mirror of
https://github.com/netwide-assembler/nasm.git
synced 2024-11-21 03:14:19 +08:00
preproc.c: Restore concat rules on context local variables
This is a backport of commits8dcbbd7af0
575d4289c9
Signed-off-by: Cyrill Gorcunov <gorcunov@gmail.com>
This commit is contained in:
parent
0457bcbf2e
commit
6cdc900d8d
164
preproc.c
164
preproc.c
@ -212,6 +212,13 @@ enum pp_token_type {
|
||||
TOK_MAX = INT_MAX /* Keep compiler from reducing the range */
|
||||
};
|
||||
|
||||
#define PP_CONCAT_MASK(x) (1 << (x))
|
||||
|
||||
struct tokseq_match {
|
||||
int mask_head;
|
||||
int mask_tail;
|
||||
};
|
||||
|
||||
struct Token {
|
||||
Token *next;
|
||||
char *text;
|
||||
@ -3642,12 +3649,14 @@ static int find_cc(Token * t)
|
||||
return i;
|
||||
}
|
||||
|
||||
static bool paste_tokens(Token **head, bool handle_paste_tokens)
|
||||
static bool paste_tokens(Token **head, const struct tokseq_match *m,
|
||||
int mnum, bool handle_paste_tokens)
|
||||
{
|
||||
Token **tail, *t, *tt;
|
||||
Token **paste_head;
|
||||
bool did_paste = false;
|
||||
char *tmp;
|
||||
int i;
|
||||
|
||||
/* Now handle token pasting... */
|
||||
paste_head = NULL;
|
||||
@ -3663,51 +3672,6 @@ static bool paste_tokens(Token **head, bool handle_paste_tokens)
|
||||
tail = &t->next;
|
||||
}
|
||||
break;
|
||||
case TOK_ID:
|
||||
case TOK_PREPROC_ID:
|
||||
case TOK_NUMBER:
|
||||
case TOK_FLOAT:
|
||||
{
|
||||
size_t len = 0;
|
||||
char *tmp, *p;
|
||||
|
||||
while (tt && (tt->type == TOK_ID || tt->type == TOK_PREPROC_ID ||
|
||||
tt->type == TOK_NUMBER || tt->type == TOK_FLOAT ||
|
||||
tt->type == TOK_OTHER)) {
|
||||
len += strlen(tt->text);
|
||||
tt = tt->next;
|
||||
}
|
||||
|
||||
/*
|
||||
* Now tt points to the first token after
|
||||
* the potential paste area...
|
||||
*/
|
||||
if (tt != t->next) {
|
||||
/* We have at least two tokens... */
|
||||
len += strlen(t->text);
|
||||
p = tmp = nasm_malloc(len+1);
|
||||
|
||||
while (t != tt) {
|
||||
strcpy(p, t->text);
|
||||
p = strchr(p, '\0');
|
||||
t = delete_Token(t);
|
||||
}
|
||||
|
||||
t = *tail = tokenize(tmp);
|
||||
nasm_free(tmp);
|
||||
|
||||
while (t->next) {
|
||||
tail = &t->next;
|
||||
t = t->next;
|
||||
}
|
||||
t->next = tt; /* Attach the remaining token chain */
|
||||
|
||||
did_paste = true;
|
||||
}
|
||||
paste_head = tail;
|
||||
tail = &t->next;
|
||||
break;
|
||||
}
|
||||
case TOK_PASTE: /* %+ */
|
||||
if (handle_paste_tokens) {
|
||||
/* Zap %+ and whitespace tokens to the right */
|
||||
@ -3721,7 +3685,6 @@ static bool paste_tokens(Token **head, bool handle_paste_tokens)
|
||||
tt = t->next;
|
||||
while (tok_type_(tt, TOK_WHITESPACE))
|
||||
tt = t->next = delete_Token(tt);
|
||||
|
||||
if (tt) {
|
||||
tmp = nasm_strcat(t->text, tt->text);
|
||||
delete_Token(t);
|
||||
@ -3741,9 +3704,55 @@ static bool paste_tokens(Token **head, bool handle_paste_tokens)
|
||||
}
|
||||
/* else fall through */
|
||||
default:
|
||||
/*
|
||||
* Concatenation of tokens might look nontrivial
|
||||
* but in real it's pretty simple -- the caller
|
||||
* prepares the masks of token types to be concatenated
|
||||
* and we simply find matched sequences and slip
|
||||
* them together
|
||||
*/
|
||||
for (i = 0; i < mnum; i++) {
|
||||
if (PP_CONCAT_MASK(t->type) & m[i].mask_head) {
|
||||
size_t len = 0;
|
||||
char *tmp, *p;
|
||||
|
||||
while (tt && (PP_CONCAT_MASK(tt->type) & m[i].mask_tail)) {
|
||||
len += strlen(tt->text);
|
||||
tt = tt->next;
|
||||
}
|
||||
|
||||
/*
|
||||
* Now tt points to the first token after
|
||||
* the potential paste area...
