blob: 8e3ef096bbe390ba4d89764810ac1a56c71492e6 [file] [log] [blame]
/* CPP Library - lexical analysis.
Copyright (C) 2000-2021 Free Software Foundation, Inc.
Contributed by Per Bothner, 1994-95.
Based on CCCP program by Paul Rubin, June 1986
Adapted to ANSI C, Richard Stallman, Jan 1987
Broken out to separate file, Zack Weinberg, Mar 2000
This program is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
Free Software Foundation; either version 3, or (at your option) any
later version.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; see the file COPYING3. If not see
<http://www.gnu.org/licenses/>. */
#include "config.h"
#include "system.h"
#include "cpplib.h"
#include "internal.h"
enum spell_type
{
SPELL_OPERATOR = 0,
SPELL_IDENT,
SPELL_LITERAL,
SPELL_NONE
};
struct token_spelling
{
enum spell_type category;
const unsigned char *name;
};
static const unsigned char *const digraph_spellings[] =
{ UC"%:", UC"%:%:", UC"<:", UC":>", UC"<%", UC"%>" };
#define OP(e, s) { SPELL_OPERATOR, UC s },
#define TK(e, s) { SPELL_ ## s, UC #e },
static const struct token_spelling token_spellings[N_TTYPES] = { TTYPE_TABLE };
#undef OP
#undef TK
#define TOKEN_SPELL(token) (token_spellings[(token)->type].category)
#define TOKEN_NAME(token) (token_spellings[(token)->type].name)
static void add_line_note (cpp_buffer *, const uchar *, unsigned int);
static int skip_line_comment (cpp_reader *);
static void skip_whitespace (cpp_reader *, cppchar_t);
static void lex_string (cpp_reader *, cpp_token *, const uchar *);
static void save_comment (cpp_reader *, cpp_token *, const uchar *, cppchar_t);
static void store_comment (cpp_reader *, cpp_token *);
static void create_literal (cpp_reader *, cpp_token *, const uchar *,
unsigned int, enum cpp_ttype);
static bool warn_in_comment (cpp_reader *, _cpp_line_note *);
static int name_p (cpp_reader *, const cpp_string *);
static tokenrun *next_tokenrun (tokenrun *);
static _cpp_buff *new_buff (size_t);
/* Utility routine:
Compares, the token TOKEN to the NUL-terminated string STRING.
TOKEN must be a CPP_NAME. Returns 1 for equal, 0 for unequal. */
int
cpp_ideq (const cpp_token *token, const char *string)
{
if (token->type != CPP_NAME)
return 0;
return !ustrcmp (NODE_NAME (token->val.node.node), (const uchar *) string);
}
/* Record a note TYPE at byte POS into the current cleaned logical
line. */
static void
add_line_note (cpp_buffer *buffer, const uchar *pos, unsigned int type)
{
if (buffer->notes_used == buffer->notes_cap)
{
buffer->notes_cap = buffer->notes_cap * 2 + 200;
buffer->notes = XRESIZEVEC (_cpp_line_note, buffer->notes,
buffer->notes_cap);
}
buffer->notes[buffer->notes_used].pos = pos;
buffer->notes[buffer->notes_used].type = type;
buffer->notes_used++;
}
/* Fast path to find line special characters using optimized character
scanning algorithms. Anything complicated falls back to the slow
path below. Since this loop is very hot it's worth doing these kinds
of optimizations.
One of the paths through the ifdefs should provide
const uchar *search_line_fast (const uchar *s, const uchar *end);
Between S and END, search for \n, \r, \\, ?. Return a pointer to
the found character.
Note that the last character of the buffer is *always* a newline,
as forced by _cpp_convert_input. This fact can be used to avoid
explicitly looking for the end of the buffer. */
/* Configure gives us an ifdef test. */
#ifndef WORDS_BIGENDIAN
#define WORDS_BIGENDIAN 0
#endif
/* We'd like the largest integer that fits into a register. There's nothing
in <stdint.h> that gives us that. For most hosts this is unsigned long,
but MS decided on an LLP64 model. Thankfully when building with GCC we
can get the "real" word size. */
#ifdef __GNUC__
typedef unsigned int word_type __attribute__((__mode__(__word__)));
#else
typedef unsigned long word_type;
#endif
/* The code below is only expecting sizes 4 or 8.
