micro_asm.py revision 5009:78d53ea88c74
1# Copyright (c) 2003-2005 The Regents of The University of Michigan 2# All rights reserved. 3# 4# Redistribution and use in source and binary forms, with or without 5# modification, are permitted provided that the following conditions are 6# met: redistributions of source code must retain the above copyright 7# notice, this list of conditions and the following disclaimer; 8# redistributions in binary form must reproduce the above copyright 9# notice, this list of conditions and the following disclaimer in the 10# documentation and/or other materials provided with the distribution; 11# neither the name of the copyright holders nor the names of its 12# contributors may be used to endorse or promote products derived from 13# this software without specific prior written permission. 14# 15# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS 16# "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT 17# LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR 18# A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT 19# OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, 20# SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT 21# LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, 22# DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY 23# THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 24# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE 25# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 26# 27# Authors: Gabe Black 28 29import os 30import sys 31import re 32import string 33import traceback 34# get type names 35from types import * 36 37# Prepend the directory where the PLY lex & yacc modules are found 38# to the search path. 39sys.path[0:0] = [os.environ['M5_PLY']] 40 41from ply import lex 42from ply import yacc 43 44########################################################################## 45# 46# Base classes for use outside of the assembler 47# 48########################################################################## 49 50class Micro_Container(object): 51 def __init__(self, name): 52 self.microops = [] 53 self.name = name 54 self.directives = {} 55 self.micro_classes = {} 56 self.labels = {} 57 58 def add_microop(self, microop): 59 self.microops.append(microop) 60 61 def __str__(self): 62 string = "%s:\n" % self.name 63 for microop in self.microops: 64 string += " %s\n" % microop 65 return string 66 67class Combinational_Macroop(Micro_Container): 68 pass 69 70class Rom_Macroop(object): 71 def __init__(self, name, target): 72 self.name = name 73 self.target = target 74 75 def __str__(self): 76 return "%s: %s\n" % (self.name, self.target) 77 78class Rom(Micro_Container): 79 def __init__(self, name): 80 super(Rom, self).__init__(name) 81 self.externs = {} 82 83########################################################################## 84# 85# Support classes 86# 87########################################################################## 88 89class Label(object): 90 def __init__(self): 91 self.extern = False 92 self.name = "" 93 94class Block(object): 95 def __init__(self): 96 self.statements = [] 97 98class Statement(object): 99 def __init__(self): 100 self.is_microop = False 101 self.is_directive = False 102 self.params = "" 103 104class Microop(Statement): 105 def __init__(self): 106 super(Microop, self).__init__() 107 self.mnemonic = "" 108 self.labels = [] 109 self.is_microop = True 110 111class Directive(Statement): 112 def __init__(self): 113 super(Directive, self).__init__() 114 self.name = "" 115 self.is_directive = True 116 117########################################################################## 118# 119# Functions that handle common tasks 120# 121########################################################################## 122 123def print_error(message): 124 print 125 print "*** %s" % message 126 print 127 128def handle_statement(parser, container, statement): 129 if statement.is_microop: 130 if statement.mnemonic not in parser.microops.keys(): 131 raise Exception, "Unrecognized mnemonic: %s" % statement.mnemonic 132 parser.symbols["__microopClassFromInsideTheAssembler"] = \ 133 parser.microops[statement.mnemonic] 134 try: 135 microop = eval('__microopClassFromInsideTheAssembler(%s)' % 136 statement.params, {}, parser.symbols) 137 except: 138 print_error("Error creating microop object with mnemonic %s." % \ 139 statement.mnemonic) 140 raise 141 try: 142 for label in statement.labels: 143 container.labels[label.text] = microop 144 if label.extern: 145 container.externs[label.text] = microop 146 container.add_microop(microop) 147 except: 148 print_error("Error adding microop.") 