This is the mail archive of the
gdb-patches@sources.redhat.com
mailing list for the GDB project.
[cplus] Fix up a few performance issues
- From: Daniel Jacobowitz <drow at mvista dot com>
- To: gdb-patches at sources dot redhat dot com
- Date: Fri, 19 Dec 2003 13:55:33 -0500
- Subject: [cplus] Fix up a few performance issues
A little time with a profiler does wonders. There's one more hot-spot
(which this has in common with c-exp.y, but c-exp.y is not
performance-critical, and this may be).
Right now I'm about six times slower than the demangler. I hope to cut it
to three times. I can't really time compared to the old demangler, since
these input files crash it...
--
Daniel Jacobowitz
MontaVista Software Debian GNU/Linux Developer
2003-12-19 Daniel Jacobowitz <drow@mvista.com>
* cp-names.y (%union): Remove typed_val_float.
(INT, FLOAT): Change to type comp.
(CONSTRUCTION_VTABLE): Change to untyped.
(demangler_special): Use D_COMP_CONSTRUCTION_VTABLE.
(array_indicator, exp): Update for new types of INT and FLOAT.
Handle D_COMP_LITERAL_NEG.
(parse_number): Simplify. Use D_COMP_LITERAL_NEG.
(HANDLE_SPECIAL): Define.
(tokentab_big): Remove.
(yylex): Remove unused variables. Optimize calls to strncmp
and handling of multi-word tokens.
Index: cp-names.y
===================================================================
RCS file: /cvs/src/src/gdb/Attic/cp-names.y,v
retrieving revision 1.1.2.8
diff -u -p -r1.1.2.8 cp-names.y
--- cp-names.y 19 Dec 2003 16:44:33 -0000 1.1.2.8
+++ cp-names.y 19 Dec 2003 18:31:37 -0000
@@ -187,9 +187,6 @@ void yyerror (char *);
LONGEST val;
struct d_comp *type;
} typed_val_int;
- struct {
- DOUBLEST dval;
- } typed_val_float;
const char *opname;
}
@@ -221,8 +218,8 @@ static int parse_number (char *, int, in
%type <lval> sign size int_type
-%token <typed_val_int> INT
-%token <typed_val_float> FLOAT
+%token <comp> INT
+%token <comp> FLOAT
%token <comp> NAME
%type <comp> name
@@ -245,8 +242,8 @@ static int parse_number (char *, int, in
%token FALSEKEYWORD
/* Non-C++ things we get from the demangler. */
-%token <lval> DEMANGLER_SPECIAL CONSTRUCTION_VTABLE
-%token CONSTRUCTION_IN
+%token <lval> DEMANGLER_SPECIAL
+%token CONSTRUCTION_VTABLE CONSTRUCTION_IN
%token <typed_val_int> GLOBAL
%{
@@ -336,7 +333,7 @@ demangler_special
d_left ($$) = $2;
d_right ($$) = NULL; }
| CONSTRUCTION_VTABLE start CONSTRUCTION_IN start
- { $$ = d_make_comp (di, $1, $2, $4); }
+ { $$ = d_make_comp (di, D_COMP_CONSTRUCTION_VTABLE, $2, $4); }
| GLOBAL
{ $$ = d_make_empty (di, $1.val);
d_left ($$) = $1.type;
@@ -672,14 +669,8 @@ array_indicator : '[' ']'
d_left ($$) = NULL;
}
| '[' INT ']'
- { struct d_comp *i;
- /* FIXME: Blatant memory leak. */
- char *buf = malloc (24);
- sprintf (buf, "%d", (int) $2.val);
- i = d_make_name (di, buf, strlen (buf));
- i = d_make_comp (di, D_COMP_LITERAL, $2.type, i);
- $$ = d_make_empty (di, D_COMP_ARRAY_TYPE);
- d_left ($$) = i;
+ { $$ = d_make_empty (di, D_COMP_ARRAY_TYPE);
+ d_left ($$) = $2;
}
/* Details of this approach inspired by the G++ < 3.4 parser. */
@@ -849,7 +840,8 @@ exp : '~' exp %prec UNARY
its type. */
exp : '(' type ')' exp %prec UNARY
- { if ($4->type == D_COMP_LITERAL)
+ { if ($4->type == D_COMP_LITERAL
+ || $4->type == D_COMP_LITERAL_NEG)
{
$$ = $4;
d_left ($4) = $2;
@@ -986,26 +978,10 @@ exp : exp '?' exp ':' exp %prec '?'
