; + } + +if (avr_feature (env,AVR_FEATURE_3_BYTE_PC )) { +cpu_stb_data (env, env->sp--, (ret &0x0000ff )); +cpu_stb_data (env, env->sp--, (ret &0x00ff00 ) >>8 ); +cpu_stb_data (env, env->sp--, (ret &0xff0000 ) >>16 ); + }else if (avr_feature (env,AVR_FEATURE_2_BYTE_PC )) { +cpu_stb_data (env, env->sp--, (ret &0x0000ff )); +cpu_stb_data (env, env->sp--, (ret &0x00ff00 ) >>8 ); + }else { +cpu_stb_data (env, env->sp--, (ret &0x0000ff )); + } + + env->pc_w = base + vector * size; + env->sregI =0 ;/* clear Global Interrupt Flag */ + + cs->exception_index =-1 ; +}
[...]
Re: [PATCH v37 01/17] target/avr: Add outward facing interfaces and core CPU logic
In reply to Philippe Mathieu-DaudéQuoted 208 lines · Philippe Mathieu-Daudé
Hi Michael,
Quoted 53 lines · Michael Rolnik
This includes:
- CPU data structures
- object model classes and functions
- migration functions
- GDB hooksCo-developed-by Michael Rolnik <[email protected]>Co-developed-by Sarah Harris <[email protected]>Signed-off-by Michael Rolnik <[email protected]>Signed-off-by Sarah Harris <[email protected]>Signed-off-by Michael Rolnik <[email protected]>Acked-by Igor Mammedov <[email protected]>--- target/avr/cpu-param.h | 37 +++ target/avr/cpu-qom.h | 54 ++++ target/avr/cpu.h | 253 ++++++++++++++++++ target/avr/cpu.c | 576 +++++++++++++++++++++++++++++++++++++++++ target/avr/gdbstub.c | 85 ++++++ target/avr/machine.c | 121 +++++++++ gdb-xml/avr-cpu.xml | 49 ++++ 7 files changed, 1175 insertions(+)create mode 100644
target/avr/cpu-param.h
create mode 100644target/avr/cpu-qom.h
create mode 100644target/avr/cpu.h
create mode 100644target/avr/cpu.c
create mode 100644target/avr/gdbstub.c
create mode 100644target/avr/machine.c
create mode 100644gdb-xml/avr-cpu.xmltarget/avr/cpu-param.h new file +17
-0@@ -0,0 +1,37 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see target/avr/cpu-param.h line 27
Quoted 10 lines
target/avr/cpu-param.h new file +10
-0+* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +#ifndef AVR_CPU_PARAM_H +#define AVR_CPU_PARAM_H 1 + +#define TARGET_LONG_BITS 32 +/* +* TARGET_PAGE_BITS cannot be more than 8 bits because +* 1. all IO registers occupy [0x0000 .. 0x00ff] address range, and they
Quoted 94 lines
target/avr/cpu-param.h new file +10
-0+* should be implemented as a device and not memory +* 2. SRAM starts at the address 0x0100 +*/ +#define TARGET_PAGE_BITS 8 +#define TARGET_PHYS_ADDR_SPACE_BITS 24 +#define TARGET_VIRT_ADDR_SPACE_BITS 24 +#define NB_MMU_MODES 2 + + +#endif target/avr/cpu-qom.h new file +54
-0@@ -0,0 +1,54 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +#ifndef QEMU_AVR_QOM_H +#define QEMU_AVR_QOM_H + +#include "hw/core/cpu.h" + +#define TYPE_AVR_CPU "avr-cpu" + +#define AVR_CPU_CLASS (klass) \ + OBJECT_CLASS_CHECK(AVRCPUClass, (klass), TYPE_AVR_CPU) +#define AVR_CPU (obj) \ + OBJECT_CHECK(AVRCPU, (obj), TYPE_AVR_CPU) +#define AVR_CPU_GET_CLASS (obj) \ + OBJECT_GET_CLASS(AVRCPUClass, (obj), TYPE_AVR_CPU) + +/** +* AVRCPUClass: +* @parent_realize: The parent class' realize handler. +* @parent_reset: The parent class' reset handler. +* @vr: Version Register value. +* +* A AVR CPU model. +*/ +typedef struct AVRCPUClass { +/*< private >*/ +CPUClass parent_class; +/*< public >*/ +DeviceRealize parent_realize; +void (*parent_reset)(CPUState *cpu); +}AVRCPUClass ; + +typedef struct AVRCPU AVRCPU ; + + +#endif /* !defined (QEMU_AVR_CPU_QOM_H) */ target/avr/cpu.h new file +18
-0@@ -0,0 +1,253 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> target/avr/cpu.h line 28
Quoted 10 lines
target/avr/cpu.h new file +10
-0+*/ + +#ifndef QEMU_AVR_CPU_H +#define QEMU_AVR_CPU_H + +#include "cpu-qom.h" +#include "exec/cpu-defs.h" + +#define TCG_GUEST_DEFAULT_MO 0 +Please add:
#define AVR_CPU_TYPE_SUFFIX "-" TYPE_AVR_CPU #define AVR_CPU_TYPE_NAME(name) (name AVR_CPU_TYPE_SUFFIX)Quoted 9 lines
target/avr/cpu.h new file +9
-0+#define CPU_RESOLVING_TYPE TYPE_AVR_CPU + +/* +* AVR has two memory spaces, data & code. +* e.g. both have 0 address +* ST/LD instructions access data space +* LPM/SPM and instruction fetching access code memory space +*/ +#define MMU_CODE_IDX 0target/avr/cpu.h line 47
Quoted 10 lines
target/avr/cpu.h new file +10
-0+#define MMU_DATA_IDX 1 + +#define EXCP_RESET 1 +#define EXCP_INT (n) (EXCP_RESET + (n) + 1) + +/* Number of CPU registers */ +#define NUMBER_OF_CPU_REGISTERS 32 +/* Number of IO registers accessible by ld/st/in/out */ +#define NUMBER_OF_IO_REGISTERS 64 +Is the following block <...
target/avr/cpu.h
Quoted 8 lines
target/avr/cpu.h new file +8
-0+/* +* Offsets of AVR memory regions in host memory space. +* +* This is needed because the AVR has separate code and data address +* spaces that both have start from zero but have to go somewhere in +* host memory. +* +* It's also useful to know where some things are, like the IO registers.
Quoted 17 lines
target/avr/cpu.h line 67
Quoted 12 lines
target/avr/cpu.h new file +12
-0+*/ +/* Flash program memory */ +#define OFFSET_CODE 0x00000000 +/* CPU registers, IO registers, and SRAM */ +#define OFFSET_DATA 0x00800000 +/* CPU registers specifically, these are mapped at the start of data */ +#define OFFSET_CPU_REGISTERS OFFSET_DATA +/* +* IO registers, including status register, stack pointer, and memory +* mapped peripherals, mapped just after CPU registers +*/ +#define OFFSET_IO_REGISTERS (OFFSET_DATA + NUMBER_OF_CPU_REGISTERS)...> really CPU specific? This doesn't seem used by the CPU (core) code, but by the MCU devices. Maybe we can extract them into "
target/avr/addrspace.h".
These defines are used in the helper.c file
Quoted 105 lines
Can be done later.
[...]
Quoted 34 lines
target/avr/cpu.h new file +34
-0+static void avr_cpu_realizefn (DeviceState *dev,Error **errp) +{ +CPUState *cs =CPU (dev); +AVRCPUClass *mcc =AVR_CPU_GET_CLASS (dev); +Error *local_err =NULL ; + +cpu_exec_realizefn (cs, &local_err); +if (local_err !=NULL ) { +error_propagate (errp, local_err); +return ; + } +qemu_init_vcpu (cs); +cpu_reset (cs); + + mcc->parent_realize (dev, errp); +} + +static void avr_cpu_set_int (void *opaque,int irq,int level) +{ +AVRCPU *cpu = opaque; +CPUAVRState *env = &cpu->env; +CPUState *cs =CPU (cpu); + +uint64_t mask = (1ull << irq); +if (level) { + env->intsrc |= mask; +cpu_interrupt (cs,CPU_INTERRUPT_HARD ); + }else { + env->intsrc &= ~mask; +if (env->intsrc ==0 ) { +cpu_reset_interrupt (cs,CPU_INTERRUPT_HARD ); + } + } +}target/avr/cpu.h line 111
Quoted 10 lines
target/avr/cpu.h new file +10
-0+ +static void avr_cpu_initfn (Object *obj) +{ +AVRCPU *cpu =AVR_CPU (obj); + +cpu_set_cpustate_pointers (cpu); + +#ifndef CONFIG_USER_ONLY +/* Set the number of interrupts supported by the CPU. */ +qdev_init_gpio_in (DEVICE (cpu), avr_cpu_set_int,57 );Deeply buried magic number...
I'm trying to understand the following comment in
sample.c:/*
* These IRQ numbers don't match the datasheet because we're
counting from
* zero and not including reset.
*/RESET is IRQ#1. IIUC
qdev_get_gpio_in(cpu, 0)will return IRQ#2? Thenqdev_get_gpio_in(cpu, 55)will return IRQ#57?This is related to:
#define EXCP_RESET 1 #define EXCP_INT(n) (EXCP_RESET + (n) + 1)But then you use the hardware numbering:
/* Interrupt numbers used by peripherals */
#defineUSART_RXC_IRQ24
#defineUSART_DRE_IRQ25
#defineUSART_TXC_IRQ26So for USART3_TX=57, have we
qdev_get_gpio_in(cpu, 57)out of bound?Trying so trigger an abort:
#1 0x00007f2ed58e1895 in abort () at /lib64/libc.so.6 #2 0x00007f2ed58e1769 in _nl_load_domain.cold () at /lib64/libc.so.6 #3 0x00007f2ed58ef566 in annobin_assert.c_end () at /lib64/libc.so.6 #4 0x0000561fbbb37417 in qdev_get_gpio_in_named (dev=0x561fbd8b19d0,name=0x0, n=57) at
hw/core/qdev.c:478
#50x0000561fbbb37454inqdev_get_gpio_in(dev=0x561fbd8b19d0, n=57) athw/core/qdev.c:484[This has to be fixed before merging]
Also, what about other MCUs with more than 57 lines? Can we use some maximum value instead? Maybe 64?
Else we need to add a property and move that into
realize(), because we won't know the value atinit()time.Please use a #define for this value.
target/avr/cpu.h line 113
Quoted 2 lines
target/avr/cpu.h new file +2
-0+#endif +}[...]
Thanks,
Phil.
Signature · 2 lines
Best Regards, Michael Rolnik
Quoted 11 lines · Aleksandar Markovic
Quoted 3 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fullytested yet.
Quoted 1 line
However I was able to execute simple code with functions. e.g fibonacci
calculation.
Quoted 3 lines
This series of patches include a non real, sample board. No fuses support yet. PC is set to 0 at reset.
Quoted 5 lines
I have a couple of general remarks, so I am responding to the cover letter, not individual patches.
1) The licenses for Sarah devices differ than the rest - shouldn't all licenses be harmonized?
Sarah,
do you mind if use the same license I use for my code?
Quoted 14 lines
2) There is an architectural problem with peripherals. It is possible that they evolve over time, so, for example, USART could not be the same for older and newer CPUs (in principle, newer peripheral is expected to be o sort of "superset" of the older). How do you solve that problem? Right now, it may not looks serious to you, but if you don;t think about that right now, from the outset, soon the code will become so entangled, ti woudl be almost very difficult to fix it. Please think about that, how would you solve it, is there a way to pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?
Hi Aleksandar,
Please explain. I don't see any problem from CPU's perspective. as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there is sample.
Quoted 391 lines
Quoted 2 lines
Following are examples of possible usages, assuming program.elf is
compiled for AVR cpu
Quoted 3 lines
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnetwindow
Quoted 1 line
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
Quoted 6 lines
run
telent localhost 5678
3. Continious non interrupted execution with serial output into stdout runqemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -S
runavr-gdb program.elfand then within GDB shell `target remote:1234`
Quoted 3 lines
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmThank you so much for these examples!
Aleksandar
Quoted 4 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding or translation
2. CPU features which allow define the following 8bit AVR coresQuoted 8 lines
avr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows
to execute simple code
Quoted 9 lines
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of
0x000000ff) for faster address manipulatonQuoted 13 lines
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM 11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHQuoted 2 lines
changes since v4
1. target is now defined as big endian in order to optimizepush_ret/pop_ret
Quoted 14 lines
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)Quoted 6 lines
changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg inavr_cpu_gdb_read_register/avr_cpu_gdb_write_register
Quoted 1 line
2. configure the target as little endian because otherwise GDB does not
work
Quoted 5 lines
3. fixing and testing gen_push_ret/gen_pop_ret
changes since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supported yetregisters only
Quoted 13 lines
3.
MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built for all
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memoryQuoted 5 lines
changes since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed toany device
Quoted 4 lines
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try to accessregisters
Quoted 3 lines
changes since v10
1. movig back fullwr & fullrd into the commit where outb and inb wereintroduced
Quoted 12 lines
2. changing
tlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codeQuoted 12 lines
changes since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
Quoted 8 lines
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thingexists)
Quoted 5 lines
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrisQuoted 8 lines
changes since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers areaccessible
Quoted 30 lines
by LD/ST only. CPU has no interest in them
changes since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
Quoted 8 lines
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixes acouple of bugs wrt single stepping
Quoted 40 lines
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'Quoted 8 lines
changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
Quoted 8 lines
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicQuoted 7 lines
changes since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU" intothree patches
Quoted 12 lines
1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tenameQuoted 9 lines
1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERSQuoted 95 lines
7. use git commit sha1 instead of `master`` in avocado test
Michael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileSarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer peripheralsconfigure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)
Signature · 2 lines
Best Regards, Michael Rolnik
Quoted 99 lines · Michael Rolnik
I don't see why you say that the peripherals are inside the chip, there is CPU within target/avr directory and then there are some peripherals in hw directory, CPU does not depend on them. what am I missing?
On Thu, Nov 28, 2019 at 3:22 PM Aleksandar Markovic <[email protected] <mailto:[email protected]>> wrote:
On Thursday, November 28, 2019, Michael Rolnik <[email protected] <mailto:[email protected]>> wrote:
On Wed, Nov 27, 2019 at 11:06 PM Aleksandar Markovic <[email protected] <mailto:[email protected]>> wrote:
On Wed, Nov 27, 2019 at 6:53 PM Michael Rolnik <[email protected] <mailto:[email protected]>> wrote:
>
> This series of patches adds 8bit AVR cores to QEMU. > All instruction, except BREAK/DES/SPM/SPMX, are
implemented. Not fully tested yet.
> However I was able to execute simple code with functions.
e.g fibonacci calculation.
> This series of patches include a non real, sample board. > No fuses support yet. PC is set to 0 at reset. >I have a couple of general remarks, so I am responding to the cover
letter, not individual patches.1) The licenses for Sarah devices differ than the rest - shouldn't all
licenses be harmonized?Sarah,
do you mind if use the same license I use for my code?2) There is an architectural problem with peripherals. It is possible
that they evolve over time, so, for example, USART could not be the
same for older and newer CPUs (in principle, newer peripheral is expected to be o sort of "superset" of the older). How do you solve
that problem? Right now, it may not looks serious to you, but if you
don;t think about that right now, from the outset, soon the code will
become so entangled, ti woudl be almost very difficult to fix it.
Please think about that, how would you solve it, is there a way to
pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the core.
If one models, let's say an external (in the sense, it is a separate chip) ADC (analog-to-digital converter), one looks at specs, implement what is resonable possible in QEMU, plug it in in one of machines thst contains it, and that's it. That ADC remains the same, of course, whatever the surrounding system is.
In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers is very large, and both cores and peripherals evolved.
For cores, you handle differences with all these
AVR_FEATUREmacros, and this seems to be working, no significant objection from my side, and btw that was not an easy task to execute, all admiration from me.But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't have time right now to spell many examples, but I read three different specs, and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if they are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on the core it is attached to?
I hope I was clearer this time.
Aleksandar
Quoted 5 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there is sample.
If I understand Aleksandar correctly, the naming is incorrect because too generic to AVR family, why Sarah only modeled the Atmel implementation.
Renaming devices such hw/char/avr_usart.c -> hw/char/atmel_usart.c (similarly with the macros) would be enough Aleksandar?
I don't see why you say that the peripherals are inside the chip, there is CPU within target/avr directory and then there are some peripherals in hw directory, CPU does not depend on them. what am I missing?
Quoted 496 lines · Aleksandar Markovic
Quoted 41 lines · Michael Rolnik
Quoted 12 lines · Aleksandar Markovic
Quoted 3 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fullytested yet.
Quoted 1 line
However I was able to execute simple code with functions. e.g
fibonacci calculation.
Quoted 3 lines
This series of patches include a non real, sample board. No fuses support yet. PC is set to 0 at reset.
Quoted 5 lines
I have a couple of general remarks, so I am responding to the cover letter, not individual patches.
1) The licenses for Sarah devices differ than the rest - shouldn't all licenses be harmonized?
Sarah,
do you mind if use the same license I use for my code?Quoted 14 lines
2) There is an architectural problem with peripherals. It is possible that they evolve over time, so, for example, USART could not be the same for older and newer CPUs (in principle, newer peripheral is expected to be o sort of "superset" of the older). How do you solve that problem? Right now, it may not looks serious to you, but if you don;t think about that right now, from the outset, soon the code will become so entangled, ti woudl be almost very difficult to fix it. Please think about that, how would you solve it, is there a way to pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?
Quoted 5 lines
Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the core.
If one models, let's say an external (in the sense, it is a separate chip) ADC (analog-to-digital converter), one looks at specs, implement what is resonable possible in QEMU, plug it in in one of machines thst contains it, and that's it. That ADC remains the same, of course, whatever the surrounding system is.
In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers is very large, and both cores and peripherals evolved.
For cores, you handle differences with all these
AVR_FEATUREmacros, and this seems to be working, no significant objection from my side, and btw that was not an easy task to execute, all admiration from me.But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't have time right now to spell many examples, but I read three different specs, and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if they are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on the core it is attached to?
I hope I was clearer this time.
Aleksandar
Quoted 410 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there is sample.Quoted 320 lines
Quoted 2 lines
Following are examples of possible usages, assuming program.elf is
compiled for AVR cpu
Quoted 3 lines
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output intotelnet window
Quoted 1 line
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
Quoted 6 lines
run
telent localhost 5678
3. Continious non interrupted execution with serial output into stdout runqemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -S
runavr-gdb program.elfand then within GDB shell `target remote:1234`
Quoted 3 lines
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmThank you so much for these examples!
Aleksandar
Quoted 3 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding ortranslation
Quoted 9 lines
2. CPU features which allow define the following 8bit AVR cores
avr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which
allows to execute simple code
Quoted 9 lines
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of
0x000000ff) for faster address manipulatonQuoted 13 lines
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM 11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHQuoted 2 lines
changes since v4
1. target is now defined as big endian in order to optimizepush_ret/pop_ret
Quoted 14 lines
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)Quoted 6 lines
changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg inavr_cpu_gdb_read_register/avr_cpu_gdb_write_register
Quoted 1 line
2. configure the target as little endian because otherwise GDB does
not work
Quoted 5 lines
3. fixing and testing gen_push_ret/gen_pop_ret
changes since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supportedyet registers only
Quoted 4 lines
3.
MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built forall
Quoted 9 lines
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memoryQuoted 5 lines
changes since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed toany device
Quoted 4 lines
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try toaccess registers
Quoted 3 lines
changes since v10
1. movig back fullwr & fullrd into the commit where outb and inb wereintroduced
Quoted 12 lines
2. changing
tlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codeQuoted 12 lines
changes since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
Quoted 8 lines
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thingexists)
Quoted 5 lines
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrisQuoted 8 lines
changes since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers areaccessible
Quoted 30 lines
by LD/ST only. CPU has no interest in them
changes since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
Quoted 8 lines
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixesa couple of bugs wrt single stepping
Quoted 40 lines
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'Quoted 8 lines
changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
Quoted 8 lines
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicQuoted 7 lines
changes since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU"into three patches
Quoted 12 lines
1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tenameQuoted 9 lines
1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERSQuoted 23 lines
7. use git commit sha1 instead of `master`` in avocado test
Michael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileQuoted 74 lines
Quoted 2 lines
Sarah Harris (1):
target/avr: Add limited support for USART and 16 bit timerperipherals
Quoted 70 lines
configure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)--
Best Regards,
Michael Rolnik
Signature · 2 lines
Best Regards, Michael Rolnik
Quoted 6 lines · Michael Rolnik
I don't see why you say that the peripherals are inside the chip, there is CPU within target/avr directory and then there are some peripherals in hw directory, CPU does not depend on them. what am I missing?
Quoted 2 lines
I meant these peripherals are physically inside the chip together with the core.
And USART in a micricontroler from 2010 is different than USART from one from 2018.
Quoted 507 lines
Quoted 79 lines · Aleksandar Markovic
Quoted 41 lines · Michael Rolnik
Quoted 12 lines · Aleksandar Markovic
Quoted 3 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU. All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not
fully tested yet.
Quoted 1 line
However I was able to execute simple code with functions. e.g
fibonacci calculation.
Quoted 3 lines
This series of patches include a non real, sample board. No fuses support yet. PC is set to 0 at reset.
Quoted 5 lines
I have a couple of general remarks, so I am responding to the cover letter, not individual patches.
1) The licenses for Sarah devices differ than the rest - shouldn't all licenses be harmonized?
Sarah,
do you mind if use the same license I use for my code?Quoted 14 lines
2) There is an architectural problem with peripherals. It is possible that they evolve over time, so, for example, USART could not be the same for older and newer CPUs (in principle, newer peripheral is expected to be o sort of "superset" of the older). How do you solve that problem? Right now, it may not looks serious to you, but if you don;t think about that right now, from the outset, soon the code will become so entangled, ti woudl be almost very difficult to fix it. Please think about that, how would you solve it, is there a way to pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?
Quoted 5 lines
Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the core.
If one models, let's say an external (in the sense, it is a separate chip) ADC (analog-to-digital converter), one looks at specs, implement what is resonable possible in QEMU, plug it in in one of machines thst contains it, and that's it. That ADC remains the same, of course, whatever the surrounding system is.
In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers is very large, and both cores and peripherals evolved.
For cores, you handle differences with all these
AVR_FEATUREmacros, and this seems to be working, no significant objection from my side, and btw that was not an easy task to execute, all admiration from me.But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't have time right now to spell many examples, but I read three different specs, and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if they are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on the core it is attached to?
Quoted 420 lines
I hope I was clearer this time.
Aleksandar
Quoted 413 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there is sample.Quoted 349 lines
Quoted 2 lines
Following are examples of possible usages, assuming program.elf is
compiled for AVR cpu
Quoted 3 lines
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output intotelnet window
Quoted 1 line
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
Quoted 2 lines
run
telent localhost 5678
3. Continious non interrupted execution with serial output intostdout
Quoted 4 lines
run
qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -Srunavr-gdb program.elfand then within GDB shell `targetremote :1234`
Quoted 3 lines
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmThank you so much for these examples!
Aleksandar
Quoted 3 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding ortranslation
Quoted 9 lines
2. CPU features which allow define the following 8bit AVR cores
avr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which
allows to execute simple code
Quoted 9 lines
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of
0x000000ff) for faster address manipulatonQuoted 13 lines
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM 11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHQuoted 2 lines
changes since v4
1. target is now defined as big endian in order to optimizepush_ret/pop_ret
Quoted 14 lines
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)Quoted 6 lines
changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_
cpu_gdb_write_registerQuoted 1 line
2. configure the target as little endian because otherwise GDB does
not work
Quoted 5 lines
3. fixing and testing gen_push_ret/gen_pop_ret
changes since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supportedyet registers only
Quoted 4 lines
3.
MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built forall
Quoted 9 lines
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memoryQuoted 5 lines
changes since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed toany device
Quoted 4 lines
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try toaccess registers
Quoted 3 lines
changes since v10
1. movig back fullwr & fullrd into the commit where outb and inb wereintroduced
Quoted 12 lines
2. changing
tlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codeQuoted 12 lines
changes since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
Quoted 8 lines
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thingexists)
Quoted 8 lines
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrischanges since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,
write(helper_outb)Quoted 5 lines
3. implementing CBIC/SBIC as
read(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers areaccessible
Quoted 30 lines
by LD/ST only. CPU has no interest in them
changes since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
Quoted 8 lines
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, andfixes a couple of bugs wrt single stepping
Quoted 40 lines
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'Quoted 8 lines
changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
Quoted 8 lines
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicQuoted 7 lines
changes since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU"into three patches
Quoted 12 lines
1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tenameQuoted 9 lines
1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERSQuoted 23 lines
7. use git commit sha1 instead of `master`` in avocado test
Michael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileQuoted 2 lines
Sarah Harris (1):
target/avr: Add limited support for USART and 16 bit timerperipherals
Quoted 16 lines
configure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++Quoted 8 lines
hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052Quoted 48 lines
++++++++++++++++++++++++++++++
Quoted 46 lines
tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)--
Best Regards,
Michael Rolnik--
Best Regards,
Michael Rolnik
Quoted 40 lines · Michael Rolnik
Quoted 11 lines · Aleksandar Markovic
Quoted 3 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fullytested yet.
Quoted 1 line
However I was able to execute simple code with functions. e.g fibonacci
calculation.
Quoted 3 lines
This series of patches include a non real, sample board. No fuses support yet. PC is set to 0 at reset.
Quoted 5 lines
I have a couple of general remarks, so I am responding to the cover letter, not individual patches.
1) The licenses for Sarah devices differ than the rest - shouldn't all licenses be harmonized?
Sarah,
do you mind if use the same license I use for my code?Quoted 14 lines
2) There is an architectural problem with peripherals. It is possible that they evolve over time, so, for example, USART could not be the same for older and newer CPUs (in principle, newer peripheral is expected to be o sort of "superset" of the older). How do you solve that problem? Right now, it may not looks serious to you, but if you don;t think about that right now, from the outset, soon the code will become so entangled, ti woudl be almost very difficult to fix it. Please think about that, how would you solve it, is there a way to pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?
Quoted 5 lines
Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the core.
If one models, let's say an external (in the sense, it is a separate chip) ADC (analog-to-digital converter), one looks at specs, implement what is resonable possible in QEMU, plug it in in one of machines thst contains it, and that's it. That ADC remains the same, of course, whatever the surrounding system is.
In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers is very large, and both cores and peripherals evolved.
For cores, you handle differences with all these AVR_FEATURE macros, and this seems to be working, no significant objection from my side, and btw that was not an easy task to execute, all admiration from me.
