| 1 | /* |
| 2 | * Copyright(c) 2019-2022 rev.ng Labs Srl. All Rights Reserved. |
| 3 | * |
| 4 | * This program is free software; you can redistribute it and/or modify |
| 5 | * it under the terms of the GNU General Public License as published by |
| 6 | * the Free Software Foundation; either version 2 of the License, or |
| 7 | * (at your option) any later version. |
| 8 | * |
| 9 | * This program is distributed in the hope that it will be useful, |
| 10 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 11 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 12 | * GNU General Public License for more details. |
| 13 | * |
| 14 | * You should have received a copy of the GNU General Public License |
| 15 | * along with this program; if not, see <http://www.gnu.org/licenses/>. |
| 16 | */ |
| 17 | |
| 18 | #ifndef IDEF_PARSER_H |
| 19 | #define IDEF_PARSER_H |
| 20 | |
| 21 | #include <inttypes.h> |
| 22 | #include <stdio.h> |
| 23 | #include <stdbool.h> |
| 24 | #include <glib.h> |
| 25 | |
| 26 | /* Variadic macros to wrap the buffer printing functions */ |
| 27 | #define EMIT(c, ...) \ |
| 28 | do { \ |
| 29 | g_string_append_printf((c)->out_str, __VA_ARGS__); \ |
| 30 | } while (0) |
| 31 | |
| 32 | #define EMIT_SIG(c, ...) \ |
| 33 | do { \ |
| 34 | g_string_append_printf((c)->signature_str, __VA_ARGS__); \ |
| 35 | } while (0) |
| 36 | |
| 37 | #define EMIT_HEAD(c, ...) \ |
| 38 | do { \ |
| 39 | g_string_append_printf((c)->header_str, __VA_ARGS__); \ |
| 40 | } while (0) |
| 41 | |
| 42 | /** |
| 43 | * Type of register, assigned to the HexReg.type field |
| 44 | */ |
| 45 | typedef enum { GENERAL_PURPOSE, CONTROL, MODIFIER, DOTNEW } HexRegType; |
| 46 | |
| 47 | typedef enum { UNKNOWN_SIGNEDNESS, SIGNED, UNSIGNED } HexSignedness; |
| 48 | |
| 49 | /** |
| 50 | * Semantic record of the REG tokens, identifying registers |
| 51 | */ |
| 52 | typedef struct HexReg { |
| 53 | uint8_t id; /**< Identifier of the register */ |
| 54 | HexRegType type; /**< Type of the register */ |
| 55 | unsigned bit_width; /**< Bit width of the reg, 32 or 64 bits */ |
| 56 | } HexReg; |
| 57 | |
| 58 | /** |
| 59 | * Data structure, identifying a TCGv temporary value |
| 60 | */ |
| 61 | typedef struct HexTmp { |
| 62 | unsigned index; /**< Index of the TCGv temporary value */ |
| 63 | } HexTmp; |
| 64 | |
| 65 | /** |
| 66 | * Enum of the possible immediate, an immediate is a value which is known |
| 67 | * at tinycode generation time, e.g. an integer value, not a TCGv |
| 68 | */ |
| 69 | enum ImmUnionTag { |
| 70 | I, |
| 71 | VARIABLE, |
| 72 | VALUE, |
| 73 | QEMU_TMP, |
| 74 | IMM_PC, |
| 75 | IMM_CONSTEXT, |
| 76 | }; |
| 77 | |
| 78 | /** |
| 79 | * Semantic record of the IMM token, identifying an immediate constant |
| 80 | */ |
| 81 | typedef struct HexImm { |
| 82 | union { |
| 83 | char id; /**< Identifier, used when type is VARIABLE */ |
| 84 | uint64_t value; /**< Immediate value, used when type is VALUE */ |
| 85 | uint64_t index; /**< Index, used when type is QEMU_TMP */ |
| 86 | }; |
| 87 | enum ImmUnionTag type; /**< Type of the immediate */ |
| 88 | } HexImm; |
| 89 | |
| 90 | /** |
| 91 | * Semantic record of the PRED token, identifying a predicate |
| 92 | */ |
| 93 | typedef struct HexPred { |
| 94 | char id; /**< Identifier of the predicate */ |
| 95 | } HexPred; |