;
exp : INT
- { struct d_comp *i;
- /* FIXME: Blatant memory leak. */
- char *buf = malloc (24);
- sprintf (buf, "%d", (int) $1.val);
- i = d_make_name (di, buf, strlen (buf));
- $$ = d_make_comp (di, D_COMP_LITERAL, $1.type, i);
- }
;
/* Not generally allowed. */
exp : FLOAT
- { struct d_comp *i;
- /* FIXME: Blatant memory leak. */
- char *buf = malloc (24);
- sprintf (buf, "%f", (double) $1.dval);
- i = d_make_name (di, buf, strlen (buf));
- $$ = d_make_comp (di, D_COMP_LITERAL,
- d_make_builtin_type (di, &d_builtin_types['d' - 'a']),
- i);
- }
;
exp : SIZEOF '(' type ')' %prec UNARY
@@ -1172,228 +1148,107 @@ parse_number (p, len, parsed_float, puti
int parsed_float;
YYSTYPE *putithere;
{
- /* FIXME: Shouldn't these be unsigned? We don't deal with negative values
- here, and we do kind of silly things like cast to unsigned. */
- LONGEST n = 0;
- LONGEST prevn = 0;
- ULONGEST un;
-
- int i = 0;
- int c;
- int base = 10;
int unsigned_p = 0;
- int negative_p = 0;
/* Number of "L" suffixes encountered. */
int long_p = 0;
- /* We have found a "L" or "U" suffix. */
- int found_suffix = 0;
-
- ULONGEST high_bit;
struct d_comp *signed_type;
struct d_comp *unsigned_type;
+ struct d_comp *type, *name;
+ enum d_comp_type literal_type;
+
+ if (p[0] == '-')
+ {
+ literal_type = D_COMP_LITERAL_NEG;
+ p++;
+ len--;
+ }
+ else
+ literal_type = D_COMP_LITERAL;
if (parsed_float)
{
/* It's a float since it contains a point or an exponent. */
char c;
- int num = 0; /* number of tokens scanned by scanf */
- char saved_char = p[len];
- p[len] = 0; /* null-terminate the token */
- if (sizeof (putithere->typed_val_float.dval) <= sizeof (float))
- num = sscanf (p, "%g%c", (float *) &putithere->typed_val_float.dval,&c);
- else if (sizeof (putithere->typed_val_float.dval) <= sizeof (double))
- num = sscanf (p, "%lg%c", (double *) &putithere->typed_val_float.dval,&c);
- else
- {
-#ifdef SCANF_HAS_LONG_DOUBLE
- num = sscanf (p, "%Lg%c", &putithere->typed_val_float.dval,&c);
-#else
- /* Scan it into a double, then assign it to the long double.
- This at least wins with values representable in the range
- of doubles. */
- double temp;
- num = sscanf (p, "%lg%c", &temp,&c);
- putithere->typed_val_float.dval = temp;
-#endif
- }
- p[len] = saved_char; /* restore the input stream */
- if (num != 1) /* check scanf found ONLY a float ... */
- return ERROR;
+ /* The GDB lexer checks the result of scanf at this point. Not doing
+ this leaves our error checking slightly weaker but only for invalid
+ data. */
+
/* See if it has `f' or `l' suffix (float or long double). */
c = TOLOWER (p[len - 1]);
-#if 0
if (c == 'f')
- putithere->typed_val_float.type = builtin_type_float;
+ {
+ len--;
+ type = d_make_builtin_type (di, &d_builtin_types ['f' - 'a']);
+ }
else if (c == 'l')
- putithere->typed_val_float.type = builtin_type_long_double;
- else if (isdigit (c) || c == '.')
