├── .gitignore ├── LICENSE ├── README.md ├── h.c ├── hello.c ├── makefile ├── xc-tutor.c └── xc.c /.gitignore: -------------------------------------------------------------------------------- 1 | # Object files 2 | *.o 3 | *.ko 4 | *.obj 5 | *.elf 6 | 7 | # Precompiled Headers 8 | *.gch 9 | *.pch 10 | 11 | # Libraries 12 | *.lib 13 | *.a 14 | *.la 15 | *.lo 16 | 17 | # Shared objects (inc. Windows DLLs) 18 | *.dll 19 | *.so 20 | *.so.* 21 | *.dylib 22 | 23 | # Executables 24 | *.exe 25 | *.out 26 | *.app 27 | *.i*86 28 | *.x86_64 29 | *.hex 30 | 31 | # Debug files 32 | *.dSYM/ 33 | -------------------------------------------------------------------------------- /LICENSE: -------------------------------------------------------------------------------- 1 | GNU GENERAL PUBLIC LICENSE 2 | Version 2, June 1991 3 | 4 | Copyright (C) 1989, 1991 Free Software Foundation, Inc., 5 | 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA 6 | Everyone is permitted to copy and distribute verbatim copies 7 | of this license document, but changing it is not allowed. 8 | 9 | Preamble 10 | 11 | The licenses for most software are designed to take away your 12 | freedom to share and change it. By contrast, the GNU General Public 13 | License is intended to guarantee your freedom to share and change free 14 | software--to make sure the software is free for all its users. This 15 | General Public License applies to most of the Free Software 16 | Foundation's software and to any other program whose authors commit to 17 | using it. (Some other Free Software Foundation software is covered by 18 | the GNU Lesser General Public License instead.) You can apply it to 19 | your programs, too. 20 | 21 | When we speak of free software, we are referring to freedom, not 22 | price. Our General Public Licenses are designed to make sure that you 23 | have the freedom to distribute copies of free software (and charge for 24 | this service if you wish), that you receive source code or can get it 25 | if you want it, that you can change the software or use pieces of it 26 | in new free programs; and that you know you can do these things. 27 | 28 | To protect your rights, we need to make restrictions that forbid 29 | anyone to deny you these rights or to ask you to surrender the rights. 30 | These restrictions translate to certain responsibilities for you if you 31 | distribute copies of the software, or if you modify it. 32 | 33 | For example, if you distribute copies of such a program, whether 34 | gratis or for a fee, you must give the recipients all the rights that 35 | you have. You must make sure that they, too, receive or can get the 36 | source code. And you must show them these terms so they know their 37 | rights. 38 | 39 | We protect your rights with two steps: (1) copyright the software, and 40 | (2) offer you this license which gives you legal permission to copy, 41 | distribute and/or modify the software. 42 | 43 | Also, for each author's protection and ours, we want to make certain 44 | that everyone understands that there is no warranty for this free 45 | software. If the software is modified by someone else and passed on, we 46 | want its recipients to know that what they have is not the original, so 47 | that any problems introduced by others will not reflect on the original 48 | authors' reputations. 49 | 50 | Finally, any free program is threatened constantly by software 51 | patents. We wish to avoid the danger that redistributors of a free 52 | program will individually obtain patent licenses, in effect making the 53 | program proprietary. To prevent this, we have made it clear that any 54 | patent must be licensed for everyone's free use or not licensed at all. 55 | 56 | The precise terms and conditions for copying, distribution and 57 | modification follow. 58 | 59 | GNU GENERAL PUBLIC LICENSE 60 | TERMS AND CONDITIONS FOR COPYING, DISTRIBUTION AND MODIFICATION 61 | 62 | 0. This License applies to any program or other work which contains 63 | a notice placed by the copyright holder saying it may be distributed 64 | under the terms of this General Public License. The "Program", below, 65 | refers to any such program or work, and a "work based on the Program" 66 | means either the Program or any derivative work under copyright law: 67 | that is to say, a work containing the Program or a portion of it, 68 | either verbatim or with modifications and/or translated into another 69 | language. (Hereinafter, translation is included without limitation in 70 | the term "modification".) Each licensee is addressed as "you". 71 | 72 | Activities other than copying, distribution and modification are not 73 | covered by this License; they are outside its scope. The act of 74 | running the Program is not restricted, and the output from the Program 75 | is covered only if its contents constitute a work based on the 76 | Program (independent of having been made by running the Program). 77 | Whether that is true depends on what the Program does. 78 | 79 | 1. You may copy and distribute verbatim copies of the Program's 80 | source code as you receive it, in any medium, provided that you 81 | conspicuously and appropriately publish on each copy an appropriate 82 | copyright notice and disclaimer of warranty; keep intact all the 83 | notices that refer to this License and to the absence of any warranty; 84 | and give any other recipients of the Program a copy of this License 85 | along with the Program. 86 | 87 | You may charge a fee for the physical act of transferring a copy, and 88 | you may at your option offer warranty protection in exchange for a fee. 89 | 90 | 2. You may modify your copy or copies of the Program or any portion 91 | of it, thus forming a work based on the Program, and copy and 92 | distribute such modifications or work under the terms of Section 1 93 | above, provided that you also meet all of these conditions: 94 | 95 | a) You must cause the modified files to carry prominent notices 96 | stating that you changed the files and the date of any change. 97 | 98 | b) You must cause any work that you distribute or publish, that in 99 | whole or in part contains or is derived from the Program or any 100 | part thereof, to be licensed as a whole at no charge to all third 101 | parties under the terms of this License. 102 | 103 | c) If the modified program normally reads commands interactively 104 | when run, you must cause it, when started running for such 105 | interactive use in the most ordinary way, to print or display an 106 | announcement including an appropriate copyright notice and a 107 | notice that there is no warranty (or else, saying that you provide 108 | a warranty) and that users may redistribute the program under 109 | these conditions, and telling the user how to view a copy of this 110 | License. (Exception: if the Program itself is interactive but 111 | does not normally print such an announcement, your work based on 112 | the Program is not required to print an announcement.) 113 | 114 | These requirements apply to the modified work as a whole. If 115 | identifiable sections of that work are not derived from the Program, 116 | and can be reasonably considered independent and separate works in 117 | themselves, then this License, and its terms, do not apply to those 118 | sections when you distribute them as separate works. But when you 119 | distribute the same sections as part of a whole which is a work based 120 | on the Program, the distribution of the whole must be on the terms of 121 | this License, whose permissions for other licensees extend to the 122 | entire whole, and thus to each and every part regardless of who wrote it. 123 | 124 | Thus, it is not the intent of this section to claim rights or contest 125 | your rights to work written entirely by you; rather, the intent is to 126 | exercise the right to control the distribution of derivative or 127 | collective works based on the Program. 128 | 129 | In addition, mere aggregation of another work not based on the Program 130 | with the Program (or with a work based on the Program) on a volume of 131 | a storage or distribution medium does not bring the other work under 132 | the scope of this License. 133 | 134 | 3. You may copy and distribute the Program (or a work based on it, 135 | under Section 2) in object code or executable form under the terms of 136 | Sections 1 and 2 above provided that you also do one of the following: 137 | 138 | a) Accompany it with the complete corresponding machine-readable 139 | source code, which must be distributed under the terms of Sections 140 | 1 and 2 above on a medium customarily used for software interchange; or, 141 | 142 | b) Accompany it with a written offer, valid for at least three 143 | years, to give any third party, for a charge no more than your 144 | cost of physically performing source distribution, a complete 145 | machine-readable copy of the corresponding source code, to be 146 | distributed under the terms of Sections 1 and 2 above on a medium 147 | customarily used for software interchange; or, 148 | 149 | c) Accompany it with the information you received as to the offer 150 | to distribute corresponding source code. (This alternative is 151 | allowed only for noncommercial distribution and only if you 152 | received the program in object code or executable form with such 153 | an offer, in accord with Subsection b above.) 154 | 155 | The source code for a work means the preferred form of the work for 156 | making modifications to it. For an executable work, complete source 157 | code means all the source code for all modules it contains, plus any 158 | associated interface definition files, plus the scripts used to 159 | control compilation and installation of the executable. However, as a 160 | special exception, the source code distributed need not include 161 | anything that is normally distributed (in either source or binary 162 | form) with the major components (compiler, kernel, and so on) of the 163 | operating system on which the executable runs, unless that component 164 | itself accompanies the executable. 165 | 166 | If distribution of executable or object code is made by offering 167 | access to copy from a designated place, then offering equivalent 168 | access to copy the source code from the same place counts as 169 | distribution of the source code, even though third parties are not 170 | compelled to copy the source along with the object code. 171 | 172 | 4. You may not copy, modify, sublicense, or distribute the Program 173 | except as expressly provided under this License. Any attempt 174 | otherwise to copy, modify, sublicense or distribute the Program is 175 | void, and will automatically terminate your rights under this License. 176 | However, parties who have received copies, or rights, from you under 177 | this License will not have their licenses terminated so long as such 178 | parties remain in full compliance. 179 | 180 | 5. You are not required to accept this License, since you have not 181 | signed it. However, nothing else grants you permission to modify or 182 | distribute the Program or its derivative works. These actions are 183 | prohibited by law if you do not accept this License. Therefore, by 184 | modifying or distributing the Program (or any work based on the 185 | Program), you indicate your acceptance of this License to do so, and 186 | all its terms and conditions for copying, distributing or modifying 187 | the Program or works based on it. 188 | 189 | 6. Each time you redistribute the Program (or any work based on the 190 | Program), the recipient automatically receives a license from the 191 | original licensor to copy, distribute or modify the Program subject to 192 | these terms and conditions. You may not impose any further 193 | restrictions on the recipients' exercise of the rights granted herein. 194 | You are not responsible for enforcing compliance by third parties to 195 | this License. 196 | 197 | 7. If, as a consequence of a court judgment or allegation of patent 198 | infringement or for any other reason (not limited to patent issues), 199 | conditions are imposed on you (whether by court order, agreement or 200 | otherwise) that contradict the conditions of this License, they do not 201 | excuse you from the conditions of this License. If you cannot 202 | distribute so as to satisfy simultaneously your obligations under this 203 | License and any other pertinent obligations, then as a consequence you 204 | may not distribute the Program at all. For example, if a patent 205 | license would not permit royalty-free redistribution of the Program by 206 | all those who receive copies directly or indirectly through you, then 207 | the only way you could satisfy both it and this License would be to 208 | refrain entirely from distribution of the Program. 