But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't have time right now to spell many examples, but I read three different specs, and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if they are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on the core it is attached to?
I hope I was clearer this time.
Aleksandar
Quoted 409 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there is sample.Quoted 296 lines
Quoted 2 lines
Following are examples of possible usages, assuming program.elf is
compiled for AVR cpu
Quoted 3 lines
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnetwindow
Quoted 1 line
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
Quoted 6 lines
run
telent localhost 5678
3. Continious non interrupted execution with serial output into stdout runqemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -S
runavr-gdb program.elfand then within GDB shell `target remote:1234`
Quoted 3 lines
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmThank you so much for these examples!
Aleksandar
Quoted 3 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding ortranslation
Quoted 9 lines
2. CPU features which allow define the following 8bit AVR cores
avr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows
to execute simple code
Quoted 9 lines
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of
0x000000ff) for faster address manipulatonQuoted 13 lines
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM 11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHQuoted 2 lines
changes since v4
1. target is now defined as big endian in order to optimizepush_ret/pop_ret
Quoted 14 lines
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)Quoted 6 lines
changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_
cpu_gdb_write_registerQuoted 1 line
2. configure the target as little endian because otherwise GDB does not
work
Quoted 5 lines
3. fixing and testing gen_push_ret/gen_pop_ret
changes since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supported yetregisters only
Quoted 4 lines
3.
MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built forall
Quoted 9 lines
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memoryQuoted 5 lines
changes since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed toany device
Quoted 4 lines
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try to accessregisters
Quoted 3 lines
changes since v10
1. movig back fullwr & fullrd into the commit where outb and inb wereintroduced
Quoted 12 lines
2. changing
tlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codeQuoted 12 lines
changes since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
Quoted 8 lines
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thingexists)
Quoted 5 lines
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrisQuoted 8 lines
changes since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers areaccessible
Quoted 30 lines
by LD/ST only. CPU has no interest in them
changes since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
Quoted 8 lines
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixesa couple of bugs wrt single stepping
Quoted 40 lines
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'Quoted 8 lines
changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
Quoted 8 lines
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicQuoted 7 lines
changes since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU" intothree patches
Quoted 12 lines
1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tenameQuoted 9 lines
1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer toQuoted 97 lines
MAINTAINERSQuoted 95 lines
7. use git commit sha1 instead of `master`` in avocado test
Michael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileSarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer peripheralsconfigure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)--
Best Regards,
Michael Rolnik
Quoted 114 lines · Philippe Mathieu-Daudé
Quoted 105 lines · Michael Rolnik
I don't see why you say that the peripherals are inside the chip, there is CPU within target/avr directory and then there are some peripherals in hw directory, CPU does not depend on them. what am I missing?
On Thu, Nov 28, 2019 at 3:22 PM Aleksandar Markovic < [email protected] <mailto:[email protected]>> wrote:
On Thursday, November 28, 2019, Michael Rolnik <[email protected] <mailto:[email protected]>> wrote:
On Wed, Nov 27, 2019 at 11:06 PM Aleksandar Markovic <[email protected] <mailto:[email protected]>> wrote:
On Wed, Nov 27, 2019 at 6:53 PM Michael Rolnik <[email protected] <mailto:[email protected]>> wrote:
>
> This series of patches adds 8bit AVR cores to QEMU. > All instruction, except BREAK/DES/SPM/SPMX, are
implemented. Not fully tested yet.
> However I was able to execute simple code with functions.
e.g fibonacci calculation.
> This series of patches include a non real, sample board. > No fuses support yet. PC is set to 0 at reset. >I have a couple of general remarks, so I am responding to the cover
letter, not individual patches.1) The licenses for Sarah devices differ than the rest - shouldn't all
licenses be harmonized?Sarah,
do you mind if use the same license I use for my code?2) There is an architectural problem with peripherals. It is possible
that they evolve over time, so, for example, USART could not be the
same for older and newer CPUs (in principle, newer peripheral
is
expected to be o sort of "superset" of the older). How do you solve
that problem? Right now, it may not looks serious to you, but if you
don;t think about that right now, from the outset, soon the code will
become so entangled, ti woudl be almost very difficult to fix it.
Please think about that, how would you solve it, is there a way to
pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the
core.If one models, let's say an external (in the sense, it is a separate chip) ADC (analog-to-digital converter), one looks at specs, implement what is resonable possible in QEMU, plug it in in one of machines thst contains it, and that's it. That ADC remains the same, of course, whatever the surrounding system is.
In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers is very large, and both cores and peripherals
evolved.For cores, you handle differences with all these
AVR_FEATUREmacros, and this seems to be working, no significant objection from my side, and btw that was not an easy task to execute, all admiration from me.But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't have time right now to spell many examples, but I read three different specs, and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if they are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on the core it is attached to?
I hope I was clearer this time.
Aleksandar
Quoted 6 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there is sample.
Quoted 7 lines
If I understand Aleksandar correctly, the naming is incorrect because too generic to AVR family, why Sarah only modeled the Atmel implementation.
Renaming devices such
hw/char/avr_usart.c->hw/char/atmel_usart.c(similarly with the macros) would be enough Aleksandar?
Some renaming could help, perhaps not quite like the one above, but my point (which I find hard to believe I can't explain to you) is that peripherals inside the chip evolved over time, as starkly opposed to external peripherals that are set in stone...
I will rename them.
Quoted 133 lines · Aleksandar Markovic
Quoted 115 lines · Philippe Mathieu-Daudé
Quoted 106 lines · Michael Rolnik
I don't see why you say that the peripherals are inside the chip, there is CPU within target/avr directory and then there are some peripherals in hw directory, CPU does not depend on them. what am I missing?
On Thu, Nov 28, 2019 at 3:22 PM Aleksandar Markovic < [email protected] <mailto:[email protected]>> wrote:
On Thursday, November 28, 2019, Michael Rolnik <[email protected] <mailto:[email protected]>> wrote:
On Wed, Nov 27, 2019 at 11:06 PM Aleksandar Markovic <[email protected] <mailto:[email protected]>> wrote:
On Wed, Nov 27, 2019 at 6:53 PM Michael Rolnik <[email protected] <mailto:[email protected]>> wrote:
>
> This series of patches adds 8bit AVR cores to QEMU. > All instruction, except BREAK/DES/SPM/SPMX, are
implemented. Not fully tested yet.
> However I was able to execute simple code with functions.
e.g fibonacci calculation.
> This series of patches include a non real, sample board. > No fuses support yet. PC is set to 0 at reset. >I have a couple of general remarks, so I am responding to the cover
letter, not individual patches.1) The licenses for Sarah devices differ than the rest - shouldn't all
licenses be harmonized?Sarah,
do you mind if use the same license I use for my code?2) There is an architectural problem with peripherals. It is possible
that they evolve over time, so, for example, USART could not be the
same for older and newer CPUs (in principle, newer
peripheral is
expected to be o sort of "superset" of the older). How do you solve
that problem? Right now, it may not looks serious to you, but if you
don;t think about that right now, from the outset, soon the code will
become so entangled, ti woudl be almost very difficult to fix it.
Please think about that, how would you solve it, is there a way to
pass the information on the currently emulated CPU to the
code
covering a peripheral, and provide a different behaviour?Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the
core.If one models, let's say an external (in the sense, it is a separate chip) ADC (analog-to-digital converter), one looks at specs, implement what is resonable possible in QEMU, plug it in in one of machines thst contains it, and that's it. That ADC remains the same, of course, whatever the surrounding system is.
In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers is very large, and both cores and peripherals
evolved.For cores, you handle differences with all these
AVR_FEATUREmacros, and this seems to be working, no significant objection from my side, and btw that was not an easy task to execute, all admiration from me.But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't have time right now to spell many examples, but I read three different specs, and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if they are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on the core it is attached to?
I hope I was clearer this time.
Aleksandar
Quoted 6 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there is sample.Quoted 7 lines
If I understand Aleksandar correctly, the naming is incorrect because too generic to AVR family, why Sarah only modeled the Atmel implementation.
Renaming devices such
hw/char/avr_usart.c->hw/char/atmel_usart.c(similarly with the macros) would be enough Aleksandar?Some renaming could help, perhaps not quite like the one above, but my point (which I find hard to believe I can't explain to you) is that peripherals inside the chip evolved over time, as starkly opposed to external peripherals that are set in stone...
Signature · 2 lines
Best Regards, Michael Rolnik
Quoted 1 line · Michael Rolnik
I will rename them.
Please wait comments from Richard before a version respin.
Quoted 1 line · On Thu, Nov 28, 2019 at 3:41 PM Aleksandar Markovic…
[...]
Quoted 16 lines
If I understand Aleksandar correctly, the naming is incorrect because too generic to AVR family, why Sarah only modeled the Atmel implementation.
Renaming devices such
hw/char/avr_usart.c->hw/char/atmel_usart.c(similarly with the macros) would be enough Aleksandar?Some renaming could help, perhaps not quite like the one above, but my point (which I find hard to believe I can't explain to you) is that peripherals inside the chip evolved over time, as starkly opposed to external peripherals that are set in stone...
Quoted 17 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fully tested yet.
However I was able to execute simple code with functions. e.g fibonacci calculation.
This series of patches include a non real, sample board.
No fuses support yet. PC is set to 0 at reset.Following are examples of possible usages, assuming program.elf is compiled for AVR cpu
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnet window
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
runtelent localhost 5678
Quoted 10 lines
3. Continious non interrupted execution with serial output into stdout run
qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -S
runavr-gdb program.elfand then within GDB shell `target remote
:1234`
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asm
What is still missing is a documentstion patch. You need to browse a bit QEMU online documentation https://qemu.weilnetz.de/doc/qemu-doc.html and find the right place and content for AVR. Source for documentation is within the tree, in one of .texi or .rst files.
Please also split Sarah's patch into three, not only for reviewing purpose, but for the sake of future readers.
Yours,
Aleksandar
Quoted 364 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding or translation
2. CPU features which allow define the following 8bit AVR coresavr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows to execute simple code
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of0x000000ff) for faster address manipulaton
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM
11.fixing MULS/MULSU not to modify inputs' content
12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHchanges since v4
1. target is now defined as big endian in order to optimize push_ret/pop_ret
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_cpu_gdb_write_register
2. configure the target as little endian because otherwise GDB does not work
3. fixing and testing gen_push_ret/gen_pop_retchanges since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supported yet registers only
3.MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built for all
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memorychanges since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed to any device
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try to access registerschanges since v10
1. movig back fullwr & fullrd into the commit where outb and inb were introduced
2. changingtlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codechanges since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions 12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thing exists)
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrischanges since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers are accessible
by LD/ST only. CPU has no interest in themchanges since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixes a
couple of bugs wrt single stepping
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicchanges since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU" into three patches 1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tename1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to MAINTAINERS 7. use git commit sha1 instead of `master`` in avocado testMichael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileSarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer peripheralsconfigure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)
Quoted 6 lines · Philippe Mathieu-Daudé
Quoted 4 lines · Michael Rolnik
I will rename them.
Renaming alone won't solve anything.
Quoted 26 lines
Please wait comments from Richard before a version respin.
On Thu, Nov 28, 2019 at 3:41 PM Aleksandar Markovic <
Quoted 2 lines
[email protected] <mailto:[email protected]>> wrote:
[...]
Quoted 17 lines
If I understand Aleksandar correctly, the naming is incorrect because too generic to AVR family, why Sarah only modeled the Atmel implementation.
Renaming devices such
hw/char/avr_usart.c->hw/char/atmel_usart.c(similarly with the macros) would be enough Aleksandar?Some renaming could help, perhaps not quite like the one above, but my point (which I find hard to believe I can't explain to you) is that peripherals inside the chip evolved over time, as starkly opposed to external peripherals that are set in stone...
Quoted 8 lines · Aleksandar Markovic
Quoted 6 lines · Michael Rolnik
I don't see why you say that the peripherals are inside the chip, there is CPU within target/avr directory and then there are some peripherals in hw directory, CPU does not depend on them. what am I missing?
Quoted 2 lines
Quoted 5 lines
I meant these peripherals are physically inside the chip together with the core.
And USART in a micricontroler from 2010 is different than USART from one from 2018.
Won't these be different chip parts? Or at least revs of the part?
I think broadly the difference between SoC devices is handled by handling versioning in the board models - the board being in this case a CPU core + a bunch of SoC components + the actual board itself.
All the target/cpu stuff needs to deal with is actual architectural revs (c.f. target/arm/cpu[64].c).
Quoted 509 lines
Quoted 509 lines
Quoted 79 lines · Aleksandar Markovic
Quoted 41 lines · Michael Rolnik
Quoted 12 lines · Aleksandar Markovic
Quoted 3 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU. All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not
fully tested yet.
Quoted 1 line
However I was able to execute simple code with functions. e.g
fibonacci calculation.
Quoted 3 lines
This series of patches include a non real, sample board. No fuses support yet. PC is set to 0 at reset.
Quoted 5 lines
I have a couple of general remarks, so I am responding to the cover letter, not individual patches.
1) The licenses for Sarah devices differ than the rest - shouldn't all licenses be harmonized?
Sarah,
do you mind if use the same license I use for my code?Quoted 14 lines
2) There is an architectural problem with peripherals. It is possible that they evolve over time, so, for example, USART could not be the same for older and newer CPUs (in principle, newer peripheral is expected to be o sort of "superset" of the older). How do you solve that problem? Right now, it may not looks serious to you, but if you don;t think about that right now, from the outset, soon the code will become so entangled, ti woudl be almost very difficult to fix it. Please think about that, how would you solve it, is there a way to pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?
Quoted 5 lines
Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the core.
If one models, let's say an external (in the sense, it is a separate chip) ADC (analog-to-digital converter), one looks at specs, implement what is resonable possible in QEMU, plug it in in one of machines thst contains it, and that's it. That ADC remains the same, of course, whatever the surrounding system is.
In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers is very large, and both cores and peripherals evolved.
For cores, you handle differences with all these
AVR_FEATUREmacros, and this seems to be working, no significant objection from my side, and btw that was not an easy task to execute, all admiration from me.But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't have time right now to spell many examples, but I read three different specs, and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if they are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on the core it is attached to?
Quoted 420 lines
I hope I was clearer this time.
Aleksandar
Quoted 413 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there is sample.Quoted 349 lines
Quoted 2 lines
Following are examples of possible usages, assuming program.elf is
compiled for AVR cpu
Quoted 3 lines
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output intotelnet window
Quoted 1 line
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
Quoted 2 lines
run
telent localhost 5678
3. Continious non interrupted execution with serial output intostdout
Quoted 4 lines
run
qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -Srunavr-gdb program.elfand then within GDB shell `targetremote :1234`
Quoted 3 lines
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmThank you so much for these examples!
Aleksandar
Quoted 3 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding ortranslation
Quoted 9 lines
2. CPU features which allow define the following 8bit AVR cores
avr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which
allows to execute simple code
Quoted 9 lines
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of
0x000000ff) for faster address manipulatonQuoted 13 lines
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM 11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHQuoted 2 lines
changes since v4
1. target is now defined as big endian in order to optimizepush_ret/pop_ret
Quoted 14 lines
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)Quoted 6 lines
changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_
cpu_gdb_write_registerQuoted 1 line
2. configure the target as little endian because otherwise GDB does
not work
Quoted 5 lines
3. fixing and testing gen_push_ret/gen_pop_ret
changes since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supportedyet registers only
Quoted 4 lines
3.
MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built forall
Quoted 9 lines
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memoryQuoted 5 lines
changes since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed toany device
Quoted 4 lines
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try toaccess registers
Quoted 3 lines
changes since v10
1. movig back fullwr & fullrd into the commit where outb and inb wereintroduced
Quoted 12 lines
2. changing
tlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codeQuoted 12 lines
changes since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
Quoted 8 lines
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thingexists)
Quoted 8 lines
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrischanges since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,
write(helper_outb)Quoted 5 lines
3. implementing CBIC/SBIC as
read(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers areaccessible
Quoted 30 lines
by LD/ST only. CPU has no interest in them
changes since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
Quoted 8 lines
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, andfixes a couple of bugs wrt single stepping
Quoted 40 lines
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'Quoted 8 lines
changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
Quoted 8 lines
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicQuoted 7 lines
changes since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU"into three patches
Quoted 12 lines
1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tenameQuoted 9 lines
1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERSQuoted 23 lines
7. use git commit sha1 instead of `master`` in avocado test
Michael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileQuoted 2 lines
Sarah Harris (1):
target/avr: Add limited support for USART and 16 bit timerperipherals
Quoted 16 lines
configure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++Quoted 8 lines
hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052Quoted 48 lines
++++++++++++++++++++++++++++++
Quoted 46 lines
tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)--
Best Regards,
Michael Rolnik--
Best Regards,
Michael Rolnik
Signature · 1 line
Alex Bennée
Quoted 21 lines · Alex Bennée
Quoted 1 line · Aleksandar Markovic
On Thursday, November 28, 2019, Michael Rolnik <[email protected]>
Quoted 2 lines · wrote:
Quoted 2 lines
I don't see why you say that the peripherals are inside the chip, there
is
Quoted 1 line
Quoted 1 line
CPU within target/avr directory and then there are some peripherals in
hw
Quoted 5 lines
Quoted 4 lines
directory, CPU does not depend on them. what am I missing?
Quoted 2 lines
I meant these peripherals are physically inside the chip together with
the
Quoted 4 lines
core.
And USART in a micricontroler from 2010 is different than USART from one from 2018.
Quoted 10 lines
Won't these be different chip parts? Or at least revs of the part?
I think broadly the difference between SoC devices is handled by handling versioning in the board models - the board being in this case a CPU core + a bunch of SoC components + the actual board itself.
All the target/cpu stuff needs to deal with is actual architectural revs (c.f. target/arm/cpu[64].c).
This sounds like a very good way of dealing with this.
I don't want to force Michael to implement some of such cases before integration, but just to think about such cases - for future improvements and developments.
Alex, I appreciate your advice, very nice of you!
Aleksandar
Quoted 536 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines · Aleksandar Markovic
On Thursday, November 28, 2019, Michael Rolnik <[email protected]>
Quoted 14 lines · wrote:
Quoted 14 lines
Quoted 14 lines
Quoted 14 lines
Quoted 8 lines · Aleksandar Markovic
Quoted 3 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU. All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not
fully tested yet.
Quoted 1 line
However I was able to execute simple code with functions. e.g
fibonacci calculation.
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
This series of patches include a non real, sample board. No fuses support yet. PC is set to 0 at reset.
I have a couple of general remarks, so I am responding to the cover letter, not individual patches.
1) The licenses for Sarah devices differ than the rest - shouldn't
all
Quoted 20 lines
Quoted 20 lines
Quoted 20 lines
Quoted 20 lines
Quoted 1 line
licenses be harmonized?
Sarah,
do you mind if use the same license I use for my code?Quoted 14 lines
2) There is an architectural problem with peripherals. It is possible that they evolve over time, so, for example, USART could not be the same for older and newer CPUs (in principle, newer peripheral is expected to be o sort of "superset" of the older). How do you solve that problem? Right now, it may not looks serious to you, but if you don;t think about that right now, from the outset, soon the code will become so entangled, ti woudl be almost very difficult to fix it. Please think about that, how would you solve it, is there a way to pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 5 lines
Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the
core.
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
If one models, let's say an external (in the sense, it is a separate chip) ADC (analog-to-digital converter), one looks at specs, implement
what
Quoted 1 line
Quoted 1 line
Quoted 1 line
is resonable possible in QEMU, plug it in in one of machines thst
contains
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
it, and that's it. That ADC remains the same, of course, whatever the surrounding system is.
In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers is very large, and both cores and peripherals evolved.
For cores, you handle differences with all these
AVR_FEATUREmacros,and
Quoted 1 line
Quoted 1 line
Quoted 1 line
this seems to be working, no significant objection from my side, and
btw
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
that was not an easy task to execute, all admiration from me.
But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't
have
Quoted 1 line
Quoted 1 line
Quoted 1 line
time right now to spell many examples, but I read three different
specs,
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if
they
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on the
core
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
it is attached to?
I hope I was clearer this time.
Aleksandar
Quoted 3 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there issample.
Quoted 83 lines
Quoted 83 lines
Quoted 83 lines
Quoted 83 lines
Quoted 83 lines
Quoted 5 lines
Following are examples of possible usages, assuming program.elf is
compiled for AVR cpu
Quoted 3 lines
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output intotelnet window
Quoted 1 line
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
Quoted 2 lines
run
telent localhost 5678
3. Continious non interrupted execution with serial output intostdout
Quoted 4 lines
run
qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -Srunavr-gdb program.elfand then within GDB shell `targetremote :1234`
Quoted 3 lines
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmThank you so much for these examples!
Aleksandar
Quoted 3 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding ortranslation
Quoted 9 lines
2. CPU features which allow define the following 8bit AVR cores
avr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which
allows to execute simple code
Quoted 9 lines
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of
0x000000ff) for faster address manipulatonQuoted 13 lines
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM 11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHQuoted 2 lines
changes since v4
1. target is now defined as big endian in order to optimizepush_ret/pop_ret
Quoted 14 lines
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg inavr_cpu_gdb_read_register/avr_
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
cpu_gdb_write_registerQuoted 1 line
2. configure the target as little endian because otherwise GDB does
not work
Quoted 2 lines
3. fixing and testing gen_push_ret/gen_pop_ret
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
changes since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supportedyet registers only
Quoted 4 lines
3.
MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be builtfor
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
all
Quoted 9 lines
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memoryQuoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
changes since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routedto
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
any device
Quoted 4 lines
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try toaccess registers
Quoted 3 lines
changes since v10
1. movig back fullwr & fullrd into the commit where outb and inbwere
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
introduced
Quoted 12 lines
2. changing
tlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codeQuoted 12 lines
changes since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such athing
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
exists)
Quoted 3 lines
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in mycode
Quoted 99 lines
Quoted 99 lines
Quoted 99 lines
Quoted 99 lines
Quoted 99 lines
Quoted 5 lines
5. moving avr configuration, within configure file, before chris
changes since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,
write(helper_outb)Quoted 5 lines
3. implementing CBIC/SBIC as
read(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers areaccessible
Quoted 30 lines
by LD/ST only. CPU has no interest in them
changes since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
Quoted 8 lines
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, andfixes a couple of bugs wrt single stepping
Quoted 48 lines
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructionQuoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
translation - CPU main translation funcions' commit
Quoted 3 lines
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, exceptsreg Z
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
changes since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicchanges since v34
1. rebasechanges since v35
1. rebase
2. useNANOSECONDS_PER_SECONDinstead of 1000000000 in
avr_timer16.cQuoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 1 line
3. split "target/avr: Register AVR support with the rest of QEMU"
into three patches
Quoted 12 lines
1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tenameQuoted 9 lines
1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERSQuoted 2 lines
7. use git commit sha1 instead of `master`` in avocado test
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Michael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU ControlInstructions
Quoted 93 lines
Quoted 93 lines
Quoted 92 lines
Quoted 91 lines
Quoted 37 lines
Quoted 10 lines
target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileQuoted 2 lines
Sarah Harris (1):
target/avr: Add limited support for USART and 16 bit timerperipherals
Quoted 16 lines
configure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++Quoted 8 lines
hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052Quoted 48 lines
++++++++++++++++++++++++++++++
Quoted 46 lines
tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)--
Best Regards,
Michael RolnikQuoted 4 lines
Quoted 4 lines
--
Best Regards,
Michael Rolnik--
Alex Bennée
Quoted 8 lines · Philippe Mathieu-Daudé
Quoted 2 lines · Michael Rolnik
I will rename them.
Please wait comments from Richard before a version respin.
Everything went well last 10 or so days, Michael and Sarah were responsive, the code and series got slowly improved more and more, but there was this disruption by your idea to "take over" the series with implementation of "real boards", rather than leave Michael doing improvements by himself, based on our feedback, like in a regular process of review... There are some pending quite reasonable and simple review items from me, Michael should continue working on them... But now he is told to wait... Shouldn't it be some better way?
On Thu, Nov 28, 2019 at 3:41 PM Aleksandar Markovic <
Quoted 23 lines
Quoted 2 lines
[email protected] <mailto:[email protected]>> wrote:
[...]
Quoted 17 lines
If I understand Aleksandar correctly, the naming is incorrect because too generic to AVR family, why Sarah only modeled the Atmel implementation.
Renaming devices such
hw/char/avr_usart.c->hw/char/atmel_usart.c(similarly with the macros) would be enough Aleksandar?Some renaming could help, perhaps not quite like the one above, but my point (which I find hard to believe I can't explain to you) is that peripherals inside the chip evolved over time, as starkly opposed to external peripherals that are set in stone...
Re: [PATCH v37 17/17] target/avr: Update MAINTAINERS file
In reply to Michael RolnikQuoted 12 lines · Michael Rolnik
Include AVR maintaners in
MAINTAINERSfileSigned-off-by Michael Rolnik <[email protected]>--- MAINTAINERS | 11 +++++++++++ 1 file changed, 11 insertions(+)MAINTAINERS +0
-0
MAINTAINERS S: Maintained · line 170
Quoted 8 lines
MAINTAINERS +5
-0@@ -163,6 +163,17 @@ S: Maintained F: hw/arm/smmu* F: include/hw/arm/smmu* +AVR TCG CPUs +M: Michael Rolnik <[email protected]> +R: Sarah Harris <[email protected]> +S: Maintained +F: target/avr/
^ This is the architectural part section
v This part should go under a new 'AVR Machines' section.
(See 'Alpha Machines' for example).
Quoted 10 lines
MAINTAINERS +6
-0+F: hw/misc/avr_mask.c +F: hw/char/avr_usart.c +F: hw/timer/avr_timer16.c +F: hw/avr/ +F: tests/acceptance/machine_avr6.py + CRIS TCG CPUs M: Edgar E. Iglesias <[email protected]> S: Maintained
Quoted 2 lines
Sarah,
do you mind if use the same license I use for my code?
I'm happy to use the same license.
Kind regards,
Sarah Harris
Quoted 443 lines · On Thu, 28 Nov 2019 14:28:19 +0200 Michael Rolnik <[email protected]> wrote:
Quoted 11 lines · Aleksandar Markovic
Quoted 3 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fullytested yet.
Quoted 1 line
However I was able to execute simple code with functions. e.g fibonacci
calculation.
Quoted 3 lines
This series of patches include a non real, sample board. No fuses support yet. PC is set to 0 at reset.
Quoted 5 lines
I have a couple of general remarks, so I am responding to the cover letter, not individual patches.
