micropython/py/asmxtensa.h
Damien George 1396a026be py: Add support to save native, viper and asm code to .mpy files.
This commit adds support for saving and loading .mpy files that contain
native code (native, viper and inline-asm).  A lot of the ground work was
already done for this in the form of removing pointers from generated
native code.  The changes here are mainly to link in qstr values to the
native code, and change the format of .mpy files to contain native code
blocks (possibly mixed with bytecode).

A top-level summary:

- @micropython.native, @micropython.viper and @micropython.asm_thumb/
  asm_xtensa are now allowed in .py files when compiling to .mpy, and they
  work transparently to the user.

- Entire .py files can be compiled to native via mpy-cross -X emit=native
  and for the most part the generated .mpy files should work the same as
  their bytecode version.

- The .mpy file format is changed to 1) specify in the header if the file
  contains native code and if so the architecture (eg x86, ARMV7M, Xtensa);
  2) for each function block the kind of code is specified (bytecode,
  native, viper, asm).

- When native code is loaded from a .mpy file the native code must be
  modified (in place) to link qstr values in, just like bytecode (see
  py/persistentcode.c:arch_link_qstr() function).

In addition, this now defines a public, native ABI for dynamically loadable
native code generated by other languages, like C.
2019-03-08 15:53:05 +11:00

332 lines
15 KiB
C

/*
* This file is part of the MicroPython project, http://micropython.org/
*
* The MIT License (MIT)
*
* Copyright (c) 2016 Damien P. George
*
* 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 MICROPY_INCLUDED_PY_ASMXTENSA_H
#define MICROPY_INCLUDED_PY_ASMXTENSA_H
#include "py/misc.h"
#include "py/asmbase.h"
// calling conventions:
// up to 6 args in a2-a7
// return value in a2
// PC stored in a0
// stack pointer is a1, stack full descending, is aligned to 16 bytes
// callee save: a1, a12, a13, a14, a15
// caller save: a3
#define ASM_XTENSA_REG_A0 (0)
#define ASM_XTENSA_REG_A1 (1)
#define ASM_XTENSA_REG_A2 (2)
#define ASM_XTENSA_REG_A3 (3)
#define ASM_XTENSA_REG_A4 (4)
#define ASM_XTENSA_REG_A5 (5)
#define ASM_XTENSA_REG_A6 (6)
#define ASM_XTENSA_REG_A7 (7)
#define ASM_XTENSA_REG_A8 (8)
#define ASM_XTENSA_REG_A9 (9)
#define ASM_XTENSA_REG_A10 (10)
#define ASM_XTENSA_REG_A11 (11)
#define ASM_XTENSA_REG_A12 (12)
#define ASM_XTENSA_REG_A13 (13)
#define ASM_XTENSA_REG_A14 (14)
#define ASM_XTENSA_REG_A15 (15)
// for bccz
#define ASM_XTENSA_CCZ_EQ (0)
#define ASM_XTENSA_CCZ_NE (1)
// for bcc and setcc
#define ASM_XTENSA_CC_NONE (0)
#define ASM_XTENSA_CC_EQ (1)
#define ASM_XTENSA_CC_LT (2)
#define ASM_XTENSA_CC_LTU (3)
#define ASM_XTENSA_CC_ALL (4)
#define ASM_XTENSA_CC_BC (5)
#define ASM_XTENSA_CC_ANY (8)
#define ASM_XTENSA_CC_NE (9)
#define ASM_XTENSA_CC_GE (10)
#define ASM_XTENSA_CC_GEU (11)
#define ASM_XTENSA_CC_NALL (12)
#define ASM_XTENSA_CC_BS (13)
// macros for encoding instructions (little endian versions)
#define ASM_XTENSA_ENCODE_RRR(op0, op1, op2, r, s, t) \