- putithere->typed_val_float.type = builtin_type_double;
+ {
+ len--;
+ type = d_make_builtin_type (di, &d_builtin_types ['e' - 'a']);
+ }
+ else if (ISDIGIT (c) || c == '.')
+ type = d_make_builtin_type (di, &d_builtin_types ['d' - 'a']);
else
return ERROR;
-#endif
+
+ name = d_make_name (di, p, len);
+ putithere->comp = d_make_comp (di, literal_type, type, name);
return FLOAT;
}
- if (p[0] == '-')
- {
- negative_p = 1;
- p++;
- len--;
- }
- else
- negative_p = 0;
+ /* This treats 0x1 and 1 as different literals. We also do not
+ automatically generate unsigned types. */
- /* Handle base-switching prefixes 0x, 0t, 0d, 0 */
- if (p[0] == '0')
- switch (p[1])
- {
- case 'x':
- case 'X':
- if (len >= 3)
- {
- p += 2;
- base = 16;
- len -= 2;
- }
- break;
-
- case 't':
- case 'T':
- case 'd':
- case 'D':
- if (len >= 3)
- {
- p += 2;
- base = 10;
- len -= 2;
- }
- break;
-
- default:
- base = 8;
- break;
- }
-
- while (len-- > 0)
+ long_p = 0;
+ unsigned_p = 0;
+ while (len > 0)
{
- c = *p++;
- if (c >= 'A' && c <= 'Z')
- c += 'a' - 'A';
- if (c != 'l' && c != 'u')
- n *= base;
- if (c >= '0' && c <= '9')
+ if (p[len - 1] == 'l' || p[len - 1] == 'L')
{
- if (found_suffix)
- return ERROR;
- n += i = c - '0';
+ len--;
+ long_p++;
+ continue;
}
- else
+ if (p[len - 1] == 'u' || p[len - 1] == 'U')
{
- if (base > 10 && c >= 'a' && c <= 'f')
- {
- if (found_suffix)
- return ERROR;
- n += i = c - 'a' + 10;
- }
- else if (c == 'l')
- {
- ++long_p;
- found_suffix = 1;
- }
- else if (c == 'u')
- {
- unsigned_p = 1;
- found_suffix = 1;
- }
- else
- return ERROR; /* Char not a digit */
- }
- if (i >= base)
- return ERROR; /* Invalid digit in this base */
-
- /* Portably test for overflow (only works for nonzero values, so make
- a second check for zero). FIXME: Can't we just make n and prevn
- unsigned and avoid this? */
- if (c != 'l' && c != 'u' && (prevn >= n) && n != 0)
- unsigned_p = 1; /* Try something unsigned */
-
- /* Portably test for unsigned overflow.