Die at compile-time if this expectation is violated. */
typedef char check_word_type_size
[(sizeof(word_type) == 8 || sizeof(word_type) == 4) * 2 - 1];
/* Return X with the first N bytes forced to values that won't match one
of the interesting characters. Note that NUL is not interesting. */
static inline word_type
acc_char_mask_misalign (word_type val, unsigned int n)
{
word_type mask = -1;
if (WORDS_BIGENDIAN)
mask >>= n * 8;
else
mask <<= n * 8;
return val & mask;
}
/* Return X replicated to all byte positions within WORD_TYPE. */
static inline word_type
acc_char_replicate (uchar x)
{
word_type ret;
ret = (x << 24) | (x << 16) | (x << 8) | x;
if (sizeof(word_type) == 8)
ret = (ret << 16 << 16) | ret;
return ret;
}
/* Return non-zero if some byte of VAL is (probably) C. */
static inline word_type
acc_char_cmp (word_type val, word_type c)
{
#if defined(__GNUC__) && defined(__alpha__)
/* We can get exact results using a compare-bytes instruction.
Get (val == c) via (0 >= (val ^ c)). */
return __builtin_alpha_cmpbge (0, val ^ c);
#else
word_type magic = 0x7efefefeU;
if (sizeof(word_type) == 8)
magic = (magic << 16 << 16) | 0xfefefefeU;
magic |= 1;
val ^= c;
return ((val + magic) ^ ~val) & ~magic;
#endif
}
/* Given the result of acc_char_cmp is non-zero, return the index of
the found character. If this was a false positive, return -1. */
static inline int
acc_char_index (word_type cmp ATTRIBUTE_UNUSED,
word_type val ATTRIBUTE_UNUSED)
{
#if defined(__GNUC__) && defined(__alpha__) && !WORDS_BIGENDIAN
/* The cmpbge instruction sets *bits* of the result corresponding to
matches in the bytes with no false positives. */
return __builtin_ctzl (cmp);
#else
unsigned int i;
/* ??? It would be nice to force unrolling here,
and have all of these constants folded. */
for (i = 0; i < sizeof(word_type); ++i)
{
uchar c;
if (WORDS_BIGENDIAN)
c = (val >> (sizeof(word_type) - i - 1) * 8) & 0xff;
else
c = (val >> i * 8) & 0xff;
if (c == '\n' || c == '\r' || c == '\\' || c == '?')
return i;
}
return -1;
#endif
}
/* A version of the fast scanner using bit fiddling techniques.
For 32-bit words, one would normally perform 16 comparisons and
16 branches. With this algorithm one performs 24 arithmetic
operations and one branch. Whether this is faster with a 32-bit
word size is going to be somewhat system dependent.
For 64-bit words, we eliminate twice the number of comparisons
and branches without increasing the number of arithmetic operations.
It's almost certainly going to be a win with 64-bit word size. */
static const uchar * search_line_acc_char (const uchar *, const uchar *)
ATTRIBUTE_UNUSED;
static const uchar *
search_line_acc_char (const uchar *s, const uchar *end ATTRIBUTE_UNUSED)
{
const word_type repl_nl = acc_char_replicate ('\n');
const word_type repl_cr = acc_char_replicate ('\r');
const word_type repl_bs = acc_char_replicate ('\\');
const word_type repl_qm = acc_char_replicate ('?');
unsigned int misalign;
const word_type *p;
word_type val, t;
/* Align the buffer. Mask out any bytes from before the beginning. */
p = (word_type *)((uintptr_t)s & -sizeof(word_type));
val = *p;
misalign = (uintptr_t)s & (sizeof(word_type) - 1);
if (misalign)
val = acc_char_mask_misalign (val, misalign);
/* Main loop. */
while (1)
{
t = acc_char_cmp (val, repl_nl);
t |= acc_char_cmp (val, repl_cr);
t |= acc_char_cmp (val, repl_bs);
t |= acc_char_cmp (val, repl_qm);
if (__builtin_expect (t != 0, 0))
{
int i = acc_char_index (t, val);
if (i >= 0)
return (const uchar *)p + i;
}
val = *++p;
}
}
/* Disable on Solaris 2/x86 until the following problem can be properly
autoconfed:
The Solaris 10+ assembler tags objects with the instruction set
extensions used, so SSE4.2 executables cannot run on machines that
don't support that extension. */
#if (GCC_VERSION >= 4005) && (__GNUC__ >= 5 || !defined(__PIC__)) && (defined(__i386__) || defined(__x86_64__)) && !(defined(__sun__) && defined(__svr4__))
/* Replicated character data to be shared between implementations.