((((uint32_t)op2) << 20) | (((uint32_t)op1) << 16) | ((r) << 12) | ((s) << 8) | ((t) << 4) | (op0))
#define ASM_XTENSA_ENCODE_RRI4(op0, op1, r, s, t, imm4) \
(((imm4) << 20) | ((op1) << 16) | ((r) << 12) | ((s) << 8) | ((t) << 4) | (op0))
#define ASM_XTENSA_ENCODE_RRI8(op0, r, s, t, imm8) \
((((uint32_t)imm8) << 16) | ((r) << 12) | ((s) << 8) | ((t) << 4) | (op0))
#define ASM_XTENSA_ENCODE_RI16(op0, t, imm16) \
(((imm16) << 8) | ((t) << 4) | (op0))
#define ASM_XTENSA_ENCODE_RSR(op0, op1, op2, rs, t) \
(((op2) << 20) | ((op1) << 16) | ((rs) << 8) | ((t) << 4) | (op0))
#define ASM_XTENSA_ENCODE_CALL(op0, n, offset) \
(((offset) << 6) | ((n) << 4) | (op0))
#define ASM_XTENSA_ENCODE_CALLX(op0, op1, op2, r, s, m, n) \
((((uint32_t)op2) << 20) | (((uint32_t)op1) << 16) | ((r) << 12) | ((s) << 8) | ((m) << 6) | ((n) << 4) | (op0))
#define ASM_XTENSA_ENCODE_BRI8(op0, r, s, m, n, imm8) \
(((imm8) << 16) | ((r) << 12) | ((s) << 8) | ((m) << 6) | ((n) << 4) | (op0))
#define ASM_XTENSA_ENCODE_BRI12(op0, s, m, n, imm12) \
(((imm12) << 12) | ((s) << 8) | ((m) << 6) | ((n) << 4) | (op0))
#define ASM_XTENSA_ENCODE_RRRN(op0, r, s, t) \
(((r) << 12) | ((s) << 8) | ((t) << 4) | (op0))
#define ASM_XTENSA_ENCODE_RI7(op0, s, imm7) \
((((imm7) & 0xf) << 12) | ((s) << 8) | ((imm7) & 0x70) | (op0))
typedef struct _asm_xtensa_t {
mp_asm_base_t base;
uint32_t cur_const;
uint32_t num_const;
uint32_t *const_table;
uint32_t stack_adjust;
} asm_xtensa_t;
void asm_xtensa_end_pass(asm_xtensa_t *as);
void asm_xtensa_entry(asm_xtensa_t *as, int num_locals);
void asm_xtensa_exit(asm_xtensa_t *as);
void asm_xtensa_op16(asm_xtensa_t *as, uint16_t op);
void asm_xtensa_op24(asm_xtensa_t *as, uint32_t op);
// raw instructions
static inline void asm_xtensa_op_add_n(asm_xtensa_t *as, uint reg_dest, uint reg_src_a, uint reg_src_b) {
asm_xtensa_op16(as, ASM_XTENSA_ENCODE_RRRN(10, reg_dest, reg_src_a, reg_src_b));
}
static inline void asm_xtensa_op_addi(asm_xtensa_t *as, uint reg_dest, uint reg_src, int imm8) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRI8(2, 12, reg_src, reg_dest, imm8 & 0xff));
}
static inline void asm_xtensa_op_and(asm_xtensa_t *as, uint reg_dest, uint reg_src_a, uint reg_src_b) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRR(0, 0, 1, reg_dest, reg_src_a, reg_src_b));
}
static inline void asm_xtensa_op_bcc(asm_xtensa_t *as, uint cond, uint reg_src1, uint reg_src2, int32_t rel8) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRI8(7, cond, reg_src1, reg_src2, rel8 & 0xff));
}
static inline void asm_xtensa_op_bccz(asm_xtensa_t *as, uint cond, uint reg_src, int32_t rel12) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_BRI12(6, reg_src, cond, 1, rel12 & 0xfff));
}
static inline void asm_xtensa_op_call0(asm_xtensa_t *as, int32_t rel18) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_CALL(5, 0, rel18 & 0x3ffff));
}
static inline void asm_xtensa_op_callx0(asm_xtensa_t *as, uint reg) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_CALLX(0, 0, 0, 0, reg, 3, 0));
}