1) The licenses for Sarah devices differ than the rest - shouldn't all licenses be harmonized?
Sarah,
do you mind if use the same license I use for my code?Quoted 14 lines
2) There is an architectural problem with peripherals. It is possible that they evolve over time, so, for example, USART could not be the same for older and newer CPUs (in principle, newer peripheral is expected to be o sort of "superset" of the older). How do you solve that problem? Right now, it may not looks serious to you, but if you don;t think about that right now, from the outset, soon the code will become so entangled, ti woudl be almost very difficult to fix it. Please think about that, how would you solve it, is there a way to pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?
Hi Aleksandar,
Please explain. I don't see any problem from CPU's perspective. as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there is sample.Quoted 391 lines
Quoted 2 lines
Following are examples of possible usages, assuming program.elf is
compiled for AVR cpu
Quoted 3 lines
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnetwindow
Quoted 1 line
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
Quoted 6 lines
run
telent localhost 5678
3. Continious non interrupted execution with serial output into stdout runqemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -S
runavr-gdb program.elfand then within GDB shell `target remote:1234`
Quoted 3 lines
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmThank you so much for these examples!
Aleksandar
Quoted 4 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding or translation
2. CPU features which allow define the following 8bit AVR coresQuoted 8 lines
avr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows
to execute simple code
Quoted 9 lines
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of
0x000000ff) for faster address manipulatonQuoted 13 lines
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM 11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHQuoted 2 lines
changes since v4
1. target is now defined as big endian in order to optimizepush_ret/pop_ret
Quoted 14 lines
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)Quoted 6 lines
changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg inavr_cpu_gdb_read_register/avr_cpu_gdb_write_register
Quoted 1 line
2. configure the target as little endian because otherwise GDB does not
work
Quoted 5 lines
3. fixing and testing gen_push_ret/gen_pop_ret
changes since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supported yetregisters only
Quoted 13 lines
3.
MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built for all
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memoryQuoted 5 lines
changes since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed toany device
Quoted 4 lines
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try to accessregisters
Quoted 3 lines
changes since v10
1. movig back fullwr & fullrd into the commit where outb and inb wereintroduced
Quoted 12 lines
2. changing
tlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codeQuoted 12 lines
changes since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
Quoted 8 lines
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thingexists)
Quoted 5 lines
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrisQuoted 8 lines
changes since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers areaccessible
Quoted 30 lines
by LD/ST only. CPU has no interest in them
changes since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
Quoted 8 lines
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixes acouple of bugs wrt single stepping
Quoted 40 lines
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'Quoted 8 lines
changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
Quoted 8 lines
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicQuoted 7 lines
changes since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU" intothree patches
Quoted 12 lines
1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tenameQuoted 9 lines
1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERSQuoted 95 lines
7. use git commit sha1 instead of `master`` in avocado test
Michael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileSarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer peripheralsconfigure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)Signature · 2 lines
Best Regards, Michael Rolnik
Quoted 21 lines · Alex Bennée
Quoted 1 line · Aleksandar Markovic
On Thursday, November 28, 2019, Michael Rolnik <[email protected]>
Quoted 2 lines · wrote:
Quoted 2 lines
I don't see why you say that the peripherals are inside the chip, there
is
Quoted 1 line
Quoted 1 line
CPU within target/avr directory and then there are some peripherals in
hw
Quoted 5 lines
Quoted 4 lines
directory, CPU does not depend on them. what am I missing?
Quoted 2 lines
I meant these peripherals are physically inside the chip together with
the
Quoted 4 lines
core.
And USART in a micricontroler from 2010 is different than USART from one from 2018.
Quoted 10 lines
Won't these be different chip parts? Or at least revs of the part?
I think broadly the difference between SoC devices is handled by handling versioning in the board models - the board being in this case a CPU core + a bunch of SoC components + the actual board itself.
All the target/cpu stuff needs to deal with is actual architectural revs (c.f. target/arm/cpu[64].c).
Michael, please take a look at this link:
https://web-engineering.info/node/28
"Part number" seems to be the thing you'll need to incorporate in the overall AVR support in QEMU in order to get the correct emulation of in-chip peripherals - as Alex actually hinted. (In my opinion, you don't need it now, in the initial version, but by no means in near future.)
Have a great weekend!
Quoted 536 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines · Aleksandar Markovic
On Thursday, November 28, 2019, Michael Rolnik <[email protected]>
Quoted 14 lines · wrote:
Quoted 14 lines
Quoted 14 lines
Quoted 14 lines
Quoted 8 lines · Aleksandar Markovic
Quoted 3 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU. All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not
fully tested yet.
Quoted 1 line
However I was able to execute simple code with functions. e.g
fibonacci calculation.
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
This series of patches include a non real, sample board. No fuses support yet. PC is set to 0 at reset.
I have a couple of general remarks, so I am responding to the cover letter, not individual patches.
1) The licenses for Sarah devices differ than the rest - shouldn't
all
Quoted 20 lines
Quoted 20 lines
Quoted 20 lines
Quoted 20 lines
Quoted 1 line
licenses be harmonized?
Sarah,
do you mind if use the same license I use for my code?Quoted 14 lines
2) There is an architectural problem with peripherals. It is possible that they evolve over time, so, for example, USART could not be the same for older and newer CPUs (in principle, newer peripheral is expected to be o sort of "superset" of the older). How do you solve that problem? Right now, it may not looks serious to you, but if you don;t think about that right now, from the outset, soon the code will become so entangled, ti woudl be almost very difficult to fix it. Please think about that, how would you solve it, is there a way to pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 5 lines
Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the
core.
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
If one models, let's say an external (in the sense, it is a separate chip) ADC (analog-to-digital converter), one looks at specs, implement
what
Quoted 1 line
Quoted 1 line
Quoted 1 line
is resonable possible in QEMU, plug it in in one of machines thst
contains
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
it, and that's it. That ADC remains the same, of course, whatever the surrounding system is.
In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers is very large, and both cores and peripherals evolved.
For cores, you handle differences with all these
AVR_FEATUREmacros,and
Quoted 1 line
Quoted 1 line
Quoted 1 line
this seems to be working, no significant objection from my side, and
btw
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
that was not an easy task to execute, all admiration from me.
But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't
have
Quoted 1 line
Quoted 1 line
Quoted 1 line
time right now to spell many examples, but I read three different
specs,
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if
they
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on the
core
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
it is attached to?
I hope I was clearer this time.
Aleksandar
Quoted 3 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there issample.
Quoted 83 lines
Quoted 83 lines
Quoted 83 lines
Quoted 83 lines
Quoted 83 lines
Quoted 5 lines
Following are examples of possible usages, assuming program.elf is
compiled for AVR cpu
Quoted 3 lines
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output intotelnet window
Quoted 1 line
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
Quoted 2 lines
run
telent localhost 5678
3. Continious non interrupted execution with serial output intostdout
Quoted 4 lines
run
qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -Srunavr-gdb program.elfand then within GDB shell `targetremote :1234`
Quoted 3 lines
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmThank you so much for these examples!
Aleksandar
Quoted 3 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding ortranslation
Quoted 9 lines
2. CPU features which allow define the following 8bit AVR cores
avr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which
allows to execute simple code
Quoted 9 lines
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of
0x000000ff) for faster address manipulatonQuoted 13 lines
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM 11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHQuoted 2 lines
changes since v4
1. target is now defined as big endian in order to optimizepush_ret/pop_ret
Quoted 14 lines
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg inavr_cpu_gdb_read_register/avr_
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
cpu_gdb_write_registerQuoted 1 line
2. configure the target as little endian because otherwise GDB does
not work
Quoted 2 lines
3. fixing and testing gen_push_ret/gen_pop_ret
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
changes since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supportedyet registers only
Quoted 4 lines
3.
MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be builtfor
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
all
Quoted 9 lines
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memoryQuoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
changes since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routedto
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
any device
Quoted 4 lines
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try toaccess registers
Quoted 3 lines
changes since v10
1. movig back fullwr & fullrd into the commit where outb and inbwere
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
introduced
Quoted 12 lines
2. changing
tlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codeQuoted 12 lines
changes since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such athing
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
exists)
Quoted 3 lines
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in mycode
Quoted 99 lines
Quoted 99 lines
Quoted 99 lines
Quoted 99 lines
Quoted 99 lines
Quoted 5 lines
5. moving avr configuration, within configure file, before chris
changes since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,
write(helper_outb)Quoted 5 lines
3. implementing CBIC/SBIC as
read(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers areaccessible
Quoted 30 lines
by LD/ST only. CPU has no interest in them
changes since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
Quoted 8 lines
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, andfixes a couple of bugs wrt single stepping
Quoted 48 lines
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructionQuoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
translation - CPU main translation funcions' commit
Quoted 3 lines
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, exceptsreg Z
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
changes since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicchanges since v34
1. rebasechanges since v35
1. rebase
2. useNANOSECONDS_PER_SECONDinstead of 1000000000 in
avr_timer16.cQuoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 1 line
3. split "target/avr: Register AVR support with the rest of QEMU"
into three patches
Quoted 12 lines
1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tenameQuoted 9 lines
1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERSQuoted 2 lines
7. use git commit sha1 instead of `master`` in avocado test
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Michael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU ControlInstructions
Quoted 93 lines
Quoted 93 lines
Quoted 92 lines
Quoted 91 lines
Quoted 37 lines
Quoted 10 lines
target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileQuoted 2 lines
Sarah Harris (1):
target/avr: Add limited support for USART and 16 bit timerperipherals
Quoted 16 lines
configure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++Quoted 8 lines
hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052Quoted 48 lines
++++++++++++++++++++++++++++++
Quoted 46 lines
tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)--
Best Regards,
Michael RolnikQuoted 4 lines
Quoted 4 lines
--
Best Regards,
Michael Rolnik--
Alex Bennée
Quoted 36 lines · Aleksandar Markovic
Quoted 21 lines · Alex Bennée
Quoted 1 line · Aleksandar Markovic
On Thursday, November 28, 2019, Michael Rolnik <[email protected]>
Quoted 2 lines · wrote:
Quoted 2 lines
I don't see why you say that the peripherals are inside the chip,
there is
Quoted 1 line
Quoted 1 line
CPU within target/avr directory and then there are some peripherals in
hw
Quoted 5 lines
Quoted 4 lines
directory, CPU does not depend on them. what am I missing?
Quoted 2 lines
I meant these peripherals are physically inside the chip together with
the
Quoted 4 lines
core.
And USART in a micricontroler from 2010 is different than USART from one from 2018.
Quoted 10 lines
Won't these be different chip parts? Or at least revs of the part?
I think broadly the difference between SoC devices is handled by handling versioning in the board models - the board being in this case a CPU core + a bunch of SoC components + the actual board itself.
All the target/cpu stuff needs to deal with is actual architectural revs (c.f. target/arm/cpu[64].c).
Quoted 10 lines
Michael, please take a look at this link:
https://web-engineering.info/node/28
"Part number" seems to be the thing you'll need to incorporate in the overall AVR support in QEMU in order to get the correct emulation of in-chip peripherals - as Alex actually hinted. (In my opinion, you don't need it now, in the initial version, but by no means in near future.)
Here is the list of "part numbers" for 212 (!! a mind-blowing number) of AVR microcontrollers:
https://github.com/felias-fogg/RescueAVR/blob/master/all.list
Quoted 545 lines
Have a great weekend!
Quoted 542 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines · Aleksandar Markovic
On Thursday, November 28, 2019, Michael Rolnik <[email protected]>
Quoted 14 lines · wrote:
Quoted 14 lines
Quoted 14 lines
Quoted 14 lines
Quoted 8 lines · Aleksandar Markovic
Quoted 3 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU. All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not
fully tested yet.
Quoted 1 line
However I was able to execute simple code with functions. e.g
fibonacci calculation.
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 3 lines
This series of patches include a non real, sample board. No fuses support yet. PC is set to 0 at reset.
I have a couple of general remarks, so I am responding to the cover letter, not individual patches.
1) The licenses for Sarah devices differ than the rest - shouldn't
all
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 1 line
licenses be harmonized?
Sarah,
do you mind if use the same license I use for my code?Quoted 4 lines
2) There is an architectural problem with peripherals. It is
possible
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
that they evolve over time, so, for example, USART could not be the same for older and newer CPUs (in principle, newer peripheral is expected to be o sort of "superset" of the older). How do you solve that problem? Right now, it may not looks serious to you, but if you don;t think about that right now, from the outset, soon the code
will
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 10 lines
Quoted 5 lines
become so entangled, ti woudl be almost very difficult to fix it. Please think about that, how would you solve it, is there a way to pass the information on the currently emulated CPU to the code covering a peripheral, and provide a different behaviour?
Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the
core.
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
If one models, let's say an external (in the sense, it is a separate chip) ADC (analog-to-digital converter), one looks at specs,
implement what
Quoted 1 line
Quoted 1 line
Quoted 1 line
is resonable possible in QEMU, plug it in in one of machines thst
contains
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
it, and that's it. That ADC remains the same, of course, whatever the surrounding system is.
In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers
is
Quoted 3 lines
Quoted 3 lines
Quoted 3 lines
very large, and both cores and peripherals evolved.
For cores, you handle differences with all these
AVR_FEATUREmacros,and
Quoted 1 line
Quoted 1 line
Quoted 1 line
this seems to be working, no significant objection from my side, and
btw
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
that was not an easy task to execute, all admiration from me.
But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't
have
Quoted 1 line
Quoted 1 line
Quoted 1 line
time right now to spell many examples, but I read three different
specs,
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if
they
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on
the core
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
it is attached to?
I hope I was clearer this time.
Aleksandar
Quoted 3 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there issample.
Quoted 83 lines
Quoted 83 lines
Quoted 83 lines
Quoted 83 lines
Quoted 83 lines
Quoted 5 lines
Following are examples of possible usages, assuming program.elf is
compiled for AVR cpu
Quoted 3 lines
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output intotelnet window
Quoted 1 line
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
Quoted 2 lines
run
telent localhost 5678
3. Continious non interrupted execution with serial output intostdout
Quoted 4 lines
run
qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -Srunavr-gdb program.elfand then within GDB shell `targetremote :1234`
Quoted 3 lines
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmThank you so much for these examples!
Aleksandar
Quoted 3 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding ortranslation
Quoted 9 lines
2. CPU features which allow define the following 8bit AVR cores
avr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which
allows to execute simple code
Quoted 9 lines
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of
0x000000ff) for faster address manipulatonQuoted 13 lines
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM 11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHQuoted 2 lines
changes since v4
1. target is now defined as big endian in order to optimizepush_ret/pop_ret
Quoted 14 lines
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg inavr_cpu_gdb_read_register/avr_
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
Quoted 2 lines
cpu_gdb_write_registerQuoted 1 line
2. configure the target as little endian because otherwise GDB
does
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
Quoted 6 lines
not work
Quoted 5 lines
3. fixing and testing gen_push_ret/gen_pop_ret
changes since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for nonsupported
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
yet registers only
Quoted 4 lines
3.
MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be builtfor
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
Quoted 10 lines
all
Quoted 9 lines
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memoryQuoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
changes since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are notrouted to
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
any device
Quoted 4 lines
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try toaccess registers
Quoted 3 lines
changes since v10
1. movig back fullwr & fullrd into the commit where outb and inbwere
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
introduced
Quoted 12 lines
2. changing
tlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codeQuoted 12 lines
changes since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such athing
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
exists)
Quoted 3 lines
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in mycode
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 5 lines
5. moving avr configuration, within configure file, before chris
changes since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,
write(helper_outb)Quoted 5 lines
3. implementing CBIC/SBIC as
read(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registersare
Quoted 88 lines
Quoted 88 lines
Quoted 88 lines
Quoted 88 lines
Quoted 88 lines
accessible
Quoted 30 lines
by LD/ST only. CPU has no interest in them
changes since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
Quoted 8 lines
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, andfixes a couple of bugs wrt single stepping
Quoted 48 lines
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructionQuoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
Quoted 4 lines
translation - CPU main translation funcions' commit
Quoted 3 lines
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, exceptsreg Z
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
changes since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicchanges since v34
1. rebasechanges since v35
1. rebase
2. useNANOSECONDS_PER_SECONDinstead of 1000000000 in
avr_timer16.cQuoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 26 lines
Quoted 1 line
3. split "target/avr: Register AVR support with the rest of QEMU"
into three patches
Quoted 12 lines
1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tenameQuoted 9 lines
1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERSQuoted 2 lines
7. use git commit sha1 instead of `master`` in avocado test
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Quoted 11 lines
Michael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU ControlInstructions
Quoted 93 lines
Quoted 93 lines
Quoted 92 lines
Quoted 91 lines
Quoted 37 lines
Quoted 10 lines
target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileQuoted 2 lines
Sarah Harris (1):
target/avr: Add limited support for USART and 16 bit timerperipherals
Quoted 16 lines
configure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++Quoted 8 lines
hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052Quoted 48 lines
++++++++++++++++++++++++++++++
Quoted 46 lines
tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)--
Best Regards,
Michael RolnikQuoted 4 lines
Quoted 4 lines
--
Best Regards,
Michael Rolnik--
Alex Bennée
Re: [PATCH v37 05/17] target/avr: Add instruction translation - Arithmetic and Logic Instructions
In reply to Michael RolnikOn Wednesday, November 27, 2019, Michael Rolnik <[email protected]> wrote:
+
Quoted 23 lines
target/avr/translate.c +3
-0+ +/* +* Performs the logical AND between the contents of register Rd and register
+ * Rr and places the result in the destination register Rd.
+ */target/avr/translate.c +16
-0+static bool trans_AND (DisasContext *ctx,arg_AND *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_and_tl (R , Rd, Rr);/* Rd = Rd and Rr */ + +/* Vf */ +tcg_gen_movi_tl (cpu_Vf,0 );/* Vf = 0 */ + +/* Zf */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +gen_ZNSf (R );
target/avr/translate.c static bool avr_have_feature(DisasContext *ctx, int feature) · line 555
Quoted 10 lines
target/avr/translate.c +10
-0+ +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + +
According to specs:
... the chips in question have cores with 16 GPRs (that correspond to GPRs R16-R31 of more common AVR cores with 32 GPRs).
How do you handle such cores?
I don't see here anything preventing usage of all 32 GPRs, resulting, of course, in an inaccurate emulation.
Thanks,
Aleksandar
Re: [PATCH v37 12/17] target/avr: Add example board configuration
In reply to Michael RolnikQuoted 11 lines · Michael Rolnik
A simple board setup that configures an AVR CPU to run a given firmware image.
This is all that's useful to implement without peripheral emulation as AVR CPUs include a lot of on-board peripherals.NOTE: this is not a real board !!!!
NOTE: it's used for CPU testing!!!!Signed-off-by Michael Rolnik <[email protected]>Reviewed-by Aleksandar Markovic <[email protected]>
Quoted 10 lines
--- hw/avr/sample.c | 282 +++++++++++++++++++++++++++++++++++++++++++ hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + 4 files changed, 290 insertions(+) create mode 100644 hw/avr/sample.c create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs
Michael, hi.
I just need a clarification here:
- What will happen if this patch is removed? Would boot and Avocado tests work? What else in general wouldn't work or be available? What was, in fact, the ultimate motivation for you to insert this patch?
Thanks,
Aleksandar
Quoted 339 lines
hw/avr/sample.c new file +24
-0@@ -0,0 +1,282 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +/* +* NOTE: +* This is not a real AVR board, this is an example! +* The CPU is an approximation of an ATmega2560, but is missing various
+ * built-in peripherals.
+ *
+ * This example board loads provided binary file into flash memory
and
+ * executes it from0x00000000address in the code memory space.
+ *
+ * Currently used for AVR CPU validation
+ *
+ */hw/avr/sample.c new file +24
-0+ +#include "qemu/osdep.h" +#include "qapi/error.h" +#include "qemu-common.h" +#include "cpu.h" +#include "hw/hw.h" +#include "sysemu/sysemu.h" +#include "sysemu/qtest.h" +#include "ui/console.h" +#include "hw/boards.h" +#include "hw/loader.h" +#include "qemu/error-report.h" +#include "exec/address-spaces.h" +#include "include/hw/sysbus.h" +#include "include/hw/char/avr_usart.h" +#include "include/hw/timer/avr_timer16.h" +#include "include/hw/misc/avr_mask.h" +#include "elf.h" +#include "hw/misc/unimp.h" + +#define SIZE_FLASH 0x00040000 +#define SIZE_SRAM 0x00002000 +/* +* Size of additional "external" memory, as if the AVR were configured to use
+ * an external RAM chip.
+ * Note that the configuration registers that normally enable this feature are
+ * unimplemented.