209 | 210 | If any portion of this section is held invalid or unenforceable under 211 | any particular circumstance, the balance of the section is intended to 212 | apply and the section as a whole is intended to apply in other 213 | circumstances. 214 | 215 | It is not the purpose of this section to induce you to infringe any 216 | patents or other property right claims or to contest validity of any 217 | such claims; this section has the sole purpose of protecting the 218 | integrity of the free software distribution system, which is 219 | implemented by public license practices. Many people have made 220 | generous contributions to the wide range of software distributed 221 | through that system in reliance on consistent application of that 222 | system; it is up to the author/donor to decide if he or she is willing 223 | to distribute software through any other system and a licensee cannot 224 | impose that choice. 225 | 226 | This section is intended to make thoroughly clear what is believed to 227 | be a consequence of the rest of this License. 228 | 229 | 8. If the distribution and/or use of the Program is restricted in 230 | certain countries either by patents or by copyrighted interfaces, the 231 | original copyright holder who places the Program under this License 232 | may add an explicit geographical distribution limitation excluding 233 | those countries, so that distribution is permitted only in or among 234 | countries not thus excluded. In such case, this License incorporates 235 | the limitation as if written in the body of this License. 236 | 237 | 9. The Free Software Foundation may publish revised and/or new versions 238 | of the General Public License from time to time. Such new versions will 239 | be similar in spirit to the present version, but may differ in detail to 240 | address new problems or concerns. 241 | 242 | Each version is given a distinguishing version number. If the Program 243 | specifies a version number of this License which applies to it and "any 244 | later version", you have the option of following the terms and conditions 245 | either of that version or of any later version published by the Free 246 | Software Foundation. If the Program does not specify a version number of 247 | this License, you may choose any version ever published by the Free Software 248 | Foundation. 249 | 250 | 10. If you wish to incorporate parts of the Program into other free 251 | programs whose distribution conditions are different, write to the author 252 | to ask for permission. For software which is copyrighted by the Free 253 | Software Foundation, write to the Free Software Foundation; we sometimes 254 | make exceptions for this. Our decision will be guided by the two goals 255 | of preserving the free status of all derivatives of our free software and 256 | of promoting the sharing and reuse of software generally. 257 | 258 | NO WARRANTY 259 | 260 | 11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY 261 | FOR THE PROGRAM, TO THE EXTENT PERMITTED BY APPLICABLE LAW. EXCEPT WHEN 262 | OTHERWISE STATED IN WRITING THE COPYRIGHT HOLDERS AND/OR OTHER PARTIES 263 | PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY OF ANY KIND, EITHER EXPRESSED 264 | OR IMPLIED, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF 265 | MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE. THE ENTIRE RISK AS 266 | TO THE QUALITY AND PERFORMANCE OF THE PROGRAM IS WITH YOU. SHOULD THE 267 | PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF ALL NECESSARY SERVICING, 268 | REPAIR OR CORRECTION. 269 | 270 | 12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING 271 | WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MAY MODIFY AND/OR 272 | REDISTRIBUTE THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, 273 | INCLUDING ANY GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING 274 | OUT OF THE USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED 275 | TO LOSS OF DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY 276 | YOU OR THIRD PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER 277 | PROGRAMS), EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE 278 | POSSIBILITY OF SUCH DAMAGES. 279 | 280 | END OF TERMS AND CONDITIONS 281 | 282 | How to Apply These Terms to Your New Programs 283 | 284 | If you develop a new program, and you want it to be of the greatest 285 | possible use to the public, the best way to achieve this is to make it 286 | free software which everyone can redistribute and change under these terms. 287 | 288 | To do so, attach the following notices to the program. It is safest 289 | to attach them to the start of each source file to most effectively 290 | convey the exclusion of warranty; and each file should have at least 291 | the "copyright" line and a pointer to where the full notice is found. 292 | 293 | {description} 294 | Copyright (C) {year} {fullname} 295 | 296 | This program is free software; you can redistribute it and/or modify 297 | it under the terms of the GNU General Public License as published by 298 | the Free Software Foundation; either version 2 of the License, or 299 | (at your option) any later version. 300 | 301 | This program is distributed in the hope that it will be useful, 302 | but WITHOUT ANY WARRANTY; without even the implied warranty of 303 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the 304 | GNU General Public License for more details. 305 | 306 | You should have received a copy of the GNU General Public License along 307 | with this program; if not, write to the Free Software Foundation, Inc., 308 | 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA. 309 | 310 | Also add information on how to contact you by electronic and paper mail. 311 | 312 | If the program is interactive, make it output a short notice like this 313 | when it starts in an interactive mode: 314 | 315 | Gnomovision version 69, Copyright (C) year name of author 316 | Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. 317 | This is free software, and you are welcome to redistribute it 318 | under certain conditions; type `show c' for details. 319 | 320 | The hypothetical commands `show w' and `show c' should show the appropriate 321 | parts of the General Public License. Of course, the commands you use may 322 | be called something other than `show w' and `show c'; they could even be 323 | mouse-clicks or menu items--whatever suits your program. 324 | 325 | You should also get your employer (if you work as a programmer) or your 326 | school, if any, to sign a "copyright disclaimer" for the program, if 327 | necessary. Here is a sample; alter the names: 328 | 329 | Yoyodyne, Inc., hereby disclaims all copyright interest in the program 330 | `Gnomovision' (which makes passes at compilers) written by James Hacker. 331 | 332 | {signature of Ty Coon}, 1 April 1989 333 | Ty Coon, President of Vice 334 | 335 | This General Public License does not permit incorporating your program into 336 | proprietary programs. If your program is a subroutine library, you may 337 | consider it more useful to permit linking proprietary applications with the 338 | library. If this is what you want to do, use the GNU Lesser General 339 | Public License instead of this License. 340 | 341 | -------------------------------------------------------------------------------- /README.md: -------------------------------------------------------------------------------- 1 | C interpreter that interpretes itself. 2 | 3 | # How to Run the Code 4 | 5 | File `xc.c` is the original one and `xc-tutor.c` is the one that I make for 6 | the tutorial step by step. 7 | 8 | ``` 9 | gcc -o xc xc.c (you may need the -m32 option on 64bit machines) 10 | ./xc hello.c 11 | ./xc -s hello.c 12 | 13 | ./xc c4.c hello.c 14 | ./xc c4.c c4.c hello.c 15 | ``` 16 | 17 | # About 18 | 19 | This project is inspired by [c4](https://github.com/rswier/c4) and is largely 20 | based on it. 21 | 22 | However, I rewrited them all to make it more understable and help myself to 23 | understand it. 24 | 25 | Despite the complexity we saw in books about compiler design, writing one is 26 | not that hard. You don't need that much theory though they will help for 27 | better understanding the logic behind the code. 28 | 29 | Also I write a series of article about how this compiler is built(in Chinese though): 30 | 31 | 1. [手把手教你构建 C 语言编译器（0）——前言](http://lotabout.me/2015/write-a-C-interpreter-0/) 32 | 2. [手把手教你构建 C 语言编译器（1）——设计](http://lotabout.me/2015/write-a-C-interpreter-1/) 33 | 3. [手把手教你构建 C 语言编译器（2）——虚拟机](http://lotabout.me/2015/write-a-C-interpreter-2/) 34 | 4. [手把手教你构建 C 语言编译器（3）——词法分析器](http://lotabout.me/2015/write-a-C-interpreter-3/) 35 | 4. [手把手教你构建 C 语言编译器（4）——递归下降](http://lotabout.me/2016/write-a-C-interpreter-4/) 36 | 5. [手把手教你构建 C 语言编译器（5）——变量定义](http://lotabout.me/2016/write-a-C-interpreter-5/) 37 | 6. [手把手教你构建 C 语言编译器（6）——函数定义](http://lotabout.me/2016/write-a-C-interpreter-6/) 38 | 7. [手把手教你构建 C 语言编译器（7）——语句](http://lotabout.me/2016/write-a-C-interpreter-7/) 39 | 8. [手把手教你构建 C 语言编译器（8）——表达式](http://lotabout.me/2016/write-a-C-interpreter-8/) 40 | 0. [手把手教你构建 C 语言编译器（9）——总结](http://lotabout.me/2016/write-a-C-interpreter-9/) 41 | 42 | # Resources 43 | 44 | Further Reading: 45 | 46 | - [Let's Build a Compiler](http://compilers.iecc.com/crenshaw/): An excellent 47 | starting material for building compiler. 48 | 49 | 50 | Forks: 51 | 52 | - [A fork that implement debugger for xc.c](https://github.com/descent/write-a-C-interpreter) 53 | 54 | 55 | # Licence 56 | 57 | The original code is licenced with GPL2, so this code will use the same 58 | licence. 59 | -------------------------------------------------------------------------------- /h.c: -------------------------------------------------------------------------------- 1 | #include 2 | 3 | int main(int argc, char **argv) 4 | { 5 | int i,j; 6 | j = 1 + 2; 7 | //printf("hello\n"); 8 | printf("hello %d\n", j); 9 | // printf("xx %d\n", i); 10 | return 2; 11 | } 12 | -------------------------------------------------------------------------------- /hello.c: -------------------------------------------------------------------------------- 1 | #include 2 | 3 | int fibonacci(int i) { 4 | if (i <= 1) { 5 | return 1; 6 | } 7 | return fibonacci(i-1) + fibonacci(i-2); 8 | } 9 | 10 | int main() 11 | { 12 | int i; 13 | i = 0; 14 | while (i <= 10) { 15 | printf("fibonacci(%2d) = %d\n", i, fibonacci(i)); 16 | i = i + 1; 17 | } 18 | return 0; 19 | } 20 | -------------------------------------------------------------------------------- /makefile: -------------------------------------------------------------------------------- 1 | CC=gcc 2 | CFLAGS = -m32 -g 3 | xc-tutor: xc-tutor.o 4 | $(CC) $(CFLAGS) -o $@ $< 5 | xc-tutor.o: xc-tutor.c 6 | $(CC) $(CFLAGS) -c $< 7 | clean: 8 | rm -rf *.o xc-tutor 9 | -------------------------------------------------------------------------------- /xc-tutor.c: -------------------------------------------------------------------------------- 1 | // this file is used for tutorial to build the compiler step by step 2 | 3 | #include 4 | #include 5 | #include 6 | #include 7 | 8 | 9 | #include 10 | #include 11 | #include 12 | #include 13 | 14 | #define SUPPORT_DEBUG 15 | 16 | #ifdef SUPPORT_DEBUG 17 | int print_spec_text(int *cur_text); 18 | void print_text(); 19 | void print_data(); 20 | void print_symbol_table(); 21 | void print_source_code(); 22 | void debug_usage(); 23 | #define MAX_BREAK_POINT 10 24 | unsigned int break_points[MAX_BREAK_POINT]; 25 | int last_bp; 26 | int meet_break_point(int *pc); 27 | #endif 28 | 29 | int token; // current token 30 | int token_val; // value of current token (mainly for number) 31 | char *src_begin; 32 | char *src, *old_src; // pointer to source code string; 33 | int poolsize; // default size of text/data/stack 34 | int line; // line number 35 | int *begin_text; 36 | int *begin_stack; 37 | signed char *begin_data; 38 | int *text, // text segment 39 | *old_text, // for dump text segment 40 | *stack; // stack 41 | signed char *data; // data segment 42 | int *pc, *bp, *sp, ax, cycle; // virtual machine registers 43 | int *current_id, // current parsed ID 44 | *symbols; // symbol table 45 | int *idmain; // the `main` function 46 | 47 | // instructions 48 | enum { LEA ,IMM ,JMP ,CALL,JZ ,JNZ ,ENT ,ADJ ,LEV ,LI ,LC ,SI ,SC ,PUSH, 49 | OR ,XOR ,AND ,EQ ,NE ,LT ,GT ,LE ,GE ,SHL ,SHR ,ADD ,SUB ,MUL ,DIV ,MOD , 50 | OPEN,READ,CLOS,PRTF,MALC,MSET,MCMP,EXIT }; 51 | 52 | #ifdef SUPPORT_DEBUG 53 | const char *inst_str[] = 54 | { 55 | "LEA","IMM","JMP","CALL","JZ ","JNZ","ENT","ADJ","LEV","LI ","LC ","SI ","SC ","PUSH"," OR ","XOR","AND","EQ ","NE ","LT ","GT ","LE ","GE ","SHL","SHR","ADD","SUB","MUL","DIV","MOD"," OPEN","READ","CLOS","PRTF","MALC","MSET","MCMP","EXIT" 56 | }; 57 | 58 | #define INST_LEN (sizeof(inst_str)/sizeof(char*)) 59 | 60 | int inst_has_argu(int inst) 61 | { 62 | if ( (LEA <= inst) && (inst <= ADJ)) 63 | return 1; 64 | else 65 | return 0; 66 | } 67 | 68 | const char* inst_2_str(int inst) 69 | { 70 | if ( (0 <= inst) && (inst < INST_LEN)) 71 | { 72 | return inst_str[inst]; 73 | } 74 | else 75 | return 0; 76 | } 77 | #endif 78 | 79 | // tokens and classes (operators last and in precedence order) 80 | enum { 81 | Num = 128, Fun, Sys, Glo, Loc, Id, 82 | Char, Else, Enum, If, Int, Return, Sizeof, While, 83 | Assign, Cond, Lor, Lan, Or, Xor, And, Eq, Ne, Lt, Gt, Le, Ge, Shl, Shr, Add, Sub, Mul, Div, Mod, Inc, Dec, Brak 84 | }; 85 | 86 | #ifdef SUPPORT_DEBUG 87 | const char *class_str_array[] = 88 | { 89 | "Num", "Fun", "Sys", "Glo", "Loc", "Id", "Char", "Else", "Enum", "If", "Int", "Return", "Sizeof", "While", "Assign", "Cond", "Lor", "Lan", "Or", "Xor", "And", "Eq", "Ne", "Lt", "Gt", "Le", "Ge", "Shl", "Shr", "Add", "Sub", "Mul", "Div", "Mod", "Inc", "Dec", "Brak" 90 | }; 91 | 92 | const char *class_str(int cls) 93 | { 94 | int class_index = cls-Num; 95 | 96 | if (0 <= class_index && class_index <= Brak) 97 | return class_str_array[class_index]; 98 | else 99 | return "no such class"; 100 | } 101 | #endif 102 | 103 | 104 | // fields of identifier 105 | enum {Token, Hash, Name, Type, Class, Value, BType, BClass, BValue, IdSize}; 106 | 107 | // types of variable/function 108 | enum { CHAR, INT, PTR }; 109 | 110 | #ifdef SUPPORT_DEBUG 111 | const char *type_str_array[] = {"CHAR", "INT", "PTR"}; 112 | const char * type_str(int type) 113 | { 114 | if (0 <= type && type <= PTR) 115 | return type_str_array[type]; 116 | else 117 | return "no such type"; 118 | } 119 | #endif 120 | 121 | int basetype; // the type of a declaration, make it global for convenience 122 | int expr_type; // the type of an expression 123 | 124 | // function frame 125 | // 126 | // 0: arg 1 127 | // 1: arg 2 128 | // 2: arg 3 129 | // 3: return address 130 | // 4: old bp pointer <- index_of_bp 131 | // 5: local var 1 132 | // 6: local var 2 133 | int index_of_bp; // index of bp pointer on stack 134 | 135 | void next() { 136 | char *last_pos; 137 | int hash; 138 | 139 | while (token = *src) { 140 | ++src; 141 | 142 | // parse token here 143 | if (token == '\n') { 144 | ++line; 145 | } 146 | else if (token == '#') { 147 | // skip macro, because we will not support it 148 | while (*src != 0 && *src != '\n') { 149 | src++; 150 | } 151 | } 152 | else if ((token >= 'a' && token <= 'z') || (token >= 'A' && token <= 'Z') || (token == '_')) { 153 | 154 | // parse identifier 155 | last_pos = src - 1; 156 | hash = token; 157 | 158 | while ((*src >= 'a' && *src <= 'z') || (*src >= 'A' && *src <= 'Z') || (*src >= '0' && *src <= '9') || (*src == '_')) { 159 | hash = hash * 147 + *src; 160 | src++; 161 | } 162 | 163 | // look for existing identifier, linear search 164 | current_id = symbols; 165 | while (current_id[Token]) { 166 | if (current_id[Hash] == hash && !memcmp((char *)current_id[Name], last_pos, src - last_pos)) { 167 | //found one, return 168 | token = current_id[Token]; 169 | return; 170 | } 171 | current_id = current_id + IdSize; 172 | } 173 | 174 | 175 | // store new ID 176 | current_id[Name] = (int)last_pos; 177 | current_id[Hash] = hash; 178 | token = current_id[Token] = Id; 179 | return; 180 | } 181 | else if (token >= '0' && token <= '9') { 182 | // parse number, three kinds: dec(123) hex(0x123) oct(017) 183 | token_val = token - '0'; 184 | if (token_val > 0) { 185 | // dec, starts with [1-9] 186 | while (*src >= '0' && *src <= '9') { 187 | token_val = token_val*10 + *src++ - '0'; 188 | } 189 | } else { 190 | // starts with 0 191 | if (*src == 'x' || *src == 'X') { 192 | //hex 193 | token = *++src; 194 | while ((token >= '0' && token <= '9') || (token >= 'a' && token <= 'f') || (token >= 'A' && token <= 'F')) { 195 | token_val = token_val * 16 + (token & 15) + (token >= 'A' ? 9 : 0); 196 | token = *++src; 197 | } 198 | } else { 199 | // oct 200 | while (*src >= '0' && *src <= '7') { 201 | token_val = token_val*8 + *src++ - '0'; 202 | } 203 | } 204 | } 205 | 206 | token = Num; 207 | return; 208 | } 209 | else if (token == '"' || token == '\'') { 210 | // parse string literal, currently, the only supported escape 211 | // character is '\n', store the string literal into data. 212 | last_pos = data; 213 | while (*src != 0 && *src != token) { 214 | token_val = *src++; 215 | if (token_val == '\\') { 216 | // escape character 217 | token_val = *src++; 218 | if (token_val == 'n') { 219 | token_val = '\n'; 220 | } 221 | } 222 | 223 | if (token == '"') { 224 | *data++ = token_val; 225 | } 226 | } 227 | 228 | src++; 229 | // if it is a single character, return Num token 230 | if (token == '"') { 231 | token_val = (int)last_pos; 232 | } else { 233 | token = Num; 234 | } 235 | 236 | return; 237 | } 238 | else if (token == '/') { 239 | if (*src == '/') { 240 | // skip comments 241 | while (*src != 0 && *src != '\n') { 242 | ++src; 243 | } 244 | } else { 245 | // divide operator 246 | token = Div; 247 | return; 248 | } 249 | } 250 | else if (token == '=') { 251 | // parse '==' and '=' 252 | if (*src == '=') { 253 | src ++; 254 | token = Eq; 255 | } else { 256 | token = Assign; 257 | } 258 | return; 259 | } 260 | else if (token == '+') { 261 | // parse '+' and '++' 262 | if (*src == '+') { 263 | src ++; 264 | token = Inc; 265 | } else { 266 | token = Add; 267 | } 268 | return; 269 | } 270 | else if (token == '-') { 271 | // parse '-' and '--' 272 | if (*src == '-') { 273 | src ++; 274 | token = Dec; 275 | } else { 276 | token = Sub; 277 | } 278 | return; 279 | } 280 | else if (token == '!') { 281 | // parse '!=' 282 | if (*src == '=') { 283 | src++; 284 | token = Ne; 285 | } 286 | return; 287 | } 288 | else if (token == '<') { 289 | // parse '<=', '<<' or '<' 290 | if (*src == '=') { 291 | src ++; 292 | token = Le; 293 | } else if (*src == '<') { 294 | src ++; 295 | token = Shl; 296 | } else { 297 | token = Lt; 298 | } 299 | return; 300 | } 301 | else if (token == '>') { 302 | // parse '>=', '>>' or '>' 303 | if (*src == '=') { 304 | src ++; 305 | token = Ge; 306 | } else if (*src == '>') { 307 | src ++; 308 | token = Shr; 309 | } else { 310 | token = Gt; 311 | } 312 | return; 313 | } 314 | else if (token == '|') { 315 | // parse '|' or '||' 316 | if (*src == '|') { 317 | src ++; 318 | token = Lor; 319 | } else { 320 | token = Or; 321 | } 322 | return; 323 | } 324 | else if (token == '&') { 325 | // parse '&' and '&&' 326 | if (*src == '&') { 327 | src ++; 328 | token = Lan; 329 | } else { 330 | token = And; 331 | } 332 | return; 333 | } 334 | else if (token == '^') { 335 | token = Xor; 336 | return; 337 | } 338 | else if (token == '%') { 339 | token = Mod; 340 | return; 341 | } 342 | else if (token == '*') { 343 | token = Mul; 344 | return; 345 | } 346 | else if (token == '[') { 347 | token = Brak; 348 | return; 349 | } 350 | else if (token == '?') { 351 | token = Cond; 352 | return; 353 | } 354 | else if (token == '~' || token == ';' || token == '{' || token == '}' || token == '(' || token == ')' || token == ']' || token == ',' || token == ':') { 355 | // directly return the character as token; 356 | return; 357 | } 358 | } 359 | return; 360 | } 361 | 362 | void match(int tk) { 363 | if (token == tk) { 364 | next(); 365 | } else { 366 | printf("%d: expected token: %d\n", line, tk); 367 | exit(-1); 368 | } 369 | } 370 | 371 | void expression(int level) { 372 | // expressions have various format. 373 | // but majorly can be divided into two parts: unit and operator 374 | // for example `(char) *a[10] = (int *) func(b > 0 ? 10 : 20); 375 | // `a[10]` is an unit while `*` is an operator. 376 | // `func(...)` in total is an unit. 377 | // so we should first parse those unit and unary operators 378 | // and then the binary ones 379 | // 380 | // also the expression can be in the following types: 381 | // 382 | // 1. unit_unary ::= unit | unit unary_op | unary_op unit 383 | // 2. expr ::= unit_unary (bin_op unit_unary ...) 384 | 385 | // unit_unary() 386 | int *id; 387 | int tmp; 388 | int *addr; 389 | { 390 | if (!token) { 391 | printf("%d: unexpected token EOF of expression\n", line); 392 | exit(-1); 393 | } 394 | if (token == Num) { 395 | match(Num); 396 | 397 | // emit code 398 | *++text = IMM; 399 | *++text = token_val; 400 | expr_type = INT; 401 | } 402 | else if (token == '"') { 403 | // continous string "abc" "abc" 404 | 405 | 406 | // emit code 407 | *++text = IMM; 408 | *++text = token_val; 409 | 410 | match('"'); 411 | // store the rest strings 412 | while (token == '"') { 413 | match('"'); 414 | } 415 | 416 | // append the end of string character '\0', all the data are default 417 | // to 0, so just move data one position forward. 418 | data = (char *)(((int)data + sizeof(int)) & (-sizeof(int))); 419 | // *data = 0; 420 | expr_type = PTR; 421 | } 422 | else if (token == Sizeof) { 423 | // sizeof is actually an unary operator 424 | // now only `sizeof(int)`, `sizeof(char)` and `sizeof(*...)` are 425 | // supported. 426 | match(Sizeof); 427 | match('('); 428 | expr_type = INT; 429 | 430 | if (token == Int) { 431 | match(Int); 432 | } else if (token == Char) { 433 | match(Char); 434 | expr_type = CHAR; 435 | } 436 | 437 | while (token == Mul) { 438 | match(Mul); 439 | expr_type = expr_type + PTR; 440 | } 441 | 442 | match(')'); 443 | 444 | // emit code 445 | *++text = IMM; 446 | *++text = (expr_type == CHAR) ? sizeof(char) : sizeof(int); 447 | 448 | expr_type = INT; 449 | } 450 | else if (token == Id) { 451 | // there are several type when occurs to Id 452 | // but this is unit, so it can only be 453 | // 1. function call 454 | // 2. Enum variable 455 | // 3. global/local variable 456 | match(Id); 457 | 458 | id = current_id; 459 | 460 | if (token == '(') { 461 | // function call 462 | match('('); 463 | 464 | // pass in arguments 465 | tmp = 0; // number of arguments 466 | while (token != ')') { 467 | expression(Assign); 468 | *++text = PUSH; 469 | tmp ++; 470 | 471 | if (token == ',') { 472 | match(','); 473 | } 474 | 475 | } 476 | match(')'); 477 | 478 | // emit code 479 | if (id[Class] == Sys) { 480 | // system functions 481 | *++text = id[Value]; 482 | } 483 | else if (id[Class] == Fun) { 484 | // function call 485 | *++text = CALL; 486 | *++text = id[Value]; 487 | } 488 | else { 489 | printf("%d: bad function call\n", line); 490 | exit(-1); 491 | } 492 | 493 | // clean the stack for arguments 494 | if (tmp > 0) { 495 | *++text = ADJ; 496 | *++text = tmp; 497 | } 498 | expr_type = id[Type]; 499 | } 500 | else if (id[Class] == Num) { 501 | // enum variable 502 | *++text = IMM; 503 | *++text = id[Value]; 504 | expr_type = INT; 505 | } 506 | else { 507 | // variable 508 | if (id[Class] == Loc) { 509 | *++text = LEA; 510 | *++text = index_of_bp - id[Value]; 511 | } 512 | else if (id[Class] == Glo) { 513 | *++text = IMM; 514 | *++text = id[Value]; 515 | } 516 | else { 517 | printf("%d: undefined variable\n", line); 518 | exit(-1); 519 | } 520 | 521 | // emit code, default behaviour is to load the value of the 522 | // address which is stored in `ax` 523 | expr_type = id[Type]; 524 | *++text = (expr_type == Char) ? LC : LI; 525 | } 526 | } 527 | else if (token == '(') { 528 | // cast or parenthesis 529 | match('('); 530 | if (token == Int || token == Char) { 531 | tmp = (token == Char) ? CHAR : INT; // cast type 532 | match(token); 533 | while (token == Mul) { 534 | match(Mul); 535 | tmp = tmp + PTR; 536 | } 537 | 538 | match(')'); 539 | 540 | expression(Inc); // cast has precedence as Inc(++) 541 | 542 | expr_type = tmp; 543 | } else { 544 | // normal parenthesis 545 | expression(Assign); 546 | match(')'); 547 | } 548 | } 549 | else if (token == Mul) { 550 | // dereference * 551 | match(Mul); 552 | expression(Inc); // dereference has the same precedence as Inc(++) 553 | 554 | if (expr_type >= PTR) { 555 | expr_type = expr_type - PTR; 556 | } else { 557 | printf("%d: bad dereference\n", line); 558 | exit(-1); 559 | } 560 | 561 | *++text = (expr_type == CHAR) ? LC : LI; 562 | } 563 | else if (token == And) { 564 | // get the address of 565 | match(And); 566 | expression(Inc); // get the address of 567 | if (*text == LC || *text == LI) { 568 | text --; 569 | } else { 570 | printf("%d: bad address of\n", line); 571 | exit(-1); 572 | } 573 | 574 | expr_type = expr_type + PTR; 575 | } 576 | else if (token == '!') { 577 | // not 578 | match('!'); 579 | expression(Inc); 580 | 581 | // emit code, use == 0 582 | *++text = PUSH; 583 | *++text = IMM; 584 | *++text = 0; 585 | *++text = EQ; 586 | 587 | expr_type = INT; 588 | } 589 | else if (token == '~') { 590 | // bitwise not 591 | match('~'); 592 | expression(Inc); 593 | 594 | // emit code, use XOR -1 595 | *++text = PUSH; 596 | *++text = IMM; 597 | *++text = -1; 598 | *++text = XOR; 599 | 600 | expr_type = INT; 601 | } 602 | else if (token == Add) { 603 | // +var, do nothing 604 | match(Add); 605 | expression(Inc); 606 | 607 | expr_type = INT; 608 | } 609 | else if (token == Sub) { 610 | // -var 611 | match(Sub); 612 | 613 | if (token == Num) { 614 | *++text = IMM; 615 | *++text = -token_val; 616 | match(Num); 617 | } else { 618 | 619 | *++text = IMM; 620 | *++text = -1; 621 | *++text = PUSH; 622 | expression(Inc); 623 | *++text = MUL; 624 | } 625 | 626 | expr_type = INT; 627 | } 628 | else if (token == Inc || token == Dec) { 629 | tmp = token; 630 | match(token); 631 | expression(Inc); 632 | if (*text == LC) { 633 | *text = PUSH; // to duplicate the address 634 | *++text = LC; 635 | } else if (*text == LI) { 636 | *text = PUSH; 637 | *++text = LI; 638 | } else { 639 | printf("%d: bad lvalue of pre-increment\n", line); 640 | exit(-1); 641 | } 642 | *++text = PUSH; 643 | *++text = IMM; 644 | *++text = (expr_type > PTR) ? sizeof(int) : sizeof(char); 645 | *++text = (tmp == Inc) ? ADD : SUB; 646 | *++text = (expr_type == CHAR) ? SC : SI; 647 | } 648 | else { 649 | printf("%d: bad expression\n", line); 650 | exit(-1); 651 | } 652 | } 653 | 654 | // binary operator and postfix operators. 