static inline void asm_xtensa_op_j(asm_xtensa_t *as, int32_t rel18) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_CALL(6, 0, rel18 & 0x3ffff));
}
static inline void asm_xtensa_op_jx(asm_xtensa_t *as, uint reg) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_CALLX(0, 0, 0, 0, reg, 2, 2));
}
static inline void asm_xtensa_op_l8ui(asm_xtensa_t *as, uint reg_dest, uint reg_base, uint byte_offset) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRI8(2, 0, reg_base, reg_dest, byte_offset & 0xff));
}
static inline void asm_xtensa_op_l16ui(asm_xtensa_t *as, uint reg_dest, uint reg_base, uint half_word_offset) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRI8(2, 1, reg_base, reg_dest, half_word_offset & 0xff));
}
static inline void asm_xtensa_op_l32i(asm_xtensa_t *as, uint reg_dest, uint reg_base, uint word_offset) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRI8(2, 2, reg_base, reg_dest, word_offset & 0xff));
}
static inline void asm_xtensa_op_l32i_n(asm_xtensa_t *as, uint reg_dest, uint reg_base, uint word_offset) {
asm_xtensa_op16(as, ASM_XTENSA_ENCODE_RRRN(8, word_offset & 0xf, reg_base, reg_dest));
}
static inline void asm_xtensa_op_l32r(asm_xtensa_t *as, uint reg_dest, uint32_t op_off, uint32_t dest_off) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RI16(1, reg_dest, ((dest_off - ((op_off + 3) & ~3)) >> 2) & 0xffff));
}
static inline void asm_xtensa_op_mov_n(asm_xtensa_t *as, uint reg_dest, uint reg_src) {
asm_xtensa_op16(as, ASM_XTENSA_ENCODE_RRRN(13, 0, reg_src, reg_dest));
}
static inline void asm_xtensa_op_movi(asm_xtensa_t *as, uint reg_dest, int32_t imm12) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRI8(2, 10, (imm12 >> 8) & 0xf, reg_dest, imm12 & 0xff));
}
static inline void asm_xtensa_op_movi_n(asm_xtensa_t *as, uint reg_dest, int imm4) {
asm_xtensa_op16(as, ASM_XTENSA_ENCODE_RI7(12, reg_dest, imm4));
}
static inline void asm_xtensa_op_mull(asm_xtensa_t *as, uint reg_dest, uint reg_src_a, uint reg_src_b) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRR(0, 2, 8, reg_dest, reg_src_a, reg_src_b));
}
static inline void asm_xtensa_op_or(asm_xtensa_t *as, uint reg_dest, uint reg_src_a, uint reg_src_b) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRR(0, 0, 2, reg_dest, reg_src_a, reg_src_b));
}
static inline void asm_xtensa_op_ret_n(asm_xtensa_t *as) {
asm_xtensa_op16(as, ASM_XTENSA_ENCODE_RRRN(13, 15, 0, 0));
}
static inline void asm_xtensa_op_s8i(asm_xtensa_t *as, uint reg_src, uint reg_base, uint byte_offset) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRI8(2, 4, reg_base, reg_src, byte_offset & 0xff));
}
static inline void asm_xtensa_op_s16i(asm_xtensa_t *as, uint reg_src, uint reg_base, uint half_word_offset) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRI8(2, 5, reg_base, reg_src, half_word_offset & 0xff));
}
static inline void asm_xtensa_op_s32i(asm_xtensa_t *as, uint reg_src, uint reg_base, uint word_offset) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRI8(2, 6, reg_base, reg_src, word_offset & 0xff));
}
static inline void asm_xtensa_op_s32i_n(asm_xtensa_t *as, uint reg_src, uint reg_base, uint word_offset) {