+ */hw/avr/sample.c new file +3
-0+#define SIZE_EXMEM 0x00000000 + +/* Offsets of peripherals in emulated memory space (i.e. not host addresses) */
hw/avr/sample.c new file +128
-0+#define PRR0_BASE 0x64 +#define PRR1_BASE 0x65 +#define USART_BASE 0xc0 +#define TIMER1_BASE 0x80 +#define TIMER1_IMSK_BASE 0x6f +#define TIMER1_IFR_BASE 0x36 + +/* Interrupt numbers used by peripherals */ +#define USART_RXC_IRQ 24 +#define USART_DRE_IRQ 25 +#define USART_TXC_IRQ 26 + +#define TIMER1_CAPT_IRQ 15 +#define TIMER1_COMPA_IRQ 16 +#define TIMER1_COMPB_IRQ 17 +#define TIMER1_COMPC_IRQ 18 +#define TIMER1_OVF_IRQ 19 + +/* Power reduction */ +#define PRR1_BIT_PRTIM5 0x05/* Timer/Counter5 */ +#define PRR1_BIT_PRTIM4 0x04/* Timer/Counter4 */ +#define PRR1_BIT_PRTIM3 0x03/* Timer/Counter3 */ +#define PRR1_BIT_PRUSART3 0x02/* USART3 */ +#define PRR1_BIT_PRUSART2 0x01/* USART2 */ +#define PRR1_BIT_PRUSART1 0x00/* USART1 */ + +#define PRR0_BIT_PRTWI 0x06/* TWI */ +#define PRR0_BIT_PRTIM2 0x05/* Timer/Counter2 */ +#define PRR0_BIT_PRTIM0 0x04/* Timer/Counter0 */ +#define PRR0_BIT_PRTIM1 0x03/* Timer/Counter1 */ +#define PRR0_BIT_PRSPI 0x02/* Serial Peripheral Interface */ +#define PRR0_BIT_PRUSART0 0x01/* USART0 */ +#define PRR0_BIT_PRADC 0x00/* ADC */ + +typedef struct { +MachineClass parent; +}SampleMachineClass ; + +typedef struct { +MachineState parent; +MemoryRegion *ram; +MemoryRegion *flash; +AVRUsartState *usart0; +AVRTimer16State *timer1; +AVRMaskState *prr[2 ]; +}SampleMachineState ; + +#define TYPE_SAMPLE_MACHINE MACHINE_TYPE_NAME("sample") + +#define SAMPLE_MACHINE (obj) \ + OBJECT_CHECK(SampleMachineState, obj, TYPE_SAMPLE_MACHINE) +#define SAMPLE_MACHINE_GET_CLASS (obj) \ + OBJECT_GET_CLASS(SampleMachineClass, obj, TYPE_SAMPLE_MACHINE) +#define SAMPLE_MACHINE_CLASS (klass) \ + OBJECT_CLASS_CHECK(SampleMachineClass, klass, TYPE_SAMPLE_MACHINE) + +static void sample_init (MachineState *machine) +{ +SampleMachineState *sms =SAMPLE_MACHINE (machine); +MemoryRegion *system_memory =get_system_memory (); +AVRCPU *cpu; +const char *firmware =NULL ; +const char *filename; +int bytes_loaded; +SysBusDevice *busdev; +DeviceState *cpudev; + + system_memory =get_system_memory (); + sms->ram =g_new (MemoryRegion,1 ); + sms->flash =g_new (MemoryRegion,1 ); + + cpu =AVR_CPU (cpu_create (machine->cpu_type)); + cpudev =DEVICE (cpu); + + +memory_region_init_rom (sms->flash,NULL ,"avr.flash" ,SIZE_FLASH , + &error_fatal); +memory_region_add_subregion (system_memory,OFFSET_CODE , sms->flash); + +/* following are atmel2560 device */ +create_unimplemented_device ("usart 3" ,OFFSET_DATA +0x0130 ,0x0007 ); +create_unimplemented_device ("timer-counter-16bit 5" , +OFFSET_DATA +0x0120 ,0x000e ); +create_unimplemented_device ("gpio L" ,OFFSET_DATA +0x0109 ,0x0003 ); +create_unimplemented_device ("gpio K" ,OFFSET_DATA +0x0106 ,0x0003 ); +create_unimplemented_device ("gpio J" ,OFFSET_DATA +0x0103 ,0x0003 ); +create_unimplemented_device ("gpio H" ,OFFSET_DATA +0x0100 ,0x0003 ); +create_unimplemented_device ("usart 2" ,OFFSET_DATA +0x00d0 ,0x0007 ); +create_unimplemented_device ("usart 1" ,OFFSET_DATA +0x00c8 ,0x0007 ); +create_unimplemented_device ("usart 0" ,OFFSET_DATA +0x00c0 ,0x0007 ); +create_unimplemented_device ("twi" ,OFFSET_DATA +0x00b8 ,0x0006 ); +create_unimplemented_device ("timer-counter-async-8bit 2" , +OFFSET_DATA +0x00b0 ,0x0007 ); +create_unimplemented_device ("timer-counter-16bit 4" , +OFFSET_DATA +0x00a0 ,0x000e ); +create_unimplemented_device ("timer-counter-16bit 3" , +OFFSET_DATA +0x0090 ,0x000e ); +create_unimplemented_device ("timer-counter-16bit 1" , +OFFSET_DATA +0x0080 ,0x000e ); +create_unimplemented_device ("ac / adc" , +OFFSET_DATA +0x0078 ,0x0008 ); +create_unimplemented_device ("ext-mem-iface" , +OFFSET_DATA +0x0074 ,0x0002 ); +create_unimplemented_device ("int-controller" , +OFFSET_DATA +0x0068 ,0x000c ); +create_unimplemented_device ("sys" , +OFFSET_DATA +0x0060 ,0x0007 ); +create_unimplemented_device ("spi" , +OFFSET_DATA +0x004c ,0x0003 ); +create_unimplemented_device ("ext-mem-iface" , +OFFSET_DATA +0x004a ,0x0002 ); +create_unimplemented_device ("timer-counter-pwm-8bit 0" , +OFFSET_DATA +0x0043 ,0x0006 ); +create_unimplemented_device ("ext-mem-iface" , +OFFSET_DATA +0x003e ,0x0005 ); +create_unimplemented_device ("int-controller" , +OFFSET_DATA +0x0035 ,0x0009 ); +create_unimplemented_device ("gpio G" ,OFFSET_DATA +0x0032 ,0x0003 ); +create_unimplemented_device ("gpio F" ,OFFSET_DATA +0x002f ,0x0003 ); +create_unimplemented_device ("gpio E" ,OFFSET_DATA +0x002c ,0x0003 ); +create_unimplemented_device ("gpio D" ,OFFSET_DATA +0x0029 ,0x0003 ); +create_unimplemented_device ("gpio C" ,OFFSET_DATA +0x0026 ,0x0003 ); +create_unimplemented_device ("gpio B" ,OFFSET_DATA +0x0023 ,0x0003 ); +create_unimplemented_device ("gpio A" ,OFFSET_DATA +0x0020 ,0x0003 ); + +memory_region_allocate_system_memory ( + sms->ram,NULL ,"avr.ram" ,SIZE_SRAM +SIZE_EXMEM ); + memory_region_add_subregion(system_memory, OFFSET_DATA + 0x200,
sms->ram);hw/avr/sample.c new file +16
-0+ +/* Power Reduction built-in peripheral */ + sms->prr[0 ] =AVR_MASK (sysbus_create_simple (TYPE_AVR_MASK , +OFFSET_DATA +PRR0_BASE ,NULL )); + sms->prr[1 ] =AVR_MASK (sysbus_create_simple (TYPE_AVR_MASK , +OFFSET_DATA +PRR1_BASE ,NULL )); + +/* USART 0 built-in peripheral */ + sms->usart0 =AVR_USART (object_new (TYPE_AVR_USART )); + busdev =SYS_BUS_DEVICE (sms->usart0); +qdev_prop_set_chr (DEVICE (sms->usart0),"chardev" ,serial_hd (0 )); +object_property_set_bool (OBJECT (sms->usart0), true,"realized" , + &error_fatal); +sysbus_mmio_map (busdev,0 ,OFFSET_DATA +USART_BASE ); +/* +* These IRQ numbers don't match the datasheet because we're counting from
hw/avr/sample.c new file +3
-0+* zero and not including reset. +*/ + sysbus_connect_irq(busdev, 0, qdev_get_gpio_in(cpudev,
USART_RXC_IRQ));
+ sysbus_connect_irq(busdev, 1, qdev_get_gpio_in(cpudev,USART_DRE_IRQ));
+ sysbus_connect_irq(busdev, 2, qdev_get_gpio_in(cpudev,USART_TXC_IRQ));hw/avr/sample.c new file +12
-0+sysbus_connect_irq (SYS_BUS_DEVICE (sms->prr[1 ]),PRR1_BIT_PRUSART1 , +qdev_get_gpio_in (DEVICE (sms->usart0),0 )); + +/* Timer 1 built-in periphal */ + sms->timer1 =AVR_TIMER16 (object_new (TYPE_AVR_TIMER16 )); +object_property_set_bool (OBJECT (sms->timer1), true,"realized" , + &error_fatal); + busdev =SYS_BUS_DEVICE (sms->timer1); +sysbus_mmio_map (busdev,0 ,OFFSET_DATA +TIMER1_BASE ); +sysbus_mmio_map (busdev,1 ,OFFSET_DATA +TIMER1_IMSK_BASE ); +sysbus_mmio_map (busdev,2 ,OFFSET_DATA +TIMER1_IFR_BASE ); + sysbus_connect_irq(busdev, 0, qdev_get_gpio_in(cpudev,
TIMER1_CAPT_IRQ));
+ sysbus_connect_irq(busdev, 1, qdev_get_gpio_in(cpudev,TIMER1_COMPA_IRQ));
+ sysbus_connect_irq(busdev, 2, qdev_get_gpio_in(cpudev,TIMER1_COMPB_IRQ));
+ sysbus_connect_irq(busdev, 3, qdev_get_gpio_in(cpudev,TIMER1_COMPC_IRQ));
+ sysbus_connect_irq(busdev, 4, qdev_get_gpio_in(cpudev,TIMER1_OVF_IRQ));hw/avr/sample.c new file +14
-0+sysbus_connect_irq (SYS_BUS_DEVICE (sms->prr[0 ]),PRR0_BIT_PRTIM1 , +qdev_get_gpio_in (DEVICE (sms->timer1),0 )); + +/* Load firmware (contents of flash) trying to auto-detect format */ + firmware = machine->firmware; +if (firmware !=NULL ) { + filename =qemu_find_file (QEMU_FILE_TYPE_BIOS , firmware); +if (filename ==NULL ) { +error_report ("Unable to find %s" , firmware); +exit (1 ); + } + + bytes_loaded = load_elf( + filename, NULL, NULL, NULL, NULL, NULL, NULL, 0, EM_NONE, 0,0);
hw/avr/sample.c new file +40
-0+if (bytes_loaded <0 ) { + bytes_loaded =load_image_targphys ( + filename,OFFSET_CODE ,SIZE_FLASH ); + } +if (bytes_loaded <0 ) { +error_report ( +"Unable to load firmware image %s as ELF or raw binary" , + firmware); +exit (1 ); + } + } +} + +static void sample_class_init (ObjectClass *oc,void *data) +{ +MachineClass *mc =MACHINE_CLASS (oc); + + mc->desc ="AVR sample/example board (ATmega2560)" ; + mc->init = sample_init; + mc->default_cpus =1 ; + mc->min_cpus = mc->default_cpus; + mc->max_cpus = mc->default_cpus; + mc->default_cpu_type ="avr6-avr-cpu" ;/* ATmega2560. */ + mc->is_default =1 ; +} + +static const TypeInfo sample_info = { + .name =TYPE_SAMPLE_MACHINE , + .parent =TYPE_MACHINE , + .instance_size =sizeof (SampleMachineState), + .class_size =sizeof (SampleMachineClass), + .class_init = sample_class_init, +}; + +static void sample_machine_init (void ) +{ +type_register_static (&sample_info); +} + +type_init (sample_machine_init);hw/Kconfig +1
-0@@ -44,6 +44,7 @@ source watchdog/Kconfig # arch Kconfig source arm/Kconfig source alpha/Kconfig +source avr/Kconfig source cris/Kconfig source hppa/Kconfig source i386/Kconfighw/avr/Kconfig new file +6
-0@@ -0,0 +1,6 @@ +config AVR_SAMPLE + bool + select AVR_TIMER16 + select AVR_USART + select AVR_MASK + select UNIMPhw/avr/Makefile.objs new file +1
-1@@ -0,0 +1 @@ +obj-y += sample.o --2.17.2 (Apple Git-113)
Quoted 17 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fully tested yet.
However I was able to execute simple code with functions. e.g fibonacci calculation.
This series of patches include a non real, sample board.
No fuses support yet. PC is set to 0 at reset.Following are examples of possible usages, assuming program.elf is compiled for AVR cpu
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnet window
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
runtelent localhost 5678
Quoted 10 lines
3. Continious non interrupted execution with serial output into stdout run
qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -S
runavr-gdb program.elfand then within GDB shell `target remote
:1234`
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asm
Hi, Michael.
Can you explain to me the mechanisms of recognition of what core/microcontroller QEMU is supposed to emulate in the examples above?
Yours, Aleksandar
Quoted 363 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding or translation
2. CPU features which allow define the following 8bit AVR coresavr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows to execute simple code
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of0x000000ff) for faster address manipulaton
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM
11.fixing MULS/MULSU not to modify inputs' content
12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHchanges since v4
1. target is now defined as big endian in order to optimize push_ret/pop_ret
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_cpu_gdb_write_register
2. configure the target as little endian because otherwise GDB does not work
3. fixing and testing gen_push_ret/gen_pop_retchanges since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supported yet registers only
3.MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built for all
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memorychanges since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed to any device
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try to access registerschanges since v10
1. movig back fullwr & fullrd into the commit where outb and inb were introduced
2. changingtlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codechanges since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions 12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thing exists)
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrischanges since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers are accessible
by LD/ST only. CPU has no interest in themchanges since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixes a
couple of bugs wrt single stepping
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicchanges since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU" into three patches 1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tename1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to MAINTAINERS 7. use git commit sha1 instead of `master`` in avocado testMichael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileSarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer peripheralsconfigure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)
Re: [PATCH v37 17/17] target/avr: Update MAINTAINERS file
In reply to Michael RolnikQuoted 16 lines · Michael Rolnik
Include AVR maintaners in
MAINTAINERSfileSigned-off-by Michael Rolnik <[email protected]>--- MAINTAINERS | 11 +++++++++++ 1 file changed, 11 insertions(+)MAINTAINERS +1
-0@@ -163,6 +163,17 @@ S: Maintained F: hw/arm/smmu* F: include/hw/arm/smmu* +AVR TCG CPUs
MAINTAINERS S: Maintained · line 176
Quoted 10 lines
MAINTAINERS +10
-0+M: Michael Rolnik <[email protected]> +R: Sarah Harris <[email protected]> +S: Maintained +F: target/avr/ +F: hw/misc/avr_mask.c +F: hw/char/avr_usart.c +F: hw/timer/avr_timer16.c +F: hw/avr/ +F: tests/acceptance/machine_avr6.py +
This is from summary from the cover letter:
create mode 100644 default-configs/avr-softmmu.mak
create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py
All such new files must be covered by this change in MAINTAINERS. Please doublecheck.
For example,
include/hw/timer/avr_timer16.h
seems to be not covered.
Thanks,
Aleksandar
Quoted 7 lines
MAINTAINERS +0
-1CRIS TCG CPUs M: Edgar E. Iglesias <[email protected]> S: Maintained --2.17.2 (Apple Git-113)
Re: [PATCH v37 05/17] target/avr: Add instruction translation - Arithmetic and Logic Instructions
In reply to Aleksandar MarkovicQuoted 41 lines · Aleksandar Markovic
On Wednesday, November 27, 2019, Michael Rolnik <[email protected]> wrote:
+
Quoted 23 lines
target/avr/translate.c +3
-0+ +/* +* Performs the logical AND between the contents of register Rd and register
+ * Rr and places the result in the destination register Rd.
+ */target/avr/translate.c +16
-0+static bool trans_AND (DisasContext *ctx,arg_AND *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_and_tl (R , Rd, Rr);/* Rd = Rd and Rr */ + +/* Vf */ +tcg_gen_movi_tl (cpu_Vf,0 );/* Vf = 0 */ + +/* Zf */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +gen_ZNSf (R );target/avr/translate.c static bool avr_have_feature(DisasContext *ctx, int feature) · line 555
Quoted 10 lines
target/avr/translate.c +10
-0+ +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + +According to specs:
Quoted 7 lines
http://ww1.microchip.com/downloads/en/devicedoc/atmel-42505-8-bit-avr-
microcontrollers-attiny102-attiny104_datasheet.pdf... the chips in question have cores with 16 GPRs (that correspond to GPRs R16-R31 of more common AVR cores with 32 GPRs).
There are more examples;
Also ATtiny20, ATtiny40.
How do you handle such cores?
Quoted 7 lines
I don't see here anything preventing usage of all 32 GPRs, resulting, of course, in an inaccurate emulation.
Thanks,
Aleksandar
Re: [PATCH v37 12/17] target/avr: Add example board configuration
In reply to Aleksandar MarkovicHi Aleksandar.
I guess no testing would be possible.
My previous series of patches (about 2 years ago, before Sarah joined) did not contain any peripherals, there were only the CPU and the sample board, I used it to test instructions.
Somebody complained that there are no peripherals and that was where I stopped to care as I did (and still don't) want to model devices. I believed that others would join and add devices. then two years passed by, Sarah implemented the timer and the UART devices.
And here we are.
Regards,
Michael Rolnik
Quoted 378 lines · Aleksandar Markovic
Quoted 11 lines · Michael Rolnik
A simple board setup that configures an AVR CPU to run a given firmware image.
This is all that's useful to implement without peripheral emulation as AVR CPUs include a lot of on-board peripherals.NOTE: this is not a real board !!!!
NOTE: it's used for CPU testing!!!!Signed-off-by Michael Rolnik <[email protected]>Reviewed-by Aleksandar Markovic <[email protected]>Quoted 10 lines
--- hw/avr/sample.c | 282 +++++++++++++++++++++++++++++++++++++++++++ hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + 4 files changed, 290 insertions(+) create mode 100644 hw/avr/sample.c create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objsMichael, hi.
I just need a clarification here:
- What will happen if this patch is removed? Would boot and Avocado tests work? What else in general wouldn't work or be available? What was, in fact, the ultimate motivation for you to insert this patch?
Thanks,
AleksandarQuoted 339 lines
hw/avr/sample.c new file +24
-0@@ -0,0 +1,282 @@ +/* +* QEMU AVR CPU +* +* Copyright (c) 2019 Michael Rolnik +* +* This library is free software; you can redistribute it and/or +* modify it under the terms of the GNU Lesser General Public +* License as published by the Free Software Foundation; either +* version 2.1 of the License, or (at your option) any later version. +* +* This library is distributed in the hope that it will be useful, +* but WITHOUT ANY WARRANTY; without even the implied warranty of +* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +* Lesser General Public License for more details. +* +* You should have received a copy of the GNU Lesser General Public +* License along with this library; if not, see +* <http://www.gnu.org/licenses/lgpl-2.1.html> +*/ + +/* +* NOTE: +* This is not a real AVR board, this is an example! +* The CPU is an approximation of an ATmega2560, but is missing various
+ * built-in peripherals.
+ *
+ * This example board loads provided binary file into flash memory
and
+ * executes it from0x00000000address in the code memory space.
+ *
+ * Currently used for AVR CPU validation
+ *
+ */hw/avr/sample.c new file +24
-0+ +#include "qemu/osdep.h" +#include "qapi/error.h" +#include "qemu-common.h" +#include "cpu.h" +#include "hw/hw.h" +#include "sysemu/sysemu.h" +#include "sysemu/qtest.h" +#include "ui/console.h" +#include "hw/boards.h" +#include "hw/loader.h" +#include "qemu/error-report.h" +#include "exec/address-spaces.h" +#include "include/hw/sysbus.h" +#include "include/hw/char/avr_usart.h" +#include "include/hw/timer/avr_timer16.h" +#include "include/hw/misc/avr_mask.h" +#include "elf.h" +#include "hw/misc/unimp.h" + +#define SIZE_FLASH 0x00040000 +#define SIZE_SRAM 0x00002000 +/* +* Size of additional "external" memory, as if the AVR were configured to use
+ * an external RAM chip.
+ * Note that the configuration registers that normally enable this feature are
+ * unimplemented.
+ */hw/avr/sample.c new file +3
-0+#define SIZE_EXMEM 0x00000000 + +/* Offsets of peripherals in emulated memory space (i.e. not host addresses) */
hw/avr/sample.c new file +128
-0+#define PRR0_BASE 0x64 +#define PRR1_BASE 0x65 +#define USART_BASE 0xc0 +#define TIMER1_BASE 0x80 +#define TIMER1_IMSK_BASE 0x6f +#define TIMER1_IFR_BASE 0x36 + +/* Interrupt numbers used by peripherals */ +#define USART_RXC_IRQ 24 +#define USART_DRE_IRQ 25 +#define USART_TXC_IRQ 26 + +#define TIMER1_CAPT_IRQ 15 +#define TIMER1_COMPA_IRQ 16 +#define TIMER1_COMPB_IRQ 17 +#define TIMER1_COMPC_IRQ 18 +#define TIMER1_OVF_IRQ 19 + +/* Power reduction */ +#define PRR1_BIT_PRTIM5 0x05/* Timer/Counter5 */ +#define PRR1_BIT_PRTIM4 0x04/* Timer/Counter4 */ +#define PRR1_BIT_PRTIM3 0x03/* Timer/Counter3 */ +#define PRR1_BIT_PRUSART3 0x02/* USART3 */ +#define PRR1_BIT_PRUSART2 0x01/* USART2 */ +#define PRR1_BIT_PRUSART1 0x00/* USART1 */ + +#define PRR0_BIT_PRTWI 0x06/* TWI */ +#define PRR0_BIT_PRTIM2 0x05/* Timer/Counter2 */ +#define PRR0_BIT_PRTIM0 0x04/* Timer/Counter0 */ +#define PRR0_BIT_PRTIM1 0x03/* Timer/Counter1 */ +#define PRR0_BIT_PRSPI 0x02/* Serial Peripheral Interface */ +#define PRR0_BIT_PRUSART0 0x01/* USART0 */ +#define PRR0_BIT_PRADC 0x00/* ADC */ + +typedef struct { +MachineClass parent; +}SampleMachineClass ; + +typedef struct { +MachineState parent; +MemoryRegion *ram; +MemoryRegion *flash; +AVRUsartState *usart0; +AVRTimer16State *timer1; +AVRMaskState *prr[2 ]; +}SampleMachineState ; + +#define TYPE_SAMPLE_MACHINE MACHINE_TYPE_NAME("sample") + +#define SAMPLE_MACHINE (obj) \ + OBJECT_CHECK(SampleMachineState, obj, TYPE_SAMPLE_MACHINE) +#define SAMPLE_MACHINE_GET_CLASS (obj) \ + OBJECT_GET_CLASS(SampleMachineClass, obj, TYPE_SAMPLE_MACHINE) +#define SAMPLE_MACHINE_CLASS (klass) \ + OBJECT_CLASS_CHECK(SampleMachineClass, klass, TYPE_SAMPLE_MACHINE) + +static void sample_init (MachineState *machine) +{ +SampleMachineState *sms =SAMPLE_MACHINE (machine); +MemoryRegion *system_memory =get_system_memory (); +AVRCPU *cpu; +const char *firmware =NULL ; +const char *filename; +int bytes_loaded; +SysBusDevice *busdev; +DeviceState *cpudev; + + system_memory =get_system_memory (); + sms->ram =g_new (MemoryRegion,1 ); + sms->flash =g_new (MemoryRegion,1 ); + + cpu =AVR_CPU (cpu_create (machine->cpu_type)); + cpudev =DEVICE (cpu); + + +memory_region_init_rom (sms->flash,NULL ,"avr.flash" ,SIZE_FLASH , + &error_fatal); +memory_region_add_subregion (system_memory,OFFSET_CODE , sms->flash); + +/* following are atmel2560 device */ +create_unimplemented_device ("usart 3" ,OFFSET_DATA +0x0130 ,0x0007 ); +create_unimplemented_device ("timer-counter-16bit 5" , +OFFSET_DATA +0x0120 ,0x000e ); +create_unimplemented_device ("gpio L" ,OFFSET_DATA +0x0109 ,0x0003 ); +create_unimplemented_device ("gpio K" ,OFFSET_DATA +0x0106 ,0x0003 ); +create_unimplemented_device ("gpio J" ,OFFSET_DATA +0x0103 ,0x0003 ); +create_unimplemented_device ("gpio H" ,OFFSET_DATA +0x0100 ,0x0003 ); +create_unimplemented_device ("usart 2" ,OFFSET_DATA +0x00d0 ,0x0007 ); +create_unimplemented_device ("usart 1" ,OFFSET_DATA +0x00c8 ,0x0007 ); +create_unimplemented_device ("usart 0" ,OFFSET_DATA +0x00c0 ,0x0007 ); +create_unimplemented_device ("twi" ,OFFSET_DATA +0x00b8 ,0x0006 ); +create_unimplemented_device ("timer-counter-async-8bit 2" , +OFFSET_DATA +0x00b0 ,0x0007 ); +create_unimplemented_device ("timer-counter-16bit 4" , +OFFSET_DATA +0x00a0 ,0x000e ); +create_unimplemented_device ("timer-counter-16bit 3" , +OFFSET_DATA +0x0090 ,0x000e ); +create_unimplemented_device ("timer-counter-16bit 1" , +OFFSET_DATA +0x0080 ,0x000e ); +create_unimplemented_device ("ac / adc" , +OFFSET_DATA +0x0078 ,0x0008 ); +create_unimplemented_device ("ext-mem-iface" , +OFFSET_DATA +0x0074 ,0x0002 ); +create_unimplemented_device ("int-controller" , +OFFSET_DATA +0x0068 ,0x000c ); +create_unimplemented_device ("sys" , +OFFSET_DATA +0x0060 ,0x0007 ); +create_unimplemented_device ("spi" , +OFFSET_DATA +0x004c ,0x0003 ); +create_unimplemented_device ("ext-mem-iface" , +OFFSET_DATA +0x004a ,0x0002 ); +create_unimplemented_device ("timer-counter-pwm-8bit 0" , +OFFSET_DATA +0x0043 ,0x0006 ); +create_unimplemented_device ("ext-mem-iface" , +OFFSET_DATA +0x003e ,0x0005 ); +create_unimplemented_device ("int-controller" , +OFFSET_DATA +0x0035 ,0x0009 ); +create_unimplemented_device ("gpio G" ,OFFSET_DATA +0x0032 ,0x0003 ); +create_unimplemented_device ("gpio F" ,OFFSET_DATA +0x002f ,0x0003 ); +create_unimplemented_device ("gpio E" ,OFFSET_DATA +0x002c ,0x0003 ); +create_unimplemented_device ("gpio D" ,OFFSET_DATA +0x0029 ,0x0003 ); +create_unimplemented_device ("gpio C" ,OFFSET_DATA +0x0026 ,0x0003 ); +create_unimplemented_device ("gpio B" ,OFFSET_DATA +0x0023 ,0x0003 ); +create_unimplemented_device ("gpio A" ,OFFSET_DATA +0x0020 ,0x0003 ); + +memory_region_allocate_system_memory ( + sms->ram,NULL ,"avr.ram" ,SIZE_SRAM +SIZE_EXMEM ); + memory_region_add_subregion(system_memory, OFFSET_DATA + 0x200,
sms->ram);hw/avr/sample.c new file +16
-0+ +/* Power Reduction built-in peripheral */ + sms->prr[0 ] =AVR_MASK (sysbus_create_simple (TYPE_AVR_MASK , +OFFSET_DATA +PRR0_BASE ,NULL )); + sms->prr[1 ] =AVR_MASK (sysbus_create_simple (TYPE_AVR_MASK , +OFFSET_DATA +PRR1_BASE ,NULL )); + +/* USART 0 built-in peripheral */ + sms->usart0 =AVR_USART (object_new (TYPE_AVR_USART )); + busdev =SYS_BUS_DEVICE (sms->usart0); +qdev_prop_set_chr (DEVICE (sms->usart0),"chardev" ,serial_hd (0 )); +object_property_set_bool (OBJECT (sms->usart0), true,"realized" , + &error_fatal); +sysbus_mmio_map (busdev,0 ,OFFSET_DATA +USART_BASE ); +/* +* These IRQ numbers don't match the datasheet because we're counting from
hw/avr/sample.c new file +3
-0+* zero and not including reset. +*/ + sysbus_connect_irq(busdev, 0, qdev_get_gpio_in(cpudev,
USART_RXC_IRQ));
+ sysbus_connect_irq(busdev, 1, qdev_get_gpio_in(cpudev,USART_DRE_IRQ));
+ sysbus_connect_irq(busdev, 2, qdev_get_gpio_in(cpudev,USART_TXC_IRQ));hw/avr/sample.c new file +12
-0+sysbus_connect_irq (SYS_BUS_DEVICE (sms->prr[1 ]),PRR1_BIT_PRUSART1 , +qdev_get_gpio_in (DEVICE (sms->usart0),0 )); + +/* Timer 1 built-in periphal */ + sms->timer1 =AVR_TIMER16 (object_new (TYPE_AVR_TIMER16 )); +object_property_set_bool (OBJECT (sms->timer1), true,"realized" , + &error_fatal); + busdev =SYS_BUS_DEVICE (sms->timer1); +sysbus_mmio_map (busdev,0 ,OFFSET_DATA +TIMER1_BASE ); +sysbus_mmio_map (busdev,1 ,OFFSET_DATA +TIMER1_IMSK_BASE ); +sysbus_mmio_map (busdev,2 ,OFFSET_DATA +TIMER1_IFR_BASE ); + sysbus_connect_irq(busdev, 0, qdev_get_gpio_in(cpudev,
TIMER1_CAPT_IRQ));
+ sysbus_connect_irq(busdev, 1, qdev_get_gpio_in(cpudev,TIMER1_COMPA_IRQ));
+ sysbus_connect_irq(busdev, 2, qdev_get_gpio_in(cpudev,TIMER1_COMPB_IRQ));
+ sysbus_connect_irq(busdev, 3, qdev_get_gpio_in(cpudev,TIMER1_COMPC_IRQ));
+ sysbus_connect_irq(busdev, 4, qdev_get_gpio_in(cpudev,TIMER1_OVF_IRQ));hw/avr/sample.c new file +14
-0+sysbus_connect_irq (SYS_BUS_DEVICE (sms->prr[0 ]),PRR0_BIT_PRTIM1 , +qdev_get_gpio_in (DEVICE (sms->timer1),0 )); + +/* Load firmware (contents of flash) trying to auto-detect format */ + firmware = machine->firmware; +if (firmware !=NULL ) { + filename =qemu_find_file (QEMU_FILE_TYPE_BIOS , firmware); +if (filename ==NULL ) { +error_report ("Unable to find %s" , firmware); +exit (1 ); + } + + bytes_loaded = load_elf( + filename, NULL, NULL, NULL, NULL, NULL, NULL, 0, EM_NONE, 0,0);
hw/avr/sample.c new file +40
-0+if (bytes_loaded <0 ) { + bytes_loaded =load_image_targphys ( + filename,OFFSET_CODE ,SIZE_FLASH ); + } +if (bytes_loaded <0 ) { +error_report ( +"Unable to load firmware image %s as ELF or raw binary" , + firmware); +exit (1 ); + } + } +} + +static void sample_class_init (ObjectClass *oc,void *data) +{ +MachineClass *mc =MACHINE_CLASS (oc); + + mc->desc ="AVR sample/example board (ATmega2560)" ; + mc->init = sample_init; + mc->default_cpus =1 ; + mc->min_cpus = mc->default_cpus; + mc->max_cpus = mc->default_cpus; + mc->default_cpu_type ="avr6-avr-cpu" ;/* ATmega2560. */ + mc->is_default =1 ; +} + +static const TypeInfo sample_info = { + .name =TYPE_SAMPLE_MACHINE , + .parent =TYPE_MACHINE , + .instance_size =sizeof (SampleMachineState), + .class_size =sizeof (SampleMachineClass), + .class_init = sample_class_init, +}; + +static void sample_machine_init (void ) +{ +type_register_static (&sample_info); +} + +type_init (sample_machine_init);hw/Kconfig +1
-0@@ -44,6 +44,7 @@ source watchdog/Kconfig # arch Kconfig source arm/Kconfig source alpha/Kconfig +source avr/Kconfig source cris/Kconfig source hppa/Kconfig source i386/Kconfighw/avr/Kconfig new file +6
-0@@ -0,0 +1,6 @@ +config AVR_SAMPLE + bool + select AVR_TIMER16 + select AVR_USART + select AVR_MASK + select UNIMPhw/avr/Makefile.objs new file +1
-1@@ -0,0 +1 @@ +obj-y += sample.o --2.17.2 (Apple Git-113)
Signature · 2 lines
Best Regards, Michael Rolnik
Re: [PATCH v37 05/17] target/avr: Add instruction translation - Arithmetic and Logic Instructions
In reply to Michael RolnikQuoted 5 lines · Michael Rolnik
Hi Aleksandar.
thanks for pointing that out I was not aware of that. I can fix it.