655 | { 656 | while (token >= level) { 657 | // handle according to current operator's precedence 658 | tmp = expr_type; 659 | if (token == Assign) { 660 | // var = expr; 661 | match(Assign); 662 | if (*text == LC || *text == LI) { 663 | *text = PUSH; // save the lvalue's pointer 664 | } else { 665 | printf("%d: bad lvalue in assignment\n", line); 666 | exit(-1); 667 | } 668 | expression(Assign); 669 | 670 | expr_type = tmp; 671 | *++text = (expr_type == CHAR) ? SC : SI; 672 | } 673 | else if (token == Cond) { 674 | // expr ? a : b; 675 | match(Cond); 676 | *++text = JZ; 677 | addr = ++text; 678 | expression(Assign); 679 | if (token == ':') { 680 | match(':'); 681 | } else { 682 | printf("%d: missing colon in conditional\n", line); 683 | exit(-1); 684 | } 685 | *addr = (int)(text + 3); 686 | *++text = JMP; 687 | addr = ++text; 688 | expression(Cond); 689 | *addr = (int)(text + 1); 690 | } 691 | else if (token == Lor) { 692 | // logic or 693 | match(Lor); 694 | *++text = JNZ; 695 | addr = ++text; 696 | expression(Lan); 697 | *addr = (int)(text + 1); 698 | expr_type = INT; 699 | } 700 | else if (token == Lan) { 701 | // logic and 702 | match(Lan); 703 | *++text = JZ; 704 | addr = ++text; 705 | expression(Or); 706 | *addr = (int)(text + 1); 707 | expr_type = INT; 708 | } 709 | else if (token == Or) { 710 | // bitwise or 711 | match(Or); 712 | *++text = PUSH; 713 | expression(Xor); 714 | *++text = OR; 715 | expr_type = INT; 716 | } 717 | else if (token == Xor) { 718 | // bitwise xor 719 | match(Xor); 720 | *++text = PUSH; 721 | expression(And); 722 | *++text = XOR; 723 | expr_type = INT; 724 | } 725 | else if (token == And) { 726 | // bitwise and 727 | match(And); 728 | *++text = PUSH; 729 | expression(Eq); 730 | *++text = AND; 731 | expr_type = INT; 732 | } 733 | else if (token == Eq) { 734 | // equal == 735 | match(Eq); 736 | *++text = PUSH; 737 | expression(Ne); 738 | *++text = EQ; 739 | expr_type = INT; 740 | } 741 | else if (token == Ne) { 742 | // not equal != 743 | match(Ne); 744 | *++text = PUSH; 745 | expression(Lt); 746 | *++text = NE; 747 | expr_type = INT; 748 | } 749 | else if (token == Lt) { 750 | // less than 751 | match(Lt); 752 | *++text = PUSH; 753 | expression(Shl); 754 | *++text = LT; 755 | expr_type = INT; 756 | } 757 | else if (token == Gt) { 758 | // greater than 759 | match(Gt); 760 | *++text = PUSH; 761 | expression(Shl); 762 | *++text = GT; 763 | expr_type = INT; 764 | } 765 | else if (token == Le) { 766 | // less than or equal to 767 | match(Le); 768 | *++text = PUSH; 769 | expression(Shl); 770 | *++text = LE; 771 | expr_type = INT; 772 | } 773 | else if (token == Ge) { 774 | // greater than or equal to 775 | match(Ge); 776 | *++text = PUSH; 777 | expression(Shl); 778 | *++text = GE; 779 | expr_type = INT; 780 | } 781 | else if (token == Shl) { 782 | // shift left 783 | match(Shl); 784 | *++text = PUSH; 785 | expression(Add); 786 | *++text = SHL; 787 | expr_type = INT; 788 | } 789 | else if (token == Shr) { 790 | // shift right 791 | match(Shr); 792 | *++text = PUSH; 793 | expression(Add); 794 | *++text = SHR; 795 | expr_type = INT; 796 | } 797 | else if (token == Add) { 798 | // add 799 | match(Add); 800 | *++text = PUSH; 801 | expression(Mul); 802 | 803 | expr_type = tmp; 804 | if (expr_type > PTR) { 805 | // pointer type, and not `char *` 806 | *++text = PUSH; 807 | *++text = IMM; 808 | *++text = sizeof(int); 809 | *++text = MUL; 810 | } 811 | *++text = ADD; 812 | } 813 | else if (token == Sub) { 814 | // sub 815 | match(Sub); 816 | *++text = PUSH; 817 | expression(Mul); 818 | if (tmp > PTR && tmp == expr_type) { 819 | // pointer subtraction 820 | *++text = SUB; 821 | *++text = PUSH; 822 | *++text = IMM; 823 | *++text = sizeof(int); 824 | *++text = DIV; 825 | expr_type = INT; 826 | } else if (tmp > PTR) { 827 | // pointer movement 828 | *++text = PUSH; 829 | *++text = IMM; 830 | *++text = sizeof(int); 831 | *++text = MUL; 832 | *++text = SUB; 833 | expr_type = tmp; 834 | } else { 835 | // numeral subtraction 836 | *++text = SUB; 837 | expr_type = tmp; 838 | } 839 | } 840 | else if (token == Mul) { 841 | // multiply 842 | match(Mul); 843 | *++text = PUSH; 844 | expression(Inc); 845 | *++text = MUL; 846 | expr_type = tmp; 847 | } 848 | else if (token == Div) { 849 | // divide 850 | match(Div); 851 | *++text = PUSH; 852 | expression(Inc); 853 | *++text = DIV; 854 | expr_type = tmp; 855 | } 856 | else if (token == Mod) { 857 | // Modulo 858 | match(Mod); 859 | *++text = PUSH; 860 | expression(Inc); 861 | *++text = MOD; 862 | expr_type = tmp; 863 | } 864 | else if (token == Inc || token == Dec) { 865 | // postfix inc(++) and dec(--) 866 | // we will increase the value to the variable and decrease it 867 | // on `ax` to get its original value. 868 | if (*text == LI) { 869 | *text = PUSH; 870 | *++text = LI; 871 | } 872 | else if (*text == LC) { 873 | *text = PUSH; 874 | *++text = LC; 875 | } 876 | else { 877 | printf("%d: bad value in increment\n", line); 878 | exit(-1); 879 | } 880 | 881 | *++text = PUSH; 882 | *++text = IMM; 883 | *++text = (expr_type > PTR) ? sizeof(int) : sizeof(char); 884 | *++text = (token == Inc) ? ADD : SUB; 885 | *++text = (expr_type == CHAR) ? SC : SI; 886 | *++text = PUSH; 887 | *++text = IMM; 888 | *++text = (expr_type > PTR) ? sizeof(int) : sizeof(char); 889 | *++text = (token == Inc) ? SUB : ADD; 890 | match(token); 891 | } 892 | else if (token == Brak) { 893 | // array access var[xx] 894 | match(Brak); 895 | *++text = PUSH; 896 | expression(Assign); 897 | match(']'); 898 | 899 | if (tmp > PTR) { 900 | // pointer, `not char *` 901 | *++text = PUSH; 902 | *++text = IMM; 903 | *++text = sizeof(int); 904 | *++text = MUL; 905 | } 906 | else if (tmp < PTR) { 907 | printf("%d: pointer type expected\n", line); 908 | exit(-1); 909 | } 910 | expr_type = tmp - PTR; 911 | *++text = ADD; 912 | *++text = (expr_type == CHAR) ? LC : LI; 913 | } 914 | else { 915 | printf("%d: compiler error, token = %d\n", line, token); 916 | exit(-1); 917 | } 918 | } 919 | } 920 | } 921 | 922 | void statement() { 923 | // there are 6 kinds of statements here: 924 | // 1. if (...) [else ] 925 | // 2. while (...) 926 | // 3. { } 927 | // 4. return xxx; 928 | // 5. ; 929 | // 6. expression; (expression end with semicolon) 930 | 931 | int *a, *b; // bess for branch control 932 | 933 | if (token == If) { 934 | // if (...) [else ] 935 | // 936 | // if (...) 937 | // JZ a 938 | // 939 | // else: JMP b 940 | // a: a: 941 | // 942 | // b: b: 943 | // 944 | // 945 | match(If); 946 | match('('); 947 | expression(Assign); // parse condition 948 | match(')'); 949 | 950 | // emit code for if 951 | *++text = JZ; 952 | b = ++text; 953 | 954 | statement(); // parse statement 955 | if (token == Else) { // parse else 956 | match(Else); 957 | 958 | // emit code for JMP B 959 | *b = (int)(text + 3); 960 | *++text = JMP; 961 | b = ++text; 962 | 963 | statement(); 964 | } 965 | 966 | *b = (int)(text + 1); 967 | } 968 | else if (token == While) { 969 | // 970 | // a: a: 971 | // while () 972 | // JZ b 973 | // 974 | // JMP a 975 | // b: b: 976 | match(While); 977 | 978 | a = text + 1; 979 | 980 | match('('); 981 | expression(Assign); 982 | match(')'); 983 | 984 | *++text = JZ; 985 | b = ++text; 986 | 987 | statement(); 988 | 989 | *++text = JMP; 990 | *++text = (int)a; 991 | *b = (int)(text + 1); 992 | } 993 | else if (token == '{') { 994 | // { ... } 995 | match('{'); 996 | 997 | while (token != '}') { 998 | statement(); 999 | } 1000 | 1001 | match('}'); 1002 | } 1003 | else if (token == Return) { 1004 | // return [expression]; 1005 | match(Return); 1006 | 1007 | if (token != ';') { 1008 | expression(Assign); 1009 | } 1010 | 1011 | match(';'); 1012 | 1013 | // emit code for return 1014 | *++text = LEV; 1015 | } 1016 | else if (token == ';') { 1017 | // empty statement 1018 | match(';'); 1019 | } 1020 | else { 1021 | // a = b; or function_call(); 1022 | expression(Assign); 1023 | match(';'); 1024 | } 1025 | } 1026 | 1027 | void function_parameter() { 1028 | int type; 1029 | int params; 1030 | params = 0; 1031 | while (token != ')') { 1032 | // int name, ... 1033 | type = INT; 1034 | if (token == Int) { 1035 | match(Int); 1036 | } else if (token == Char) { 1037 | type = CHAR; 1038 | match(Char); 1039 | } 1040 | 1041 | // pointer type 1042 | while (token == Mul) { 1043 | match(Mul); 1044 | type = type + PTR; 1045 | } 1046 | 1047 | // parameter name 1048 | if (token != Id) { 1049 | printf("%d: bad parameter declaration\n", line); 1050 | exit(-1); 1051 | } 1052 | if (current_id[Class] == Loc) { 1053 | printf("%d: duplicate parameter declaration\n", line); 1054 | exit(-1); 1055 | } 1056 | 1057 | match(Id); 1058 | // store the local variable 1059 | current_id[BClass] = current_id[Class]; current_id[Class] = Loc; 1060 | current_id[BType] = current_id[Type]; current_id[Type] = type; 1061 | current_id[BValue] = current_id[Value]; current_id[Value] = params++; // index of current parameter 1062 | 1063 | if (token == ',') { 1064 | match(','); 1065 | } 1066 | } 1067 | index_of_bp = params+1; 1068 | } 1069 | 1070 | void function_body() { 1071 | // type func_name (...) {...} 1072 | // -->| |<-- 1073 | 1074 | // ... { 1075 | // 1. local declarations 1076 | // 2. statements 1077 | // } 1078 | 1079 | int pos_local; // position of local variables on the stack. 1080 | int type; 1081 | pos_local = index_of_bp; 1082 | 1083 | while (token == Int || token == Char) { 1084 | // local variable declaration, just like global ones. 1085 | basetype = (token == Int) ? INT : CHAR; 1086 | match(token); 1087 | 1088 | while (token != ';') { 1089 | type = basetype; 1090 | while (token == Mul) { 1091 | match(Mul); 1092 | type = type + PTR; 1093 | } 1094 | 1095 | if (token != Id) { 1096 | // invalid declaration 1097 | printf("%d: bad local declaration\n", line); 1098 | exit(-1); 1099 | } 1100 | if (current_id[Class] == Loc) { 1101 | // identifier exists 1102 | printf("%d: duplicate local declaration\n", line); 1103 | exit(-1); 1104 | } 1105 | match(Id); 1106 | 1107 | // store the local variable 1108 | current_id[BClass] = current_id[Class]; current_id[Class] = Loc; 1109 | current_id[BType] = current_id[Type]; current_id[Type] = type; 1110 | current_id[BValue] = current_id[Value]; current_id[Value] = ++pos_local; // index of current parameter 1111 | 1112 | if (token == ',') { 1113 | match(','); 1114 | } 1115 | } 1116 | match(';'); 1117 | } 1118 | 1119 | // save the stack size for local variables 1120 | *++text = ENT; 1121 | *++text = pos_local - index_of_bp; 1122 | 1123 | // statements 1124 | while (token != '}') { 1125 | statement(); 1126 | } 1127 | 1128 | // emit code for leaving the sub function 1129 | *++text = LEV; 1130 | } 1131 | 1132 | void function_declaration() { 1133 | // type func_name (...) {...} 1134 | // | this part 1135 | 1136 | match('('); 1137 | function_parameter(); 1138 | match(')'); 1139 | match('{'); 1140 | function_body(); 1141 | //match('}'); 1142 | 1143 | // unwind local variable declarations for all local variables. 1144 | current_id = symbols; 1145 | while (current_id[Token]) { 1146 | if (current_id[Class] == Loc) { 1147 | current_id[Class] = current_id[BClass]; 1148 | current_id[Type] = current_id[BType]; 1149 | current_id[Value] = current_id[BValue]; 1150 | } 1151 | current_id = current_id + IdSize; 1152 | } 1153 | } 1154 | 1155 | void enum_declaration() { 1156 | // parse enum [id] { a = 1, b = 3, ...