Recall that outside of a context with vector support we can't
define compatible vector types, therefore these are all defined
in terms of raw characters. */
static const char repl_chars[4][16] __attribute__((aligned(16))) = {
{ '\n', '\n', '\n', '\n', '\n', '\n', '\n', '\n',
'\n', '\n', '\n', '\n', '\n', '\n', '\n', '\n' },
{ '\r', '\r', '\r', '\r', '\r', '\r', '\r', '\r',
'\r', '\r', '\r', '\r', '\r', '\r', '\r', '\r' },
{ '\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\',
'\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\' },
{ '?', '?', '?', '?', '?', '?', '?', '?',
'?', '?', '?', '?', '?', '?', '?', '?' },
};
/* A version of the fast scanner using MMX vectorized byte compare insns.
This uses the PMOVMSKB instruction which was introduced with "MMX2",
which was packaged into SSE1; it is also present in the AMD MMX
extension. Mark the function as using "sse" so that we emit a real
"emms" instruction, rather than the 3dNOW "femms" instruction. */
static const uchar *
#ifndef __SSE__
__attribute__((__target__("sse")))
#endif
search_line_mmx (const uchar *s, const uchar *end ATTRIBUTE_UNUSED)
{
typedef char v8qi __attribute__ ((__vector_size__ (8)));
typedef int __m64 __attribute__ ((__vector_size__ (8), __may_alias__));
const v8qi repl_nl = *(const v8qi *)repl_chars[0];
const v8qi repl_cr = *(const v8qi *)repl_chars[1];
const v8qi repl_bs = *(const v8qi *)repl_chars[2];
const v8qi repl_qm = *(const v8qi *)repl_chars[3];
unsigned int misalign, found, mask;
const v8qi *p;
v8qi data, t, c;
/* Align the source pointer. While MMX doesn't generate unaligned data
faults, this allows us to safely scan to the end of the buffer without
reading beyond the end of the last page. */
misalign = (uintptr_t)s & 7;
p = (const v8qi *)((uintptr_t)s & -8);
data = *p;
/* Create a mask for the bytes that are valid within the first
16-byte block. The Idea here is that the AND with the mask
within the loop is "free", since we need some AND or TEST
insn in order to set the flags for the branch anyway. */
mask = -1u << misalign;
/* Main loop processing 8 bytes at a time. */
goto start;
do
{
data = *++p;
mask = -1;
start:
t = __builtin_ia32_pcmpeqb(data, repl_nl);
c = __builtin_ia32_pcmpeqb(data, repl_cr);
t = (v8qi) __builtin_ia32_por ((__m64)t, (__m64)c);
c = __builtin_ia32_pcmpeqb(data, repl_bs);
t = (v8qi) __builtin_ia32_por ((__m64)t, (__m64)c);
c = __builtin_ia32_pcmpeqb(data, repl_qm);
t = (v8qi) __builtin_ia32_por ((__m64)t, (__m64)c);
found = __builtin_ia32_pmovmskb (t);
found &= mask;
}
while (!found);
__builtin_ia32_emms ();
/* FOUND contains 1 in bits for which we matched a relevant
character. Conversion to the byte index is trivial. */
found = __builtin_ctz(found);
return (const uchar *)p + found;
}
/* A version of the fast scanner using SSE2 vectorized byte compare insns. */
static const uchar *
#ifndef __SSE2__
__attribute__((__target__("sse2")))
#endif
search_line_sse2 (const uchar *s, const uchar *end ATTRIBUTE_UNUSED)
{
typedef char v16qi __attribute__ ((__vector_size__ (16)));
const v16qi repl_nl = *(const v16qi *)repl_chars[0];
const v16qi repl_cr = *(const v16qi *)repl_chars[1];
const v16qi repl_bs = *(const v16qi *)repl_chars[2];
const v16qi repl_qm = *(const v16qi *)repl_chars[3];
unsigned int misalign, found, mask;
const v16qi *p;
v16qi data, t;
/* Align the source pointer. */
misalign = (uintptr_t)s & 15;
p = (const v16qi *)((uintptr_t)s & -16);
data = *p;
/* Create a mask for the bytes that are valid within the first
16-byte block. The Idea here is that the AND with the mask
within the loop is "free", since we need some AND or TEST
insn in order to set the flags for the branch anyway. */
mask = -1u << misalign;
/* Main loop processing 16 bytes at a time. */
goto start;
do
{
data = *++p;
mask = -1;
start:
t = data == repl_nl;
t |= data == repl_cr;
t |= data == repl_bs;
t |= data == repl_qm;
found = __builtin_ia32_pmovmskb128 (t);
found &= mask;
}
while (!found);