149 raise 150 elif statement.is_directive: 151 if statement.name not in container.directives.keys(): 152 raise Exception, "Unrecognized directive: %s" % statement.name 153 parser.symbols["__directiveFunctionFromInsideTheAssembler"] = \ 154 container.directives[statement.name] 155 try: 156 eval('__directiveFunctionFromInsideTheAssembler(%s)' % 157 statement.params, {}, parser.symbols) 158 except: 159 print_error("Error executing directive.") 160 print container.directives 161 raise 162 else: 163 raise Exception, "Didn't recognize the type of statement", statement 164 165########################################################################## 166# 167# Lexer specification 168# 169########################################################################## 170 171# Error handler. Just call exit. Output formatted to work under 172# Emacs compile-mode. Optional 'print_traceback' arg, if set to True, 173# prints a Python stack backtrace too (can be handy when trying to 174# debug the parser itself). 175def error(lineno, string, print_traceback = False): 176 # Print a Python stack backtrace if requested. 177 if (print_traceback): 178 traceback.print_exc() 179 if lineno != 0: 180 line_str = "%d:" % lineno 181 else: 182 line_str = "" 183 sys.exit("%s %s" % (line_str, string)) 184 185reserved = ('DEF', 'MACROOP', 'ROM', 'EXTERN') 186 187tokens = reserved + ( 188 # identifier 189 'ID', 190 # arguments for microops and directives 191 'PARAMS', 192 193 'LPAREN', 'RPAREN', 194 'LBRACE', 'RBRACE', 195 'COLON', 'SEMI', 'DOT', 196 'NEWLINE' 197 ) 198 199# New lines are ignored at the top level, but they end statements in the 200# assembler 201states = ( 202 ('asm', 'exclusive'), 203 ('params', 'exclusive'), 204) 205 206reserved_map = { } 207for r in reserved: 208 reserved_map[r.lower()] = r 209 210# Ignore comments 211def t_ANY_COMMENT(t): 212 r'\#[^\n]*(?=\n)' 213 214def t_ANY_MULTILINECOMMENT(t): 215 r'/\*([^/]|((?<!\*)/))*\*/' 216 217# A colon marks the end of a label. It should follow an ID which will 218# put the lexer in the "params" state. Seeing the colon will put it back 219# in the "asm" state since it knows it saw a label and not a mnemonic. 220def t_params_COLON(t): 221 r':' 222 t.lexer.begin('asm') 223 return t 224 225# Parameters are a string of text which don't contain an unescaped statement 226# statement terminator, ie a newline or semi colon. 227def t_params_PARAMS(t): 228 r'([^\n;\\]|(\\[\n;\\]))+' 229 t.lineno += t.value.count('\n') 230 unescapeParamsRE = re.compile(r'(\\[\n;\\])') 231 def unescapeParams(mo): 232 val = mo.group(0) 233 return val[1] 234 t.value = unescapeParamsRE.sub(unescapeParams, t.value) 235 t.lexer.begin('asm') 236 return t 237 238# An "ID" in the micro assembler is either a label, directive, or mnemonic 239# If it's either a directive or a mnemonic, it will be optionally followed by 240# parameters. If it's a label, the following colon will make the lexer stop 241# looking for parameters. 242def t_asm_ID(t): 243 r'[A-Za-z_]\w*' 244 t.type = reserved_map.get(t.value, 'ID') 245 t.lexer.begin('params') 246 return t 247 248# If there is a label and you're -not- in the assembler (which would be caught 249# above), don't start looking for parameters. 250def t_ANY_ID(t): 251 r'[A-Za-z_]\w*' 252 t.type = reserved_map.get(t.value, 'ID') 253 return t 254 255# Braces enter and exit micro assembly 256def t_INITIAL_LBRACE(t): 257 r'\{' 258 t.lexer.begin('asm') 259 return t 260 261def t_asm_RBRACE(t): 262 r'\}' 263 t.lexer.begin('INITIAL') 264 return t 265 266# At the top level, keep track of newlines only for line counting. 267def t_INITIAL_NEWLINE(t): 268 r'\n+' 269 t.lineno += t.value.count('\n') 270 271# In the micro assembler, do line counting but also return a token. The 272# token is needed by the parser to detect the end of a statement. 273def t_asm_NEWLINE(t): 274 r'\n+' 275 t.lineno += t.value.count('\n') 276 return t 277 278# A newline or semi colon when looking for params signals that the statement 279# is over and the lexer should go back to looking for regular assembly. 280def t_params_NEWLINE(t): 281 r'\n+' 282 t.lineno += t.value.count('\n') 283 t.lexer.begin('asm') 284 return t 285 286def t_params_SEMI(t): 287 r';' 288 t.lexer.begin('asm') 289 return t 290 291# Basic regular expressions to pick out simple tokens 292t_ANY_LPAREN = r'\(' 293t_ANY_RPAREN = r'\)' 294t_ANY_SEMI = r';' 295t_ANY_DOT = r'\.' 296 297t_ANY_ignore = ' \t\x0c' 298 299def t_ANY_error(t): 300 error(t.lineno, "illegal character '%s'" % t.value[0]) 301 t.skip(1) 302 303########################################################################## 304# 305# Parser specification 306# 307########################################################################## 308 309# Start symbol for a file which may have more than one macroop or rom 310# specification. 