} 1157 | int i; 1158 | i = 0; 1159 | while (token != '}') { 1160 | if (token != Id) { 1161 | printf("%d: bad enum identifier %d\n", line, token); 1162 | exit(-1); 1163 | } 1164 | next(); 1165 | if (token == Assign) { 1166 | // like {a=10} 1167 | next(); 1168 | if (token != Num) { 1169 | printf("%d: bad enum initializer\n", line); 1170 | exit(-1); 1171 | } 1172 | i = token_val; 1173 | next(); 1174 | } 1175 | 1176 | current_id[Class] = Num; 1177 | current_id[Type] = INT; 1178 | current_id[Value] = i++; 1179 | 1180 | if (token == ',') { 1181 | next(); 1182 | } 1183 | } 1184 | } 1185 | 1186 | void global_declaration() { 1187 | // global_declaration ::= enum_decl | variable_decl | function_decl 1188 | // 1189 | // enum_decl ::= 'enum' [id] '{' id ['=' 'num'] {',' id ['=' 'num'} '}' 1190 | // 1191 | // variable_decl ::= type {'*'} id { ',' {'*'} id } ';' 1192 | // 1193 | // function_decl ::= type {'*'} id '(' parameter_decl ')' '{' body_decl '}' 1194 | 1195 | 1196 | int type; // tmp, actual type for variable 1197 | int i; // tmp 1198 | 1199 | basetype = INT; 1200 | 1201 | // parse enum, this should be treated alone. 1202 | if (token == Enum) { 1203 | // enum [id] { a = 10, b = 20, ... } 1204 | match(Enum); 1205 | if (token != '{') { 1206 | match(Id); // skip the [id] part 1207 | } 1208 | if (token == '{') { 1209 | // parse the assign part 1210 | match('{'); 1211 | enum_declaration(); 1212 | match('}'); 1213 | } 1214 | 1215 | match(';'); 1216 | return; 1217 | } 1218 | 1219 | // parse type information 1220 | if (token == Int) { 1221 | match(Int); 1222 | } 1223 | else if (token == Char) { 1224 | match(Char); 1225 | basetype = CHAR; 1226 | } 1227 | 1228 | // parse the comma seperated variable declaration. 1229 | while (token != ';' && token != '}') { 1230 | type = basetype; 1231 | // parse pointer type, note that there may exist `int ****x;` 1232 | while (token == Mul) { 1233 | match(Mul); 1234 | type = type + PTR; 1235 | } 1236 | 1237 | if (token != Id) { 1238 | // invalid declaration 1239 | printf("%d: bad global declaration\n", line); 1240 | exit(-1); 1241 | } 1242 | if (current_id[Class]) { 1243 | // identifier exists 1244 | printf("%d: duplicate global declaration\n", line); 1245 | exit(-1); 1246 | } 1247 | match(Id); 1248 | current_id[Type] = type; 1249 | 1250 | if (token == '(') { 1251 | current_id[Class] = Fun; 1252 | current_id[Value] = (int)(text + 1); // the memory address of function 1253 | function_declaration(); 1254 | } else { 1255 | // variable declaration 1256 | current_id[Class] = Glo; // global variable 1257 | current_id[Value] = (int)data; // assign memory address 1258 | data = data + sizeof(int); 1259 | // *data = 0; 1260 | } 1261 | 1262 | if (token == ',') { 1263 | match(','); 1264 | } 1265 | } 1266 | next(); 1267 | } 1268 | 1269 | void program() { 1270 | // get next token 1271 | next(); 1272 | while (token > 0) { 1273 | global_declaration(); 1274 | } 1275 | } 1276 | 1277 | void show_regs() 1278 | { 1279 | printf("ax: %#x(%d)\n", ax, ax); 1280 | printf("sp: %p\n", sp); 1281 | printf("bp: %p\n", bp); 1282 | printf("pc: %p\n", pc); 1283 | } 1284 | 1285 | int is_data(unsigned int addr) 1286 | { 1287 | if (((unsigned int)begin_data <= addr) && (addr < (unsigned int)data)) 1288 | return 1; 1289 | else 1290 | return 0; 1291 | } 1292 | 1293 | int is_text(unsigned int addr) 1294 | { 1295 | if (((unsigned int)begin_text <= addr) && (addr < (unsigned int)text)) 1296 | return 1; 1297 | else 1298 | return 0; 1299 | } 1300 | 1301 | int is_stack(unsigned int addr) 1302 | { 1303 | if (((unsigned int)begin_stack <= addr) && (addr < (unsigned int)begin_stack + poolsize)) 1304 | return 1; 1305 | else 1306 | return 0; 1307 | } 1308 | 1309 | int run_debug_func(char *cmd_line) 1310 | { 1311 | switch (cmd_line[0]) 1312 | { 1313 | case 'q': 1314 | { 1315 | exit(1); 1316 | } 1317 | case 'd': 1318 | { 1319 | print_data(); 1320 | break; 1321 | } 1322 | case 't': 1323 | { 1324 | print_symbol_table(); 1325 | break; 1326 | } 1327 | case 'h': 1328 | case '?': 1329 | { 1330 | debug_usage(); 1331 | break; 1332 | } 1333 | case 'l': 1334 | { 1335 | print_source_code(); 1336 | break; 1337 | } 1338 | case 'e': 1339 | { 1340 | print_text(); 1341 | break; 1342 | } 1343 | case 'b': // break pointer, ex: b 0xf756301c 1344 | { 1345 | unsigned int addr; 1346 | char cmd; 1347 | 1348 | sscanf(cmd_line, "%c %x\n", &cmd, &addr); 1349 | if (last_bp < MAX_BREAK_POINT) 1350 | { 1351 | break_points[last_bp] = addr; 1352 | printf("set break_points[%d]: %x\n", last_bp, addr); 1353 | ++last_bp; 1354 | } 1355 | else 1356 | { 1357 | printf("exceed %d break points\n", MAX_BREAK_POINT); 1358 | } 1359 | break; 1360 | } 1361 | case 'x': // show data segment content, ex: x 0xf756301c 1362 | { 1363 | char cmd_str[5]; 1364 | 1365 | switch (cmd_line[1]) 1366 | { 1367 | case 's': 1368 | { 1369 | break; 1370 | } 1371 | case 'c': 1372 | { 1373 | break; 1374 | } 1375 | case 'i': 1376 | { 1377 | break; 1378 | } 1379 | case 'x': 1380 | { 1381 | break; 1382 | } 1383 | default: 1384 | { 1385 | printf("xxx\n"); 1386 | return -1; 1387 | } 1388 | 1389 | } 1390 | unsigned int addr; 1391 | 1392 | sscanf(cmd_line, "%s %x\n", cmd_str, &addr); 1393 | printf("cmd: %s, addr: %#x\n", cmd_str, addr); 1394 | #if 1 1395 | is_text(addr); 1396 | is_stack(addr); 1397 | //if (((unsigned int)begin_data <= addr) && (addr < (unsigned int)data)) 1398 | if (is_data(addr) ) 1399 | { 1400 | if (cmd_line[1] == 's') 1401 | printf("data seg: %s\n", (char *)addr); 1402 | if (cmd_line[1] == 'i') 1403 | printf("data seg: %#d\n", *(int *)addr); 1404 | if (cmd_line[1] == 'x') 1405 | printf("data seg: %#x\n", *(int *)addr); 1406 | } 1407 | else if (is_text(addr) ) 1408 | printf("text seg: %#x(%d)\n", *((int *)addr), *((int *)addr)); 1409 | else if (is_stack(addr) ) 1410 | printf("stack area: %#x(%d)\n", *((int *)addr), *((int *)addr)); 1411 | else 1412 | { 1413 | printf("%x is not in \ntext segment (%p ~ %p)\ndata segment (%p ~ %p)\nstack range (%p ~ %p)\n", addr, begin_text, text, begin_data, data, begin_stack, begin_stack + poolsize); 1414 | 1415 | } 1416 | 1417 | //printf("%#x(%d)\n", *((int *)addr)); 1418 | #endif 1419 | break; 1420 | } 1421 | case 'r': // show registers 1422 | { 1423 | show_regs(); 1424 | break; 1425 | } 1426 | default: 1427 | { 1428 | break; 1429 | } 1430 | 1431 | } // end switch (input_str[0]) 1432 | return 0; 1433 | } 1434 | 1435 | void debug_usage() 1436 | { 1437 | printf("command\n"); 1438 | printf("s: step\n"); 1439 | printf("q: quit\n"); 1440 | printf("c: continue\n"); 1441 | printf("r: print all register content\n"); 1442 | printf("d: print data\n"); 1443 | printf("e: print text\n"); 1444 | printf("l: print source code\n"); 1445 | printf("t: print symbol table\n"); 1446 | printf("xs address: print text/data segment stack area content as string\n"); 1447 | printf("xc address: not yet complete. print text/data segment stack area content as char\n"); 1448 | printf("xi address: print text/data segment stack area content as int\n"); 1449 | printf("xx address: print text/data segment stack area content as hex\n"); 1450 | printf("b address: set breakpoint, max breakpoint is %d\n", MAX_BREAK_POINT); 1451 | } 1452 | 1453 | #define INPUT_SIZE 20 1454 | int eval() { 1455 | int line=0, ch; 1456 | unsigned int break_point = 0; 1457 | int continue_run = 0; 1458 | int meet_bp = 0; 1459 | char input_str[INPUT_SIZE+1]; 1460 | int op, *tmp; 1461 | while (1) 1462 | { 1463 | #ifdef SUPPORT_DEBUG 1464 | print_spec_text(pc); 1465 | //if (break_point == (unsigned int)pc || continue_run == 0) 1466 | 1467 | debug_input: 1468 | while(continue_run == 0) 1469 | { 1470 | printf("%d ## debug> ", line++); 1471 | fgets(input_str, INPUT_SIZE, stdin); 1472 | 1473 | if (input_str[0] == 's' || input_str[0] == '\n') 1474 | { 1475 | continue_run = 0; 1476 | break; 1477 | } 1478 | 1479 | if (input_str[0] == 'c') 1480 | { 1481 | continue_run = 1; 1482 | break; 1483 | } 1484 | else 1485 | { 1486 | run_debug_func(input_str); 1487 | continue_run = 0; 1488 | } 1489 | } 1490 | 1491 | if (meet_bp == 0 && meet_break_point(pc)) 1492 | { 1493 | meet_bp = 1; 1494 | // printf("meet break pointer: %x\n", (unsigned int)pc); 1495 | continue_run = 0; 1496 | goto debug_input; 1497 | } 1498 | else 1499 | { 1500 | meet_bp = 0; 1501 | } 1502 | 1503 | #endif 1504 | op = *pc++; // get next operation code 1505 | 1506 | if (op == IMM) {ax = *pc++;} // load immediate value to ax 1507 | else if (op == LC) {ax = *(char *)ax;} // load character to ax, address in ax 1508 | else if (op == LI) {ax = *(int *)ax;} // load integer to ax, address in ax 1509 | else if (op == SC) {ax = *(char *)*sp++ = ax;} // save character to address, value in ax, address on stack 1510 | else if (op == SI) {*(int *)*sp++ = ax;} // save integer to address, value in ax, address on stack 1511 | else if (op == PUSH) {*--sp = ax;} // push the value of ax onto the stack 1512 | else if (op == JMP) {pc = (int *)*pc;} // jump to the address 1513 | else if (op == JZ) {pc = ax ? pc + 1 : (int *)*pc;} // jump if ax is zero 1514 | else if (op == JNZ) {pc = ax ? (int *)*pc : pc + 1;} // jump if ax is zero 1515 | else if (op == CALL) {*--sp = (int)(pc+1); pc = (int *)*pc;} // call subroutine 1516 | //else if (op == RET) {pc = (int *)*sp++;} // return from subroutine; 1517 | else if (op == ENT) {*--sp = (int)bp; bp = sp; sp = sp - *pc++;} // make new stack frame 1518 | else if (op == ADJ) {sp = sp + *pc++;} // add esp, 1519 | else if (op == LEV) {sp = bp; bp = (int *)*sp++; pc = (int *)*sp++;} // restore call frame and PC 1520 | else if (op == ENT) {*--sp = (int)bp; bp = sp; sp = sp - *pc++;} // make new stack frame 1521 | else if (op == ADJ) {sp = sp + *pc++;} // add esp, 1522 | else if (op == LEV) {sp = bp; bp = (int *)*sp++; pc = (int *)*sp++;} // restore call frame and PC 1523 | else if (op == LEA) {ax = (int)(bp + *pc++);} // load address for arguments. 1524 | 1525 | else if (op == OR) ax = *sp++ | ax; 1526 | else if (op == XOR) ax = *sp++ ^ ax; 1527 | else if (op == AND) ax = *sp++ & ax; 1528 | else if (op == EQ) ax = *sp++ == ax; 1529 | else if (op == NE) ax = *sp++ != ax; 1530 | else if (op == LT) ax = *sp++ < ax; 1531 | else if (op == LE) ax = *sp++ <= ax; 1532 | else if (op == GT) ax = *sp++ > ax; 1533 | else if (op == GE) ax = *sp++ >= ax; 1534 | else if (op == SHL) ax = *sp++ << ax; 1535 | else if (op == SHR) ax = *sp++ >> ax; 1536 | else if (op == ADD) ax = *sp++ + ax; 1537 | else if (op == SUB) ax = *sp++ - ax; 1538 | else if (op == MUL) ax = *sp++ * ax; 1539 | else if (op == DIV) ax = *sp++ / ax; 1540 | else if (op == MOD) ax = *sp++ % ax; 1541 | 1542 | 1543 | else if (op == EXIT) { printf("exit(%d)", *sp); return *sp;} 1544 | else if (op == OPEN) { ax = open((char *)sp[1], sp[0]); } 1545 | else if (op == CLOS) { ax = close(*sp);} 1546 | else if (op == READ) { ax = read(sp[2], (char *)sp[1], *sp); } 1547 | else if (op == PRTF) { tmp = sp + pc[1]; ax = printf((char *)tmp[-1], tmp[-2], tmp[-3], tmp[-4], tmp[-5], tmp[-6]); } 1548 | else if (op == MALC) { ax = (int)malloc(*sp);} 1549 | else if (op == MSET) { ax = (int)memset((char *)sp[2], sp[1], *sp);} 1550 | else if (op == MCMP) { ax = memcmp((char *)sp[2], (char *)sp[1], *sp);} 1551 | else { 1552 | printf("unknown instruction:%d\n", op); 1553 | return -1; 1554 | } 1555 | } 1556 | return 0; 1557 | } 1558 | 1559 | #ifdef SUPPORT_DEBUG 1560 | int meet_break_point(int *pc) 1561 | { 1562 | int i; 1563 | unsigned int cur_addr = (unsigned int)pc; 1564 | //printf("cur_addr: %x\n", cur_addr); 1565 | 1566 | for (i = 0 ; i < last_bp ; ++i) 1567 | { 1568 | //printf("cur break_points[%d]: %x\n", i, break_points[i]); 1569 | if (cur_addr == break_points[i]) 1570 | { 1571 | printf("break_points[%d]: %x\n", i, break_points[i]); 1572 | return 1; 1573 | } 1574 | } 1575 | return 0; 1576 | } 1577 | 1578 | #ifdef SUPPORT_DEBUG 1579 | void print_symbol_table() 1580 | { 1581 | int *cur_id; 1582 | cur_id = symbols; 1583 | 1584 | printf("symbol table:\n"); 1585 | 1586 | while(cur_id[Token]) 1587 | { 1588 | printf("cur_id[Name]: %s\n", cur_id[Name]); 1589 | printf("cur_id[Hash]: %#x\n", cur_id[Hash]); 1590 | printf("cur_id[Type]: %s (%d)\n", type_str(cur_id[Type]), cur_id[Type]); 1591 | printf("cur_id[Class]: %s (%d)\n", class_str(cur_id[Class]), cur_id[Class]); 1592 | printf("cur_id[Value]: %#x (%d)\n", cur_id[Value], cur_id[Value]); 1593 | cur_id = cur_id + IdSize; 1594 | } 1595 | } 1596 | #endif 1597 | 1598 | // return 1: has argument 1599 | // return 0: has no argument 1600 | int print_spec_text(int *cur_text) 1601 | { 1602 | int has_argu=0; 1603 | 1604 | const char* inst_str = inst_2_str(*cur_text); 1605 | has_argu = inst_has_argu(*cur_text); 1606 | 1607 | if (inst_str) 1608 | printf("addr %p ## %s", cur_text, inst_str); 1609 | else 1610 | printf("addr %p ## %x", cur_text, *cur_text); 1611 | 1612 | if (has_argu) 1613 | { 1614 | ++cur_text; 1615 | printf(" %#x(%d)", *cur_text, *cur_text); 1616 | } 1617 | printf("\n"); 1618 | return has_argu; 1619 | } 1620 | 1621 | void print_source_code() 1622 | { 1623 | printf("%s\n", src_begin); 1624 | } 1625 | 1626 | void print_text() 1627 | { 1628 | printf("text segment:\n"); 1629 | 1630 | int *cur_text = begin_text+1; 1631 | int i; 1632 | int has_argu=0; 1633 | while (cur_text != text) 1634 | { 1635 | has_argu = print_spec_text(cur_text); 1636 | if (has_argu) 1637 | cur_text += 2; 1638 | else 1639 | ++cur_text; 1640 | #if 0 1641 | const char* inst_str = inst_2_str(*cur_text); 1642 | has_argu = inst_has_argu(*cur_text); 1643 | if (*cur_text == -1) 1644 | { 1645 | ++cur_text; 1646 | continue; 1647 | } 1648 | if (inst_str) 1649 | printf("addr %p ## %s\n", cur_text, inst_str); 1650 | else 1651 | printf("addr %p ## %x\n", cur_text, *cur_text); 1652 | 1653 | if (has_argu) 1654 | { 1655 | ++cur_text; 1656 | printf("addr %p ## %#x (%d)\n", cur_text, *cur_text, *cur_text); 1657 | has_argu = 0; 1658 | } 1659 | #endif 1660 | } 1661 | 1662 | } 1663 | 1664 | void print_data() 1665 | { 1666 | signed char *cur_data = begin_data; 1667 | int print_addr = 1; 1668 | 1669 | printf("data segment:\n"); 1670 | 1671 | while(cur_data != data) 1672 | { 1673 | if (*cur_data != 0) 1674 | { 1675 | if (print_addr) 1676 | { 1677 | printf("%p: ", cur_data); 1678 | print_addr = 0; 1679 | } 1680 | printf("%c", *cur_data); 1681 | } 1682 | else 1683 | { 1684 | printf("\n"); 1685 | print_addr = 1; 1686 | } 1687 | ++cur_data; 1688 | } 1689 | } 1690 | #endif 1691 | 1692 | int main(int argc, char **argv) 1693 | { 1694 | 1695 | int i, fd; 1696 | int *tmp; 1697 | 1698 | argc--; 1699 | argv++; 1700 | 1701 | poolsize = 256 * 1024; // arbitrary size 1702 | line = 1; 1703 | 1704 | if ((fd = open(*argv, 0)) < 0) { 1705 | printf("could not open(%s)\n", *argv); 1706 | return -1; 1707 | } 1708 | 1709 | // allocate memory for virtual machine 1710 | if (!