/* FOUND contains 1 in bits for which we matched a relevant
character. Conversion to the byte index is trivial. */
found = __builtin_ctz(found);
return (const uchar *)p + found;
}
#ifdef HAVE_SSE4
/* A version of the fast scanner using SSE 4.2 vectorized string insns. */
static const uchar *
#ifndef __SSE4_2__
__attribute__((__target__("sse4.2")))
#endif
search_line_sse42 (const uchar *s, const uchar *end)
{
typedef char v16qi __attribute__ ((__vector_size__ (16)));
static const v16qi search = { '\n', '\r', '?', '\\' };
uintptr_t si = (uintptr_t)s;
uintptr_t index;
/* Check for unaligned input. */
if (si & 15)
{
v16qi sv;
if (__builtin_expect (end - s < 16, 0)
&& __builtin_expect ((si & 0xfff) > 0xff0, 0))
{
/* There are less than 16 bytes left in the buffer, and less
than 16 bytes left on the page. Reading 16 bytes at this
point might generate a spurious page fault. Defer to the
SSE2 implementation, which already handles alignment. */
return search_line_sse2 (s, end);
}
/* ??? The builtin doesn't understand that the PCMPESTRI read from
memory need not be aligned. */
sv = __builtin_ia32_loaddqu ((const char *) s);
index = __builtin_ia32_pcmpestri128 (search, 4, sv, 16, 0);
if (__builtin_expect (index < 16, 0))
goto found;
/* Advance the pointer to an aligned address. We will re-scan a
few bytes, but we no longer need care for reading past the
end of a page, since we're guaranteed a match. */
s = (const uchar *)((si + 15) & -16);
}
/* Main loop, processing 16 bytes at a time. */
#ifdef __GCC_ASM_FLAG_OUTPUTS__
while (1)
{
char f;
/* By using inline assembly instead of the builtin,
we can use the result, as well as the flags set. */
__asm ("%vpcmpestri\t$0, %2, %3"
: "=c"(index), "=@ccc"(f)
: "m"(*s), "x"(search), "a"(4), "d"(16));
if (f)
break;
s += 16;
}
#else
s -= 16;
/* By doing the whole loop in inline assembly,
we can make proper use of the flags set. */
__asm ( ".balign 16\n"
"0: add $16, %1\n"
" %vpcmpestri\t$0, (%1), %2\n"
" jnc 0b"
: "=&c"(index), "+r"(s)
: "x"(search), "a"(4), "d"(16));
#endif
found:
return s + index;
}
#else
/* Work around out-dated assemblers without sse4 support. */
#define search_line_sse42 search_line_sse2
#endif
/* Check the CPU capabilities. */
#include "../gcc/config/i386/cpuid.h"
typedef const uchar * (*search_line_fast_type) (const uchar *, const uchar *);
static search_line_fast_type search_line_fast;
#define HAVE_init_vectorized_lexer 1
static inline void
init_vectorized_lexer (void)
{
unsigned dummy, ecx = 0, edx = 0;
search_line_fast_type impl = search_line_acc_char;
int minimum = 0;
#if defined(__SSE4_2__)
minimum = 3;
#elif defined(__SSE2__)
minimum = 2;
#elif defined(__SSE__)
minimum = 1;
#endif
if (minimum == 3)
impl = search_line_sse42;
else if (__get_cpuid (1, &dummy, &dummy, &ecx, &edx) || minimum == 2)
{
if (minimum == 3 || (ecx & bit_SSE4_2))
impl = search_line_sse42;
else if (minimum == 2 || (edx & bit_SSE2))
impl = search_line_sse2;
else if (minimum == 1 || (edx & bit_SSE))
impl = search_line_mmx;
}
else if (__get_cpuid (0x80000001, &dummy, &dummy, &dummy, &edx))
{
if (minimum == 1
|| (edx & (bit_MMXEXT | bit_CMOV)) == (bit_MMXEXT | bit_CMOV))
impl = search_line_mmx;
}
search_line_fast = impl;
}
#elif (GCC_VERSION >= 4005) && defined(_ARCH_PWR8) && defined(__ALTIVEC__)
/* A vection of the fast scanner using AltiVec vectorized byte compares
and VSX unaligned loads (when VSX is available). This is otherwise
the same as the AltiVec version. */
ATTRIBUTE_NO_SANITIZE_UNDEFINED
static const uchar *
search_line_fast (const uchar *s, const uchar *end ATTRIBUTE_UNUSED)
{
typedef __attribute__((altivec(vector))) unsigned char vc;
const vc repl_nl = {
'\n', '\n', '\n', '\n', '\n', '\n', '\n', '\n',
'\n', '\n', '\n', '\n', '\n', '\n', '\n', '\n'
};
const vc repl_cr = {
'\r', '\r', '\r', '\r', '\r', '\r', '\r', '\r',
'\r', '\r', '\r', '\r', '\r', '\r', '\r', '\r'
};
const vc repl_bs = {
'\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\',
