platform_external_dtc/convert-dtsv0-lexer.l

236 lines
4.3 KiB
Text
Raw Normal View History

/* SPDX-License-Identifier: GPL-2.0-or-later */
/*
* (C) Copyright David Gibson <dwg@au1.ibm.com>, IBM Corporation. 2005, 2008.
*/
%option noyywrap nounput noinput never-interactive
%x BYTESTRING
%x PROPNODENAME
PROPNODECHAR [a-zA-Z0-9,._+*#?@-]
PATHCHAR ({PROPNODECHAR}|[/])
LABEL [a-zA-Z_][a-zA-Z0-9_]*
dtc: Clean up lexing of include files Currently we scan the /include/ directive as two tokens, the "/include/" keyword itself, then the string giving the file name to include. We use a special scanner state to keep the two linked together, and use the scanner state stack to keep track of the original state while we're parsing the two /include/ tokens. This does mean that we need to enable the 'stack' option in flex, which results in a not-easily-suppressed warning from the flex boilerplate code. This is mildly irritating. However, this two-token scanning of the /include/ directive also has some extremely strange edge cases, because there are a variety of tokens recognized in all scanner states, including INCLUDE. For example the following strange dts file: /include/ /dts-v1/; / { /* ... */ }; Will be processed successfully with the /include/ being effectively ignored: the '/dts-v1/' and ';' are recognized even in INCLUDE state, then the ';' transitions us to PROPNODENAME state, throwing away INCLUDE, and the previous state is never popped off the stack. Or for another example this construct: foo /include/ = "somefile.dts" will be parsed as though it were: foo = /include/ "somefile.dts" Again, the '=' is scanned without leaving INCLUDE state, then the next string triggers the include logic. And finally, we use a different regexp for the string with the included filename than the normal string regexpt, which is also potentially weird. This patch, therefore, cleans up the lexical handling of the /include/ directive. Instead of the INCLUDE state, we instead scan the whole include directive, both keyword and filename as a single token. This does mean a bit more complexity in extracting the filename out of yytext, but I think it's worth it to avoid the strageness described above. It also means it's no longer possible to put a comment between the /include/ and the filename, but I'm really not very worried about breaking files using such a strange construct.
2008-06-26 09:08:57 +02:00
STRING \"([^\\"]|\\.)*\"
WS [[:space:]]
COMMENT "/*"([^*]|\*+[^*/])*\*+"/"
LINECOMMENT "//".*\n
GAP ({WS}|{COMMENT}|{LINECOMMENT})*
%{
2008-06-26 03:03:49 +02:00
#include <string.h>
#include <stdlib.h>
#include <stdarg.h>
2008-06-26 03:03:49 +02:00
#include <errno.h>
#include <assert.h>
#include <fnmatch.h>
2008-06-26 03:03:49 +02:00
#include "srcpos.h"
#include "util.h"
static int v1_tagged; /* = 0 */
static int cbase = 16;
static int saw_hyphen; /* = 0 */
static unsigned long long last_val;
static char *last_name; /* = NULL */
static const struct {
const char *pattern;
int obase, width;
} guess_table[] = {
{ "*-frequency", 10, 0 },
{ "num-*", 10, 0 },
{ "#*-cells", 10, 0 },
{ "*cache-line-size", 10, 0 },
{ "*cache-block-size", 10, 0 },
{ "*cache-size", 10, 0 },
{ "*cache-sets", 10, 0 },
{ "cell-index", 10, 0 },
{ "bank-width", 10, 0 },
{ "*-fifo-size", 10, 0 },
{ "*-frame-size", 10, 0 },
{ "*-channel", 10, 0 },
{ "current-speed", 10, 0 },
{ "phy-map", 16, 8 },
{ "dcr-reg", 16, 3 },
{ "reg", 16, 8 },
{ "ranges", 16, 8},
};
%}
%%
dtc: Clean up lexing of include files Currently we scan the /include/ directive as two tokens, the "/include/" keyword itself, then the string giving the file name to include. We use a special scanner state to keep the two linked together, and use the scanner state stack to keep track of the original state while we're parsing the two /include/ tokens. This does mean that we need to enable the 'stack' option in flex, which results in a not-easily-suppressed warning from the flex boilerplate code. This is mildly irritating. However, this two-token scanning of the /include/ directive also has some extremely strange edge cases, because there are a variety of tokens recognized in all scanner states, including INCLUDE. For example the following strange dts file: /include/ /dts-v1/; / { /* ... */ }; Will be processed successfully with the /include/ being effectively ignored: the '/dts-v1/' and ';' are recognized even in INCLUDE state, then the ';' transitions us to PROPNODENAME state, throwing away INCLUDE, and the previous state is never popped off the stack. Or for another example this construct: foo /include/ = "somefile.dts" will be parsed as though it were: foo = /include/ "somefile.dts" Again, the '=' is scanned without leaving INCLUDE state, then the next string triggers the include logic. And finally, we use a different regexp for the string with the included filename than the normal string regexpt, which is also potentially weird. This patch, therefore, cleans up the lexical handling of the /include/ directive. Instead of the INCLUDE state, we instead scan the whole include directive, both keyword and filename as a single token. This does mean a bit more complexity in extracting the filename out of yytext, but I think it's worth it to avoid the strageness described above. It also means it's no longer possible to put a comment between the /include/ and the filename, but I'm really not very worried about breaking files using such a strange construct.