That'd be great!
Thanks,
Aleksandar
Quoted 82 lines
Regards,
Michael RolnikQuoted 54 lines · Aleksandar Markovic
Quoted 42 lines · Aleksandar Markovic
On Wednesday, November 27, 2019, Michael Rolnik <[email protected]> wrote:
+
Quoted 23 lines
target/avr/translate.c +3
-0+ +/* +* Performs the logical AND between the contents of register Rd and register
+ * Rr and places the result in the destination register Rd.
+ */target/avr/translate.c +16
-0+static bool trans_AND (DisasContext *ctx,arg_AND *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_and_tl (R , Rd, Rr);/* Rd = Rd and Rr */ + +/* Vf */ +tcg_gen_movi_tl (cpu_Vf,0 );/* Vf = 0 */ + +/* Zf */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +gen_ZNSf (R );target/avr/translate.c static bool avr_have_feature(DisasContext *ctx, int feature) · line 555
Quoted 10 lines
target/avr/translate.c +10
-0+ +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + +According to specs:
Quoted 7 lines
http://ww1.microchip.com/downloads/en/devicedoc/atmel-42505-8-bit-avr-
microcontrollers-attiny102-attiny104_datasheet.pdf... the chips in question have cores with 16 GPRs (that correspond to GPRs R16-R31 of more common AVR cores with 32 GPRs).
Quoted 16 lines
There are more examples;
http://ww1.microchip.com/downloads/en/DeviceDoc/atmel-8127-avr-8-bit-
microcontroller-attiny4-attiny5-attiny9-attiny10_datasheet.pdfAlso ATtiny20, ATtiny40.
How do you handle such cores?
Quoted 7 lines
I don't see here anything preventing usage of all 32 GPRs, resulting, of course, in an inaccurate emulation.
Thanks,
Aleksandar--
Best Regards,
Michael Rolnik
There is *-cpu *option where you can specify what CPU you want, if this option is not specified avr6 (avr6-avr-cpu) is chosen.
*./avr-softmmu/qemu-system-avr -cpu help*
avr1-avr-cpu
avr2-avr-cpu
avr25-avr-cpu
avr3-avr-cpu
avr31-avr-cpu
avr35-avr-cpu
avr4-avr-cpu
avr5-avr-cpu
avr51-avr-cpu
avr6-avr-cpu
xmega2-avr-cpu
xmega4-avr-cpu
xmega5-avr-cpu
xmega6-avr-cpu
xmega7-avr-cpu
Regards,
Michael Rolnik
Quoted 403 lines · Aleksandar Markovic
Quoted 17 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fully tested yet.
However I was able to execute simple code with functions. e.g fibonacci calculation.
This series of patches include a non real, sample board.
No fuses support yet. PC is set to 0 at reset.Following are examples of possible usages, assuming program.elf is compiled for AVR cpu
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnet window
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
runtelent localhost 5678Quoted 10 lines
3. Continious non interrupted execution with serial output into stdout run
qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -S
runavr-gdb program.elfand then within GDB shell `target remote
:1234`
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmHi, Michael.
Can you explain to me the mechanisms of recognition of what core/microcontroller QEMU is supposed to emulate in the examples above?
Yours, Aleksandar
Quoted 363 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding or translation
2. CPU features which allow define the following 8bit AVR coresavr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows to execute simple code
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of0x000000ff) for faster address manipulaton
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM
11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHchanges since v4
1. target is now defined as big endian in order to optimize push_ret/pop_ret
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_cpu_gdb_write_register
2. configure the target as little endian because otherwise GDB does not work
3. fixing and testing gen_push_ret/gen_pop_retchanges since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supported yet registers only
3.MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built for all
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memorychanges since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed to any device
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try to access registerschanges since v10
1. movig back fullwr & fullrd into the commit where outb and inb were introduced
2. changingtlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codechanges since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions 12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thing exists)
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrischanges since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers are accessible
by LD/ST only. CPU has no interest in themchanges since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixes a
couple of bugs wrt single stepping
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicchanges since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU" into three patches 1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tename1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to MAINTAINERS 7. use git commit sha1 instead of `master`` in avocado testMichael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileSarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer peripheralsconfigure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)
Signature · 2 lines
Best Regards, Michael Rolnik
Re: [PATCH v37 05/17] target/avr: Add instruction translation - Arithmetic and Logic Instructions
In reply to Aleksandar MarkovicHi Aleksandar.
thanks for pointing that out I was not aware of that. I can fix it.
Regards,
Michael Rolnik
Quoted 70 lines · Aleksandar Markovic
Quoted 43 lines · Aleksandar Markovic
On Wednesday, November 27, 2019, Michael Rolnik <[email protected]> wrote:
+
Quoted 23 lines
target/avr/translate.c +3
-0+ +/* +* Performs the logical AND between the contents of register Rd and register
+ * Rr and places the result in the destination register Rd.
+ */target/avr/translate.c +16
-0+static bool trans_AND (DisasContext *ctx,arg_AND *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_and_tl (R , Rd, Rr);/* Rd = Rd and Rr */ + +/* Vf */ +tcg_gen_movi_tl (cpu_Vf,0 );/* Vf = 0 */ + +/* Zf */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +gen_ZNSf (R );target/avr/translate.c static bool avr_have_feature(DisasContext *ctx, int feature) · line 555
Quoted 10 lines
target/avr/translate.c +10
-0+ +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + +According to specs:
Quoted 6 lines
... the chips in question have cores with 16 GPRs (that correspond to GPRs R16-R31 of more common AVR cores with 32 GPRs).
There are more examples;
Also ATtiny20, ATtiny40.
How do you handle such cores?
Quoted 7 lines
I don't see here anything preventing usage of all 32 GPRs, resulting, of course, in an inaccurate emulation.
Thanks,
Aleksandar
Signature · 2 lines
Best Regards, Michael Rolnik
Re: [PATCH v37 05/17] target/avr: Add instruction translation - Arithmetic and Logic Instructions
Quoted 6 lines · Michael Rolnik
Hi Aleksandar.
thanks for pointing that out I was not aware of that. I can fix it.
Hey, Michael,
Please take alook at this:
https://en.m.wikipedia.org/wiki/ATtiny_microcontroller_comparison_chart
It looks that all cases of AVR 16-gpr-registers cores belong to the group "avrtiny10",
that actually you may want to add to your solution. Just a hint.
Best regards,
Aleksandar
Regards,
Quoted 80 lines
Michael Rolnik
Quoted 54 lines · Aleksandar Markovic
Quoted 42 lines · Aleksandar Markovic
On Wednesday, November 27, 2019, Michael Rolnik <[email protected]> wrote:
+
Quoted 23 lines
target/avr/translate.c +3
-0+ +/* +* Performs the logical AND between the contents of register Rd and register
+ * Rr and places the result in the destination register Rd.
+ */target/avr/translate.c +16
-0+static bool trans_AND (DisasContext *ctx,arg_AND *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_and_tl (R , Rd, Rr);/* Rd = Rd and Rr */ + +/* Vf */ +tcg_gen_movi_tl (cpu_Vf,0 );/* Vf = 0 */ + +/* Zf */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +gen_ZNSf (R );target/avr/translate.c static bool avr_have_feature(DisasContext *ctx, int feature) · line 555
Quoted 10 lines
target/avr/translate.c +10
-0+ +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + +According to specs:
Quoted 7 lines
http://ww1.microchip.com/downloads/en/devicedoc/atmel-42505-8-bit-avr-
microcontrollers-attiny102-attiny104_datasheet.pdf... the chips in question have cores with 16 GPRs (that correspond to GPRs R16-R31 of more common AVR cores with 32 GPRs).
Quoted 16 lines
There are more examples;
http://ww1.microchip.com/downloads/en/DeviceDoc/atmel-8127-avr-8-bit-
microcontroller-attiny4-attiny5-attiny9-attiny10_datasheet.pdfAlso ATtiny20, ATtiny40.
How do you handle such cores?
Quoted 7 lines
I don't see here anything preventing usage of all 32 GPRs, resulting, of course, in an inaccurate emulation.
Thanks,
Aleksandar--
Best Regards,
Michael Rolnik
Quoted 8 lines
Renaming devices such
hw/char/avr_usart.c->hw/char/atmel_usart.c(similarly with the macros) would be enough Aleksandar?On Thursday, November 28, 2019, Michael Rolnik <[email protected]> wrote:
I will rename them.
AVR is the name of a microcontroller lineup, and Atmel is the name of the company that started producing them. Atmel was recently acquired by Microchip, so thw word Atmel now does not even exist in new specs.
Taking this into account, I don't think renaming
hw/char/avr_usart.c -> hw/char/atmel_usart.c
is not appropriate. Renaming macros, too. The current names are fine, for now.
A separate but related naming question will show up later in future, when we, let's say, want to implement two different version of a peripheral (let's say USART), one as specified for older microcontrollers, and one for newer.
But, OK, let's leave that for future.
Regards,
Aleksandar
Quoted 144 lines
Quoted 130 lines · Aleksandar Markovic
Quoted 117 lines · Philippe Mathieu-Daudé
Quoted 108 lines · Michael Rolnik
I don't see why you say that the peripherals are inside the chip, there is CPU within target/avr directory and then there are some peripherals in hw directory, CPU does not depend on them. what am I missing?
On Thu, Nov 28, 2019 at 3:22 PM Aleksandar Markovic < [email protected] <mailto:[email protected]>> wrote:
On Thursday, November 28, 2019, Michael Rolnik <[email protected] <mailto:[email protected]>> wrote:
On Wed, Nov 27, 2019 at 11:06 PM Aleksandar Markovic <[email protected] <mailto:[email protected]>> wrote:
On Wed, Nov 27, 2019 at 6:53 PM Michael Rolnik <[email protected] <mailto:[email protected]>> wrote:
>
> This series of patches adds 8bit AVR cores to QEMU. > All instruction, except BREAK/DES/SPM/SPMX, are
implemented. Not fully tested yet.
> However I was able to execute simple code with functions.
e.g fibonacci calculation.
> This series of patches include a non real, sample board. > No fuses support yet. PC is set to 0 at reset. >I have a couple of general remarks, so I am responding to the cover
letter, not individual patches.1) The licenses for Sarah devices differ than the rest - shouldn't all
licenses be harmonized?Sarah,
do you mind if use the same license I use for my code?2) There is an architectural problem with peripherals. It is possible
that they evolve over time, so, for example, USART could not be the
same for older and newer CPUs (in principle, newer
peripheral is
expected to be o sort of "superset" of the older). How do you solve
that problem? Right now, it may not looks serious to you, but if you
don;t think about that right now, from the outset, soon the code will
become so entangled, ti woudl be almost very difficult to fix it.
Please think about that, how would you solve it, is there a way to
pass the information on the currently emulated CPU to the
code
covering a peripheral, and provide a different behaviour?Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the
core.If one models, let's say an external (in the sense, it is a separate chip) ADC (analog-to-digital converter), one looks at specs, implement what is resonable possible in QEMU, plug it in in one of machines thst contains it, and that's it. That ADC remains the same, of course, whatever the surrounding system is.
In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers is very large, and both cores and peripherals
evolved.For cores, you handle differences with all these
AVR_FEATUREmacros, and this seems to be working, no significant objection from my side, and btw that was not an easy task to execute, all admiration from
me.But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't have time right now to spell many examples, but I read three different specs, and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if they are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on the core it is attached to?
I hope I was clearer this time.
Aleksandar
Quoted 6 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there is sample.Quoted 7 lines
If I understand Aleksandar correctly, the naming is incorrect because too generic to AVR family, why Sarah only modeled the Atmel implementation.
Renaming devices such
hw/char/avr_usart.c->hw/char/atmel_usart.c(similarly with the macros) would be enough Aleksandar?Quoted 5 lines
Some renaming could help, perhaps not quite like the one above, but my point (which I find hard to believe I can't explain to you) is that peripherals inside the chip evolved over time, as starkly opposed to external peripherals that are set in stone...
--
Best Regards,
Michael Rolnik
Quoted 12 lines · Aleksandar Markovic
Renaming devices such
hw/char/avr_usart.c->hw/char/atmel_usart.cQuoted 6 lines
(similarly with the macros) would be enough Aleksandar?
On Thursday, November 28, 2019, Michael Rolnik <[email protected]> wrote:
I will rename them.
AVR is the name of a microcontroller lineup, and Atmel is the name of the company that started producing them. Atmel was recently acquired by Microchip, so thw word Atmel now does not even exist in new specs.
Quoted 7 lines
Taking this into account, I don't think renaming
hw/char/avr_usart.c->hw/char/atmel_usart.cis not appropriate.
I meant to say the renaming is not appropriate. Sorry for confusion.
Quoted 168 lines
Renaming macros, too. The current names are fine, for now.
A separate but related naming question will show up later in future, when we, let's say, want to implement two different version of a peripheral (let's say USART), one as specified for older microcontrollers, and one for newer.
But, OK, let's leave that for future.
Regards,
AleksandarQuoted 151 lines
Quoted 137 lines · Aleksandar Markovic
Quoted 124 lines · Philippe Mathieu-Daudé
Quoted 115 lines · Michael Rolnik
I don't see why you say that the peripherals are inside the chip, there is CPU within target/avr directory and then there are some peripherals in hw directory, CPU does not depend on them. what am I missing?
On Thu, Nov 28, 2019 at 3:22 PM Aleksandar Markovic < [email protected] <mailto:[email protected]>> wrote:
On Thursday, November 28, 2019, Michael Rolnik <[email protected] <mailto:[email protected]>> wrote:
On Wed, Nov 27, 2019 at 11:06 PM Aleksandar Markovic <[email protected] <mailto:[email protected]>> wrote:
On Wed, Nov 27, 2019 at 6:53 PM Michael Rolnik <[email protected] <mailto:[email protected]>> wrote:
>
> This series of patches adds 8bit AVR cores to QEMU. > All instruction, except BREAK/DES/SPM/SPMX, are
implemented. Not fully tested yet.
> However I was able to execute simple code with
functions.
e.g fibonacci calculation.
> This series of patches include a non real, sample board. > No fuses support yet. PC is set to 0 at reset. >I have a couple of general remarks, so I am responding to the cover
letter, not individual patches.1) The licenses for Sarah devices differ than the rest - shouldn't all
licenses be harmonized?Sarah,
do you mind if use the same license I use for my code?2) There is an architectural problem with peripherals. It
is
possible
that they evolve over time, so, for example, USART could
not
be the
same for older and newer CPUs (in principle, newer
peripheral is
expected to be o sort of "superset" of the older). How do you solve
that problem? Right now, it may not looks serious to you, but if you
don;t think about that right now, from the outset, soon the code will
become so entangled, ti woudl be almost very difficult to fix it.
Please think about that, how would you solve it, is there a way to
pass the information on the currently emulated CPU to the
code
covering a peripheral, and provide a different behaviour?Hi Aleksandar,
Please explain.
My concern is about peripherals inside the chip, together with the
core.If one models, let's say an external (in the sense, it is a
separate
chip) ADC (analog-to-digital converter), one looks at specs, implement what is resonable possible in QEMU, plug it in in one of machines thst contains it, and that's it. That ADC remains the
same,
of course, whatever the surrounding system is.In AVR case, I think we have a phenomenon likes of which we didn't see before (at least I don't know about). Number of AVR microcontrollers is very large, and both cores and peripherals
evolved.For cores, you handle differences with all these
AVR_FEATURE
macros,
and this seems to be working, no significant objection from my
side,
and btw that was not an easy task to execute, all admiration from
me.But what about peripherals inside the chip? A peripheral with the same name and the same general area of functionality may be differently specified for microcontrollers from 2010 and 2018. By the difference I don't mean starting address, but the difference in behavior. I don't have time right now to spell many examples, but I read three different specs, and there are differences in USART specifications.
I am not clear what is your envisioned solution for these cases. Would you such close, but not the same, flabors of a peripheral treat as if they are two completely separate cases of a peripheral? Or would you have a single peripheral that would somehow configure itself depending on the core it is attached to?
I hope I was clearer this time.
Aleksandar
Quoted 6 lines
I don't see any problem from CPU's perspective.
as for the sample board is just a sample, I hope other people will create real models or real hw.
there was no way I could provide a CPU alone, that's why there is sample.Quoted 7 lines
If I understand Aleksandar correctly, the naming is incorrect because too generic to AVR family, why Sarah only modeled the Atmel implementation.
Renaming devices such
hw/char/avr_usart.c->hw/char/atmel_usart.c(similarly with the macros) would be enough Aleksandar?Quoted 5 lines
Some renaming could help, perhaps not quite like the one above, but my point (which I find hard to believe I can't explain to you) is that peripherals inside the chip evolved over time, as starkly opposed to external peripherals that are set in stone...
--
Best Regards,
Michael Rolnik
Re: [PATCH v37 10/17] target/avr: Add instruction disassembly function
In reply to Michael RolnikQuoted 73 lines · Michael Rolnik
Provide function disassembles executed instruction when
-d in_asmis providedExample:
./avr-softmmu/qemu-system-avr -bios free-rtos/Demo/AVR_ATMega2560_GCC/demo.elf -d in_asmwill produce something like the following``` ... IN: 0x0000014a: CALL 0x3808 IN: main 0x00003808: CALL 0x4b4 IN: vParTestInitialise 0x000004b4: LDI r24, 255 0x000004b6: STS r24, 0 0x000004b8: MULS r16, r20 0x000004ba: OUT $1, r24 0x000004bc: LDS r24, 0 0x000004be: MULS r16, r20 0x000004c0: OUT $2, r24 0x000004c2: RET ... ```Signed-off-by Michael Rolnik <[email protected]>Suggested-by Richard Henderson <[email protected]>Suggested-by Philippe Mathieu-Daudé <[email protected]>Suggested-by Aleksandar Markovic <[email protected]>Reviewed-by Philippe Mathieu-Daudé <[email protected]>--- target/avr/cpu.h | 1 + target/avr/cpu.c | 2 +- target/avr/disas.c | 228 +++++++++++++++++++++++++++++++++++++++++ target/avr/translate.c | 11 ++ 4 files changed, 241 insertions(+), 1 deletion(-) create mode 100644 target/avr/disas.ctarget/avr/cpu.h +1
-0@@ -157,6 +157,7 @@ bool avr_cpu_exec_interrupt(CPUState *cpu, int int_req);hwaddr avr_cpu_get_phys_page_debug (CPUState *cpu,vaddr addr);int avr_cpu_gdb_read_register (CPUState *cpu,uint8_t *buf,int reg);int avr_cpu_gdb_write_register (CPUState *cpu,uint8_t *buf,int reg); +int avr_print_insn (bfd_vma addr,disassemble_info *info);static inline int avr_feature (CPUAVRState *env,int feature){
target/avr/cpu.c +1
-1@@ -83,7 +83,7 @@ static void avr_cpu_reset(CPUState *cs)static void avr_cpu_disas_set_info (CPUState *cpu,disassemble_info *info) { info->mach = bfd_arch_avr; - info->print_insn =NULL ; + info->print_insn = avr_print_insn; }static void avr_cpu_realizefn (DeviceState *dev,Error **errp)target/avr/disas.c new file +0
-0
target/avr/disas.c line 4
Quoted 4 lines
target/avr/disas.c new file +4
-0@@ -0,0 +1,228 @@ +/* +* AVR disassembler +* +* Copyright (c) 2018 Richard Henderson <[email protected]>
Just a detail: since this file is created in 2019, the copyright year should be 2019 too.
+ * Copyright (c) 2019 Michael Rolnik <[email protected]>
Quoted 249 lines
target/avr/disas.c new file +223
-0+* +* This program is free software: you can redistribute it and/or modify +* it under the terms of the GNU General Public License as published by +* the Free Software Foundation, either version 2 of the License, or +* (at your option) any later version. +* +* This program is distributed in the hope that 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/>. +*/ + +#include "qemu/osdep.h" +#include "disas/dis-asm.h" +#include "qemu/bitops.h" +#include "cpu.h" + +typedef struct { +disassemble_info *info; +uint16_t next_word; +bool next_word_used; +}DisasContext ; + +static int to_regs_16_31_by_one (DisasContext *ctx,int indx) +{ +return 16 + (indx %16 ); +} + +static int to_regs_16_23_by_one (DisasContext *ctx,int indx) +{ +return 16 + (indx %8 ); +} +static int to_regs_24_30_by_two (DisasContext *ctx,int indx) +{ +return 24 + (indx %4 ) *2 ; +} +static int to_regs_00_30_by_two (DisasContext *ctx,int indx) +{ +return (indx %16 ) *2 ; +} + +static uint16_t next_word (DisasContext *ctx) +{ + ctx->next_word_used = true; +return ctx->next_word; +} + +static int append_16 (DisasContext *ctx,int x) +{ +return x <<16 |next_word (ctx); +} + + +/* Include the auto-generated decoder. */ +static bool decode_insn (DisasContext *ctx,uint16_t insn); +#include "decode_insn.inc.c" + +#define output (mnemonic, format, ...) \ + (pctx->info->fprintf_func(pctx->info->stream, "%-9s " format, \ + mnemonic, ##__VA_ARGS__)) + +int avr_print_insn (bfd_vma addr,disassemble_info *info) +{ +DisasContext ctx; +DisasContext *pctx = &ctx; +bfd_byte buffer[4 ]; +uint16_t insn; +int status; + + ctx.info = info; + + status = info->read_memory_func (addr, buffer,4 , info); +if (status !=0 ) { + info->memory_error_func (status, addr, info); +return -1 ; + } + insn =bfd_getl16 (buffer); + ctx.next_word =bfd_getl16 (buffer +2 ); + ctx.next_word_used = false; + +if (!decode_insn (&ctx, insn)) { +output (".db" ,"0x%02x, 0x%02x" , buffer[0 ], buffer[1 ]); + } + +return ctx.next_word_used ?4 :2 ; +} + + +#define INSN (opcode, format, ...) \ +static bool trans_##opcode(DisasContext *pctx, arg_##opcode * a) \ +{ \ + output(#opcode, format, ##__VA_ARGS__); \ + return true; \ +} + +#define INSN_MNEMONIC (opcode, mnemonic, format, ...) \ +static bool trans_##opcode(DisasContext *pctx, arg_##opcode * a) \ +{ \ + output(mnemonic, format, ##__VA_ARGS__); \ + return true; \ +} + +/* +* C Z N V S H T I +* 0 1 2 3 4 5 6 7 +*/ +static const char *brbc[] = { +"BRCC" ,"BRNE" ,"BRPL" ,"BRVC" ,"BRGE" ,"BRHC" ,"BRTC" ,"BRID" +}; + +static const char *brbs[] = { +"BRCS" ,"BREQ" ,"BRMI" ,"BRVS" ,"BRLT" ,"BRHS" ,"BRTS" ,"BRIE" +}; + +static const char *bset[] = { +"SEC" ,"SEZ" ,"SEN" ,"SEZ" ,"SES" ,"SEH" ,"SET" ,"SEI" +}; + +static const char *bclr[] = { +"CLC" ,"CLZ" ,"CLN" ,"CLZ" ,"CLS" ,"CLH" ,"CLT" ,"CLI" +}; + +INSN(ADC, "r%d, r%d", a->rd, a->rr) +INSN(ADD, "r%d, r%d", a->rd, a->rr) +INSN(ADIW, "r%d:r%d, %d", a->rd + 1, a->rd, a->imm) +INSN(AND, "r%d, r%d", a->rd, a->rr) +INSN(ANDI, "r%d, %d", a->rd, a->imm) +INSN(ASR, "r%d", a->rd) +INSN_MNEMONIC(BCLR, bclr[a->bit], "") +INSN(BLD, "r%d, %d", a->rd, a->bit) +INSN_MNEMONIC(BRBC, brbc[a->bit], ".%+d", a->imm * 2) +INSN_MNEMONIC(BRBS, brbs[a->bit], ".%+d", a->imm * 2) +INSN(BREAK, "") +INSN_MNEMONIC(BSET, bset[a->bit], "") +INSN(BST, "r%d, %d", a->rd, a->bit) +INSN(CALL, "0x%x", a->imm * 2) +INSN(CBI, "%d, %d", a->reg, a->bit) +INSN(COM, "r%d", a->rd) +INSN(CP, "r%d, r%d", a->rd, a->rr) +INSN(CPC, "r%d, r%d", a->rd, a->rr) +INSN(CPI, "r%d, %d", a->rd, a->imm) +INSN(CPSE, "r%d, r%d", a->rd, a->rr) +INSN(DEC, "r%d", a->rd) +INSN(DES, "%d", a->imm) +INSN(EICALL, "") +INSN(EIJMP, "") +INSN(ELPM1, "") +INSN(ELPM2, "r%d, Z", a->rd) +INSN(ELPMX, "r%d, Z+", a->rd) +INSN(EOR, "r%d, r%d", a->rd, a->rr) +INSN(FMUL, "r%d, r%d", a->rd, a->rr) +INSN(FMULS, "r%d, r%d", a->rd, a->rr) +INSN(FMULSU, "r%d, r%d", a->rd, a->rr) +INSN(ICALL, "") +INSN(IJMP, "") +INSN(IN, "r%d, $%d", a->rd, a->imm) +INSN(INC, "r%d", a->rd) +INSN(JMP, "0x%x", a->imm * 2) +INSN(LAC, "Z, r%d", a->rd) +INSN(LAS, "Z, r%d", a->rd) +INSN(LAT, "Z, r%d", a->rd) +INSN(LDDY, "r%d, Y+%d", a->rd, a->imm) +INSN(LDDZ, "r%d, Z+%d", a->rd, a->imm) +INSN(LDI, "r%d, %d", a->rd, a->imm) +INSN(LDS, "r%d, %d", a->rd, a->imm) +INSN(LDX1, "r%d, X", a->rd) +INSN(LDX2, "r%d, X+", a->rd) +INSN(LDX3, "r%d, -X", a->rd) +INSN(LDY2, "r%d, Y+", a->rd) +INSN(LDY3, "r%d, -Y", a->rd) +INSN(LDZ2, "r%d, Z+", a->rd) +INSN(LDZ3, "r%d, -Z", a->rd) +INSN(LPM1, "") +INSN(LPM2, "r%d, Z", a->rd) +INSN(LPMX, "r%d, Z+", a->rd) +INSN(LSR, "r%d", a->rd) +INSN(MOV, "r%d, r%d", a->rd, a->rr) +INSN(MOVW, "r%d:r%d, r%d,r:r%d", a->rd + 1, a->rd, a->rr + 1, a->rr) +INSN(MUL, "r%d, r%d", a->rd, a->rr) +INSN(MULS, "r%d, r%d", a->rd, a->rr) +INSN(MULSU, "r%d, r%d", a->rd, a->rr) +INSN(NEG, "r%d", a->rd) +INSN(NOP, "") +INSN(OR, "r%d, r%d", a->rd, a->rr) +INSN(ORI, "r%d, %d", a->rd, a->imm) +INSN(OUT, "$%d, r%d", a->imm, a->rd) +INSN(POP, "r%d", a->rd) +INSN(PUSH, "r%d", a->rd) +INSN(RCALL, ".%+d", a->imm * 2) +INSN(RET, "") +INSN(RETI, "") +INSN(RJMP, ".%+d", a->imm * 2) +INSN(ROR, "r%d", a->rd) +INSN(SBC, "r%d, r%d", a->rd, a->rr) +INSN(SBCI, "r%d, %d", a->rd, a->imm) +INSN(SBI, "$%d, %d", a->reg, a->bit) +INSN(SBIC, "$%d, %d", a->reg, a->bit) +INSN(SBIS, "$%d, %d", a->reg, a->bit) +INSN(SBIW, "r%d:r%d, %d", a->rd + 1, a->rd, a->imm) +INSN(SBRC, "r%d, %d", a->rr, a->bit) +INSN(SBRS, "r%d, %d", a->rr, a->bit) +INSN(SLEEP, "") +INSN(SPM, "") +INSN(SPMX, "Z+") +INSN(STDY, "r%d, Y+%d", a->rd, a->imm) +INSN(STDZ, "r%d, Z+%d", a->rd, a->imm) +INSN(STS, "r%d, %d", a->rd, a->imm) +INSN(STX1, "r%d, X", a->rr) +INSN(STX2, "r%d, X+", a->rr) +INSN(STX3, "r%d, -X", a->rr) +INSN(STY2, "r%d, Y+", a->rd) +INSN(STY3, "r%d, -Y", a->rd) +INSN(STZ2, "r%d, Z+", a->rd) +INSN(STZ3, "r%d, -Z", a->rd) +INSN(SUB, "r%d, r%d", a->rd, a->rr) +INSN(SUBI, "r%d, %d", a->rd, a->imm) +INSN(SWAP, "r%d", a->rd) +INSN(WDR, "") +INSN (XCH ,"Z, r%d" , a->rd) +target/avr/translate.c +11
-1@@ -3032,6 +3032,17 @@ done_generating: tb->size = (ctx.npc - pc_start) *2 ; tb->icount = num_insns; + +#ifdef DEBUG_DISAS +if (qemu_loglevel_mask (CPU_LOG_TB_IN_ASM ) + &&qemu_log_in_addr_range (tb->pc)) { +qemu_log_lock (); +qemu_log ("IN: %s\n" ,lookup_symbol (tb->pc)); +log_target_disas (cs, tb->pc, tb->size); +qemu_log ("\n" ); +qemu_log_unlock (); + } +#endif } void restore_state_to_opc(CPUAVRState *env, TranslationBlock *tb, --2.17.2 (Apple Git-113)
Re: [PATCH v37 10/17] target/avr: Add instruction disassembly function
Aleksandar.