(text = old_text = malloc(poolsize))) { 1711 | printf("could not malloc(%d) for text area\n", poolsize); 1712 | return -1; 1713 | } 1714 | begin_text = text; 1715 | 1716 | if (!(data = malloc(poolsize))) { 1717 | printf("could not malloc(%d) for data area\n", poolsize); 1718 | return -1; 1719 | } 1720 | 1721 | begin_data = data; 1722 | 1723 | if (!(stack = malloc(poolsize))) { 1724 | printf("could not malloc(%d) for stack area\n", poolsize); 1725 | return -1; 1726 | } 1727 | 1728 | begin_stack = stack; 1729 | 1730 | if (!(symbols = malloc(poolsize))) { 1731 | printf("could not malloc(%d) for symbol table\n", poolsize); 1732 | return -1; 1733 | } 1734 | 1735 | #ifdef SUPPORT_DEBUG 1736 | printf("text: %p\n", text); 1737 | printf("data: %p\n", data); 1738 | printf("stack: %p\n", stack); 1739 | #endif 1740 | 1741 | memset(text, 0, poolsize); 1742 | memset(data, 0, poolsize); 1743 | memset(stack, 0, poolsize); 1744 | memset(symbols, 0, poolsize); 1745 | bp = sp = (int *)((int)stack + poolsize); 1746 | ax = 0; 1747 | 1748 | src = "char else enum if int return sizeof while " 1749 | "open read close printf malloc memset memcmp exit void main"; 1750 | 1751 | // add keywords to symbol table 1752 | i = Char; 1753 | while (i <= While) { 1754 | next(); 1755 | current_id[Token] = i++; 1756 | } 1757 | 1758 | // add library to symbol table 1759 | i = OPEN; 1760 | while (i <= EXIT) { 1761 | next(); 1762 | current_id[Class] = Sys; 1763 | current_id[Type] = INT; 1764 | current_id[Value] = i++; 1765 | } 1766 | 1767 | next(); current_id[Token] = Char; // handle void type 1768 | next(); idmain = current_id; // keep track of main 1769 | 1770 | 1771 | // read the source file 1772 | if ((fd = open(*argv, 0)) < 0) { 1773 | printf("could not open(%s)\n", *argv); 1774 | return -1; 1775 | } 1776 | 1777 | if (!(src = old_src = malloc(poolsize))) { 1778 | printf("could not malloc(%d) for source area\n", poolsize); 1779 | return -1; 1780 | } 1781 | // read the source file 1782 | if ((i = read(fd, src, poolsize-1)) <= 0) { 1783 | printf("read() returned %d\n", i); 1784 | return -1; 1785 | } 1786 | src[i] = 0; // add EOF character 1787 | src_begin = src; 1788 | close(fd); 1789 | 1790 | program(); 1791 | 1792 | if (!(pc = (int *)idmain[Value])) { 1793 | printf("main() not defined\n"); 1794 | return -1; 1795 | } 1796 | 1797 | // setup stack 1798 | sp = (int *)((int)stack + poolsize); 1799 | 1800 | #ifdef SUPPORT_DEBUG 1801 | printf("sp: %x\n", sp); 1802 | #endif 1803 | 1804 | *--sp = EXIT; // call exit if main returns 1805 | *--sp = PUSH; tmp = sp; 1806 | *--sp = argc; 1807 | *--sp = (int)argv; 1808 | *--sp = (int)tmp; 1809 | 1810 | #ifdef SUPPORT_DEBUG 1811 | //print_text(); 1812 | //print_data(); 1813 | // print_symbol_table(); 1814 | #endif 1815 | return eval(); 1816 | } 1817 | -------------------------------------------------------------------------------- /xc.c: -------------------------------------------------------------------------------- 1 | #include 2 | #include 3 | #include 4 | #include 5 | 6 | int debug; // print the executed instructions 7 | int assembly; // print out the assembly and source 8 | 9 | int token; // current token 10 | 11 | // instructions 12 | enum { LEA ,IMM ,JMP ,CALL,JZ ,JNZ ,ENT ,ADJ ,LEV ,LI ,LC ,SI ,SC ,PUSH, 13 | OR ,XOR ,AND ,EQ ,NE ,LT ,GT ,LE ,GE ,SHL ,SHR ,ADD ,SUB ,MUL ,DIV ,MOD , 14 | OPEN,READ,CLOS,PRTF,MALC,MSET,MCMP,EXIT }; 15 | 16 | // tokens and classes (operators last and in precedence order) 17 | // copied from c4 18 | enum { 19 | Num = 128, Fun, Sys, Glo, Loc, Id, 20 | Char, Else, Enum, If, Int, Return, Sizeof, While, 21 | Assign, Cond, Lor, Lan, Or, Xor, And, Eq, Ne, Lt, Gt, Le, Ge, Shl, Shr, Add, Sub, Mul, Div, Mod, Inc, Dec, Brak 22 | }; 23 | 24 | // fields of identifier 25 | enum {Token, Hash, Name, Type, Class, Value, BType, BClass, BValue, IdSize}; 26 | 27 | 28 | // types of variable/function 29 | enum { CHAR, INT, PTR }; 30 | 31 | // type of declaration. 32 | enum {Global, Local}; 33 | 34 | int *text, // text segment 35 | *stack;// stack 36 | int * old_text; // for dump text segment 37 | char *data; // data segment 38 | int *idmain; 39 | 40 | char *src, *old_src; // pointer to source code string; 41 | 42 | int poolsize; // default size of text/data/stack 43 | int *pc, *bp, *sp, ax, cycle; // virtual machine registers 44 | 45 | int *current_id, // current parsed ID 46 | *symbols, // symbol table 47 | line, // line number of source code 48 | token_val; // value of current token (mainly for number) 49 | 50 | int basetype; // the type of a declaration, make it global for convenience 51 | int expr_type; // the type of an expression 52 | 53 | // function frame 54 | // 55 | // 0: arg 1 56 | // 1: arg 2 57 | // 2: arg 3 58 | // 3: return address 59 | // 4: old bp pointer <- index_of_bp 60 | // 5: local var 1 61 | // 6: local var 2 62 | int index_of_bp; // index of bp pointer on stack 63 | 64 | void next() { 65 | char *last_pos; 66 | int hash; 67 | 68 | while (token = *src) { 69 | ++src; 70 | 71 | if (token == '\n') { 72 | if (assembly) { 73 | // print compile info 74 | printf("%d: %.*s", line, src-old_src, old_src); 75 | old_src = src; 76 | 77 | while (old_text < text) { 78 | printf("%8.4s", & "LEA ,IMM ,JMP ,CALL,JZ ,JNZ ,ENT ,ADJ ,LEV ,LI ,LC ,SI ,SC ,PUSH," 79 | "OR ,XOR ,AND ,EQ ,NE ,LT ,GT ,LE ,GE ,SHL ,SHR ,ADD ,SUB ,MUL ,DIV ,MOD ," 80 | "OPEN,READ,CLOS,PRTF,MALC,MSET,MCMP,EXIT"[*++old_text * 5]); 81 | 82 | if (*old_text <= ADJ) 83 | printf(" %d\n", *++old_text); 84 | else 85 | printf("\n"); 86 | } 87 | } 88 | ++line; 89 | } 90 | else if (token == '#') { 91 | // skip macro, because we will not support it 92 | while (*src != 0 && *src != '\n') { 93 | src++; 94 | } 95 | } 96 | else if ((token >= 'a' && token <= 'z') || (token >= 'A' && token <= 'Z') || (token == '_')) { 97 | 98 | // parse identifier 99 | last_pos = src - 1; 100 | hash = token; 101 | 102 | while ((*src >= 'a' && *src <= 'z') || (*src >= 'A' && *src <= 'Z') || (*src >= '0' && *src <= '9') || (*src == '_')) { 103 | hash = hash * 147 + *src; 104 | src++; 105 | } 106 | 107 | // look for existing identifier, linear search 108 | current_id = symbols; 109 | while (current_id[Token]) { 110 | if (current_id[Hash] == hash && !memcmp((char *)current_id[Name], last_pos, src - last_pos)) { 111 | //found one, return 112 | token = current_id[Token]; 113 | return; 114 | } 115 | current_id = current_id + IdSize; 116 | } 117 | 118 | 119 | // store new ID 120 | current_id[Name] = (int)last_pos; 121 | current_id[Hash] = hash; 122 | token = current_id[Token] = Id; 123 | return; 124 | } 125 | else if (token >= '0' && token <= '9') { 126 | // parse number, three kinds: dec(123) hex(0x123) oct(017) 127 | token_val = token - '0'; 128 | if (token_val > 0) { 129 | // dec, starts with [1-9] 130 | while (*src >= '0' && *src <= '9') { 131 | token_val = token_val*10 + *src++ - '0'; 132 | } 133 | } else { 134 | // starts with number 0 135 | if (*src == 'x' || *src == 'X') { 136 | //hex 137 | token = *++src; 138 | while ((token >= '0' && token <= '9') || (token >= 'a' && token <= 'f') || (token >= 'A' && token <= 'F')) { 139 | token_val = token_val * 16 + (token & 15) + (token >= 'A' ? 9 : 0); 140 | token = *++src; 141 | } 142 | } else { 143 | // oct 144 | while (*src >= '0' && *src <= '7') { 145 | token_val = token_val*8 + *src++ - '0'; 146 | } 147 | } 148 | } 149 | 150 | token = Num; 151 | return; 152 | } 153 | else if (token == '/') { 154 | if (*src == '/') { 155 | // skip comments 156 | while (*src != 0 && *src != '\n') { 157 | ++src; 158 | } 159 | } else { 160 | // divide operator 161 | token = Div; 162 | return; 163 | } 164 | } 165 | else if (token == '"' || token == '\'') { 166 | // parse string literal, currently, the only supported escape 167 | // character is '\n', store the string literal into data. 168 | last_pos = data; 169 | while (*src != 0 && *src != token) { 170 | token_val = *src++; 171 | if (token_val == '\\') { 172 | // escape character 173 | token_val = *src++; 174 | if (token_val == 'n') { 175 | token_val = '\n'; 176 | } 177 | } 178 | 179 | if (token == '"') { 180 | *data++ = token_val; 181 | } 182 | } 183 | 184 | src++; 185 | // if it is a single character, return Num token 186 | if (token == '"') { 187 | token_val = (int)last_pos; 188 | } else { 189 | token = Num; 190 | } 191 | 192 | return; 193 | } 194 | else if (token == '=') { 195 | // parse '==' and '=' 196 | if (*src == '=') { 197 | src ++; 198 | token = Eq; 199 | } else { 200 | token = Assign; 201 | } 202 | return; 203 | } 204 | else if (token == '+') { 205 | // parse '+' and '++' 206 | if (*src == '+') { 207 | src ++; 208 | token = Inc; 209 | } else { 210 | token = Add; 211 | } 212 | return; 213 | } 214 | else if (token == '-') { 215 | // parse '-' and '--' 216 | if (*src == '-') { 217 | src ++; 218 | token = Dec; 219 | } else { 220 | token = Sub; 221 | } 222 | return; 223 | } 224 | else if (token == '!') { 225 | // parse '!=' 226 | if (*src == '=') { 227 | src++; 228 | token = Ne; 229 | } 230 | return; 231 | } 232 | else if (token == '<') { 233 | // parse '<=', '<<' or '<' 234 | if (*src == '=') { 235 | src ++; 236 | token = Le; 237 | } else if (*src == '<') { 238 | src ++; 239 | token = Shl; 240 | } else { 241 | token = Lt; 242 | } 243 | return; 244 | } 245 | else if (token == '>') { 246 | // parse '>=', '>>' or '>' 247 | if (*src == '=') { 248 | src ++; 249 | token = Ge; 250 | } else if (*src == '>') { 251 | src ++; 252 | token = Shr; 253 | } else { 254 | token = Gt; 255 | } 256 | return; 257 | } 258 | else if (token == '|') { 259 | // parse '|' or '||' 260 | if (*src == '|') { 261 | src ++; 262 | token = Lor; 263 | } else { 264 | token = Or; 265 | } 266 | return; 267 | } 268 | else if (token == '&') { 269 | // parse '&' and '&&' 270 | if (*src == '&') { 271 | src ++; 272 | token = Lan; 273 | } else { 274 | token = And; 275 | } 276 | return; 277 | } 278 | else if (token == '^') { 279 | token = Xor; 280 | return; 281 | } 282 | else if (token == '%') { 283 | token = Mod; 284 | return; 285 | } 286 | else if (token == '*') { 287 | token = Mul; 288 | return; 289 | } 290 | else if (token == '[') { 291 | token = Brak; 292 | return; 293 | } 294 | else if (token == '?') { 295 | token = Cond; 296 | return; 297 | } 298 | else if (token == '~' || token == ';' || token == '{' || token == '}' || token == '(' || token == ')' || token == ']' || token == ',' || token == ':') { 299 | // directly return the character as token; 300 | return; 301 | } 302 | } 303 | } 304 | 305 | void match(int tk) { 306 | if (token == tk) { 307 | next(); 308 | } else { 309 | printf("%d: expected token: %d\n", line, tk); 310 | exit(-1); 311 | } 312 | } 313 | 314 | 315 | void expression(int level) { 316 | // expressions have various format. 317 | // but majorly can be divided into two parts: unit and operator 318 | // for example `(char) *a[10] = (int *) func(b > 0 ? 10 : 20); 319 | // `a[10]` is an unit while `*` is an operator. 320 | // `func(...)` in total is an unit. 321 | // so we should first parse those unit and unary operators 322 | // and then the binary ones 323 | // 324 | // also the expression can be in the following types: 325 | // 326 | // 1. unit_unary ::= unit | unit unary_op | unary_op unit 327 | // 2. expr ::= unit_unary (bin_op unit_unary ...) 