2008-06-26 09:08:57 +02:00
<*>"/include/"{GAP}{STRING} ECHO;
<*>\"([^\\"]|\\.)*\" ECHO;
<*>"/dts-v1/" {
die("Input dts file is already version 1\n");
}
<*>"/memreserve/" {
if (!v1_tagged) {
fprintf(yyout, "/dts-v1/;\n\n");
v1_tagged = 1;
}
ECHO;
BEGIN(INITIAL);
}
<*>{LABEL}: ECHO;
<INITIAL>[bodh]# {
if (*yytext == 'b')
cbase = 2;
else if (*yytext == 'o')
cbase = 8;
else if (*yytext == 'd')
cbase = 10;
else
cbase = 16;
}
<INITIAL>[0-9a-fA-F]+ {
unsigned long long val;
int obase = 16, width = 0;
unsigned int i;
val = strtoull(yytext, NULL, cbase);
if (saw_hyphen)
val = val - last_val + 1;
if (last_name) {
for (i = 0; i < ARRAY_SIZE(guess_table); i++)
if (fnmatch(guess_table[i].pattern,
last_name, 0) == 0) {
obase = guess_table[i].obase;
width = guess_table[i].width;
}
} else {
obase = 16;
width = 16;
}
if (cbase != 16)
obase = cbase;
switch (obase) {
case 2:
case 16:
fprintf(yyout, "0x%0*llx", width, val);
break;
case 8:
fprintf(yyout, "0%0*llo", width, val);
break;
case 10:
fprintf(yyout, "%*llu", width, val);
break;
}
cbase = 16;
last_val = val;
saw_hyphen = 0;
}
\&{LABEL} ECHO;
"&{/"{PATHCHAR}+\} ECHO;
<INITIAL>"&/"{PATHCHAR}+ fprintf(yyout, "&{/%s}", yytext + 2);
<BYTESTRING>[0-9a-fA-F]{2} ECHO;
<BYTESTRING>"]" {
ECHO;
BEGIN(INITIAL);
}
<PROPNODENAME>{PROPNODECHAR}+ {
ECHO;
last_name = xstrdup(yytext);
BEGIN(INITIAL);
}
dtc: Clean up lexing of include files Currently we scan the /include/ directive as two tokens, the "/include/" keyword itself, then the string giving the file name to include. We use a special scanner state to keep the two linked together, and use the scanner state stack to keep track of the original state while we're parsing the two /include/ tokens. This does mean that we need to enable the 'stack' option in flex, which results in a not-easily-suppressed warning from the flex boilerplate code. This is mildly irritating. However, this two-token scanning of the /include/ directive also has some extremely strange edge cases, because there are a variety of tokens recognized in all scanner states, including INCLUDE. For example the following strange dts file: /include/ /dts-v1/; / { /* ... */ }; Will be processed successfully with the /include/ being effectively ignored: the '/dts-v1/' and ';' are recognized even in INCLUDE state, then the ';' transitions us to PROPNODENAME state, throwing away INCLUDE, and the previous state is never popped off the stack. Or for another example this construct: foo /include/ = "somefile.dts" will be parsed as though it were: foo = /include/ "somefile.dts" Again, the '=' is scanned without leaving INCLUDE state, then the next string triggers the include logic. And finally, we use a different regexp for the string with the included filename than the normal string regexpt, which is also potentially weird. This patch, therefore, cleans up the lexical handling of the /include/ directive. Instead of the INCLUDE state, we instead scan the whole include directive, both keyword and filename as a single token. This does mean a bit more complexity in extracting the filename out of yytext, but I think it's worth it to avoid the strageness described above. It also means it's no longer possible to put a comment between the /include/ and the filename, but I'm really not very worried about breaking files using such a strange construct.
2008-06-26 09:08:57 +02:00
<*>{GAP} ECHO;
<*>- { /* Hack to convert old style memreserves */
saw_hyphen = 1;
fprintf(yyout, " ");
}
<*>. {
if (!v1_tagged) {
fprintf(yyout, "/dts-v1/;\n\n");
v1_tagged = 1;
}
ECHO;
if (yytext[0] == '[') {
BEGIN(BYTESTRING);
}
if ((yytext[0] == '{')
|| (yytext[0] == ';')) {
BEGIN(PROPNODENAME);
}
}
%%
/* Usage related data. */
static const char usage_synopsis[] = "convert-dtsv0 [options] <v0 dts file>...";
static const char usage_short_opts[] = "" USAGE_COMMON_SHORT_OPTS;
static struct option const usage_long_opts[] = {
USAGE_COMMON_LONG_OPTS
};
static const char * const usage_opts_help[] = {
USAGE_COMMON_OPTS_HELP
};
static void convert_file(const char *fname)
{
const char suffix[] = "v1";
int len = strlen(fname);
char *newname;
newname = xmalloc(len + sizeof(suffix));
memcpy(newname, fname, len);
memcpy(newname + len, suffix, sizeof(suffix));
yyin = fopen(fname, "r");
if (!yyin)
die("Couldn't open input file %s: %s\n",
fname, strerror(errno));
yyout = fopen(newname, "w");
if (!yyout)
die("Couldn't open output file %s: %s\n",
newname, strerror(errno));
while(yylex())
;
free(newname);
}
int main(int argc, char *argv[])
{
int opt;
int i;
while ((opt = util_getopt_long()) != EOF) {
switch (opt) {
case_USAGE_COMMON_FLAGS
}
}
if (argc < 2)
usage("missing filename");
for (i = 1; i < argc; i++) {
fprintf(stderr, "Converting %s from dts v0 to dts v1\n", argv[i]);
convert_file(argv[i]);
}
exit(0);
}