'\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\'
};
const vc repl_qm = {
'?', '?', '?', '?', '?', '?', '?', '?',
'?', '?', '?', '?', '?', '?', '?', '?',
};
const vc zero = { 0 };
vc data, t;
/* Main loop processing 16 bytes at a time. */
do
{
vc m_nl, m_cr, m_bs, m_qm;
data = __builtin_vec_vsx_ld (0, s);
s += 16;
m_nl = (vc) __builtin_vec_cmpeq(data, repl_nl);
m_cr = (vc) __builtin_vec_cmpeq(data, repl_cr);
m_bs = (vc) __builtin_vec_cmpeq(data, repl_bs);
m_qm = (vc) __builtin_vec_cmpeq(data, repl_qm);
t = (m_nl | m_cr) | (m_bs | m_qm);
/* T now contains 0xff in bytes for which we matched one of the relevant
characters. We want to exit the loop if any byte in T is non-zero.
Below is the expansion of vec_any_ne(t, zero). */
}
while (!__builtin_vec_vcmpeq_p(/*__CR6_LT_REV*/3, t, zero));
/* Restore s to to point to the 16 bytes we just processed. */
s -= 16;
{
#define N (sizeof(vc) / sizeof(long))
union {
vc v;
/* Statically assert that N is 2 or 4. */
unsigned long l[(N == 2 || N == 4) ? N : -1];
} u;
unsigned long l, i = 0;
u.v = t;
/* Find the first word of T that is non-zero. */
switch (N)
{
case 4:
l = u.l[i++];
if (l != 0)
break;
s += sizeof(unsigned long);
l = u.l[i++];
if (l != 0)
break;
s += sizeof(unsigned long);
/* FALLTHRU */
case 2:
l = u.l[i++];
if (l != 0)
break;
s += sizeof(unsigned long);
l = u.l[i];
}
/* L now contains 0xff in bytes for which we matched one of the
relevant characters. We can find the byte index by finding
its bit index and dividing by 8. */
#ifdef __BIG_ENDIAN__
l = __builtin_clzl(l) >> 3;
#else
l = __builtin_ctzl(l) >> 3;
#endif
return s + l;
#undef N
}
}
#elif (GCC_VERSION >= 4005) && defined(__ALTIVEC__) && defined (__BIG_ENDIAN__)
/* A vection of the fast scanner using AltiVec vectorized byte compares.
This cannot be used for little endian because vec_lvsl/lvsr are
deprecated for little endian and the code won't work properly. */
/* ??? Unfortunately, attribute(target("altivec")) is not yet supported,
so we can't compile this function without -maltivec on the command line
(or implied by some other switch). */
static const uchar *
search_line_fast (const uchar *s, const uchar *end ATTRIBUTE_UNUSED)
{
typedef __attribute__((altivec(vector))) unsigned char vc;
const vc repl_nl = {
'\n', '\n', '\n', '\n', '\n', '\n', '\n', '\n',
'\n', '\n', '\n', '\n', '\n', '\n', '\n', '\n'
};
const vc repl_cr = {
'\r', '\r', '\r', '\r', '\r', '\r', '\r', '\r',
'\r', '\r', '\r', '\r', '\r', '\r', '\r', '\r'
};
const vc repl_bs = {
'\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\',
'\\', '\\', '\\', '\\', '\\', '\\', '\\', '\\'
};
const vc repl_qm = {
'?', '?', '?', '?', '?', '?', '?', '?',
'?', '?', '?', '?', '?', '?', '?', '?',
};
const vc ones = {
-1, -1, -1, -1, -1, -1, -1, -1,
-1, -1, -1, -1, -1, -1, -1, -1,
};
const vc zero = { 0 };
vc data, mask, t;
/* Altivec loads automatically mask addresses with -16. This lets us
issue the first load as early as possible. */
data = __builtin_vec_ld(0, (const vc *)s);
/* Discard bytes before the beginning of the buffer. Do this by
beginning with all ones and shifting in zeros according to the
mis-alignment. The LVSR instruction pulls the exact shift we
want from the address. */
mask = __builtin_vec_lvsr(0, s);
mask = __builtin_vec_perm(zero, ones, mask);
data &= mask;
/* While altivec loads mask addresses, we still need to align S so
that the offset we compute at the end is correct. */
s = (const uchar *)((uintptr_t)s & -16);
/* Main loop processing 16 bytes at a time. */
goto start;
do
{
vc m_nl, m_cr, m_bs, m_qm;
s += 16;
data = __builtin_vec_ld(0, (const vc *)s);
start:
m_nl = (vc) __builtin_vec_cmpeq(data, repl_nl);
m_cr = (vc) __builtin_vec_cmpeq(data, repl_cr);
m_bs = (vc) __builtin_vec_cmpeq(data, repl_bs);
m_qm = (vc) __builtin_vec_cmpeq(data, repl_qm);
t = (m_nl | m_cr) | (m_bs | m_qm);
/* T now contains 0xff in bytes for which we matched one of the relevant
characters. We want to exit the loop if any byte in T is non-zero.