If this code is going to be merge in 2019 I should modify al the copyrights, right. or should I put 2020 in?
Regards,
Michael Rolnik
Quoted 337 lines · Aleksandar Markovic
Quoted 74 lines · Michael Rolnik
Provide function disassembles executed instruction when
-d in_asmis providedExample:
./avr-softmmu/qemu-system-avr -bios free-rtos/Demo/AVR_ATMega2560_GCC/demo.elf -d in_asmwill produce something like the following``` ... IN: 0x0000014a: CALL 0x3808 IN: main 0x00003808: CALL 0x4b4 IN: vParTestInitialise 0x000004b4: LDI r24, 255 0x000004b6: STS r24, 0 0x000004b8: MULS r16, r20 0x000004ba: OUT $1, r24 0x000004bc: LDS r24, 0 0x000004be: MULS r16, r20 0x000004c0: OUT $2, r24 0x000004c2: RET ... ```Signed-off-by Michael Rolnik <[email protected]>Suggested-by Richard Henderson <[email protected]>Suggested-by Philippe Mathieu-Daudé <[email protected]>Suggested-by Aleksandar Markovic <[email protected]>Reviewed-by Philippe Mathieu-Daudé <[email protected]>--- target/avr/cpu.h | 1 + target/avr/cpu.c | 2 +- target/avr/disas.c | 228 +++++++++++++++++++++++++++++++++++++++++ target/avr/translate.c | 11 ++ 4 files changed, 241 insertions(+), 1 deletion(-) create mode 100644 target/avr/disas.ctarget/avr/cpu.h +1
-0@@ -157,6 +157,7 @@ bool avr_cpu_exec_interrupt(CPUState *cpu, int int_req);hwaddr avr_cpu_get_phys_page_debug (CPUState *cpu,vaddr addr);int avr_cpu_gdb_read_register (CPUState *cpu,uint8_t *buf,int reg);int avr_cpu_gdb_write_register (CPUState *cpu,uint8_t *buf,int reg); +int avr_print_insn (bfd_vma addr,disassemble_info *info);static inline int avr_feature (CPUAVRState *env,int feature){
target/avr/cpu.c +1
-1@@ -83,7 +83,7 @@ static void avr_cpu_reset(CPUState *cs)static void avr_cpu_disas_set_info (CPUState *cpu,disassemble_info *info) { info->mach = bfd_arch_avr; - info->print_insn =NULL ; + info->print_insn = avr_print_insn; }static void avr_cpu_realizefn (DeviceState *dev,Error **errp)target/avr/disas.c new file +0
-0target/avr/disas.c line 4
Quoted 4 lines
target/avr/disas.c new file +4
-0@@ -0,0 +1,228 @@ +/* +* AVR disassembler +* +* Copyright (c) 2018 Richard Henderson <[email protected]> Just a detail: since this file is created in 2019, the copyright year should be 2019 too.
+ * Copyright (c) 2019 Michael Rolnik <[email protected]>
Quoted 249 lines
target/avr/disas.c new file +223
-0+* +* This program is free software: you can redistribute it and/or modify +* it under the terms of the GNU General Public License as published by +* the Free Software Foundation, either version 2 of the License, or +* (at your option) any later version. +* +* This program is distributed in the hope that 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/>. +*/ + +#include "qemu/osdep.h" +#include "disas/dis-asm.h" +#include "qemu/bitops.h" +#include "cpu.h" + +typedef struct { +disassemble_info *info; +uint16_t next_word; +bool next_word_used; +}DisasContext ; + +static int to_regs_16_31_by_one (DisasContext *ctx,int indx) +{ +return 16 + (indx %16 ); +} + +static int to_regs_16_23_by_one (DisasContext *ctx,int indx) +{ +return 16 + (indx %8 ); +} +static int to_regs_24_30_by_two (DisasContext *ctx,int indx) +{ +return 24 + (indx %4 ) *2 ; +} +static int to_regs_00_30_by_two (DisasContext *ctx,int indx) +{ +return (indx %16 ) *2 ; +} + +static uint16_t next_word (DisasContext *ctx) +{ + ctx->next_word_used = true; +return ctx->next_word; +} + +static int append_16 (DisasContext *ctx,int x) +{ +return x <<16 |next_word (ctx); +} + + +/* Include the auto-generated decoder. */ +static bool decode_insn (DisasContext *ctx,uint16_t insn); +#include "decode_insn.inc.c" + +#define output (mnemonic, format, ...) \ + (pctx->info->fprintf_func(pctx->info->stream, "%-9s " format, \ + mnemonic, ##__VA_ARGS__)) + +int avr_print_insn (bfd_vma addr,disassemble_info *info) +{ +DisasContext ctx; +DisasContext *pctx = &ctx; +bfd_byte buffer[4 ]; +uint16_t insn; +int status; + + ctx.info = info; + + status = info->read_memory_func (addr, buffer,4 , info); +if (status !=0 ) { + info->memory_error_func (status, addr, info); +return -1 ; + } + insn =bfd_getl16 (buffer); + ctx.next_word =bfd_getl16 (buffer +2 ); + ctx.next_word_used = false; + +if (!decode_insn (&ctx, insn)) { +output (".db" ,"0x%02x, 0x%02x" , buffer[0 ], buffer[1 ]); + } + +return ctx.next_word_used ?4 :2 ; +} + + +#define INSN (opcode, format, ...) \ +static bool trans_##opcode(DisasContext *pctx, arg_##opcode * a) \ +{ \ + output(#opcode, format, ##__VA_ARGS__); \ + return true; \ +} + +#define INSN_MNEMONIC (opcode, mnemonic, format, ...) \ +static bool trans_##opcode(DisasContext *pctx, arg_##opcode * a) \ +{ \ + output(mnemonic, format, ##__VA_ARGS__); \ + return true; \ +} + +/* +* C Z N V S H T I +* 0 1 2 3 4 5 6 7 +*/ +static const char *brbc[] = { +"BRCC" ,"BRNE" ,"BRPL" ,"BRVC" ,"BRGE" ,"BRHC" ,"BRTC" ,"BRID" +}; + +static const char *brbs[] = { +"BRCS" ,"BREQ" ,"BRMI" ,"BRVS" ,"BRLT" ,"BRHS" ,"BRTS" ,"BRIE" +}; + +static const char *bset[] = { +"SEC" ,"SEZ" ,"SEN" ,"SEZ" ,"SES" ,"SEH" ,"SET" ,"SEI" +}; + +static const char *bclr[] = { +"CLC" ,"CLZ" ,"CLN" ,"CLZ" ,"CLS" ,"CLH" ,"CLT" ,"CLI" +}; + +INSN(ADC, "r%d, r%d", a->rd, a->rr) +INSN(ADD, "r%d, r%d", a->rd, a->rr) +INSN(ADIW, "r%d:r%d, %d", a->rd + 1, a->rd, a->imm) +INSN(AND, "r%d, r%d", a->rd, a->rr) +INSN(ANDI, "r%d, %d", a->rd, a->imm) +INSN(ASR, "r%d", a->rd) +INSN_MNEMONIC(BCLR, bclr[a->bit], "") +INSN(BLD, "r%d, %d", a->rd, a->bit) +INSN_MNEMONIC(BRBC, brbc[a->bit], ".%+d", a->imm * 2) +INSN_MNEMONIC(BRBS, brbs[a->bit], ".%+d", a->imm * 2) +INSN(BREAK, "") +INSN_MNEMONIC(BSET, bset[a->bit], "") +INSN(BST, "r%d, %d", a->rd, a->bit) +INSN(CALL, "0x%x", a->imm * 2) +INSN(CBI, "%d, %d", a->reg, a->bit) +INSN(COM, "r%d", a->rd) +INSN(CP, "r%d, r%d", a->rd, a->rr) +INSN(CPC, "r%d, r%d", a->rd, a->rr) +INSN(CPI, "r%d, %d", a->rd, a->imm) +INSN(CPSE, "r%d, r%d", a->rd, a->rr) +INSN(DEC, "r%d", a->rd) +INSN(DES, "%d", a->imm) +INSN(EICALL, "") +INSN(EIJMP, "") +INSN(ELPM1, "") +INSN(ELPM2, "r%d, Z", a->rd) +INSN(ELPMX, "r%d, Z+", a->rd) +INSN(EOR, "r%d, r%d", a->rd, a->rr) +INSN(FMUL, "r%d, r%d", a->rd, a->rr) +INSN(FMULS, "r%d, r%d", a->rd, a->rr) +INSN(FMULSU, "r%d, r%d", a->rd, a->rr) +INSN(ICALL, "") +INSN(IJMP, "") +INSN(IN, "r%d, $%d", a->rd, a->imm) +INSN(INC, "r%d", a->rd) +INSN(JMP, "0x%x", a->imm * 2) +INSN(LAC, "Z, r%d", a->rd) +INSN(LAS, "Z, r%d", a->rd) +INSN(LAT, "Z, r%d", a->rd) +INSN(LDDY, "r%d, Y+%d", a->rd, a->imm) +INSN(LDDZ, "r%d, Z+%d", a->rd, a->imm) +INSN(LDI, "r%d, %d", a->rd, a->imm) +INSN(LDS, "r%d, %d", a->rd, a->imm) +INSN(LDX1, "r%d, X", a->rd) +INSN(LDX2, "r%d, X+", a->rd) +INSN(LDX3, "r%d, -X", a->rd) +INSN(LDY2, "r%d, Y+", a->rd) +INSN(LDY3, "r%d, -Y", a->rd) +INSN(LDZ2, "r%d, Z+", a->rd) +INSN(LDZ3, "r%d, -Z", a->rd) +INSN(LPM1, "") +INSN(LPM2, "r%d, Z", a->rd) +INSN(LPMX, "r%d, Z+", a->rd) +INSN(LSR, "r%d", a->rd) +INSN(MOV, "r%d, r%d", a->rd, a->rr) +INSN(MOVW, "r%d:r%d, r%d,r:r%d", a->rd + 1, a->rd, a->rr + 1, a->rr) +INSN(MUL, "r%d, r%d", a->rd, a->rr) +INSN(MULS, "r%d, r%d", a->rd, a->rr) +INSN(MULSU, "r%d, r%d", a->rd, a->rr) +INSN(NEG, "r%d", a->rd) +INSN(NOP, "") +INSN(OR, "r%d, r%d", a->rd, a->rr) +INSN(ORI, "r%d, %d", a->rd, a->imm) +INSN(OUT, "$%d, r%d", a->imm, a->rd) +INSN(POP, "r%d", a->rd) +INSN(PUSH, "r%d", a->rd) +INSN(RCALL, ".%+d", a->imm * 2) +INSN(RET, "") +INSN(RETI, "") +INSN(RJMP, ".%+d", a->imm * 2) +INSN(ROR, "r%d", a->rd) +INSN(SBC, "r%d, r%d", a->rd, a->rr) +INSN(SBCI, "r%d, %d", a->rd, a->imm) +INSN(SBI, "$%d, %d", a->reg, a->bit) +INSN(SBIC, "$%d, %d", a->reg, a->bit) +INSN(SBIS, "$%d, %d", a->reg, a->bit) +INSN(SBIW, "r%d:r%d, %d", a->rd + 1, a->rd, a->imm) +INSN(SBRC, "r%d, %d", a->rr, a->bit) +INSN(SBRS, "r%d, %d", a->rr, a->bit) +INSN(SLEEP, "") +INSN(SPM, "") +INSN(SPMX, "Z+") +INSN(STDY, "r%d, Y+%d", a->rd, a->imm) +INSN(STDZ, "r%d, Z+%d", a->rd, a->imm) +INSN(STS, "r%d, %d", a->rd, a->imm) +INSN(STX1, "r%d, X", a->rr) +INSN(STX2, "r%d, X+", a->rr) +INSN(STX3, "r%d, -X", a->rr) +INSN(STY2, "r%d, Y+", a->rd) +INSN(STY3, "r%d, -Y", a->rd) +INSN(STZ2, "r%d, Z+", a->rd) +INSN(STZ3, "r%d, -Z", a->rd) +INSN(SUB, "r%d, r%d", a->rd, a->rr) +INSN(SUBI, "r%d, %d", a->rd, a->imm) +INSN(SWAP, "r%d", a->rd) +INSN(WDR, "") +INSN (XCH ,"Z, r%d" , a->rd) +target/avr/translate.c +11
-1@@ -3032,6 +3032,17 @@ done_generating: tb->size = (ctx.npc - pc_start) *2 ; tb->icount = num_insns; + +#ifdef DEBUG_DISAS +if (qemu_loglevel_mask (CPU_LOG_TB_IN_ASM ) + &&qemu_log_in_addr_range (tb->pc)) { +qemu_log_lock (); +qemu_log ("IN: %s\n" ,lookup_symbol (tb->pc)); +log_target_disas (cs, tb->pc, tb->size); +qemu_log ("\n" ); +qemu_log_unlock (); + } +#endif } void restore_state_to_opc(CPUAVRState *env, TranslationBlock *tb, --2.17.2 (Apple Git-113)
Signature · 2 lines
Best Regards, Michael Rolnik
Re: [PATCH v37 05/17] target/avr: Add instruction translation - Arithmetic and Logic Instructions
In reply to Aleksandar MarkovicAleksandar.
I could not find what happens if an instruction with unsupported registers is executed. So, I am leaving this tiny core for later.
Regards,
Michael Rolnik
Quoted 106 lines · Aleksandar Markovic
Quoted 6 lines · Michael Rolnik
Hi Aleksandar.
thanks for pointing that out I was not aware of that. I can fix it.
Hey, Michael,
Please take alook at this:
https://en.m.wikipedia.org/wiki/ATtiny_microcontroller_comparison_chart
It looks that all cases of AVR 16-gpr-registers cores belong to the group "avrtiny10", that actually you may want to add to your solution. Just a hint.
Best regards,
AleksandarRegards,
Quoted 80 lines
Michael Rolnik
Quoted 54 lines · Aleksandar Markovic
Quoted 43 lines · Aleksandar Markovic
On Wednesday, November 27, 2019, Michael Rolnik <[email protected]> wrote:
+
Quoted 23 lines
target/avr/translate.c +3
-0+ +/* +* Performs the logical AND between the contents of register Rd and register
+ * Rr and places the result in the destination register Rd.
+ */target/avr/translate.c +16
-0+static bool trans_AND (DisasContext *ctx,arg_AND *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_and_tl (R , Rd, Rr);/* Rd = Rd and Rr */ + +/* Vf */ +tcg_gen_movi_tl (cpu_Vf,0 );/* Vf = 0 */ + +/* Zf */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */ + +gen_ZNSf (R );target/avr/translate.c static bool avr_have_feature(DisasContext *ctx, int feature) · line 555
Quoted 10 lines
target/avr/translate.c +10
-0+ +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + +According to specs:
Quoted 6 lines
... the chips in question have cores with 16 GPRs (that correspond to GPRs R16-R31 of more common AVR cores with 32 GPRs).
Quoted 16 lines
There are more examples;
Also ATtiny20, ATtiny40.
How do you handle such cores?
Quoted 7 lines
I don't see here anything preventing usage of all 32 GPRs, resulting, of course, in an inaccurate emulation.
Thanks,
Aleksandar--
Best Regards,
Michael Rolnik
Signature · 2 lines
Best Regards, Michael Rolnik
Re: [PATCH v37 05/17] target/avr: Add instruction translation - Arithmetic and Logic Instructions
Quoted 6 lines · Michael Rolnik
Aleksandar.
I could not find what happens if an instruction with unsupported registers is executed. So, I am leaving this tiny core for later.
No problem with me. You already have instruction support for a rich variety of cores. These can certainly added it in future.
May I suggest, then, a following cosmetic change: In order for a reader to get a brtter "first glance" feeling for emulations of ALL instructions that involve updating status register, I suggest that the code blocks like this one:
+
+ /* Vf */
+ tcg_gen_movi_tl(cpu_Vf, 0); /* Vf = 0 */
+
+ /* Zf */
+ tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_Zf, R, 0); /* Zf = R == 0 */
+
+ gen_ZNSf(R);
+
+ /* R */
+ tcg_gen_mov_tl(Rd, R);
+
be replaced with this one:
+
+ /* update status register */ + tcg_gen_movi_tl(cpu_Vf, 0); /* Vf = 0 */ + tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_Zf, R, 0); /* Zf = R == 0 */ + gen_ZNSf(R); + tcg_gen_mov_tl(Rd, R);
+
The idea is to distinguish visually better the portions for updating status register from the central part of the operation (usually marked by "/* op */" comment. The code also gets more compact, which looks like a good thing too.
Aleksandar
Quoted 119 lines
Regards,
Michael RolnikQuoted 20 lines · Aleksandar Markovic
Quoted 6 lines · Michael Rolnik
Hi Aleksandar.
thanks for pointing that out I was not aware of that. I can fix it.
Hey, Michael,
Please take alook at this:
https://en.m.wikipedia.org/wiki/ATtiny_microcontroller_comparison_chart
It looks that all cases of AVR 16-gpr-registers cores belong to the group "avrtiny10", that actually you may want to add to your solution. Just a hint.
Quoted 88 lines
Best regards,
AleksandarRegards,
Quoted 81 lines
Michael Rolnik
Quoted 55 lines · Aleksandar Markovic
Quoted 43 lines · Aleksandar Markovic
On Wednesday, November 27, 2019, Michael Rolnik <[email protected]> wrote:
+
Quoted 24 lines
target/avr/translate.c +3
-0+ +/* +* Performs the logical AND between the contents of register Rd and register
+ * Rr and places the result in the destination register Rd.
+ */target/avr/translate.c +14
-0+static bool trans_AND (DisasContext *ctx,arg_AND *a) +{ +TCGv Rd = cpu_r[a->rd]; +TCGv Rr = cpu_r[a->rr]; +TCGv R =tcg_temp_new_i32 (); + +/* op */ +tcg_gen_and_tl (R , Rd, Rr);/* Rd = Rd and Rr */ + +/* Vf */ +tcg_gen_movi_tl (cpu_Vf,0 );/* Vf = 0 */ + +/* Zf */ +tcg_gen_setcondi_tl (TCG_COND_EQ , cpu_Zf,R ,0 );/* Zf = R == 0 */
target/avr/translate.c +2
-0+ +gen_ZNSf (R );target/avr/translate.c static bool avr_have_feature(DisasContext *ctx, int feature) · line 266
Quoted 10 lines
target/avr/translate.c +10
-0+ +/* R */ +tcg_gen_mov_tl (Rd,R ); + +tcg_temp_free_i32 (R ); + +return true; +} + +According to specs:
Quoted 7 lines
http://ww1.microchip.com/downloads/en/devicedoc/atmel-42505-8-bit-avr-
microcontrollers-attiny102-attiny104_datasheet.pdf... the chips in question have cores with 16 GPRs (that correspond to GPRs R16-R31 of more common AVR cores with 32 GPRs).
Quoted 16 lines
There are more examples;
http://ww1.microchip.com/downloads/en/DeviceDoc/atmel-8127-avr-8-bit-
microcontroller-attiny4-attiny5-attiny9-attiny10_datasheet.pdfAlso ATtiny20, ATtiny40.
How do you handle such cores?
Quoted 7 lines
I don't see here anything preventing usage of all 32 GPRs, resulting, of course, in an inaccurate emulation.
Thanks,
Aleksandar--
Best Regards,
Michael Rolnik--
Best Regards,
Michael Rolnik
Re: [PATCH v37 05/17] target/avr: Add instruction translation - Arithmetic and Logic Instructions
Review
Quoted 10 lines · Aleksandar Markovic
(quoted diff) +7
-0+ + /* update status register */ + tcg_gen_movi_tl(cpu_Vf, 0); /* Vf = 0 */ + tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_Zf, R, 0); /* Zf = R == 0 */ + gen_ZNSf(R); + tcg_gen_mov_tl(Rd, R); +
My email client drives me crazy, all lines from the above segment should be consecutive, without any empty lines in between. If you see empty lines, that is because of my email client.
Quoted 130 lines
The idea is to distinguish visually better the portions for updating status register from the central part of the operation (usually marked by "/* op */" comment. The code also gets more compact, which looks like a good thing too.
Aleksandar
Quoted 121 lines
Regards,
Michael RolnikQuoted 21 lines · Aleksandar Markovic
Quoted 6 lines · Michael Rolnik
Hi Aleksandar.
thanks for pointing that out I was not aware of that. I can fix it.
Hey, Michael,
Please take alook at this:
https://en.m.wikipedia.org/wiki/ATtiny_microcontroller_comparison_chart
It looks that all cases of AVR 16-gpr-registers cores belong to the group "avrtiny10", that actually you may want to add to your solution. Just a hint.
Quoted 89 lines
Best regards,
AleksandarRegards,
Quoted 82 lines
Michael Rolnik
Quoted 57 lines · Aleksandar Markovic
Quoted 43 lines · Aleksandar Markovic
On Wednesday, November 27, 2019, Michael Rolnik <[email protected]> wrote:
+
Quoted 24 lines
(quoted diff) +3
-0+ +/* + * Performs the logical AND between the contents of register Rdand register
+ * Rr and places the result in the destination register Rd.
+ */(quoted diff) +14
-0+static bool trans_AND(DisasContext *ctx, arg_AND *a) +{ + TCGv Rd = cpu_r[a->rd]; + TCGv Rr = cpu_r[a->rr]; + TCGv R = tcg_temp_new_i32(); + + /* op */ + tcg_gen_and_tl(R, Rd, Rr); /* Rd = Rd and Rr */ + + /* Vf */ + tcg_gen_movi_tl(cpu_Vf, 0); /* Vf = 0 */ + + /* Zf */ + tcg_gen_setcondi_tl(TCG_COND_EQ, cpu_Zf, R, 0); /* Zf = R == 0*/
(quoted diff) +2
-0+ + gen_ZNSf(R);Review
Quoted 10 lines
(quoted diff) +10
-0+ + /* R */ + tcg_gen_mov_tl(Rd, R); + + tcg_temp_free_i32(R); + + return true; +} + +According to specs:
Quoted 7 lines
http://ww1.microchip.com/downloads/en/devicedoc/atmel-42505-
8-bit-avr-microcontrollers-attiny102-attiny104_datasheet.pdf... the chips in question have cores with 16 GPRs (that correspond to GPRs R16-R31 of more common AVR cores with 32 GPRs).
There are more examples;
Quoted 15 lines
http://ww1.microchip.com/downloads/en/DeviceDoc/atmel-8127-
avr-8-bit-microcontroller-attiny4-attiny5-attiny9- attiny10_datasheet.pdfAlso ATtiny20, ATtiny40.
How do you handle such cores?
Quoted 7 lines
I don't see here anything preventing usage of all 32 GPRs, resulting, of course, in an inaccurate emulation.