|
||||
*/
|
||||
if (tt != t->next) {
|
||||
/* We have at least two tokens... */
|
||||
len += strlen(t->text);
|
||||
p = tmp = nasm_malloc(len+1);
|
||||
while (t != tt) {
|
||||
strcpy(p, t->text);
|
||||
p = strchr(p, '\0');
|
||||
t = delete_Token(t);
|
||||
}
|
||||
t = *tail = tokenize(tmp);
|
||||
nasm_free(tmp);
|
||||
while (t->next) {
|
||||
tail = &t->next;
|
||||
t = t->next;
|
||||
}
|
||||
t->next = tt; /* Attach the remaining token chain */
|
||||
did_paste = true;
|
||||
}
|
||||
paste_head = tail;
|
||||
tail = &t->next;
|
||||
break;
|
||||
}
|
||||
}
|
||||
if (i >= mnum) { /* no match */
|
||||
tail = &t->next;
|
||||
if (!tok_type_(t->next, TOK_WHITESPACE))
|
||||
paste_head = tail;
|
||||
}
|
||||
break;
|
||||
}
|
||||
}
|
||||
@ -3977,35 +3986,6 @@ static Token *expand_mmac_params(Token * tline)
|
||||
}
|
||||
delete_Token(t);
|
||||
changed = true;
|
||||
} else if (tline->type == TOK_PREPROC_ID &&
|
||||
tline->text[0] == '%' &&
|
||||
tline->text[1] == '$' &&
|
||||
!tok_type_(tline->next, TOK_WHITESPACE) &&
|
||||
(tok_type_(tline->next, TOK_ID) ||
|
||||
tok_type_(tline->next, TOK_PREPROC_ID) ||
|
||||
tok_type_(tline->next, TOK_NUMBER) ||
|
||||
tok_type_(tline->next, TOK_OTHER) ||
|
||||
tok_type_(tline->next, TOK_FLOAT))) {
|
||||
/*
|
||||
* In a sake of backward compatibility we allow
|
||||
* to expand local single macro that early before
|
||||
* pasting token code have place
|
||||
*
|
||||
* NOTE: that new code MUST use %+ macro to obtain
|
||||
* same result
|
||||
*/
|
||||
t = tline;
|
||||
tline = tline->next;
|
||||
tt = tokenize(t->text);
|
||||
tt = expand_smacro(tt);
|
||||
*tail = tt;
|
||||
while (tt) {
|
||||
tt->a.mac = NULL;
|
||||
tail = &tt->next;
|
||||
tt = tt->next;
|
||||
}
|
||||
delete_Token(t);
|
||||
changed = true;
|
||||
} else {
|
||||
t = *tail = tline;
|
||||
tline = tline->next;
|
||||
@ -4015,8 +3995,23 @@ static Token *expand_mmac_params(Token * tline)
|
||||
}
|
||||
*tail = NULL;
|
||||
|
||||
if (changed)
|
||||
paste_tokens(&thead, false);
|
||||
if (changed) {
|
||||
const struct tokseq_match t[] = {
|
||||
{
|
||||
PP_CONCAT_MASK(TOK_ID) |
|
||||
PP_CONCAT_MASK(TOK_FLOAT), /* head */
|
||||
PP_CONCAT_MASK(TOK_ID) |
|
||||
PP_CONCAT_MASK(TOK_NUMBER) |
|
||||
PP_CONCAT_MASK(TOK_FLOAT) |
|
||||
PP_CONCAT_MASK(TOK_OTHER) /* tail */
|
||||
},
|
||||
{
|
||||
PP_CONCAT_MASK(TOK_NUMBER), /* head */
|
||||
PP_CONCAT_MASK(TOK_NUMBER) /* tail */
|
||||
}
|
||||
};
|
||||
paste_tokens(&thead, t, ARRAY_SIZE(t), false);
|
||||
}
|
||||
|
||||
return thead;
|
||||
}
|
||||
@ -4327,7 +4322,17 @@ again:
|
||||
* Also we look for %+ tokens and concatenate the tokens before and after
|
||||
* them (without white spaces in between).
|
||||
*/
|
||||
if (expanded && paste_tokens(&thead, true)) {
|
||||
if (expanded) {
|
||||
const struct tokseq_match t[] = {
|
||||
{
|
||||
PP_CONCAT_MASK(TOK_ID) |
|
||||
PP_CONCAT_MASK(TOK_PREPROC_ID), /* head */
|
||||
PP_CONCAT_MASK(TOK_ID) |
|
||||
PP_CONCAT_MASK(TOK_PREPROC_ID) |
|
||||
PP_CONCAT_MASK(TOK_NUMBER) /* tail */
|
||||
}
|
||||
};
|
||||
if (paste_tokens(&thead, t, ARRAY_SIZE(t), true)) {
|
||||
/*
|
||||
* If we concatenated something, *and* we had previously expanded
|
||||
* an actual macro, scan the lines again for macros...
|
||||
@ -4336,6 +4341,7 @@ again:
|
||||
expanded = false;
|
||||
goto again;
|
||||
}
|
||||
}
|
||||
|
||||
err:
|
||||
if (org_tline) {
|
||||
|
Loading…
Reference in New Issue
Block a user