Below is the expansion of vec_any_ne(t, zero). */
}
while (!__builtin_vec_vcmpeq_p(/*__CR6_LT_REV*/3, t, zero));
{
#define N (sizeof(vc) / sizeof(long))
union {
vc v;
/* Statically assert that N is 2 or 4. */
unsigned long l[(N == 2 || N == 4) ? N : -1];
} u;
unsigned long l, i = 0;
u.v = t;
/* Find the first word of T that is non-zero. */
switch (N)
{
case 4:
l = u.l[i++];
if (l != 0)
break;
s += sizeof(unsigned long);
l = u.l[i++];
if (l != 0)
break;
s += sizeof(unsigned long);
/* FALLTHROUGH */
case 2:
l = u.l[i++];
if (l != 0)
break;
s += sizeof(unsigned long);
l = u.l[i];
}
/* L now contains 0xff in bytes for which we matched one of the
relevant characters. We can find the byte index by finding
its bit index and dividing by 8. */
l = __builtin_clzl(l) >> 3;
return s + l;
#undef N
}
}
#elif defined (__ARM_NEON) && defined (__ARM_64BIT_STATE)
#include "arm_neon.h"
/* This doesn't have to be the exact page size, but no system may use
a size smaller than this. ARMv8 requires a minimum page size of
4k. The impact of being conservative here is a small number of
cases will take the slightly slower entry path into the main
loop. */
#define AARCH64_MIN_PAGE_SIZE 4096
static const uchar *
search_line_fast (const uchar *s, const uchar *end ATTRIBUTE_UNUSED)
{
const uint8x16_t repl_nl = vdupq_n_u8 ('\n');
const uint8x16_t repl_cr = vdupq_n_u8 ('\r');
const uint8x16_t repl_bs = vdupq_n_u8 ('\\');
const uint8x16_t repl_qm = vdupq_n_u8 ('?');
const uint8x16_t xmask = (uint8x16_t) vdupq_n_u64 (0x8040201008040201ULL);
#ifdef __ARM_BIG_ENDIAN
const int16x8_t shift = {8, 8, 8, 8, 0, 0, 0, 0};
#else
const int16x8_t shift = {0, 0, 0, 0, 8, 8, 8, 8};
#endif
unsigned int found;
const uint8_t *p;
uint8x16_t data;
uint8x16_t t;
uint16x8_t m;
uint8x16_t u, v, w;
/* Align the source pointer. */
p = (const uint8_t *)((uintptr_t)s & -16);
/* Assuming random string start positions, with a 4k page size we'll take
the slow path about 0.37% of the time. */
if (__builtin_expect ((AARCH64_MIN_PAGE_SIZE
- (((uintptr_t) s) & (AARCH64_MIN_PAGE_SIZE - 1)))
< 16, 0))
{
/* Slow path: the string starts near a possible page boundary. */
uint32_t misalign, mask;
misalign = (uintptr_t)s & 15;
mask = (-1u << misalign) & 0xffff;
data = vld1q_u8 (p);
t = vceqq_u8 (data, repl_nl);
u = vceqq_u8 (data, repl_cr);
v = vorrq_u8 (t, vceqq_u8 (data, repl_bs));
w = vorrq_u8 (u, vceqq_u8 (data, repl_qm));
t = vorrq_u8 (v, w);
t = vandq_u8 (t, xmask);
m = vpaddlq_u8 (t);
m = vshlq_u16 (m, shift);
found = vaddvq_u16 (m);
found &= mask;
if (found)
return (const uchar*)p + __builtin_ctz (found);
}
else
{
data = vld1q_u8 ((const uint8_t *) s);
t = vceqq_u8 (data, repl_nl);
u = vceqq_u8 (data, repl_cr);
v = vorrq_u8 (t, vceqq_u8 (data, repl_bs));
w = vorrq_u8 (u, vceqq_u8 (data, repl_qm));
t = vorrq_u8 (v, w);
if (__builtin_expect (vpaddd_u64 ((uint64x2_t)t) != 0, 0))
goto done;
}
do
{
p += 16;
data = vld1q_u8 (p);
t = vceqq_u8 (data, repl_nl);
u = vceqq_u8 (data, repl_cr);
v = vorrq_u8 (t, vceqq_u8 (data, repl_bs));
w = vorrq_u8 (u, vceqq_u8 (data, repl_qm));
t = vorrq_u8 (v, w);
} while (!vpaddd_u64 ((uint64x2_t)t));
done:
/* Now that we've found the terminating substring, work out precisely where