328 | 329 | // unit_unary() 330 | int *id; 331 | int tmp; 332 | int *addr; 333 | { 334 | if (!token) { 335 | printf("%d: unexpected token EOF of expression\n", line); 336 | exit(-1); 337 | } 338 | if (token == Num) { 339 | match(Num); 340 | 341 | // emit code 342 | *++text = IMM; 343 | *++text = token_val; 344 | expr_type = INT; 345 | } 346 | else if (token == '"') { 347 | // continous string "abc" "abc" 348 | 349 | 350 | // emit code 351 | *++text = IMM; 352 | *++text = token_val; 353 | 354 | match('"'); 355 | // store the rest strings 356 | while (token == '"') { 357 | match('"'); 358 | } 359 | 360 | // append the end of string character '\0', all the data are default 361 | // to 0, so just move data one position forward. 362 | data = (char *)(((int)data + sizeof(int)) & (-sizeof(int))); 363 | expr_type = PTR; 364 | } 365 | else if (token == Sizeof) { 366 | // sizeof is actually an unary operator 367 | // now only `sizeof(int)`, `sizeof(char)` and `sizeof(*...)` are 368 | // supported. 369 | match(Sizeof); 370 | match('('); 371 | expr_type = INT; 372 | 373 | if (token == Int) { 374 | match(Int); 375 | } else if (token == Char) { 376 | match(Char); 377 | expr_type = CHAR; 378 | } 379 | 380 | while (token == Mul) { 381 | match(Mul); 382 | expr_type = expr_type + PTR; 383 | } 384 | 385 | match(')'); 386 | 387 | // emit code 388 | *++text = IMM; 389 | *++text = (expr_type == CHAR) ? sizeof(char) : sizeof(int); 390 | 391 | expr_type = INT; 392 | } 393 | else if (token == Id) { 394 | // there are several type when occurs to Id 395 | // but this is unit, so it can only be 396 | // 1. function call 397 | // 2. Enum variable 398 | // 3. global/local variable 399 | match(Id); 400 | 401 | id = current_id; 402 | 403 | if (token == '(') { 404 | // function call 405 | match('('); 406 | 407 | // pass in arguments 408 | tmp = 0; // number of arguments 409 | while (token != ')') { 410 | expression(Assign); 411 | *++text = PUSH; 412 | tmp ++; 413 | 414 | if (token == ',') { 415 | match(','); 416 | } 417 | 418 | } 419 | match(')'); 420 | 421 | // emit code 422 | if (id[Class] == Sys) { 423 | // system functions 424 | *++text = id[Value]; 425 | } 426 | else if (id[Class] == Fun) { 427 | // function call 428 | *++text = CALL; 429 | *++text = id[Value]; 430 | } 431 | else { 432 | printf("%d: bad function call\n", line); 433 | exit(-1); 434 | } 435 | 436 | // clean the stack for arguments 437 | if (tmp > 0) { 438 | *++text = ADJ; 439 | *++text = tmp; 440 | } 441 | expr_type = id[Type]; 442 | } 443 | else if (id[Class] == Num) { 444 | // enum variable 445 | *++text = IMM; 446 | *++text = id[Value]; 447 | expr_type = INT; 448 | } 449 | else { 450 | // variable 451 | if (id[Class] == Loc) { 452 | *++text = LEA; 453 | *++text = index_of_bp - id[Value]; 454 | } 455 | else if (id[Class] == Glo) { 456 | *++text = IMM; 457 | *++text = id[Value]; 458 | } 459 | else { 460 | printf("%d: undefined variable\n", line); 461 | exit(-1); 462 | } 463 | 464 | // emit code, default behaviour is to load the value of the 465 | // address which is stored in `ax` 466 | expr_type = id[Type]; 467 | *++text = (expr_type == Char) ? LC : LI; 468 | } 469 | } 470 | else if (token == '(') { 471 | // cast or parenthesis 472 | match('('); 473 | if (token == Int || token == Char) { 474 | tmp = (token == Char) ? CHAR : INT; // cast type 475 | match(token); 476 | while (token == Mul) { 477 | match(Mul); 478 | tmp = tmp + PTR; 479 | } 480 | 481 | match(')'); 482 | 483 | expression(Inc); // cast has precedence as Inc(++) 484 | 485 | expr_type = tmp; 486 | } else { 487 | // normal parenthesis 488 | expression(Assign); 489 | match(')'); 490 | } 491 | } 492 | else if (token == Mul) { 493 | // dereference * 494 | match(Mul); 495 | expression(Inc); // dereference has the same precedence as Inc(++) 496 | 497 | if (expr_type >= PTR) { 498 | expr_type = expr_type - PTR; 499 | } else { 500 | printf("%d: bad dereference\n", line); 501 | exit(-1); 502 | } 503 | 504 | *++text = (expr_type == CHAR) ? LC : LI; 505 | } 506 | else if (token == And) { 507 | // get the address of 508 | match(And); 509 | expression(Inc); // get the address of 510 | if (*text == LC || *text == LI) { 511 | text --; 512 | } else { 513 | printf("%d: bad address of\n", line); 514 | exit(-1); 515 | } 516 | 517 | expr_type = expr_type + PTR; 518 | } 519 | else if (token == '!') { 520 | // not 521 | match('!'); 522 | expression(Inc); 523 | 524 | // emit code, use == 0 525 | *++text = PUSH; 526 | *++text = IMM; 527 | *++text = 0; 528 | *++text = EQ; 529 | 530 | expr_type = INT; 531 | } 532 | else if (token == '~') { 533 | // bitwise not 534 | match('~'); 535 | expression(Inc); 536 | 537 | // emit code, use XOR -1 538 | *++text = PUSH; 539 | *++text = IMM; 540 | *++text = -1; 541 | *++text = XOR; 542 | 543 | expr_type = INT; 544 | } 545 | else if (token == Add) { 546 | // +var, do nothing 547 | match(Add); 548 | expression(Inc); 549 | 550 | expr_type = INT; 551 | } 552 | else if (token == Sub) { 553 | // -var 554 | match(Sub); 555 | 556 | if (token == Num) { 557 | *++text = IMM; 558 | *++text = -token_val; 559 | match(Num); 560 | } else { 561 | 562 | *++text = IMM; 563 | *++text = -1; 564 | *++text = PUSH; 565 | expression(Inc); 566 | *++text = MUL; 567 | } 568 | 569 | expr_type = INT; 570 | } 571 | else if (token == Inc || token == Dec) { 572 | tmp = token; 573 | match(token); 574 | expression(Inc); 575 | if (*text == LC) { 576 | *text = PUSH; // to duplicate the address 577 | *++text = LC; 578 | } else if (*text == LI) { 579 | *text = PUSH; 580 | *++text = LI; 581 | } else { 582 | printf("%d: bad lvalue of pre-increment\n", line); 583 | exit(-1); 584 | } 585 | *++text = PUSH; 586 | *++text = IMM; 587 | *++text = (expr_type > PTR) ? sizeof(int) : sizeof(char); 588 | *++text = (tmp == Inc) ? ADD : SUB; 589 | *++text = (expr_type == CHAR) ? SC : SI; 590 | } 591 | else { 592 | printf("%d: bad expression\n", line); 593 | exit(-1); 594 | } 595 | } 596 | 597 | // binary operator and postfix operators. 598 | { 599 | while (token >= level) { 600 | // handle according to current operator's precedence 601 | tmp = expr_type; 602 | if (token == Assign) { 603 | // var = expr; 604 | match(Assign); 605 | if (*text == LC || *text == LI) { 606 | *text = PUSH; // save the lvalue's pointer 607 | } else { 608 | printf("%d: bad lvalue in assignment\n", line); 609 | exit(-1); 610 | } 611 | expression(Assign); 612 | 613 | expr_type = tmp; 614 | *++text = (expr_type == CHAR) ? SC : SI; 615 | } 616 | else if (token == Cond) { 617 | // expr ? a : b; 618 | match(Cond); 619 | *++text = JZ; 620 | addr = ++text; 621 | expression(Assign); 622 | if (token == ':') { 623 | match(':'); 624 | } else { 625 | printf("%d: missing colon in conditional\n", line); 626 | exit(-1); 627 | } 628 | *addr = (int)(text + 3); 629 | *++text = JMP; 630 | addr = ++text; 631 | expression(Cond); 632 | *addr = (int)(text + 1); 633 | } 634 | else if (token == Lor) { 635 | // logic or 636 | match(Lor); 637 | *++text = JNZ; 638 | addr = ++text; 639 | expression(Lan); 640 | *addr = (int)(text + 1); 641 | expr_type = INT; 642 | } 643 | else if (token == Lan) { 644 | // logic and 645 | match(Lan); 646 | *++text = JZ; 647 | addr = ++text; 648 | expression(Or); 649 | *addr = (int)(text + 1); 650 | expr_type = INT; 651 | } 652 | else if (token == Or) { 653 | // bitwise or 654 | match(Or); 655 | *++text = PUSH; 656 | expression(Xor); 657 | *++text = OR; 658 | expr_type = INT; 659 | } 660 | else if (token == Xor) { 661 | // bitwise xor 662 | match(Xor); 663 | *++text = PUSH; 664 | expression(And); 665 | *++text = XOR; 666 | expr_type = INT; 667 | } 668 | else if (token == And) { 669 | // bitwise and 670 | match(And); 671 | *++text = PUSH; 672 | expression(Eq); 673 | *++text = AND; 674 | expr_type = INT; 675 | } 676 | else if (token == Eq) { 677 | // equal == 678 | match(Eq); 679 | *++text = PUSH; 680 | expression(Ne); 681 | *++text = EQ; 682 | expr_type = INT; 683 | } 684 | else if (token == Ne) { 685 | // not equal != 686 | match(Ne); 687 | *++text = PUSH; 688 | expression(Lt); 689 | *++text = NE; 690 | expr_type = INT; 691 | } 692 | else if (token == Lt) { 693 | // less than 694 | match(Lt); 695 | *++text = PUSH; 696 | expression(Shl); 697 | *++text = LT; 698 | expr_type = INT; 699 | } 700 | else if (token == Gt) { 701 | // greater than 702 | match(Gt); 703 | *++text = PUSH; 704 | expression(Shl); 705 | *++text = GT; 706 | expr_type = INT; 707 | } 708 | else if (token == Le) { 709 | // less than or equal to 710 | match(Le); 711 | *++text = PUSH; 712 | expression(Shl); 713 | *++text = LE; 714 | expr_type = INT; 715 | } 716 | else if (token == Ge) { 717 | // greater than or equal to 718 | match(Ge); 719 | *++text = PUSH; 720 | expression(Shl); 721 | *++text = GE; 722 | expr_type = INT; 723 | } 724 | else if (token == Shl) { 725 | // shift left 726 | match(Shl); 727 | *++text = PUSH; 728 | expression(Add); 729 | *++text = SHL; 730 | expr_type = INT; 731 | } 732 | else if (token == Shr) { 733 | // shift right 734 | match(Shr); 735 | *++text = PUSH; 736 | expression(Add); 737 | *++text = SHR; 738 | expr_type = INT; 739 | } 740 | else if (token == Add) { 741 | // add 742 | match(Add); 743 | *++text = PUSH; 744 | expression(Mul); 745 | 746 | expr_type = tmp; 747 | if (expr_type > PTR) { 748 | // pointer type, and not `char *` 749 | *++text = PUSH; 750 | *++text = IMM; 751 | *++text = sizeof(int); 752 | *++text = MUL; 753 | } 754 | *++text = ADD; 755 | } 756 | else if (token == Sub) { 757 | // sub 758 | match(Sub); 759 | *++text = PUSH; 760 | expression(Mul); 761 | if (tmp > PTR && tmp == expr_type) { 762 | // pointer subtraction 763 | *++text = SUB; 764 | *++text = PUSH; 765 | *++text = IMM; 766 | *++text = sizeof(int); 767 | *++text = DIV; 768 | expr_type = INT; 769 | } else if (tmp > PTR) { 770 | // pointer movement 771 | *++text = PUSH; 772 | *++text = IMM; 773 | *++text = sizeof(int); 774 | *++text = MUL; 775 | *++text = SUB; 776 | expr_type = tmp; 777 | } else { 778 | // numeral subtraction 779 | *++text = SUB; 780 | expr_type = tmp; 781 | } 782 | } 783 | else if (token == Mul) { 784 | // multiply 785 | match(Mul); 786 | *++text = PUSH; 787 | expression(Inc); 788 | *++text = MUL; 789 | expr_type = tmp; 790 | } 791 | else if (token == Div) { 792 | // divide 793 | match(Div); 794 | *++text = PUSH; 795 | expression(Inc); 796 | *++text = DIV; 797 | expr_type = tmp; 798 | } 799 | else if (token == Mod) { 800 | // Modulo 801 | match(Mod); 802 | *++text = PUSH; 803 | expression(Inc); 804 | *++text = MOD; 805 | expr_type = tmp; 806 | } 807 | else if (token == Inc || token == Dec) { 808 | // postfix inc(++) and dec(--) 809 | // we will increase the value to the variable and decrease it 810 | // on `ax` to get its original value. 811 | if (*text == LI) { 812 | *text = PUSH; 813 | *++text = LI; 814 | } 815 | else if (*text == LC) { 816 | *text = PUSH; 817 | *++text = LC; 818 | } 819 | else { 820 | printf("%d: bad value in increment\n", line); 821 | exit(-1); 822 | } 823 | 824 | *++text = PUSH; 825 | *++text = IMM; 826 | *++text = (expr_type > PTR) ? sizeof(int) : sizeof(char); 827 | *++text = (token == Inc) ? ADD : SUB; 828 | *++text = (expr_type == CHAR) ? SC : SI; 829 | *++text = PUSH; 830 | *++text = IMM; 831 | *++text = (expr_type > PTR) ? sizeof(int) : sizeof(char); 832 | *++text = (token == Inc) ? SUB : ADD; 833 | match(token); 834 | } 835 | else if (token == Brak) { 836 | // array access var[xx] 837 | match(Brak); 838 | *++text = PUSH; 839 | expression(Assign); 840 | match(']'); 841 | 842 | if (tmp > PTR) { 843 | // pointer, `not char *` 844 | *++text = PUSH; 845 | *++text = IMM; 846 | *++text = sizeof(int); 847 | *++text = MUL; 848 | } 849 | else if (tmp < PTR) { 850 | printf("%d: pointer type expected\n", line); 851 | exit(-1); 852 | } 853 | expr_type = tmp - PTR; 854 | *++text = ADD; 855 | *++text = (expr_type == CHAR) ? LC : LI; 856 | } 857 | else { 858 | printf("%d: compiler error, token = %d\n", line, token); 859 | exit(-1); 860 | } 861 | } 862 | } 863 | } 864 | 865 | void statement() { 866 | // there are 8 kinds of statements here: 867 | // 1. if (...) [else ] 868 | // 2. while (...) 869 | // 3. { } 870 | // 4. return xxx; 871 | // 5. ; 872 | // 6. expression; (expression end with semicolon) 873 | 874 | int *a, *b; // bess for branch control 875 | 876 | if (token == If) { 877 | // if (...) [else ] 878 | // 879 | // if (...) 880 | // JZ a 881 | // 882 | // else: JMP b 883 | // a: 884 | // 885 | // b: b: 886 | // 887 | // 888 | match(If); 889 | match('('); 890 | expression(Assign); // parse condition 891 | match(')'); 892 | 893 | // emit code for if 894 | *++text = JZ; 895 | b = ++text; 896 | 897 | statement(); // parse statement 898 | if (token == Else) { // parse else 899 | match(Else); 900 | 901 | // emit code for JMP B 902 | *b = (int)(text + 3); 903 | *++text = JMP; 904 | b = ++text; 905 | 906 | statement(); 907 | } 908 | 909 | *b = (int)(text + 1); 910 | } 911 | else if (token == While) { 912 | // 913 | // a: a: 914 | // while () 915 | // JZ b 916 | // 917 | // JMP a 918 | // b: b: 919 | match(While); 920 | 921 | a = text + 1; 922 | 923 | match('('); 924 | expression(Assign); 925 | match(')'); 926 | 927 | *++text = JZ; 928 | b = ++text; 929 | 930 | statement(); 931 | 932 | *++text = JMP; 933 | *++text = (int)a; 934 | *b = (int)(text + 1); 935 | } 936 | else if (token == '{') { 937 | // { ... } 938 | match('{'); 939 | 940 | while (token != '}') { 941 | statement(); 942 | } 943 | 944 | match('}'); 945 | } 946 | else if (token == Return) { 947 | // return [expression]; 948 | match(Return); 949 | 950 | if (token != ';') { 951 | expression(Assign); 952 | } 953 | 954 | match(';'); 955 | 956 | // emit code for return 957 | *++text = LEV; 958 | } 959 | else if (token == ';') { 960 | // empty statement 961 | match(';'); 962 | } 963 | else { 964 | // a = b; or function_call(); 965 | expression(Assign); 966 | match(';'); 967 | } 968 | } 969 | 970 | void enum_declaration() { 971 | // parse enum [id] { a = 1, b = 3, ...