asm_xtensa_op16(as, ASM_XTENSA_ENCODE_RRRN(9, word_offset & 0xf, reg_base, reg_src));
}
static inline void asm_xtensa_op_sll(asm_xtensa_t *as, uint reg_dest, uint reg_src) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRR(0, 1, 10, reg_dest, reg_src, 0));
}
static inline void asm_xtensa_op_sra(asm_xtensa_t *as, uint reg_dest, uint reg_src) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRR(0, 1, 11, reg_dest, 0, reg_src));
}
static inline void asm_xtensa_op_ssl(asm_xtensa_t *as, uint reg_src) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRR(0, 0, 4, 1, reg_src, 0));
}
static inline void asm_xtensa_op_ssr(asm_xtensa_t *as, uint reg_src) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRR(0, 0, 4, 0, reg_src, 0));
}
static inline void asm_xtensa_op_sub(asm_xtensa_t *as, uint reg_dest, uint reg_src_a, uint reg_src_b) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRR(0, 0, 12, reg_dest, reg_src_a, reg_src_b));
}
static inline void asm_xtensa_op_xor(asm_xtensa_t *as, uint reg_dest, uint reg_src_a, uint reg_src_b) {
asm_xtensa_op24(as, ASM_XTENSA_ENCODE_RRR(0, 0, 3, reg_dest, reg_src_a, reg_src_b));
}
// convenience functions
void asm_xtensa_j_label(asm_xtensa_t *as, uint label);
void asm_xtensa_bccz_reg_label(asm_xtensa_t *as, uint cond, uint reg, uint label);
void asm_xtensa_bcc_reg_reg_label(asm_xtensa_t *as, uint cond, uint reg1, uint reg2, uint label);
void asm_xtensa_setcc_reg_reg_reg(asm_xtensa_t *as, uint cond, uint reg_dest, uint reg_src1, uint reg_src2);
size_t asm_xtensa_mov_reg_i32(asm_xtensa_t *as, uint reg_dest, uint32_t i32);
void asm_xtensa_mov_reg_i32_optimised(asm_xtensa_t *as, uint reg_dest, uint32_t i32);
void asm_xtensa_mov_local_reg(asm_xtensa_t *as, int local_num, uint reg_src);
void asm_xtensa_mov_reg_local(asm_xtensa_t *as, uint reg_dest, int local_num);
void asm_xtensa_mov_reg_local_addr(asm_xtensa_t *as, uint reg_dest, int local_num);
void asm_xtensa_mov_reg_pcrel(asm_xtensa_t *as, uint reg_dest, uint label);
void asm_xtensa_call_ind(asm_xtensa_t *as, uint idx);
// Holds a pointer to mp_fun_table
#define ASM_XTENSA_REG_FUN_TABLE ASM_XTENSA_REG_A15
#if GENERIC_ASM_API
// The following macros provide a (mostly) arch-independent API to
// generate native code, and are used by the native emitter.
#define ASM_WORD_SIZE (4)
#define REG_RET ASM_XTENSA_REG_A2
#define REG_ARG_1 ASM_XTENSA_REG_A2
#define REG_ARG_2 ASM_XTENSA_REG_A3
#define REG_ARG_3 ASM_XTENSA_REG_A4
#define REG_ARG_4 ASM_XTENSA_REG_A5
#define REG_ARG_5 ASM_XTENSA_REG_A6
#define REG_TEMP0 ASM_XTENSA_REG_A2
#define REG_TEMP1 ASM_XTENSA_REG_A3
#define REG_TEMP2 ASM_XTENSA_REG_A4
#define REG_LOCAL_1 ASM_XTENSA_REG_A12
#define REG_LOCAL_2 ASM_XTENSA_REG_A13
#define REG_LOCAL_3 ASM_XTENSA_REG_A14
#define REG_LOCAL_NUM (3)
#define REG_FUN_TABLE ASM_XTENSA_REG_FUN_TABLE
#define ASM_T asm_xtensa_t
#define ASM_END_PASS asm_xtensa_end_pass
#define ASM_ENTRY asm_xtensa_entry
#define ASM_EXIT asm_xtensa_exit
#define ASM_JUMP asm_xtensa_j_label
#define ASM_JUMP_IF_REG_ZERO(as, reg, label, bool_test) \
asm_xtensa_bccz_reg_label(as, ASM_XTENSA_CCZ_EQ, reg, label)