| 96 | |
| 97 | /** |
| 98 | * Semantic record of the SAT token, identifying the saturate operator |
| 99 | * Note: All saturates are assumed to implicitly set overflow. |
| 100 | */ |
| 101 | typedef struct HexSat { |
| 102 | HexSignedness signedness; /**< Signedness of the sat. op. */ |
| 103 | } HexSat; |
| 104 | |
| 105 | /** |
| 106 | * Semantic record of the CAST token, identifying the cast operator |
| 107 | */ |
| 108 | typedef struct HexCast { |
| 109 | unsigned bit_width; /**< Bit width of the cast operator */ |
| 110 | HexSignedness signedness; /**< Unsigned flag for the cast operator */ |
| 111 | } HexCast; |
| 112 | |
| 113 | /** |
| 114 | * Semantic record of the EXTRACT token, identifying the cast operator |
| 115 | */ |
| 116 | typedef struct HexExtract { |
| 117 | unsigned bit_width; /**< Bit width of the extract operator */ |
| 118 | unsigned storage_bit_width; /**< Actual bit width of the extract operator */ |
| 119 | HexSignedness signedness; /**< Unsigned flag for the extract operator */ |
| 120 | } HexExtract; |
| 121 | |
| 122 | /** |
| 123 | * Semantic record of the MPY token, identifying the fMPY multiplication |
| 124 | * operator |
| 125 | */ |
| 126 | typedef struct HexMpy { |
| 127 | unsigned first_bit_width; /**< Bit width of 1st operand of fMPY */ |
| 128 | unsigned second_bit_width; /**< Bit width of 2nd operand of fMPY */ |
| 129 | HexSignedness first_signedness; /**< Signedness of 1st operand of fMPY */ |
| 130 | HexSignedness second_signedness; /**< Signedness of 2nd operand of fMPY */ |
| 131 | } HexMpy; |
| 132 | |
| 133 | /** |
| 134 | * Semantic record of the VARID token, identifying declared variables |
| 135 | * of the input language |
| 136 | */ |
| 137 | typedef struct HexVar { |
| 138 | GString *name; /**< Name of the VARID variable */ |
| 139 | } HexVar; |
| 140 | |
| 141 | /** |
| 142 | * Data structure uniquely identifying a declared VARID variable, used for |
| 143 | * keeping track of declared variable, so that any variable is declared only |
| 144 | * once, and its properties are propagated through all the subsequent instances |
| 145 | * of that variable |
| 146 | */ |
| 147 | typedef struct Var { |
| 148 | GString *name; /**< Name of the VARID variable */ |
| 149 | uint8_t bit_width; /**< Bit width of the VARID variable */ |
| 150 | HexSignedness signedness; /**< Unsigned flag for the VARID var */ |
| 151 | } Var; |
| 152 | |
| 153 | /** |
| 154 | * Enum of the possible rvalue types, used in the HexValue.type field |
| 155 | */ |
| 156 | typedef enum RvalueUnionTag { |
| 157 | REGISTER, REGISTER_ARG, TEMP, IMMEDIATE, PREDICATE, VARID |
| 158 | } RvalueUnionTag; |
| 159 | |
| 160 | /** |
| 161 | * Semantic record of the rvalue token, identifying any numeric value, |
| 162 | * immediate or register based. The rvalue tokens are combined together |
| 163 | * through the use of several operators, to encode expressions |
| 164 | */ |
| 165 | typedef struct HexValue { |
| 166 | union { |
| 167 | HexReg reg; /**< rvalue of register type */ |
| 168 | HexTmp tmp; /**< rvalue of temporary type */ |
| 169 | HexImm imm; /**< rvalue of immediate type */ |
| 170 | HexPred pred; /**< rvalue of predicate type */ |
| 171 | HexVar var; /**< rvalue of declared variable type */ |
| 172 | }; |
| 173 | RvalueUnionTag type; /**< Type of the rvalue */ |