311def p_file(t): 312 'file : opt_rom_or_macros' 313 314def p_opt_rom_or_macros_0(t): 315 'opt_rom_or_macros : ' 316 317def p_opt_rom_or_macros_1(t): 318 'opt_rom_or_macros : rom_or_macros' 319 320def p_rom_or_macros_0(t): 321 'rom_or_macros : rom_or_macro' 322 323def p_rom_or_macros_1(t): 324 'rom_or_macros : rom_or_macros rom_or_macro' 325 326def p_rom_or_macro_0(t): 327 '''rom_or_macro : rom_block 328 | macroop_def''' 329 330# Defines a section of microcode that should go in the current ROM 331def p_rom_block(t): 332 'rom_block : DEF ROM block SEMI' 333 if not t.parser.rom: 334 print_error("Rom block found, but no Rom object specified.") 335 raise TypeError, "Rom block found, but no Rom object was specified." 336 for statement in t[3].statements: 337 handle_statement(t.parser, t.parser.rom, statement) 338 t[0] = t.parser.rom 339 340# Defines a macroop that jumps to an external label in the ROM 341def p_macroop_def_0(t): 342 'macroop_def : DEF MACROOP ID LPAREN ID RPAREN SEMI' 343 if not t.parser.rom_macroop_type: 344 print_error("ROM based macroop found, but no ROM macroop class was specified.") 345 raise TypeError, "ROM based macroop found, but no ROM macroop class was specified." 346 macroop = t.parser.rom_macroop_type(t[3], t[5]) 347 t.parser.macroops[t[3]] = macroop 348 349 350# Defines a macroop that is combinationally generated 351def p_macroop_def_1(t): 352 'macroop_def : DEF MACROOP ID block SEMI' 353 try: 354 curop = t.parser.macro_type(t[3]) 355 except TypeError: 356 print_error("Error creating macroop object.") 357 raise 358 for statement in t[4].statements: 359 handle_statement(t.parser, curop, statement) 360 t.parser.macroops[t[3]] = curop 361 362# A block of statements 363def p_block(t): 364 'block : LBRACE statements RBRACE' 365 block = Block() 366 block.statements = t[2] 367 t[0] = block 368 369def p_statements_0(t): 370 'statements : statement' 371 if t[1]: 372 t[0] = [t[1]] 373 else: 374 t[0] = [] 375 376def p_statements_1(t): 377 'statements : statements statement' 378 if t[2]: 379 t[1].append(t[2]) 380 t[0] = t[1] 381 382def p_statement(t): 383 'statement : content_of_statement end_of_statement' 384 t[0] = t[1] 385 386# A statement can be a microop or an assembler directive 387def p_content_of_statement_0(t): 388 '''content_of_statement : microop 389 | directive''' 390 t[0] = t[1] 391 392# Ignore empty statements 393def p_content_of_statement_1(t): 394 'content_of_statement : ' 395 pass 396 397# Statements are ended by newlines or a semi colon 398def p_end_of_statement(t): 399 '''end_of_statement : NEWLINE 400 | SEMI''' 401 pass 402 403# Different flavors of microop to avoid shift/reduce errors 404def p_microop_0(t): 405 'microop : labels ID' 406 microop = Microop() 407 microop.labels = t[1] 408 microop.mnemonic = t[2] 409 t[0] = microop 410 411def p_microop_1(t): 412 'microop : ID' 413 microop = Microop() 414 microop.mnemonic = t[1] 415 t[0] = microop 416 417def p_microop_2(t): 418 'microop : labels ID PARAMS' 419 microop = Microop() 420 microop.labels = t[1] 421 microop.mnemonic = t[2] 422 microop.params = t[3] 423 t[0] = microop 424 425def p_microop_3(t): 426 'microop : ID PARAMS' 427 microop = Microop() 428 microop.mnemonic = t[1] 429 microop.params = t[2] 430 t[0] = microop 431 432# Labels in the microcode 433def p_labels_0(t): 434 'labels : label' 435 t[0] = [t[1]] 436 437def p_labels_1(t): 438 'labels : labels label' 439 t[1].append(t[2]) 440 t[0] = t[1] 441 442# labels on lines by themselves are attached to the following instruction. 443def p_labels_2(t): 444 'labels : labels NEWLINE' 445 t[0] = t[1] 446 447def p_label_0(t): 448 'label : ID COLON' 449 label = Label() 450 label.is_extern = False 451 label.text = t[1] 452 t[0] = label 453 454def p_label_1(t): 455 'label : EXTERN ID COLON' 456 label = Label() 457 label.is_extern = True 458 label.text = t[2] 459 t[0] = label 460 461# Directives for the macroop 462def p_directive_0(t): 463 'directive : DOT ID' 464 directive = Directive() 465 directive.name = t[2] 466 t[0] = directive 467 468def p_directive_1(t): 469 'directive : DOT ID PARAMS' 470 directive = Directive() 471 directive.name = t[2] 472 directive.params = t[3] 473 t[0] = directive 474 475# Parse error handler. Note that the argument here is the offending 476# *token*, not a grammar symbol (hence the need to use t.value) 477def p_error(t): 478 if t: 479 error(t.lineno, "syntax error at '%s'" % t.value) 480 else: 481 error(0, "unknown syntax error", True) 482 483class MicroAssembler(object): 484 485 def __init__(self, macro_type, microops, 486 rom = None, rom_macroop_type = None): 487 self.lexer = lex.lex() 488 self.parser = yacc.yacc() 489 self.parser.macro_type = macro_type 490 self.parser.macroops = {} 491 self.parser.microops = microops 492 self.parser.rom = rom 493 self.parser.rom_macroop_type = rom_macroop_type 494 self.parser.symbols = {} 495 self.symbols = self.parser.symbols 496 497 def assemble(self, asm): 498 self.parser.parse(asm, lexer=self.lexer) 499 macroops = self.parser.macroops 500 self.parser.macroops = {} 501 return macroops 502