#define ASM_JUMP_IF_REG_NONZERO(as, reg, label, bool_test) \
asm_xtensa_bccz_reg_label(as, ASM_XTENSA_CCZ_NE, reg, label)
#define ASM_JUMP_IF_REG_EQ(as, reg1, reg2, label) \
asm_xtensa_bcc_reg_reg_label(as, ASM_XTENSA_CC_EQ, reg1, reg2, label)
#define ASM_JUMP_REG(as, reg) asm_xtensa_op_jx((as), (reg))
#define ASM_CALL_IND(as, idx) asm_xtensa_call_ind((as), (idx))
#define ASM_MOV_LOCAL_REG(as, local_num, reg_src) asm_xtensa_mov_local_reg((as), (local_num), (reg_src))
#define ASM_MOV_REG_IMM(as, reg_dest, imm) asm_xtensa_mov_reg_i32_optimised((as), (reg_dest), (imm))
#define ASM_MOV_REG_IMM_FIX_U16(as, reg_dest, imm) asm_xtensa_mov_reg_i32((as), (reg_dest), (imm))
#define ASM_MOV_REG_IMM_FIX_WORD(as, reg_dest, imm) asm_xtensa_mov_reg_i32((as), (reg_dest), (imm))
#define ASM_MOV_REG_LOCAL(as, reg_dest, local_num) asm_xtensa_mov_reg_local((as), (reg_dest), (local_num))
#define ASM_MOV_REG_REG(as, reg_dest, reg_src) asm_xtensa_op_mov_n((as), (reg_dest), (reg_src))
#define ASM_MOV_REG_LOCAL_ADDR(as, reg_dest, local_num) asm_xtensa_mov_reg_local_addr((as), (reg_dest), (local_num))
#define ASM_MOV_REG_PCREL(as, reg_dest, label) asm_xtensa_mov_reg_pcrel((as), (reg_dest), (label))
#define ASM_LSL_REG_REG(as, reg_dest, reg_shift) \
do { \
asm_xtensa_op_ssl((as), (reg_shift)); \
asm_xtensa_op_sll((as), (reg_dest), (reg_dest)); \
} while (0)
#define ASM_ASR_REG_REG(as, reg_dest, reg_shift) \
do { \
asm_xtensa_op_ssr((as), (reg_shift)); \
asm_xtensa_op_sra((as), (reg_dest), (reg_dest)); \
} while (0)
#define ASM_OR_REG_REG(as, reg_dest, reg_src) asm_xtensa_op_or((as), (reg_dest), (reg_dest), (reg_src))
#define ASM_XOR_REG_REG(as, reg_dest, reg_src) asm_xtensa_op_xor((as), (reg_dest), (reg_dest), (reg_src))
#define ASM_AND_REG_REG(as, reg_dest, reg_src) asm_xtensa_op_and((as), (reg_dest), (reg_dest), (reg_src))
#define ASM_ADD_REG_REG(as, reg_dest, reg_src) asm_xtensa_op_add_n((as), (reg_dest), (reg_dest), (reg_src))
#define ASM_SUB_REG_REG(as, reg_dest, reg_src) asm_xtensa_op_sub((as), (reg_dest), (reg_dest), (reg_src))
#define ASM_MUL_REG_REG(as, reg_dest, reg_src) asm_xtensa_op_mull((as), (reg_dest), (reg_dest), (reg_src))
#define ASM_LOAD_REG_REG_OFFSET(as, reg_dest, reg_base, word_offset) asm_xtensa_op_l32i_n((as), (reg_dest), (reg_base), (word_offset))
#define ASM_LOAD8_REG_REG(as, reg_dest, reg_base) asm_xtensa_op_l8ui((as), (reg_dest), (reg_base), 0)
#define ASM_LOAD16_REG_REG(as, reg_dest, reg_base) asm_xtensa_op_l16ui((as), (reg_dest), (reg_base), 0)
#define ASM_LOAD32_REG_REG(as, reg_dest, reg_base) asm_xtensa_op_l32i_n((as), (reg_dest), (reg_base), 0)
#define ASM_STORE_REG_REG_OFFSET(as, reg_dest, reg_base, word_offset) asm_xtensa_op_s32i_n((as), (reg_dest), (reg_base), (word_offset))
#define ASM_STORE8_REG_REG(as, reg_src, reg_base) asm_xtensa_op_s8i((as), (reg_src), (reg_base), 0)
#define ASM_STORE16_REG_REG(as, reg_src, reg_base) asm_xtensa_op_s16i((as), (reg_src), (reg_base), 0)
#define ASM_STORE32_REG_REG(as, reg_src, reg_base) asm_xtensa_op_s32i_n((as), (reg_src), (reg_base), 0)
#endif // GENERIC_ASM_API
#endif // MICROPY_INCLUDED_PY_ASMXTENSA_H