- FIXME: This check is wrong; for example it doesn't find overflow
- on 0x123456789 when LONGEST is 32 bits. */
- if (c != 'l' && c != 'u' && n != 0)
- {
- if ((unsigned_p && (ULONGEST) prevn >= (ULONGEST) n))
- error ("Numeric constant too large.");
+ len--;
+ unsigned_p++;
+ continue;
}
- prevn = n;
+ break;
}
- /* An integer constant is an int, a long, or a long long. An L
- suffix forces it to be long; an LL suffix forces it to be long
- long. If not forced to a larger size, it gets the first type of
- the above that it fits in. To figure out whether it fits, we
- shift it right and see whether anything remains. Note that we
- can't shift sizeof (LONGEST) * HOST_CHAR_BIT bits or more in one
- operation, because many compilers will warn about such a shift
- (which always produces a zero result). Sometimes TARGET_INT_BIT
- or TARGET_LONG_BIT will be that big, sometimes not. To deal with
- the case where it is we just always shift the value more than
- once, with fewer bits each time. */
-
- un = (ULONGEST)n >> 2;
- if (long_p == 0
- && (un >> (TARGET_INT_BIT - 2)) == 0)
+ if (long_p == 0)
{
- high_bit = ((ULONGEST)1) << (TARGET_INT_BIT-1);
-
- /* A large decimal (not hex or octal) constant (between INT_MAX
- and UINT_MAX) is a long or unsigned long, according to ANSI,
- never an unsigned int, but this code treats it as unsigned
- int. This probably should be fixed. GCC gives a warning on
- such constants. */
-
unsigned_type = d_int_type ('j' - 'a');
signed_type = d_int_type ('i' - 'a');
}
- else if (long_p <= 1
- && (un >> (TARGET_LONG_BIT - 2)) == 0)
+ else if (long_p == 1)
{
- high_bit = ((ULONGEST)1) << (TARGET_LONG_BIT-1);
unsigned_type = d_int_type ('m' - 'a');
signed_type = d_int_type ('l' - 'a');
}
else
{
- int shift;
- if (sizeof (ULONGEST) * HOST_CHAR_BIT < TARGET_LONG_LONG_BIT)
- /* A long long does not fit in a LONGEST. */
- shift = (sizeof (ULONGEST) * HOST_CHAR_BIT - 1);
- else
- shift = (TARGET_LONG_LONG_BIT - 1);
- high_bit = (ULONGEST) 1 << shift;
unsigned_type = d_int_type ('x' - 'a');
signed_type = d_int_type ('y' - 'a');
}
- /* FIXME: overflow, et cetera. Dumb this whole function down a notch. */
- if (negative_p)
- n = -n;
-
- putithere->typed_val_int.val = n;
-
/* If the high bit of the worked out type is set then this number
has to be unsigned. */
- if (unsigned_p || (n & high_bit))
- putithere->typed_val_int.type = unsigned_type;
+ if (unsigned_p)
+ type = unsigned_type;
else
- putithere->typed_val_int.type = signed_type;
+ type = signed_type;
+
+ name = d_make_name (di, p, len);
+ putithere->comp = d_make_comp (di, literal_type, type, name);
return INT;
}
@@ -1588,26 +1443,18 @@ static const struct token tokentab2[] =
{">=", GEQ, BINOP_END}
};
-static const struct token tokentab_big[] = {
- { "global constructors keyed to", 28, GLOBAL_CONSTRUCTORS},
- { "global destructors keyed to", 27, GLOBAL_DESTRUCTORS},
- { "construction vtable for", 23, D_COMP_CONSTRUCTION_VTABLE},
- { "VTT for", 7, D_COMP_VTT},
- { "vtable for", 10, D_COMP_VTABLE},