| 174 | unsigned bit_width; /**< Bit width of the rvalue */ |
| 175 | HexSignedness signedness; /**< Unsigned flag for the rvalue */ |
| 176 | bool is_dotnew; /**< rvalue of predicate type is dotnew? */ |
| 177 | } HexValue; |
| 178 | |
| 179 | /** |
| 180 | * State of ternary operator |
| 181 | */ |
| 182 | typedef enum TernaryState { IN_LEFT, IN_RIGHT } TernaryState; |
| 183 | |
| 184 | /** |
| 185 | * Data structure used to handle side effects inside ternary operators |
| 186 | */ |
| 187 | typedef struct Ternary { |
| 188 | TernaryState state; |
| 189 | HexValue cond; |
| 190 | } Ternary; |
| 191 | |
| 192 | /** |
| 193 | * Operator type, used for referencing the correct operator when calling the |
| 194 | * gen_bin_op() function, which in turn will generate the correct code to |
| 195 | * execute the operation between the two rvalues |
| 196 | */ |
| 197 | typedef enum OpType { |
| 198 | ADD_OP, SUB_OP, MUL_OP, ASL_OP, ASR_OP, LSR_OP, ANDB_OP, ORB_OP, |
| 199 | XORB_OP, ANDL_OP, MINI_OP, MAXI_OP |
| 200 | } OpType; |
| 201 | |
| 202 | /** |
| 203 | * Data structure including instruction specific information, to be cleared |
| 204 | * out after the compilation of each instruction |
| 205 | */ |
| 206 | typedef struct Inst { |
| 207 | GString *name; /**< Name of the compiled instruction */ |
| 208 | char *code_begin; /**< Beginning of instruction input code */ |
| 209 | char *code_end; /**< End of instruction input code */ |
| 210 | unsigned tmp_count; /**< Index of the last declared TCGv temp */ |
| 211 | unsigned qemu_tmp_count; /**< Index of the last declared int temp */ |
| 212 | unsigned if_count; /**< Index of the last declared if label */ |
| 213 | unsigned error_count; /**< Number of generated errors */ |
| 214 | GArray *allocated; /**< Allocated declaredVARID vars */ |
| 215 | GArray *init_list; /**< List of initialized registers */ |
| 216 | GArray *strings; /**< Strings allocated by the instruction */ |
| 217 | } Inst; |
| 218 | |
| 219 | /** |
| 220 | * Data structure representing the whole translation context, which in a |
| 221 | * reentrant flex/bison parser just like ours is passed between the scanner |
| 222 | * and the parser, holding all the necessary information to perform the |
| 223 | * parsing, this data structure survives between the compilation of different |
| 224 | * instructions |
| 225 | */ |
| 226 | typedef struct Context { |
| 227 | void *scanner; /**< Reentrant parser state pointer */ |
| 228 | char *input_buffer; /**< Buffer containing the input code */ |
| 229 | GString *out_str; /**< String containing the output code */ |
| 230 | GString *signature_str; /**< String containing the signatures code */ |
| 231 | GString *header_str; /**< String containing the header code */ |
| 232 | FILE *defines_file; /**< FILE * of the generated header */ |
| 233 | FILE *output_file; /**< FILE * of the C output file */ |
| 234 | FILE *enabled_file; /**< FILE * of the list of enabled inst */ |
| 235 | GArray *ternary; /**< Array to track nesting of ternary ops */ |
| 236 | unsigned total_insn; /**< Number of instructions in input file */ |
| 237 | unsigned implemented_insn; /**< Instruction compiled without errors */ |
| 238 | Inst inst; /**< Parsing data of the current inst */ |
| 239 | } Context; |
| 240 | |
| 241 | #endif /* IDEF_PARSER_H */ |