Thanks,
Aleksandar--
Best Regards,
Michael Rolnik--
Best Regards,
Michael Rolnik
Quoted 21 lines · Michael Rolnik
There is *-cpu *option where you can specify what CPU you want, if this option is not specified avr6 (avr6-avr-cpu) is chosen.
*./avr-softmmu/qemu-system-avr -cpu help*
avr1-avr-cpu
avr2-avr-cpu
avr25-avr-cpu
avr3-avr-cpu
avr31-avr-cpu
avr35-avr-cpu
avr4-avr-cpu
avr5-avr-cpu
avr51-avr-cpu
avr6-avr-cpu
xmega2-avr-cpu
xmega4-avr-cpu
xmega5-avr-cpu
xmega6-avr-cpu
xmega7-avr-cpu
What happens if you specify a core via -cpu, and supply elf file compiled for another core?
Best regards,
Akeksandar
Quoted 419 lines
Regards,
Michael RolnikQuoted 32 lines · Aleksandar Markovic
Quoted 17 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fully tested yet.
However I was able to execute simple code with functions. e.g fibonacci calculation.
This series of patches include a non real, sample board.
No fuses support yet. PC is set to 0 at reset.Following are examples of possible usages, assuming program.elf is compiled for AVR cpu
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnet window
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
runtelent localhost 5678Quoted 10 lines
3. Continious non interrupted execution with serial output into stdout run
qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -S
runavr-gdb program.elfand then within GDB shell `target remote
:1234`
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmQuoted 374 lines
Hi, Michael.
Can you explain to me the mechanisms of recognition of what core/microcontroller QEMU is supposed to emulate in the examples above?
Yours, Aleksandar
Quoted 365 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding or translation
2. CPU features which allow define the following 8bit AVR coresavr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows to execute simple code
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of0x000000ff) for faster address manipulaton
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM
11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHchanges since v4
1. target is now defined as big endian in order to optimize push_ret/pop_ret
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_cpu_gdb_write_register
2. configure the target as little endian because otherwise GDB does not work
3. fixing and testing gen_push_ret/gen_pop_retchanges since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supported yet registers only
3.MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built for all
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memorychanges since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed to any device
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try to access registerschanges since v10
1. movig back fullwr & fullrd into the commit where outb and inb were introduced
2. changingtlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codechanges since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thing exists)
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrischanges since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers are accessible
by LD/ST only. CPU has no interest in themchanges since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixes a
couple of bugs wrt single stepping
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicchanges since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU" into three patches 1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tename1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERS
7. use git commit sha1 instead of `master`` in avocado testMichael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileSarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer peripheralsconfigure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)--
Best Regards,
Michael Rolnik
Quoted 25 lines · Aleksandar Markovic
Quoted 21 lines · Michael Rolnik
There is *-cpu *option where you can specify what CPU you want, if this option is not specified avr6 (avr6-avr-cpu) is chosen.
*./avr-softmmu/qemu-system-avr -cpu help*
avr1-avr-cpu
avr2-avr-cpu
avr25-avr-cpu
avr3-avr-cpu
avr31-avr-cpu
avr35-avr-cpu
avr4-avr-cpu
avr5-avr-cpu
avr51-avr-cpu
avr6-avr-cpu
xmega2-avr-cpu
xmega4-avr-cpu
xmega5-avr-cpu
xmega6-avr-cpu
xmega7-avr-cpu
Quoted 4 lines
What happens if you specify a core via -cpu, and supply elf file compiled for another core?
It looks there is some related info written in ELF header. This is from a binutils header:
(so it looks you could detect the core from elf file - do you do that detection right now?)
#define E_AVR_MACH_AVR1 1
#define E_AVR_MACH_AVR2 2
#define E_AVR_MACH_AVR25 25
#define E_AVR_MACH_AVR3 3
#define E_AVR_MACH_AVR31 31
#define E_AVR_MACH_AVR35 35
#define E_AVR_MACH_AVR4 4
#define E_AVR_MACH_AVR5 5
#define E_AVR_MACH_AVR51 51
#define E_AVR_MACH_AVR6 6
#define E_AVR_MACH_AVRTINY 100
#define E_AVR_MACH_XMEGA1 101 #define E_AVR_MACH_XMEGA2 102 #define E_AVR_MACH_XMEGA3 103 #define E_AVR_MACH_XMEGA4 104 #define E_AVR_MACH_XMEGA5 105 #define E_AVR_MACH_XMEGA6 106 #define E_AVR_MACH_XMEGA7 107
Best regards,
Quoted 427 lines
Akeksandar
Quoted 421 lines
Regards,
Michael RolnikQuoted 32 lines · Aleksandar Markovic
Quoted 17 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fully tested yet.
However I was able to execute simple code with functions. e.g fibonacci calculation.
This series of patches include a non real, sample board.
No fuses support yet. PC is set to 0 at reset.Following are examples of possible usages, assuming program.elf is compiled for AVR cpu
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnet window
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
runtelent localhost 5678Quoted 10 lines
3. Continious non interrupted execution with serial output into stdout run
qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -S
runavr-gdb program.elfand then within GDB shell `target remote
:1234`
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmQuoted 376 lines
Hi, Michael.
Can you explain to me the mechanisms of recognition of what core/microcontroller QEMU is supposed to emulate in the examples above?
Yours, Aleksandar
Quoted 367 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding or translation
2. CPU features which allow define the following 8bit AVR coresavr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows to execute simple code
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of0x000000ff) for faster address manipulaton
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM
11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHchanges since v4
1. target is now defined as big endian in order to optimize push_ret/pop_ret
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_cpu_gdb_write_register
2. configure the target as little endian because otherwise GDB does not work
3. fixing and testing gen_push_ret/gen_pop_retchanges since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supported yet registers only
3.MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built for all
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memorychanges since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed to any device
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try to access registerschanges since v10
1. movig back fullwr & fullrd into the commit where outb and inb were introduced
2. changingtlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codechanges since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thing exists)
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrischanges since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers are accessible
by LD/ST only. CPU has no interest in themchanges since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixes
a couple of bugs wrt single stepping
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicchanges since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU" into three patches 1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tename1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERS
7. use git commit sha1 instead of `master`` in avocado testMichael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileSarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer peripheralsconfigure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)--
Best Regards,
Michael Rolnik
Re: [PATCH v37 10/17] target/avr: Add instruction disassembly function
In reply to Michael RolnikQuoted 6 lines · Michael Rolnik
Aleksandar.
If this code is going to be merge in 2019 I should modify al the copyrights, right. or should I put 2020 in?
I don't have a better answer other than:
:))))))
Quoted 360 lines
Regards,
Michael RolnikQuoted 344 lines · Aleksandar Markovic
Quoted 74 lines · Michael Rolnik
Provide function disassembles executed instruction when
-d in_asmis providedExample:
./avr-softmmu/qemu-system-avr -bios free-rtos/Demo/AVR_ATMega2560_GCC/demo.elf -d in_asmwill produce something like the following``` ... IN: 0x0000014a: CALL 0x3808 IN: main 0x00003808: CALL 0x4b4 IN: vParTestInitialise 0x000004b4: LDI r24, 255 0x000004b6: STS r24, 0 0x000004b8: MULS r16, r20 0x000004ba: OUT $1, r24 0x000004bc: LDS r24, 0 0x000004be: MULS r16, r20 0x000004c0: OUT $2, r24 0x000004c2: RET ... ```Signed-off-by Michael Rolnik <[email protected]>Suggested-by Richard Henderson <[email protected]>Suggested-by Philippe Mathieu-Daudé <[email protected]>Suggested-by Aleksandar Markovic <[email protected]>Reviewed-by Philippe Mathieu-Daudé <[email protected]>--- target/avr/cpu.h | 1 + target/avr/cpu.c | 2 +- target/avr/disas.c | 228 +++++++++++++++++++++++++++++++++++++++++ target/avr/translate.c | 11 ++ 4 files changed, 241 insertions(+), 1 deletion(-) create mode 100644 target/avr/disas.ctarget/avr/cpu.h +1
-0@@ -157,6 +157,7 @@ bool avr_cpu_exec_interrupt(CPUState *cpu, int int_req);hwaddr avr_cpu_get_phys_page_debug (CPUState *cpu,vaddr addr);int avr_cpu_gdb_read_register (CPUState *cpu,uint8_t *buf,int reg);int avr_cpu_gdb_write_register (CPUState *cpu,uint8_t *buf,int reg); +int avr_print_insn (bfd_vma addr,disassemble_info *info);static inline int avr_feature (CPUAVRState *env,int feature){
target/avr/cpu.c +1
-1@@ -83,7 +83,7 @@ static void avr_cpu_reset(CPUState *cs)static void avr_cpu_disas_set_info (CPUState *cpu,disassemble_info *info) { info->mach = bfd_arch_avr; - info->print_insn =NULL ; + info->print_insn = avr_print_insn; }static void
avr_cpu_realizefn(DeviceState *dev, Error **errp)target/avr/disas.c new file +0
-0target/avr/disas.c line 4
Quoted 4 lines
target/avr/disas.c new file +4
-0@@ -0,0 +1,228 @@ +/* +* AVR disassembler +* +* Copyright (c) 2018 Richard Henderson <[email protected]> Just a detail: since this file is created in 2019, the copyright year should be 2019 too.
+ * Copyright (c) 2019 Michael Rolnik <[email protected]>
Quoted 255 lines
target/avr/disas.c new file +13
-0+* +* This program is free software: you can redistribute it and/or modify +* it under the terms of the GNU General Public License as published by +* the Free Software Foundation, either version 2 of the License, or +* (at your option) any later version. +* +* This program is distributed in the hope that 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/> .
+ */target/avr/disas.c new file +79
-0+ +#include "qemu/osdep.h" +#include "disas/dis-asm.h" +#include "qemu/bitops.h" +#include "cpu.h" + +typedef struct { +disassemble_info *info; +uint16_t next_word; +bool next_word_used; +}DisasContext ; + +static int to_regs_16_31_by_one (DisasContext *ctx,int indx) +{ +return 16 + (indx %16 ); +} + +static int to_regs_16_23_by_one (DisasContext *ctx,int indx) +{ +return 16 + (indx %8 ); +} +static int to_regs_24_30_by_two (DisasContext *ctx,int indx) +{ +return 24 + (indx %4 ) *2 ; +} +static int to_regs_00_30_by_two (DisasContext *ctx,int indx) +{ +return (indx %16 ) *2 ; +} + +static uint16_t next_word (DisasContext *ctx) +{ + ctx->next_word_used = true; +return ctx->next_word; +} + +static int append_16 (DisasContext *ctx,int x) +{ +return x <<16 |next_word (ctx); +} + + +/* Include the auto-generated decoder. */ +static bool decode_insn (DisasContext *ctx,uint16_t insn); +#include "decode_insn.inc.c" + +#define output (mnemonic, format, ...) \ + (pctx->info->fprintf_func(pctx->info->stream, "%-9s " format, \ + mnemonic, ##__VA_ARGS__)) + +int avr_print_insn (bfd_vma addr,disassemble_info *info) +{ +DisasContext ctx; +DisasContext *pctx = &ctx; +bfd_byte buffer[4 ]; +uint16_t insn; +int status; + + ctx.info = info; + + status = info->read_memory_func (addr, buffer,4 , info); +if (status !=0 ) { + info->memory_error_func (status, addr, info); +return -1 ; + } + insn =bfd_getl16 (buffer); + ctx.next_word =bfd_getl16 (buffer +2 ); + ctx.next_word_used = false; + +if (!decode_insn (&ctx, insn)) { +output (".db" ,"0x%02x, 0x%02x" , buffer[0 ], buffer[1 ]); + } + +return ctx.next_word_used ?4 :2 ; +} + + +#define INSN (opcode, format, ...) \ +static bool trans_##opcode(DisasContext *pctx, arg_##opcode * a)\
target/avr/disas.c new file +3
-0+{ \ + output(#opcode, format, ##__VA_ARGS__); \ +return true;\
target/avr/disas.c new file +3
-0+} + +#define INSN_MNEMONIC (opcode, mnemonic, format, ...)\
+static bool trans_##opcode(DisasContext *pctx, arg_##opcode * a) \target/avr/disas.c new file +2
-0+{ \ + output(mnemonic, format, ##__VA_ARGS__);\
+ return true;
\target/avr/disas.c new file +120
-0+} + +/* +* C Z N V S H T I +* 0 1 2 3 4 5 6 7 +*/ +static const char *brbc[] = { +"BRCC" ,"BRNE" ,"BRPL" ,"BRVC" ,"BRGE" ,"BRHC" ,"BRTC" ,"BRID" +}; + +static const char *brbs[] = { +"BRCS" ,"BREQ" ,"BRMI" ,"BRVS" ,"BRLT" ,"BRHS" ,"BRTS" ,"BRIE" +}; + +static const char *bset[] = { +"SEC" ,"SEZ" ,"SEN" ,"SEZ" ,"SES" ,"SEH" ,"SET" ,"SEI" +}; + +static const char *bclr[] = { +"CLC" ,"CLZ" ,"CLN" ,"CLZ" ,"CLS" ,"CLH" ,"CLT" ,"CLI" +}; + +INSN(ADC, "r%d, r%d", a->rd, a->rr) +INSN(ADD, "r%d, r%d", a->rd, a->rr) +INSN(ADIW, "r%d:r%d, %d", a->rd + 1, a->rd, a->imm) +INSN(AND, "r%d, r%d", a->rd, a->rr) +INSN(ANDI, "r%d, %d", a->rd, a->imm) +INSN(ASR, "r%d", a->rd) +INSN_MNEMONIC(BCLR, bclr[a->bit], "") +INSN(BLD, "r%d, %d", a->rd, a->bit) +INSN_MNEMONIC(BRBC, brbc[a->bit], ".%+d", a->imm * 2) +INSN_MNEMONIC(BRBS, brbs[a->bit], ".%+d", a->imm * 2) +INSN(BREAK, "") +INSN_MNEMONIC(BSET, bset[a->bit], "") +INSN(BST, "r%d, %d", a->rd, a->bit) +INSN(CALL, "0x%x", a->imm * 2) +INSN(CBI, "%d, %d", a->reg, a->bit) +INSN(COM, "r%d", a->rd) +INSN(CP, "r%d, r%d", a->rd, a->rr) +INSN(CPC, "r%d, r%d", a->rd, a->rr) +INSN(CPI, "r%d, %d", a->rd, a->imm) +INSN(CPSE, "r%d, r%d", a->rd, a->rr) +INSN(DEC, "r%d", a->rd) +INSN(DES, "%d", a->imm) +INSN(EICALL, "") +INSN(EIJMP, "") +INSN(ELPM1, "") +INSN(ELPM2, "r%d, Z", a->rd) +INSN(ELPMX, "r%d, Z+", a->rd) +INSN(EOR, "r%d, r%d", a->rd, a->rr) +INSN(FMUL, "r%d, r%d", a->rd, a->rr) +INSN(FMULS, "r%d, r%d", a->rd, a->rr) +INSN(FMULSU, "r%d, r%d", a->rd, a->rr) +INSN(ICALL, "") +INSN(IJMP, "") +INSN(IN, "r%d, $%d", a->rd, a->imm) +INSN(INC, "r%d", a->rd) +INSN(JMP, "0x%x", a->imm * 2) +INSN(LAC, "Z, r%d", a->rd) +INSN(LAS, "Z, r%d", a->rd) +INSN(LAT, "Z, r%d", a->rd) +INSN(LDDY, "r%d, Y+%d", a->rd, a->imm) +INSN(LDDZ, "r%d, Z+%d", a->rd, a->imm) +INSN(LDI, "r%d, %d", a->rd, a->imm) +INSN(LDS, "r%d, %d", a->rd, a->imm) +INSN(LDX1, "r%d, X", a->rd) +INSN(LDX2, "r%d, X+", a->rd) +INSN(LDX3, "r%d, -X", a->rd) +INSN(LDY2, "r%d, Y+", a->rd) +INSN(LDY3, "r%d, -Y", a->rd) +INSN(LDZ2, "r%d, Z+", a->rd) +INSN(LDZ3, "r%d, -Z", a->rd) +INSN(LPM1, "") +INSN(LPM2, "r%d, Z", a->rd) +INSN(LPMX, "r%d, Z+", a->rd) +INSN(LSR, "r%d", a->rd) +INSN(MOV, "r%d, r%d", a->rd, a->rr) +INSN(MOVW, "r%d:r%d, r%d,r:r%d", a->rd + 1, a->rd, a->rr + 1, a->rr) +INSN(MUL, "r%d, r%d", a->rd, a->rr) +INSN(MULS, "r%d, r%d", a->rd, a->rr) +INSN(MULSU, "r%d, r%d", a->rd, a->rr) +INSN(NEG, "r%d", a->rd) +INSN(NOP, "") +INSN(OR, "r%d, r%d", a->rd, a->rr) +INSN(ORI, "r%d, %d", a->rd, a->imm) +INSN(OUT, "$%d, r%d", a->imm, a->rd) +INSN(POP, "r%d", a->rd) +INSN(PUSH, "r%d", a->rd) +INSN(RCALL, ".%+d", a->imm * 2) +INSN(RET, "") +INSN(RETI, "") +INSN(RJMP, ".%+d", a->imm * 2) +INSN(ROR, "r%d", a->rd) +INSN(SBC, "r%d, r%d", a->rd, a->rr) +INSN(SBCI, "r%d, %d", a->rd, a->imm) +INSN(SBI, "$%d, %d", a->reg, a->bit) +INSN(SBIC, "$%d, %d", a->reg, a->bit) +INSN(SBIS, "$%d, %d", a->reg, a->bit) +INSN(SBIW, "r%d:r%d, %d", a->rd + 1, a->rd, a->imm) +INSN(SBRC, "r%d, %d", a->rr, a->bit) +INSN(SBRS, "r%d, %d", a->rr, a->bit) +INSN(SLEEP, "") +INSN(SPM, "") +INSN(SPMX, "Z+") +INSN(STDY, "r%d, Y+%d", a->rd, a->imm) +INSN(STDZ, "r%d, Z+%d", a->rd, a->imm) +INSN(STS, "r%d, %d", a->rd, a->imm) +INSN(STX1, "r%d, X", a->rr) +INSN(STX2, "r%d, X+", a->rr) +INSN(STX3, "r%d, -X", a->rr) +INSN(STY2, "r%d, Y+", a->rd) +INSN(STY3, "r%d, -Y", a->rd) +INSN(STZ2, "r%d, Z+", a->rd) +INSN(STZ3, "r%d, -Z", a->rd) +INSN(SUB, "r%d, r%d", a->rd, a->rr) +INSN(SUBI, "r%d, %d", a->rd, a->imm) +INSN(SWAP, "r%d", a->rd) +INSN(WDR, "") +INSN (XCH ,"Z, r%d" , a->rd) +target/avr/translate.c +11
-0@@ -3032,6 +3032,17 @@ done_generating: tb->size = (ctx.npc - pc_start) *2 ; tb->icount = num_insns; + +#ifdef DEBUG_DISAS +if (qemu_loglevel_mask (CPU_LOG_TB_IN_ASM ) + &&qemu_log_in_addr_range (tb->pc)) { +qemu_log_lock (); +qemu_log ("IN: %s\n" ,lookup_symbol (tb->pc)); +log_target_disas (cs, tb->pc, tb->size); +qemu_log ("\n" ); +qemu_log_unlock (); + } +#endif }Quoted 5 lines
Quoted 5 lines
target/avr/translate.c +0
-1void restore_state_to_opc(CPUAVRState *env, TranslationBlock *tb, --2.17.2 (Apple Git-113)
--
Best Regards,
Michael Rolnik
Re: [PATCH v37 10/17] target/avr: Add instruction disassembly function
In reply to Michael RolnikQuoted 6 lines · Michael Rolnik
Aleksandar.
If this code is going to be merge in 2019 I should modify al the copyrights, right. or should I put 2020 in?
Michael,
Initially I saw your message as wry sense of humor.
But, in fact, it is not a bad idea at all. Since we know that AVR support will not be in QEMU 4.2, the last release of QEMU in 2019, it looks to me the right thing is to update *all* copyright message to year 2020. Please do it.
Thank you for giving me the reason to laugh this morning! You made my day!
Aleksandar
Quoted 360 lines
Regards,
Michael RolnikQuoted 344 lines · Aleksandar Markovic
Quoted 74 lines · Michael Rolnik
Provide function disassembles executed instruction when
-d in_asmis providedExample:
./avr-softmmu/qemu-system-avr -bios free-rtos/Demo/AVR_ATMega2560_GCC/demo.elf -d in_asmwill produce something like the following``` ... IN: 0x0000014a: CALL 0x3808 IN: main 0x00003808: CALL 0x4b4 IN: vParTestInitialise 0x000004b4: LDI r24, 255 0x000004b6: STS r24, 0 0x000004b8: MULS r16, r20 0x000004ba: OUT $1, r24 0x000004bc: LDS r24, 0 0x000004be: MULS r16, r20 0x000004c0: OUT $2, r24 0x000004c2: RET ... ```Signed-off-by Michael Rolnik <[email protected]>Suggested-by Richard Henderson <[email protected]>Suggested-by Philippe Mathieu-Daudé <[email protected]>Suggested-by Aleksandar Markovic <[email protected]>Reviewed-by Philippe Mathieu-Daudé <[email protected]>--- target/avr/cpu.h | 1 + target/avr/cpu.c | 2 +- target/avr/disas.c | 228 +++++++++++++++++++++++++++++++++++++++++ target/avr/translate.c | 11 ++ 4 files changed, 241 insertions(+), 1 deletion(-) create mode 100644 target/avr/disas.ctarget/avr/cpu.h +1
-0@@ -157,6 +157,7 @@ bool avr_cpu_exec_interrupt(CPUState *cpu, int int_req);hwaddr avr_cpu_get_phys_page_debug (CPUState *cpu,vaddr addr);int avr_cpu_gdb_read_register (CPUState *cpu,uint8_t *buf,int reg);int avr_cpu_gdb_write_register (CPUState *cpu,uint8_t *buf,int reg); +int avr_print_insn (bfd_vma addr,disassemble_info *info);static inline int avr_feature (CPUAVRState *env,int feature){
target/avr/cpu.c +1
-1@@ -83,7 +83,7 @@ static void avr_cpu_reset(CPUState *cs)static void avr_cpu_disas_set_info (CPUState *cpu,disassemble_info *info) { info->mach = bfd_arch_avr; - info->print_insn =NULL ; + info->print_insn = avr_print_insn; }static void
avr_cpu_realizefn(DeviceState *dev, Error **errp)target/avr/disas.c new file +0
-0target/avr/disas.c line 4
Quoted 4 lines
target/avr/disas.c new file +4
-0@@ -0,0 +1,228 @@ +/* +* AVR disassembler +* +* Copyright (c) 2018 Richard Henderson <[email protected]> Just a detail: since this file is created in 2019, the copyright year should be 2019 too.
+ * Copyright (c) 2019 Michael Rolnik <[email protected]>
Quoted 255 lines
target/avr/disas.c new file +13
-0+* +* This program is free software: you can redistribute it and/or modify +* it under the terms of the GNU General Public License as published by +* the Free Software Foundation, either version 2 of the License, or +* (at your option) any later version. +* +* This program is distributed in the hope that 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/> .