- { "typeinfo for", 12, D_COMP_TYPEINFO},
- { "typeinfo fn for", 15, D_COMP_TYPEINFO_FN},
- { "typeinfo name for", 17, D_COMP_TYPEINFO_NAME},
- { "non-virtual thunk to", 20, D_COMP_THUNK},
- { "virtual thunk to", 16, D_COMP_VIRTUAL_THUNK},
- { "covariant return thunk to", 25, D_COMP_COVARIANT_THUNK},
- { "guard variable for", 18, D_COMP_GUARD},
- { "reference temporary for", 23, D_COMP_REFTEMP}
-};
+#define HANDLE_SPECIAL(string, len, comp) \
+ if (strncmp (tokstart, string, len) == 0) \
+ { \
+ lexptr = tokstart + len; \
+ yylval.lval = comp; \
+ return DEMANGLER_SPECIAL; \
+ }
/* Read one token, getting characters through lexptr. */
static int
-yylex ()
+yylex (void)
{
int c;
int namelen;
@@ -1617,9 +1464,6 @@ yylex ()
int tempbufindex;
static char *tempbuf;
static int tempbufsize;
- struct symbol * sym_class = NULL;
- char * token_string = NULL;
- int class_prefix = 0;
int unquoted_expr;
retry:
@@ -1630,7 +1474,9 @@ yylex ()
tokstart = lexptr;
/* See if it is a special token of length 3. */
for (i = 0; i < sizeof tokentab3 / sizeof tokentab3[0]; i++)
- if (strncmp (tokstart, tokentab3[i].operator, 3) == 0)
+ if (tokstart[0] == tokentab3[i].operator[0]
+ && tokstart[1] == tokentab3[i].operator[1]
+ && tokstart[2] == tokentab3[i].operator[2])
{
lexptr += 3;
yylval.opname = tokentab3[i].operator;
@@ -1639,28 +1485,14 @@ yylex ()
/* See if it is a special token of length 2. */
for (i = 0; i < sizeof tokentab2 / sizeof tokentab2[0]; i++)
- if (strncmp (tokstart, tokentab2[i].operator, 2) == 0)
+ if (tokstart[0] == tokentab2[i].operator[0]
+ && tokstart[1] == tokentab2[i].operator[1])
{
lexptr += 2;
yylval.opname = tokentab2[i].operator;
return tokentab2[i].token;
}
- /* For construction vtables. This is kind of hokey. */
- if (strncmp (tokstart, "-in-", 4) == 0)
- {
- lexptr += 4;
- return CONSTRUCTION_IN;
- }
-
- if (strncmp (tokstart, "(anonymous namespace)", 21) == 0)
- {
- lexptr += 21;
- yylval.comp = d_make_name (di, "(anonymous namespace)",
- sizeof "(anonymous namespace)" - 1);
- return NAME;
- }
-
switch (c = *tokstart)
{
case 0:
@@ -1702,6 +1534,15 @@ yylex ()
return INT;
case '(':
+ if (strncmp (tokstart, "(anonymous namespace)", 21) == 0)
+ {
+ lexptr += 21;
+ yylval.comp = d_make_name (di, "(anonymous namespace)",
+ sizeof "(anonymous namespace)" - 1);
+ return NAME;
+ }
+ /* FALL THROUGH */
+
case ')':
case ',':
lexptr++;
@@ -1720,6 +1561,13 @@ yylex ()
/* FALL THRU into number case. */
case '-':
+ /* For construction vtables. This is kind of hokey. */
+ if (strncmp (tokstart, "-in-", 4) == 0)
+ {
+ lexptr += 4;
+ return CONSTRUCTION_IN;
+ }
+
if (lexptr[1] < '0' || lexptr[1] > '9')
{
lexptr++;
@@ -1893,27 +1741,6 @@ yylex ()
/* We must have come across a bad character (e.g. ';'). */
error ("Invalid character '%c' in expression.", c);
- /* Tokens containing spaces. */
- for (i = 0; i < sizeof tokentab_big / sizeof tokentab_big[0]; i++)
- if (strncmp (tokstart, tokentab_big[i].operator, tokentab_big[i].token) == 0)
- {
- lexptr += tokentab_big[i].token;
