Training courses

Kernel and Embedded Linux

Bootlin training courses

Embedded Linux, kernel,
Yocto Project, Buildroot, real-time,
graphics, boot time, debugging...

Bootlin logo

Elixir Cross Referencer

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
%{
/*
 * zlexer.lex - lexical analyzer for (DNS) zone files
 * 
 * Copyright (c) 2001-2006, NLnet Labs. All rights reserved
 *
 * See LICENSE for the license.
 *
 */
/* because flex keeps having sign-unsigned compare problems that are unfixed*/
#if defined(__clang__)||(defined(__GNUC__)&&((__GNUC__ >4)||(defined(__GNUC_MINOR__)&&(__GNUC__ ==4)&&(__GNUC_MINOR__ >=2))))
#pragma GCC diagnostic ignored "-Wsign-compare"
#endif
/* ignore fallthrough warnings in the generated parse code case statements */
#if defined(__clang__)||(defined(__GNUC__)&&(__GNUC__ >=7))
#pragma GCC diagnostic ignored "-Wimplicit-fallthrough"
#endif

#include "config.h"

#include <ctype.h>
#include <errno.h>
#include <string.h>
#include <strings.h>

#include "zonec.h"
#include "dname.h"
#include "zparser.h"

#if 0
#define LEXOUT(s)  printf s /* used ONLY when debugging */
#else
#define LEXOUT(s)
#endif

enum lexer_state {
	EXPECT_OWNER,
	PARSING_OWNER,
	PARSING_TTL_CLASS_TYPE,
	PARSING_RDATA
};

static int parse_token(int token, char *yytext, enum lexer_state *lexer_state);

static YY_BUFFER_STATE include_stack[MAXINCLUDES];
static zparser_type zparser_stack[MAXINCLUDES];
static int include_stack_ptr = 0;

/*
 * Saves the file specific variables on the include stack.
 */
static void
push_parser_state(FILE *input)
{
	zparser_stack[include_stack_ptr].filename = parser->filename;
	zparser_stack[include_stack_ptr].line = parser->line;
	zparser_stack[include_stack_ptr].origin = parser->origin;
	include_stack[include_stack_ptr] = YY_CURRENT_BUFFER;
	yy_switch_to_buffer(yy_create_buffer(input, YY_BUF_SIZE));
	++include_stack_ptr;
}

/*
 * Restores the file specific variables from the include stack.
 */
static void
pop_parser_state(void)
{
	--include_stack_ptr;
	parser->filename = zparser_stack[include_stack_ptr].filename;
	parser->line = zparser_stack[include_stack_ptr].line;
	parser->origin = zparser_stack[include_stack_ptr].origin;
	yy_delete_buffer(YY_CURRENT_BUFFER);
	yy_switch_to_buffer(include_stack[include_stack_ptr]);
}

static YY_BUFFER_STATE oldstate;
/* Start string scan */
void
parser_push_stringbuf(char* str)
{
	oldstate = YY_CURRENT_BUFFER;
	yy_switch_to_buffer(yy_scan_string(str));
}

void
parser_pop_stringbuf(void)
{
	yy_delete_buffer(YY_CURRENT_BUFFER);
	yy_switch_to_buffer(oldstate);
	oldstate = NULL;
}

	static int paren_open = 0;
	static enum lexer_state lexer_state = EXPECT_OWNER;
void
parser_flush(void)
{
	YY_FLUSH_BUFFER;
	paren_open = 0;
	lexer_state = EXPECT_OWNER;
}

#ifndef yy_set_bol /* compat definition, for flex 2.4.6 */
#define yy_set_bol(at_bol) \
	{ \
		if ( ! yy_current_buffer ) \
			yy_current_buffer = yy_create_buffer( yyin, YY_BUF_SIZE ); \
		yy_current_buffer->yy_ch_buf[0] = ((at_bol)?'\n':' '); \
	}
#endif
	
%}
%option noinput
%option nounput
%{
#ifndef YY_NO_UNPUT
#define YY_NO_UNPUT 1
#endif
#ifndef YY_NO_INPUT
#define YY_NO_INPUT 1
#endif
%}

SPACE   [ \t]
LETTER  [a-zA-Z]
NEWLINE [\n\r]
ZONESTR [^ \t\n\r();.\"\$]|\\.|\\\n
CHARSTR [^ \t\n\r();.]|\\.|\\\n
QUOTE   \"
DOLLAR  \$
COMMENT ;
DOT     \.
BIT	[^\]\n]|\\.
ANY     [^\"\n\\]|\\.