} 972 | int i; 973 | i = 0; 974 | while (token != '}') { 975 | if (token != Id) { 976 | printf("%d: bad enum identifier %d\n", line, token); 977 | exit(-1); 978 | } 979 | next(); 980 | if (token == Assign) { 981 | // like {a=10} 982 | next(); 983 | if (token != Num) { 984 | printf("%d: bad enum initializer\n", line); 985 | exit(-1); 986 | } 987 | i = token_val; 988 | next(); 989 | } 990 | 991 | current_id[Class] = Num; 992 | current_id[Type] = INT; 993 | current_id[Value] = i++; 994 | 995 | if (token == ',') { 996 | next(); 997 | } 998 | } 999 | } 1000 | 1001 | void function_parameter() { 1002 | int type; 1003 | int params; 1004 | params = 0; 1005 | while (token != ')') { 1006 | // int name, ... 1007 | type = INT; 1008 | if (token == Int) { 1009 | match(Int); 1010 | } else if (token == Char) { 1011 | type = CHAR; 1012 | match(Char); 1013 | } 1014 | 1015 | // pointer type 1016 | while (token == Mul) { 1017 | match(Mul); 1018 | type = type + PTR; 1019 | } 1020 | 1021 | // parameter name 1022 | if (token != Id) { 1023 | printf("%d: bad parameter declaration\n", line); 1024 | exit(-1); 1025 | } 1026 | if (current_id[Class] == Loc) { 1027 | printf("%d: duplicate parameter declaration\n", line); 1028 | exit(-1); 1029 | } 1030 | 1031 | match(Id); 1032 | // store the local variable 1033 | current_id[BClass] = current_id[Class]; current_id[Class] = Loc; 1034 | current_id[BType] = current_id[Type]; current_id[Type] = type; 1035 | current_id[BValue] = current_id[Value]; current_id[Value] = params++; // index of current parameter 1036 | 1037 | if (token == ',') { 1038 | match(','); 1039 | } 1040 | } 1041 | index_of_bp = params+1; 1042 | } 1043 | 1044 | void function_body() { 1045 | // type func_name (...) {...} 1046 | // -->| |<-- 1047 | 1048 | // ... { 1049 | // 1. local declarations 1050 | // 2. statements 1051 | // } 1052 | 1053 | int pos_local; // position of local variables on the stack. 1054 | int type; 1055 | pos_local = index_of_bp; 1056 | 1057 | while (token == Int || token == Char) { 1058 | // local variable declaration, just like global ones. 1059 | basetype = (token == Int) ? INT : CHAR; 1060 | match(token); 1061 | 1062 | while (token != ';') { 1063 | type = basetype; 1064 | while (token == Mul) { 1065 | match(Mul); 1066 | type = type + PTR; 1067 | } 1068 | 1069 | if (token != Id) { 1070 | // invalid declaration 1071 | printf("%d: bad local declaration\n", line); 1072 | exit(-1); 1073 | } 1074 | if (current_id[Class] == Loc) { 1075 | // identifier exists 1076 | printf("%d: duplicate local declaration\n", line); 1077 | exit(-1); 1078 | } 1079 | match(Id); 1080 | 1081 | // store the local variable 1082 | current_id[BClass] = current_id[Class]; current_id[Class] = Loc; 1083 | current_id[BType] = current_id[Type]; current_id[Type] = type; 1084 | current_id[BValue] = current_id[Value]; current_id[Value] = ++pos_local; // index of current parameter 1085 | 1086 | if (token == ',') { 1087 | match(','); 1088 | } 1089 | } 1090 | match(';'); 1091 | } 1092 | 1093 | // save the stack size for local variables 1094 | *++text = ENT; 1095 | *++text = pos_local - index_of_bp; 1096 | 1097 | // statements 1098 | while (token != '}') { 1099 | statement(); 1100 | } 1101 | 1102 | // emit code for leaving the sub function 1103 | *++text = LEV; 1104 | } 1105 | 1106 | void function_declaration() { 1107 | // type func_name (...) {...} 1108 | // | this part 1109 | 1110 | match('('); 1111 | function_parameter(); 1112 | match(')'); 1113 | match('{'); 1114 | function_body(); 1115 | //match('}'); 1116 | 1117 | // unwind local variable declarations for all local variables. 1118 | current_id = symbols; 1119 | while (current_id[Token]) { 1120 | if (current_id[Class] == Loc) { 1121 | current_id[Class] = current_id[BClass]; 1122 | current_id[Type] = current_id[BType]; 1123 | current_id[Value] = current_id[BValue]; 1124 | } 1125 | current_id = current_id + IdSize; 1126 | } 1127 | } 1128 | 1129 | void global_declaration() { 1130 | // int [*]id [; | (...) {...}] 1131 | 1132 | 1133 | int type; // tmp, actual type for variable 1134 | int i; // tmp 1135 | 1136 | basetype = INT; 1137 | 1138 | // parse enum, this should be treated alone. 1139 | if (token == Enum) { 1140 | // enum [id] { a = 10, b = 20, ... } 1141 | match(Enum); 1142 | if (token != '{') { 1143 | match(Id); // skip the [id] part 1144 | } 1145 | if (token == '{') { 1146 | // parse the assign part 1147 | match('{'); 1148 | enum_declaration(); 1149 | match('}'); 1150 | } 1151 | 1152 | match(';'); 1153 | return; 1154 | } 1155 | 1156 | // parse type information 1157 | if (token == Int) { 1158 | match(Int); 1159 | } 1160 | else if (token == Char) { 1161 | match(Char); 1162 | basetype = CHAR; 1163 | } 1164 | 1165 | // parse the comma seperated variable declaration. 1166 | while (token != ';' && token != '}') { 1167 | type = basetype; 1168 | // parse pointer type, note that there may exist `int ****x;` 1169 | while (token == Mul) { 1170 | match(Mul); 1171 | type = type + PTR; 1172 | } 1173 | 1174 | if (token != Id) { 1175 | // invalid declaration 1176 | printf("%d: bad global declaration\n", line); 1177 | exit(-1); 1178 | } 1179 | if (current_id[Class]) { 1180 | // identifier exists 1181 | printf("%d: duplicate global declaration\n", line); 1182 | exit(-1); 1183 | } 1184 | match(Id); 1185 | current_id[Type] = type; 1186 | 1187 | if (token == '(') { 1188 | current_id[Class] = Fun; 1189 | current_id[Value] = (int)(text + 1); // the memory address of function 1190 | function_declaration(); 1191 | } else { 1192 | // variable declaration 1193 | current_id[Class] = Glo; // global variable 1194 | current_id[Value] = (int)data; // assign memory address 1195 | data = data + sizeof(int); 1196 | } 1197 | 1198 | if (token == ',') { 1199 | match(','); 1200 | } 1201 | } 1202 | next(); 1203 | } 1204 | 1205 | void program() { 1206 | // get next token 1207 | next(); 1208 | while (token > 0) { 1209 | global_declaration(); 1210 | } 1211 | } 1212 | 1213 | int eval() { 1214 | int op, *tmp; 1215 | cycle = 0; 1216 | while (1) { 1217 | cycle ++; 1218 | op = *pc++; // get next operation code 1219 | 1220 | // print debug info 1221 | if (debug) { 1222 | printf("%d> %.4s", cycle, 1223 | & "LEA ,IMM ,JMP ,CALL,JZ ,JNZ ,ENT ,ADJ ,LEV ,LI ,LC ,SI ,SC ,PUSH," 1224 | "OR ,XOR ,AND ,EQ ,NE ,LT ,GT ,LE ,GE ,SHL ,SHR ,ADD ,SUB ,MUL ,DIV ,MOD ," 1225 | "OPEN,READ,CLOS,PRTF,MALC,MSET,MCMP,EXIT"[op * 5]); 1226 | if (op <= ADJ) 1227 | printf(" %d\n", *pc); 1228 | else 1229 | printf("\n"); 1230 | } 1231 | if (op == IMM) {ax = *pc++;} // load immediate value to ax 1232 | else if (op == LC) {ax = *(char *)ax;} // load character to ax, address in ax 1233 | else if (op == LI) {ax = *(int *)ax;} // load integer to ax, address in ax 1234 | else if (op == SC) {ax = *(char *)*sp++ = ax;} // save character to address, value in ax, address on stack 1235 | else if (op == SI) {*(int *)*sp++ = ax;} // save integer to address, value in ax, address on stack 1236 | else if (op == PUSH) {*--sp = ax;} // push the value of ax onto the stack 1237 | else if (op == JMP) {pc = (int *)*pc;} // jump to the address 1238 | else if (op == JZ) {pc = ax ? pc + 1 : (int *)*pc;} // jump if ax is zero 1239 | else if (op == JNZ) {pc = ax ? (int *)*pc : pc + 1;} // jump if ax is zero 1240 | else if (op == CALL) {*--sp = (int)(pc+1); pc = (int *)*pc;} // call subroutine 1241 | //else if (op == RET) {pc = (int *)*sp++;} // return from subroutine; 1242 | else if (op == ENT) {*--sp = (int)bp; bp = sp; sp = sp - *pc++;} // make new stack frame 1243 | else if (op == ADJ) {sp = sp + *pc++;} // add esp, 1244 | else if (op == LEV) {sp = bp; bp = (int *)*sp++; pc = (int *)*sp++;} // restore call frame and PC 1245 | else if (op == LEA) {ax = (int)(bp + *pc++);} // load address for arguments. 1246 | 1247 | else if (op == OR) ax = *sp++ | ax; 1248 | else if (op == XOR) ax = *sp++ ^ ax; 1249 | else if (op == AND) ax = *sp++ & ax; 1250 | else if (op == EQ) ax = *sp++ == ax; 1251 | else if (op == NE) ax = *sp++ != ax; 1252 | else if (op == LT) ax = *sp++ < ax; 1253 | else if (op == LE) ax = *sp++ <= ax; 1254 | else if (op == GT) ax = *sp++ > ax; 1255 | else if (op == GE) ax = *sp++ >= ax; 1256 | else if (op == SHL) ax = *sp++ << ax; 1257 | else if (op == SHR) ax = *sp++ >> ax; 1258 | else if (op == ADD) ax = *sp++ + ax; 1259 | else if (op == SUB) ax = *sp++ - ax; 1260 | else if (op == MUL) ax = *sp++ * ax; 1261 | else if (op == DIV) ax = *sp++ / ax; 1262 | else if (op == MOD) ax = *sp++ % ax; 1263 | 1264 | else if (op == EXIT) { printf("exit(%d)", *sp); return *sp;} 1265 | else if (op == OPEN) { ax = open((char *)sp[1], sp[0]); } 1266 | else if (op == CLOS) { ax = close(*sp);} 1267 | else if (op == READ) { ax = read(sp[2], (char *)sp[1], *sp); } 1268 | else if (op == PRTF) { tmp = sp + pc[1]; ax = printf((char *)tmp[-1], tmp[-2], tmp[-3], tmp[-4], tmp[-5], tmp[-6]); } 1269 | else if (op == MALC) { ax = (int)malloc(*sp);} 1270 | else if (op == MSET) { ax = (int)memset((char *)sp[2], sp[1], *sp);} 1271 | else if (op == MCMP) { ax = memcmp((char *)sp[2], (char *)sp[1], *sp);} 1272 | else { 1273 | printf("unknown instruction:%d\n", op); 1274 | return -1; 1275 | } 1276 | } 1277 | } 1278 | 1279 | int main(int argc, char **argv) 1280 | { 1281 | int i, fd; 1282 | int *tmp; 1283 | 1284 | argc--; 1285 | argv++; 1286 | 1287 | // parse arguments 1288 | if (argc > 0 && **argv == '-' && (*argv)[1] == 's') { 1289 | assembly = 1; 1290 | --argc; 1291 | ++argv; 1292 | } 1293 | if (argc > 0 && **argv == '-' && (*argv)[1] == 'd') { 1294 | debug = 1; 1295 | --argc; 1296 | ++argv; 1297 | } 1298 | if (argc < 1) { 1299 | printf("usage: xc [-s] [-d] file ...\n"); 1300 | return -1; 1301 | } 1302 | 1303 | if ((fd = open(*argv, 0)) < 0) { 1304 | printf("could not open(%s)\n", *argv); 1305 | return -1; 1306 | } 1307 | 1308 | poolsize = 256 * 1024; // arbitrary size 1309 | line = 1; 1310 | 1311 | // allocate memory 1312 | if (!(text = malloc(poolsize))) { 1313 | printf("could not malloc(%d) for text area\n", poolsize); 1314 | return -1; 1315 | } 1316 | if (!(data = malloc(poolsize))) { 1317 | printf("could not malloc(%d) for data area\n", poolsize); 1318 | return -1; 1319 | } 1320 | if (!(stack = malloc(poolsize))) { 1321 | printf("could not malloc(%d) for stack area\n", poolsize); 1322 | return -1; 1323 | } 1324 | if (!(symbols = malloc(poolsize))) { 1325 | printf("could not malloc(%d) for symbol table\n", poolsize); 1326 | return -1; 1327 | } 1328 | 1329 | memset(text, 0, poolsize); 1330 | memset(data, 0, poolsize); 1331 | memset(stack, 0, poolsize); 1332 | memset(symbols, 0, poolsize); 1333 | 1334 | old_text = text; 1335 | 1336 | src = "char else enum if int return sizeof while " 1337 | "open read close printf malloc memset memcmp exit void main"; 1338 | 1339 | // add keywords to symbol table 1340 | i = Char; 1341 | while (i <= While) { 1342 | next(); 1343 | current_id[Token] = i++; 1344 | } 1345 | 1346 | // add library to symbol table 1347 | i = OPEN; 1348 | while (i <= EXIT) { 1349 | next(); 1350 | current_id[Class] = Sys; 1351 | current_id[Type] = INT; 1352 | current_id[Value] = i++; 1353 | } 1354 | 1355 | next(); current_id[Token] = Char; // handle void type 1356 | next(); idmain = current_id; // keep track of main 1357 | 1358 | if (!(src = old_src = malloc(poolsize))) { 1359 | printf("could not malloc(%d) for source area\n", poolsize); 1360 | return -1; 1361 | } 1362 | // read the source file 1363 | if ((i = read(fd, src, poolsize-1)) <= 0) { 1364 | printf("read() returned %d\n", i); 1365 | return -1; 1366 | } 1367 | src[i] = 0; // add EOF character 1368 | close(fd); 1369 | 1370 | program(); 1371 | 1372 | if (!(pc = (int *)idmain[Value])) { 1373 | printf("main() not defined\n"); 1374 | return -1; 1375 | } 1376 | 1377 | // dump_text(); 1378 | if (assembly) { 1379 | // only for compile 1380 | return 0; 1381 | } 1382 | 1383 | // setup stack 1384 | sp = (int *)((int)stack + poolsize); 1385 | *--sp = EXIT; // call exit if main returns 1386 | *--sp = PUSH; tmp = sp; 1387 | *--sp = argc; 1388 | *--sp = (int)argv; 1389 | *--sp = (int)tmp; 1390 | 1391 | return eval(); 1392 | } 1393 | --------------------------------------------------------------------------------