+ */target/avr/disas.c new file +79
-0+ +#include "qemu/osdep.h" +#include "disas/dis-asm.h" +#include "qemu/bitops.h" +#include "cpu.h" + +typedef struct { +disassemble_info *info; +uint16_t next_word; +bool next_word_used; +}DisasContext ; + +static int to_regs_16_31_by_one (DisasContext *ctx,int indx) +{ +return 16 + (indx %16 ); +} + +static int to_regs_16_23_by_one (DisasContext *ctx,int indx) +{ +return 16 + (indx %8 ); +} +static int to_regs_24_30_by_two (DisasContext *ctx,int indx) +{ +return 24 + (indx %4 ) *2 ; +} +static int to_regs_00_30_by_two (DisasContext *ctx,int indx) +{ +return (indx %16 ) *2 ; +} + +static uint16_t next_word (DisasContext *ctx) +{ + ctx->next_word_used = true; +return ctx->next_word; +} + +static int append_16 (DisasContext *ctx,int x) +{ +return x <<16 |next_word (ctx); +} + + +/* Include the auto-generated decoder. */ +static bool decode_insn (DisasContext *ctx,uint16_t insn); +#include "decode_insn.inc.c" + +#define output (mnemonic, format, ...) \ + (pctx->info->fprintf_func(pctx->info->stream, "%-9s " format, \ + mnemonic, ##__VA_ARGS__)) + +int avr_print_insn (bfd_vma addr,disassemble_info *info) +{ +DisasContext ctx; +DisasContext *pctx = &ctx; +bfd_byte buffer[4 ]; +uint16_t insn; +int status; + + ctx.info = info; + + status = info->read_memory_func (addr, buffer,4 , info); +if (status !=0 ) { + info->memory_error_func (status, addr, info); +return -1 ; + } + insn =bfd_getl16 (buffer); + ctx.next_word =bfd_getl16 (buffer +2 ); + ctx.next_word_used = false; + +if (!decode_insn (&ctx, insn)) { +output (".db" ,"0x%02x, 0x%02x" , buffer[0 ], buffer[1 ]); + } + +return ctx.next_word_used ?4 :2 ; +} + + +#define INSN (opcode, format, ...) \ +static bool trans_##opcode(DisasContext *pctx, arg_##opcode * a)\
target/avr/disas.c new file +3
-0+{ \ + output(#opcode, format, ##__VA_ARGS__); \ +return true;\
target/avr/disas.c new file +3
-0+} + +#define INSN_MNEMONIC (opcode, mnemonic, format, ...)\
+static bool trans_##opcode(DisasContext *pctx, arg_##opcode * a) \target/avr/disas.c new file +2
-0+{ \ + output(mnemonic, format, ##__VA_ARGS__);\
+ return true;
\target/avr/disas.c new file +120
-0+} + +/* +* C Z N V S H T I +* 0 1 2 3 4 5 6 7 +*/ +static const char *brbc[] = { +"BRCC" ,"BRNE" ,"BRPL" ,"BRVC" ,"BRGE" ,"BRHC" ,"BRTC" ,"BRID" +}; + +static const char *brbs[] = { +"BRCS" ,"BREQ" ,"BRMI" ,"BRVS" ,"BRLT" ,"BRHS" ,"BRTS" ,"BRIE" +}; + +static const char *bset[] = { +"SEC" ,"SEZ" ,"SEN" ,"SEZ" ,"SES" ,"SEH" ,"SET" ,"SEI" +}; + +static const char *bclr[] = { +"CLC" ,"CLZ" ,"CLN" ,"CLZ" ,"CLS" ,"CLH" ,"CLT" ,"CLI" +}; + +INSN(ADC, "r%d, r%d", a->rd, a->rr) +INSN(ADD, "r%d, r%d", a->rd, a->rr) +INSN(ADIW, "r%d:r%d, %d", a->rd + 1, a->rd, a->imm) +INSN(AND, "r%d, r%d", a->rd, a->rr) +INSN(ANDI, "r%d, %d", a->rd, a->imm) +INSN(ASR, "r%d", a->rd) +INSN_MNEMONIC(BCLR, bclr[a->bit], "") +INSN(BLD, "r%d, %d", a->rd, a->bit) +INSN_MNEMONIC(BRBC, brbc[a->bit], ".%+d", a->imm * 2) +INSN_MNEMONIC(BRBS, brbs[a->bit], ".%+d", a->imm * 2) +INSN(BREAK, "") +INSN_MNEMONIC(BSET, bset[a->bit], "") +INSN(BST, "r%d, %d", a->rd, a->bit) +INSN(CALL, "0x%x", a->imm * 2) +INSN(CBI, "%d, %d", a->reg, a->bit) +INSN(COM, "r%d", a->rd) +INSN(CP, "r%d, r%d", a->rd, a->rr) +INSN(CPC, "r%d, r%d", a->rd, a->rr) +INSN(CPI, "r%d, %d", a->rd, a->imm) +INSN(CPSE, "r%d, r%d", a->rd, a->rr) +INSN(DEC, "r%d", a->rd) +INSN(DES, "%d", a->imm) +INSN(EICALL, "") +INSN(EIJMP, "") +INSN(ELPM1, "") +INSN(ELPM2, "r%d, Z", a->rd) +INSN(ELPMX, "r%d, Z+", a->rd) +INSN(EOR, "r%d, r%d", a->rd, a->rr) +INSN(FMUL, "r%d, r%d", a->rd, a->rr) +INSN(FMULS, "r%d, r%d", a->rd, a->rr) +INSN(FMULSU, "r%d, r%d", a->rd, a->rr) +INSN(ICALL, "") +INSN(IJMP, "") +INSN(IN, "r%d, $%d", a->rd, a->imm) +INSN(INC, "r%d", a->rd) +INSN(JMP, "0x%x", a->imm * 2) +INSN(LAC, "Z, r%d", a->rd) +INSN(LAS, "Z, r%d", a->rd) +INSN(LAT, "Z, r%d", a->rd) +INSN(LDDY, "r%d, Y+%d", a->rd, a->imm) +INSN(LDDZ, "r%d, Z+%d", a->rd, a->imm) +INSN(LDI, "r%d, %d", a->rd, a->imm) +INSN(LDS, "r%d, %d", a->rd, a->imm) +INSN(LDX1, "r%d, X", a->rd) +INSN(LDX2, "r%d, X+", a->rd) +INSN(LDX3, "r%d, -X", a->rd) +INSN(LDY2, "r%d, Y+", a->rd) +INSN(LDY3, "r%d, -Y", a->rd) +INSN(LDZ2, "r%d, Z+", a->rd) +INSN(LDZ3, "r%d, -Z", a->rd) +INSN(LPM1, "") +INSN(LPM2, "r%d, Z", a->rd) +INSN(LPMX, "r%d, Z+", a->rd) +INSN(LSR, "r%d", a->rd) +INSN(MOV, "r%d, r%d", a->rd, a->rr) +INSN(MOVW, "r%d:r%d, r%d,r:r%d", a->rd + 1, a->rd, a->rr + 1, a->rr) +INSN(MUL, "r%d, r%d", a->rd, a->rr) +INSN(MULS, "r%d, r%d", a->rd, a->rr) +INSN(MULSU, "r%d, r%d", a->rd, a->rr) +INSN(NEG, "r%d", a->rd) +INSN(NOP, "") +INSN(OR, "r%d, r%d", a->rd, a->rr) +INSN(ORI, "r%d, %d", a->rd, a->imm) +INSN(OUT, "$%d, r%d", a->imm, a->rd) +INSN(POP, "r%d", a->rd) +INSN(PUSH, "r%d", a->rd) +INSN(RCALL, ".%+d", a->imm * 2) +INSN(RET, "") +INSN(RETI, "") +INSN(RJMP, ".%+d", a->imm * 2) +INSN(ROR, "r%d", a->rd) +INSN(SBC, "r%d, r%d", a->rd, a->rr) +INSN(SBCI, "r%d, %d", a->rd, a->imm) +INSN(SBI, "$%d, %d", a->reg, a->bit) +INSN(SBIC, "$%d, %d", a->reg, a->bit) +INSN(SBIS, "$%d, %d", a->reg, a->bit) +INSN(SBIW, "r%d:r%d, %d", a->rd + 1, a->rd, a->imm) +INSN(SBRC, "r%d, %d", a->rr, a->bit) +INSN(SBRS, "r%d, %d", a->rr, a->bit) +INSN(SLEEP, "") +INSN(SPM, "") +INSN(SPMX, "Z+") +INSN(STDY, "r%d, Y+%d", a->rd, a->imm) +INSN(STDZ, "r%d, Z+%d", a->rd, a->imm) +INSN(STS, "r%d, %d", a->rd, a->imm) +INSN(STX1, "r%d, X", a->rr) +INSN(STX2, "r%d, X+", a->rr) +INSN(STX3, "r%d, -X", a->rr) +INSN(STY2, "r%d, Y+", a->rd) +INSN(STY3, "r%d, -Y", a->rd) +INSN(STZ2, "r%d, Z+", a->rd) +INSN(STZ3, "r%d, -Z", a->rd) +INSN(SUB, "r%d, r%d", a->rd, a->rr) +INSN(SUBI, "r%d, %d", a->rd, a->imm) +INSN(SWAP, "r%d", a->rd) +INSN(WDR, "") +INSN (XCH ,"Z, r%d" , a->rd) +target/avr/translate.c +11
-0@@ -3032,6 +3032,17 @@ done_generating: tb->size = (ctx.npc - pc_start) *2 ; tb->icount = num_insns; + +#ifdef DEBUG_DISAS +if (qemu_loglevel_mask (CPU_LOG_TB_IN_ASM ) + &&qemu_log_in_addr_range (tb->pc)) { +qemu_log_lock (); +qemu_log ("IN: %s\n" ,lookup_symbol (tb->pc)); +log_target_disas (cs, tb->pc, tb->size); +qemu_log ("\n" ); +qemu_log_unlock (); + } +#endif }Quoted 5 lines
Quoted 5 lines
target/avr/translate.c +0
-1void restore_state_to_opc(CPUAVRState *env, TranslationBlock *tb, --2.17.2 (Apple Git-113)
--
Best Regards,
Michael Rolnik
No, I don't.
but I also can load and execute a binary file which does not have this information.
Quoted 490 lines · Aleksandar Markovic
Quoted 25 lines · Aleksandar Markovic
Quoted 21 lines · Michael Rolnik
There is *-cpu *option where you can specify what CPU you want, if this option is not specified avr6 (avr6-avr-cpu) is chosen.
*./avr-softmmu/qemu-system-avr -cpu help*
avr1-avr-cpu
avr2-avr-cpu
avr25-avr-cpu
avr3-avr-cpu
avr31-avr-cpu
avr35-avr-cpu
avr4-avr-cpu
avr5-avr-cpu
avr51-avr-cpu
avr6-avr-cpu
xmega2-avr-cpu
xmega4-avr-cpu
xmega5-avr-cpu
xmega6-avr-cpu
xmega7-avr-cpuQuoted 4 lines
What happens if you specify a core via -cpu, and supply elf file compiled for another core?
It looks there is some related info written in ELF header. This is from a binutils header:
(so it looks you could detect the core from elf file - do you do that detection right now?)
#define
E_AVR_MACH_AVR11
#defineE_AVR_MACH_AVR22
#defineE_AVR_MACH_AVR2525
#defineE_AVR_MACH_AVR33
#defineE_AVR_MACH_AVR3131
#defineE_AVR_MACH_AVR3535
#defineE_AVR_MACH_AVR44
#defineE_AVR_MACH_AVR55
#defineE_AVR_MACH_AVR5151
#defineE_AVR_MACH_AVR66
#defineE_AVR_MACH_AVRTINY100#define E_AVR_MACH_XMEGA1 101 #define E_AVR_MACH_XMEGA2 102 #define E_AVR_MACH_XMEGA3 103 #define E_AVR_MACH_XMEGA4 104 #define E_AVR_MACH_XMEGA5 105 #define E_AVR_MACH_XMEGA6 106 #define E_AVR_MACH_XMEGA7 107Best regards,
Quoted 428 lines
Akeksandar
Quoted 422 lines
Regards,
Michael RolnikQuoted 32 lines · Aleksandar Markovic
Quoted 17 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU.
All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fully tested yet.
However I was able to execute simple code with functions. e.g fibonacci calculation.
This series of patches include a non real, sample board.
No fuses support yet. PC is set to 0 at reset.Following are examples of possible usages, assuming program.elf is compiled for AVR cpu
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnet window
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
runtelent localhost 5678Quoted 10 lines
3. Continious non interrupted execution with serial output into stdout run
qemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -S
runavr-gdb program.elfand then within GDB shell `target remote
:1234`
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmQuoted 377 lines
Hi, Michael.
Can you explain to me the mechanisms of recognition of what core/microcontroller QEMU is supposed to emulate in the examples above?
Yours, Aleksandar
Quoted 368 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding or translation
2. CPU features which allow define the following 8bit AVR coresavr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows to execute simple code
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of0x000000ff) for faster address manipulaton
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM
11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHchanges since v4
1. target is now defined as big endian in order to optimize push_ret/pop_ret
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_cpu_gdb_write_register
2. configure the target as little endian because otherwise GDB does not work
3. fixing and testing gen_push_ret/gen_pop_retchanges since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supported yet registers only
3.MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built for all
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memorychanges since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed to any device
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try to access registerschanges since v10
1. movig back fullwr & fullrd into the commit where outb and inb were introduced
2. changingtlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codechanges since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thing exists)
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrischanges since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers are accessible
by LD/ST only. CPU has no interest in themchanges since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and fixes
a couple of bugs wrt single stepping
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicchanges since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU" into three patches 1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tename1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERS
7. use git commit sha1 instead of `master`` in avocado testMichael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileSarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer
peripheralsconfigure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)--
Best Regards,
Michael Rolnik
Signature · 2 lines
Best Regards, Michael Rolnik
Quoted 6 lines · Michael Rolnik
No, I don't.
but I also can load and execute a binary file which does not have this information.Quoted 2 lines
OK. Let's think about that for a while. I currently think you have here an opportunity to add a really clean interface from the outset of AVR support in QEMU (that even some established platforms don't have in full), which is, trust me, very important for future. And it not that difficult to implement at all. But let's both think for a while.
Best regards,
Aleksandar
Quoted 502 lines
Quoted 62 lines · Aleksandar Markovic
Quoted 26 lines · Aleksandar Markovic
Quoted 21 lines · Michael Rolnik
There is *-cpu *option where you can specify what CPU you want, if this option is not specified avr6 (avr6-avr-cpu) is chosen.
*./avr-softmmu/qemu-system-avr -cpu help*
avr1-avr-cpu
avr2-avr-cpu
avr25-avr-cpu
avr3-avr-cpu
avr31-avr-cpu
avr35-avr-cpu
avr4-avr-cpu
avr5-avr-cpu
avr51-avr-cpu
avr6-avr-cpu
xmega2-avr-cpu
xmega4-avr-cpu
xmega5-avr-cpu
xmega6-avr-cpu
xmega7-avr-cpuQuoted 4 lines
What happens if you specify a core via -cpu, and supply elf file compiled for another core?
It looks there is some related info written in ELF header. This is from a binutils header:
(so it looks you could detect the core from elf file - do you do that detection right now?)
#define
E_AVR_MACH_AVR11
#defineE_AVR_MACH_AVR22
#defineE_AVR_MACH_AVR2525
#defineE_AVR_MACH_AVR33
#defineE_AVR_MACH_AVR3131
#defineE_AVR_MACH_AVR3535
#defineE_AVR_MACH_AVR44
#defineE_AVR_MACH_AVR55
#defineE_AVR_MACH_AVR5151
#defineE_AVR_MACH_AVR66
#defineE_AVR_MACH_AVRTINY100#define E_AVR_MACH_XMEGA1 101 #define E_AVR_MACH_XMEGA2 102 #define E_AVR_MACH_XMEGA3 103 #define E_AVR_MACH_XMEGA4 104 #define E_AVR_MACH_XMEGA5 105 #define E_AVR_MACH_XMEGA6 106 #define E_AVR_MACH_XMEGA7 107Quoted 432 lines
Best regards,
Quoted 431 lines
Akeksandar
Quoted 425 lines
Regards,
Michael RolnikQuoted 33 lines · Aleksandar Markovic
Quoted 18 lines · Michael Rolnik
This series of patches adds 8bit AVR cores to QEMU. All instruction, except BREAK/DES/SPM/SPMX, are implemented. Not fully tested yet.
However I was able to execute simple code with functions. e.g fibonacci calculation.
This series of patches include a non real, sample board. No fuses support yet. PC is set to 0 at reset.Following are examples of possible usages, assuming program.elf is compiled for AVR cpu
1. Continious non interrupted execution
runqemu-system-avr -kernel program.elf
2. Continious non interrupted execution with serial output into telnet window
run `qemu-system-avr -kernel program.elf -serial
tcp::5678,server,nowait -nographic `
runtelent localhost 5678
3. Continious non interrupted execution with serial output intoQuoted 10 lines
stdout
runqemu-system-avr -kernel program.elf -serial stdio
4. Debugging wit GDB debugger
runqemu-system-avr -kernel program.elf -s -Srunavr-gdb program.elfand then within GDB shell `target
remote :1234`
5. Print out executed instructions
runqemu-system-avr -kernel program.elf -d in_asmQuoted 379 lines
Hi, Michael.
Can you explain to me the mechanisms of recognition of what core/microcontroller QEMU is supposed to emulate in the examples above?
Yours, Aleksandar
Quoted 370 lines
the patches include the following
1. just a basic 8bit AVR CPU, without instruction decoding or translation
2. CPU features which allow define the following 8bit AVR coresavr1 avr2 avr25 avr3 avr31 avr35 avr4 avr5 avr51 avr6 xmega2 xmega4 xmega5 xmega6 xmega73. a definition of sample machine with SRAM, FLASH and CPU which allows to execute simple code
4. encoding for all AVR instructions
5. interrupt handling
6. helpers for IN, OUT, SLEEP, WBR & unsupported instructions
7. a decoder which given an opcode decides what istruction it is
8. translation of AVR instruction into TCG
9. all features togetherchanges since v3
1. rampD/X/Y/Z registers are encoded as0x00ff0000(instead of0x000000ff) for faster address manipulaton
2. ffs changed to ctz32
3. duplicate code removed atavr_cpu_do_interrupt
4. using andc instead of not + and
5. fixing V flag calculation in varios instructions
6. freeing local variables in PUSH
7.tcg_const_local_i32->tcg_const_i32
8. using sextract32 instead of my implementation
9. fixing BLD instruction
10.xor(r) instead of0xff- r at COM
11.fixing MULS/MULSU not to modify inputs' content 12.using SUB for NEG
13.fixing tcg_gen_qemu_ld/st call in XCHchanges since v4
1. target is now defined as big endian in order to optimize push_ret/pop_ret
2. all style warnings are fixed
3. adding cpu_set/get_sreg functions
4. simplifyinggen_goto_tbas there is no real paging
5.env->pc->env->pc_w
6. making flag dump more compact
7. more spacing
8. renaming CODE/DATA_INDEX -> MMU_CODE/DATA_IDX
9. removingavr_set_feature
10. SPL/SPH set bug fix
11. switchingstb_phystocpu_stb_data
12. cleaning upavr_decode
13. saving sreg, rampD/X/Y/Z, eind in HW format (savevm)
14. saving CPU features (savevm)changes since v5
1. BLD bug fix
2. decoder generator is addedchages since v6
1. using cpu_get_sreg/cpu_set_sreg in avr_cpu_gdb_read_register/avr_cpu_gdb_write_register
2. configure the target as little endian because otherwise GDB does not work
3. fixing and testing gen_push_ret/gen_pop_retchanges since v7
1. folding back v6
2. logging athelper_outbandhelper_inbare done for non supported yet registers only
3.MAINTAINERSupdatedchanges since v8
1. removing hw/avr from hw/Makefile.obj as it should not be built for all
2. making linux compilable
3. testing ona. Mac, Apple LLVM version 7.0.0 b. Ubuntu 12.04, gcc 4.9.2 c. Fedora 23, gcc 5.3.14. folding back some patches
5. translation bug fixes for ORI, CPI, XOR instructions
6. propper handling of cpu register writes though memorychanges since v9
1. removing forward declarations of static functions
2. disabling debug prints
3. switching to case range instead of if else if ...
4. LD/ST IN/OUT accessing CPU maintainder registers are not routed to any device
5. commenst about sample board and sample IO device added
6. sample board description is more descriptive now
7.memory_region_allocate_system_memoryis used to create RAM
8. now there arehelper_fullrd&helper_fullwrwhen LD/ST try to access registerschanges since v10
1. movig back fullwr & fullrd into the commit where outb and inb were introduced
2. changingtlb_fillfunction signature
3. adding empty line between functions
4. adding newline on the last line of the file
5. usingtb->flagsto generae full access ST/LD instructions
6. fixing SBRC bug
7. folding back 10th commit
8. whenever a new file is introduced it's added to Makefile.objschanges since v11
1. updating to v2.7.0-rc
2. removing assignment toenv->fullaccfromgen_intermediate_codechanges since v12
1. fixing spacing
2. fixing get/put_segment functions
3. removingtarget-avr/machine.hfile
4.VMSTATE_SINGLE_TEST->VMSTATE_SINGLE
5. comment spelling
6. removinghw/avr/sample_io.c
7. char const* -> const char*
8. proper ram allocation
9. fixing breakpoint functionality.
10.env1 -> env
11.fixingavr_cpu_gdb_write_register&avr_cpu_gdb_read_registerfunctions
12.any cpu is removed
12.feature bits are not saved into vm statechanges since v13
1. rebasing to v2.7.0-rc1changes since v14
1. I made self review with git gui tool. (I did not know such a thing exists)
2. removing all double/tripple spaces
3. removing comment reference to SampleIO
4. folding back some changes, so there is not deleted lines in my code
5. moving avr configuration, within configure file, before chrischanges since v15
1. removing IO registers cache from CPU
2. implementing CBI/SBI asread(helper_inb), modify,write(helper_outb)
3. implementing CBIC/SBIC asread(helper_inb), check, branch
4. adding missingtcg_temp_free_i32fortcg_const_i32changes since v16
1. removing EXT IO registers knoledge from CPU. These registers are accessible
by LD/ST only. CPU has no interest in themchanges since v17 (by Richard Henderson)
This is Michael's v17, with some adjustments of my own:1. Fix the whitespace errors reported by "git am",
2. Replace the utf-8 characters with normal ascii,
3. Ditch the separate compilation oftranslate.c.I retained the two separate files that could be regenerated from the included cpugen program, but merged in
translate-insn.c. Not that it matters, but the code generated is about 3k smaller.changes since v18
1. moving target-avr into target/avr 2. do not call cpu_exec_initfn function from avr_cpu_initfn 3. call cpu_exec_realizefn avr_cpu_realizefn 4. do not fail sample machine creation if no rom is suplied 5. add tcg_gen_exit_tb(0) for BS_BRANCH in gen_intermediate_code 6. fix a register getters/setters in machine.c 7. changing QEMU_ARCH_AVR from 1<<17 to 1<<18changes since v19
1. use decodetree.py tool to decode instructions 2. adding USART 3. adding 16 bit timer peripherals 4. changing QEMU_ARCH_AVR from 1<<18 to 1<<20 5. renaming tlb_fill to avr_cpu_tlb_fillchanges since v20
1. use one CPU naming convention 2. merging insn16.decode & insn32.decode files 3. modifying skip next instruction mechanizm 4. translate BREAK as NOP for nowchanges since v21
1. Reorganize bstate.
This will make transition to <exec/translator.h> easier, and
fixes a couple of bugs wrt single stepping
by [email protected]
2. Drop cpc and fix page cross condition.
by [email protected]
3. Refactor checking supported/unsupported instructions
4. Addgdb-xml/avr-cpu.xmlchanges since v22
1. Rebase
2. Split long commentchanges since v23
1. remove avr_cpu_list_compare function 2. modify avr_cpu_class_by_name function 3. modify avr_cpu_list_entry function 4. modify avr_cpu_list functionchanges since v24
1. removeAVR_CPU_TYPE_NAMEmacrochanges since v25
1. fix patches. every file belong to one patch only 2. change copyright year from 2016 to 2019 3. create mask device to emulate prr0/prr1changes since v26
1. add avocado acceptence test
2. add boot serial testchanges since v27
1. list atmel2560 devices as unimplemented
2. fix sram base/sizechanges since v28
1. rebase
2. fix includes & buildchanges since v29
1. fixing ownership
2. using 'since' instread of 'added in'changes since v30
1. rebasechanges since v31
1. splitting 'Add instruction translation' commit 2. fixing typo in qapi/machine.json sicne -> since 3. removing unintended changes in configure file 4. adding Richard Henderson as a co developer to 'Add instructiontranslation - CPU main translation funcions' commit
changes since v32
1. modifycpu_get_sregto treat sreg C as other flags, except sreg Zchanges since v33
1. ensure flag C is always calculated as one bit
2. calculate flag Z as one bit, without using inverse logicchanges since v34
1. rebasechanges since v35
1. rebase 2. use NANOSECONDS_PER_SECOND instead of 1000000000 in avr_timer16.c 3. split "target/avr: Register AVR support with the rest of QEMU" into three patches 1. "target/avr: Register AVR support with the rest of QEMU" 2. "target/avr: Update build system" 3. "target/avr: Update MAINTAINERS file"4. split "target/avr: Add tests" patch into two patches
1. "target/avr: Add Avocado test"
2. "target/avr: Add boot serial test"
5. Add instruction disassembly function
6. change "since 4.2" to "since 5.0"changes since v36
1. rebase
2. tename1. NO_CPU_REGISTERS -> NUMBER_OF_CPU_REGISTERS 2. NO_IO_REGISTERS -> NUMBER_OF_IO_REGISTERS 3. to_A -> to_regs_16_31_by_one 4. to_B -> to_regs_16_23_by_one 5. to_C -> to_regs_24_30_by_two 6. to_D -> to_regs_00_30_by_two 3. add missing licences 4. add usage example (see above) 5. ass Sarah Harris <[email protected]> as a reviewer to
MAINTAINERS
7. use git commit sha1 instead of `master`` in avocado testMichael Rolnik (16):
target/avr: Add outward facing interfaces and core CPU logic target/avr: Add instruction helpers
target/avr: Add instruction decoding
target/avr: Add instruction translation - Registers definition target/avr: Add instruction translation - Arithmetic and Logic
Instructions
target/avr: Add instruction translation - Branch Instructions target/avr: Add instruction translation - Bit and Bit-test
Instructions
target/avr: Add instruction translation - MCU Control Instructions target/avr: Add instruction translation - CPU main translation
function
target/avr: Add instruction disassembly function target/avr: Add example board configuration
target/avr: Register AVR support with the rest of QEMU target/avr: Update build system
target/avr: Add boot serial test
target/avr: Add Avocado test
target/avr: UpdateMAINTAINERSfileSarah Harris (1):
target/avr: Add limited support for USART and 16 bit timer
peripheralsconfigure | 7 + default-configs/avr-softmmu.mak | 5 + qapi/machine.json | 3 +- include/disas/dis-asm.h | 6 + include/hw/char/avr_usart.h | 97 + include/hw/misc/avr_mask.h | 47 + include/hw/timer/avr_timer16.h | 97 + include/sysemu/arch_init.h | 1 + target/avr/cpu-param.h | 37 + target/avr/cpu-qom.h | 54 + target/avr/cpu.h | 254 +++ target/avr/helper.h | 29 + arch_init.c | 2 + hw/avr/sample.c | 282 +++ hw/char/avr_usart.c | 324 ++++ hw/misc/avr_mask.c | 112 ++ hw/timer/avr_timer16.c | 605 ++++++ target/avr/cpu.c | 576 ++++++ target/avr/disas.c | 228 +++ target/avr/gdbstub.c | 85 + target/avr/helper.c | 354 ++++ target/avr/machine.c | 121 ++ target/avr/translate.c | 3052 ++++++++++++++++++++++++++++++ tests/boot-serial-test.c | 10 + tests/machine-none-test.c | 1 + MAINTAINERS | 11 + gdb-xml/avr-cpu.xml | 49 + hw/Kconfig | 1 + hw/avr/Kconfig | 6 + hw/avr/Makefile.objs | 1 + hw/char/Kconfig | 3 + hw/char/Makefile.objs | 1 + hw/misc/Kconfig | 3 + hw/misc/Makefile.objs | 2 + hw/timer/Kconfig | 3 + hw/timer/Makefile.objs | 2 + target/avr/Makefile.objs | 34 + target/avr/insn.decode | 194 ++ tests/Makefile.include | 2 + tests/acceptance/machine_avr6.py | 56 + 40 files changed, 6756 insertions(+), 1 deletion(-) create mode 100644 default-configs/avr-softmmu.mak create mode 100644 include/hw/char/avr_usart.h create mode 100644 include/hw/misc/avr_mask.h create mode 100644 include/hw/timer/avr_timer16.h create mode 100644 target/avr/cpu-param.h create mode 100644 target/avr/cpu-qom.h create mode 100644 target/avr/cpu.h create mode 100644 target/avr/helper.h create mode 100644 hw/avr/sample.c create mode 100644 hw/char/avr_usart.c create mode 100644 hw/misc/avr_mask.c create mode 100644 hw/timer/avr_timer16.c create mode 100644 target/avr/cpu.c create mode 100644 target/avr/disas.c create mode 100644 target/avr/gdbstub.c create mode 100644 target/avr/helper.c create mode 100644 target/avr/machine.c create mode 100644 target/avr/translate.c create mode 100644 gdb-xml/avr-cpu.xml create mode 100644 hw/avr/Kconfig create mode 100644 hw/avr/Makefile.objs create mode 100644 target/avr/Makefile.objs create mode 100644 target/avr/insn.decode create mode 100644 tests/acceptance/machine_avr6.py--
2.17.2 (Apple Git-113)--
Best Regards,
Michael Rolnik--
Best Regards,
Michael Rolnik