%x	incl bitlabel quotedstring

%%
{SPACE}*{COMMENT}.*	/* ignore */
^{DOLLAR}TTL            { lexer_state = PARSING_RDATA; return DOLLAR_TTL; }
^{DOLLAR}ORIGIN         { lexer_state = PARSING_RDATA; return DOLLAR_ORIGIN; }

	/*
	 * Handle $INCLUDE directives.  See
	 * http://dinosaur.compilertools.net/flex/flex_12.html#SEC12.
	 */
^{DOLLAR}INCLUDE        {
	BEGIN(incl);
	/* ignore case statement fallthrough on incl<EOF> flex rule */
}
<incl>\n		|
<incl><<EOF>>		{
	int error_occurred = parser->error_occurred;
	BEGIN(INITIAL);
	zc_error("missing file name in $INCLUDE directive");
	yy_set_bol(1); /* Set beginning of line, so "^" rules match.  */
	++parser->line;
	parser->error_occurred = error_occurred;
}
<incl>.+ 		{ 	
	char *tmp;
	domain_type *origin = parser->origin;
	int error_occurred = parser->error_occurred;
	
	BEGIN(INITIAL);
	if (include_stack_ptr >= MAXINCLUDES ) {
		zc_error("includes nested too deeply, skipped (>%d)",
			 MAXINCLUDES);
	} else {
		FILE *input;

		/* Remove trailing comment.  */
		tmp = strrchr(yytext, ';');
		if (tmp) {
			*tmp = '\0';
		}
		strip_string(yytext);
		
		/* Parse origin for include file.  */
		tmp = strrchr(yytext, ' ');
		if (!tmp) {
			tmp = strrchr(yytext, '\t');
		}
		if (tmp) {
			const dname_type *dname;
			
			/* split the original yytext */
			*tmp = '\0';
			strip_string(yytext);

			dname = dname_parse(parser->region, tmp + 1);
			if (!dname) {
				zc_error("incorrect include origin '%s'",
					 tmp + 1);
			} else if (*(tmp + strlen(tmp + 1)) != '.') {
				zc_error("$INCLUDE directive requires absolute domain name");
			} else {
				origin = domain_table_insert(
					parser->db->domains, dname);
			}
		}
		
		if (strlen(yytext) == 0) {
			zc_error("missing file name in $INCLUDE directive");
		} else if (!(input = fopen(yytext, "r"))) {
			zc_error("cannot open include file '%s': %s",
				 yytext, strerror(errno));
		} else {
			/* Initialize parser for include file.  */
			char *filename = region_strdup(parser->region, yytext);
			push_parser_state(input); /* Destroys yytext.  */
			parser->filename = filename;
			parser->line = 1;
			parser->origin = origin;
			lexer_state = EXPECT_OWNER;
		}
	}

	parser->error_occurred = error_occurred;
}
<INITIAL><<EOF>>	{
	yy_set_bol(1); /* Set beginning of line, so "^" rules match.  */
	if (include_stack_ptr == 0) {
		yyterminate();
	} else {
		fclose(yyin);
		pop_parser_state();
	}
}
^{DOLLAR}{LETTER}+	{ zc_warning("Unknown directive: %s", yytext); }
{DOT}	{
	LEXOUT((". "));
	return parse_token('.', yytext, &lexer_state);
}
@	{
	LEXOUT(("@ "));
	return parse_token('@', yytext, &lexer_state);
}
\\#	{
	LEXOUT(("\\# "));
	return parse_token(URR, yytext, &lexer_state);
}
{NEWLINE}	{
	++parser->line;
	if (!paren_open) { 
		lexer_state = EXPECT_OWNER;
		LEXOUT(("NL\n"));
		return NL;
	} else {
		LEXOUT(("SP "));
		return SP;
	}
}
\(	{
	if (paren_open) {
		zc_error("nested parentheses");
		yyterminate();
	}
	LEXOUT(("( "));
	paren_open = 1;
	return SP;
}
\)	{
	if (!paren_open) {
		zc_error("closing parentheses without opening parentheses");
		yyterminate();
	}
	LEXOUT((") "));
	paren_open = 0;
	return SP;
}
{SPACE}+	{
	if (!paren_open && lexer_state == EXPECT_OWNER) {
		lexer_state = PARSING_TTL_CLASS_TYPE;
		LEXOUT(("PREV "));
		return PREV;
	}
	if (lexer_state == PARSING_OWNER) {
		lexer_state = PARSING_TTL_CLASS_TYPE;
	}
	LEXOUT(("SP "));
	return SP;
}