we need to stop. */
t = vandq_u8 (t, xmask);
m = vpaddlq_u8 (t);
m = vshlq_u16 (m, shift);
found = vaddvq_u16 (m);
return (((((uintptr_t) p) < (uintptr_t) s) ? s : (const uchar *)p)
+ __builtin_ctz (found));
}
#elif defined (__ARM_NEON)
#include "arm_neon.h"
static const uchar *
search_line_fast (const uchar *s, const uchar *end ATTRIBUTE_UNUSED)
{
const uint8x16_t repl_nl = vdupq_n_u8 ('\n');
const uint8x16_t repl_cr = vdupq_n_u8 ('\r');
const uint8x16_t repl_bs = vdupq_n_u8 ('\\');
const uint8x16_t repl_qm = vdupq_n_u8 ('?');
const uint8x16_t xmask = (uint8x16_t) vdupq_n_u64 (0x8040201008040201ULL);
unsigned int misalign, found, mask;
const uint8_t *p;
uint8x16_t data;
/* Align the source pointer. */
misalign = (uintptr_t)s & 15;
p = (const uint8_t *)((uintptr_t)s & -16);
data = vld1q_u8 (p);
/* Create a mask for the bytes that are valid within the first
16-byte block. The Idea here is that the AND with the mask
within the loop is "free", since we need some AND or TEST
insn in order to set the flags for the branch anyway. */
mask = (-1u << misalign) & 0xffff;
/* Main loop, processing 16 bytes at a time. */
goto start;
do
{
uint8x8_t l;
uint16x4_t m;
uint32x2_t n;
uint8x16_t t, u, v, w;
p += 16;
data = vld1q_u8 (p);
mask = 0xffff;
start:
t = vceqq_u8 (data, repl_nl);
u = vceqq_u8 (data, repl_cr);
v = vorrq_u8 (t, vceqq_u8 (data, repl_bs));
w = vorrq_u8 (u, vceqq_u8 (data, repl_qm));
t = vandq_u8 (vorrq_u8 (v, w), xmask);
l = vpadd_u8 (vget_low_u8 (t), vget_high_u8 (t));
m = vpaddl_u8 (l);
n = vpaddl_u16 (m);
found = vget_lane_u32 ((uint32x2_t) vorr_u64 ((uint64x1_t) n,
vshr_n_u64 ((uint64x1_t) n, 24)), 0);
found &= mask;
}
while (!found);
/* FOUND contains 1 in bits for which we matched a relevant
character. Conversion to the byte index is trivial. */
found = __builtin_ctz (found);
return (const uchar *)p + found;
}
#else
/* We only have one accelerated alternative. Use a direct call so that
we encourage inlining. */
#define search_line_fast search_line_acc_char
#endif
/* Initialize the lexer if needed. */
void
_cpp_init_lexer (void)
{
#ifdef HAVE_init_vectorized_lexer
init_vectorized_lexer ();
#endif
}
/* Returns with a logical line that contains no escaped newlines or
trigraphs. This is a time-critical inner loop. */
void
_cpp_clean_line (cpp_reader *pfile)
{
cpp_buffer *buffer;
const uchar *s;
uchar c, *d, *p;
buffer = pfile->buffer;
buffer->cur_note = buffer->notes_used = 0;
buffer->cur = buffer->line_base = buffer->next_line;
buffer->need_line = false;
s = buffer->next_line;
if (!buffer->from_stage3)
{
const uchar *pbackslash = NULL;
/* Fast path. This is the common case of an un-escaped line with
no trigraphs. The primary win here is by not writing any
data back to memory until we have to. */
while (1)
{
/* Perform an optimized search for \n, \r, \\, ?. */
s = search_line_fast (s, buffer->rlimit);
c = *s;
if (c == '\\')
{
/* Record the location of the backslash and continue. */
pbackslash = s++;
}
else if (__builtin_expect (c == '?', 0))
{
if (__builtin_expect (s[1] == '?', false)
&& _cpp_trigraph_map[s[2]])
{
/* Have a trigraph. We may or may not have to convert
it. Add a line note regardless, for -Wtrigraphs. */