- yylval.lval = tokentab_big[i].opcode;
- if (yylval.lval == D_COMP_CONSTRUCTION_VTABLE)
- return CONSTRUCTION_VTABLE;
- else if (yylval.lval == GLOBAL_CONSTRUCTORS
- || yylval.lval == GLOBAL_DESTRUCTORS)
- {
- /* Skip the trailing space, find the end of the symbol. */
- char *p = symbol_end (lexptr + 1);
- yylval.typed_val_int.val = yylval.lval;
- yylval.typed_val_int.type = d_make_name (di, lexptr + 1, p - (lexptr + 1));
- lexptr = p;
- return GLOBAL;
- }
- return DEMANGLER_SPECIAL;
- }
-
/* It's a name. See how long it is. */
namelen = 0;
for (c = tokstart[namelen];
@@ -1923,8 +1750,6 @@ yylex ()
lexptr += namelen;
- tryname:
-
/* Catch specific keywords. Should be done with a data structure. */
switch (namelen)
{
@@ -1940,7 +1765,21 @@ yylex ()
if (strncmp (tokstart, "static_cast", 11) == 0)
return STATIC_CAST;
break;
+ case 10:
+ if (strncmp (tokstart, "construction vtable for", 23) == 0)
+ {
+ lexptr = tokstart + 23;
+ return CONSTRUCTION_VTABLE;
+ }
+ break;
+ case 9:
+ HANDLE_SPECIAL ("covariant return thunk to", 25, D_COMP_COVARIANT_THUNK);
+ HANDLE_SPECIAL ("reference temporary for", 23, D_COMP_REFTEMP);
+ break;
case 8:
+ HANDLE_SPECIAL ("typeinfo for", 12, D_COMP_TYPEINFO);
+ HANDLE_SPECIAL ("typeinfo fn for", 15, D_COMP_TYPEINFO_FN);
+ HANDLE_SPECIAL ("typeinfo name for", 17, D_COMP_TYPEINFO_NAME);
if (strncmp (tokstart, "operator", 8) == 0)
return OPERATOR;
if (strncmp (tokstart, "restrict", 8) == 0)
@@ -1953,10 +1792,35 @@ yylex ()
return VOLATILE_KEYWORD;
break;
case 7:
+ HANDLE_SPECIAL ("virtual thunk to", 16, D_COMP_VIRTUAL_THUNK);
if (strncmp (tokstart, "wchar_t", 7) == 0)
return WCHAR_T;
break;
case 6:
+ if (strncmp (tokstart, "global constructors keyed to", 28) == 0)
+ {
+ char *p;
+ lexptr = tokstart + 28;
+ yylval.typed_val_int.val = GLOBAL_CONSTRUCTORS;
+ /* Skip the trailing space, find the end of the symbol. */
+ p = symbol_end (lexptr + 1);
+ yylval.typed_val_int.type = d_make_name (di, lexptr + 1, p - (lexptr + 1));
+ lexptr = p;
+ return GLOBAL;
+ }
+ if (strncmp (tokstart, "global destructors keyed to", 27) == 0)
+ {
+ char *p;
+ lexptr = tokstart + 27;
+ yylval.typed_val_int.val = GLOBAL_DESTRUCTORS;
+ /* Skip the trailing space, find the end of the symbol. */
+ p = symbol_end (lexptr + 1);
+ yylval.typed_val_int.type = d_make_name (di, lexptr + 1, p - (lexptr + 1));
+ lexptr = p;
+ return GLOBAL;
+ }
+
+ HANDLE_SPECIAL ("vtable for", 10, D_COMP_VTABLE);
if (strncmp (tokstart, "delete", 6) == 0)
return DELETE;
if (strncmp (tokstart, "struct", 6) == 0)
@@ -1969,6 +1833,7 @@ yylex ()
return DOUBLE_KEYWORD;
break;
case 5:
+ HANDLE_SPECIAL ("guard variable for", 18, D_COMP_GUARD);
if (strncmp (tokstart, "false", 5) == 0)
return FALSEKEYWORD;
if (strncmp (tokstart, "class", 5) == 0)
@@ -1997,6 +1862,8 @@ yylex ()
return TRUEKEYWORD;
break;
case 3:
+ HANDLE_SPECIAL ("VTT for", 7, D_COMP_VTT);
+ HANDLE_SPECIAL ("non-virtual thunk to", 20, D_COMP_THUNK);
if (strncmp (tokstart, "new", 3) == 0)
return NEW;
if (strncmp (tokstart, "int", 3) == 0)