	/* Bitlabels.  Strip leading and ending brackets.  */
\\\[			{ BEGIN(bitlabel); }
<bitlabel><<EOF>>	{
	zc_error("EOF inside bitlabel");
	BEGIN(INITIAL);
	yyrestart(yyin); /* this is so that lex does not give an internal err */
	yyterminate();
}
<bitlabel>{BIT}*	{ yymore(); }
<bitlabel>\n		{ ++parser->line; yymore(); }
<bitlabel>\]		{
	BEGIN(INITIAL);
	yytext[yyleng - 1] = '\0';
	return parse_token(BITLAB, yytext, &lexer_state);
}

	/* Quoted strings.  Strip leading and ending quotes.  */
{QUOTE}			{ BEGIN(quotedstring); LEXOUT(("\" ")); }
<quotedstring><<EOF>> 	{
	zc_error("EOF inside quoted string");
	BEGIN(INITIAL);
	yyrestart(yyin); /* this is so that lex does not give an internal err */
	yyterminate();
}
<quotedstring>{ANY}*	{ LEXOUT(("STR ")); yymore(); }
<quotedstring>\n 	{ ++parser->line; yymore(); }
<quotedstring>{QUOTE} {
	LEXOUT(("\" "));
	BEGIN(INITIAL);
	yytext[yyleng - 1] = '\0';
	return parse_token(STR, yytext, &lexer_state);
}

{ZONESTR}({CHARSTR})* {
	/* Any allowed word.  */
	return parse_token(STR, yytext, &lexer_state);
}
. {
	zc_error("unknown character '%c' (\\%03d) seen - is this a zonefile?",
		 (int) yytext[0], (int) yytext[0]);
}
%%

/*
 * Analyze "word" to see if it matches an RR type, possibly by using
 * the "TYPExxx" notation.  If it matches, the corresponding token is
 * returned and the TYPE parameter is set to the RR type value.
 */
static int
rrtype_to_token(const char *word, uint16_t *type) 
{
	uint16_t t = rrtype_from_string(word);
	if (t != 0) {
		rrtype_descriptor_type *entry = rrtype_descriptor_by_type(t);
		*type = t;
		return entry->token;
	}

	return 0;
}


/*
 * Remove \DDD constructs from the input. See RFC 1035, section 5.1.
 */
static size_t
zoctet(char *text) 
{
	/*
	 * s follows the string, p lags behind and rebuilds the new
	 * string
	 */
	char *s;
	char *p;
	
	for (s = p = text; *s; ++s, ++p) {
		assert(p <= s);
		if (s[0] != '\\') {
			/* Ordinary character.  */
			*p = *s;
		} else if (isdigit((unsigned char)s[1]) && isdigit((unsigned char)s[2]) && isdigit((unsigned char)s[3])) {
			/* \DDD escape.  */
			int val = (hexdigit_to_int(s[1]) * 100 +
				   hexdigit_to_int(s[2]) * 10 +
				   hexdigit_to_int(s[3]));
			if (0 <= val && val <= 255) {
				s += 3;
				*p = val;
			} else {
				zc_warning("text escape \\DDD overflow");
				*p = *++s;
			}
		} else if (s[1] != '\0') {
			/* \X where X is any character, keep X.  */
			*p = *++s;
		} else {
			/* Trailing backslash, ignore it.  */
			zc_warning("trailing backslash ignored");
			--p;
		}
	}
	*p = '\0';
	return p - text;
}

static int
parse_token(int token, char *yytext, enum lexer_state *lexer_state)
{
	size_t len;
	char *str;

	if (*lexer_state == EXPECT_OWNER) {
		*lexer_state = PARSING_OWNER;
	} else if (*lexer_state == PARSING_TTL_CLASS_TYPE) {
		const char *t;
		int token;
		uint16_t rrclass;
		
		/* type */
		token = rrtype_to_token(yytext, &yylval.type);
		if (token != 0) {
			*lexer_state = PARSING_RDATA;
			LEXOUT(("%d[%s] ", token, yytext));
			return token;
		}

		/* class */
		rrclass = rrclass_from_string(yytext);
		if (rrclass != 0) {
			yylval.klass = rrclass;
			LEXOUT(("CLASS "));
			return T_RRCLASS;
		}

		/* ttl */
		yylval.ttl = strtottl(yytext, &t);
		if (*t == '\0') {
			LEXOUT(("TTL "));
			return T_TTL;
		}
	}

	str = region_strdup(parser->rr_region, yytext);
	len = zoctet(str);

	yylval.data.str = str;
	yylval.data.len = len;
	
	LEXOUT(("%d[%s] ", token, yytext));
	return token;
}