add_line_note (buffer, s, s[2]);
if (CPP_OPTION (pfile, trigraphs))
{
/* We do, and that means we have to switch to the
slow path. */
d = (uchar *) s;
*d = _cpp_trigraph_map[s[2]];
s += 2;
goto slow_path;
}
}
/* Not a trigraph. Continue on fast-path. */
s++;
}
else
break;
}
/* This must be \r or \n. We're either done, or we'll be forced
to write back to the buffer and continue on the slow path. */
d = (uchar *) s;
if (__builtin_expect (s == buffer->rlimit, false))
goto done;
/* DOS line ending? */
if (__builtin_expect (c == '\r', false) && s[1] == '\n')
{
s++;
if (s == buffer->rlimit)
goto done;
}
if (__builtin_expect (pbackslash == NULL, true))
goto done;
/* Check for escaped newline. */
p = d;
while (is_nvspace (p[-1]))
p--;
if (p - 1 != pbackslash)
goto done;
/* Have an escaped newline; process it and proceed to
the slow path. */
add_line_note (buffer, p - 1, p != d ? ' ' : '\\');
d = p - 2;
buffer->next_line = p - 1;
slow_path:
while (1)
{
c = *++s;
*++d = c;
if (c == '\n' || c == '\r')
{
/* Handle DOS line endings. */
if (c == '\r' && s != buffer->rlimit && s[1] == '\n')
s++;
if (s == buffer->rlimit)
break;
/* Escaped? */
p = d;
while (p != buffer->next_line && is_nvspace (p[-1]))
p--;
if (p == buffer->next_line || p[-1] != '\\')
break;
add_line_note (buffer, p - 1, p != d ? ' ': '\\');
d = p - 2;
buffer->next_line = p - 1;
}
else if (c == '?' && s[1] == '?' && _cpp_trigraph_map[s[2]])
{
/* Add a note regardless, for the benefit of -Wtrigraphs. */
add_line_note (buffer, d, s[2]);
if (CPP_OPTION (pfile, trigraphs))
{
*d = _cpp_trigraph_map[s[2]];
s += 2;
}
}
}
}
else
{
while (*s != '\n' && *s != '\r')
s++;
d = (uchar *) s;
/* Handle DOS line endings. */
if (*s == '\r' && s + 1 != buffer->rlimit && s[1] == '\n')
s++;
}
done:
*d = '\n';
/* A sentinel note that should never be processed. */
add_line_note (buffer, d + 1, '\n');
buffer->next_line = s + 1;
}
/* Return true if the trigraph indicated by NOTE should be warned
about in a comment. */
static bool
warn_in_comment (cpp_reader *pfile, _cpp_line_note *note)
{
const uchar *p;
/* Within comments we don't warn about trigraphs, unless the
trigraph forms an escaped newline, as that may change
behavior. */
if (note->type != '/')
return false;
/* If -trigraphs, then this was an escaped newline iff the next note
is coincident. */
if (CPP_OPTION (pfile, trigraphs))
return note[1].pos == note->pos;
/* Otherwise, see if this forms an escaped newline. */
p = note->pos + 3;
while (is_nvspace (*p))
p++;
/* There might have been escaped newlines between the trigraph and the
newline we found. Hence the position test. */
return (*p == '\n' && p < note[1].pos);
}
/* Process the notes created by add_line_note as far as the current
location. */
void
_cpp_process_line_notes (cpp_reader *pfile, int in_comment)
{
cpp_buffer *buffer = pfile->buffer;
for (;;)
{
_cpp_line_note *note = &buffer->notes[buffer->cur_note];
unsigned int col;
if (note->pos > buffer->cur)
break;
buffer->cur_note++;
col = CPP_BUF_COLUMN (buffer, note->pos + 1);
if (note->type == '\\' || note->type == ' ')
{
if (note->type == ' ' && !in_comment)
cpp_error_with_line (pfile, CPP_DL_WARNING, pfile->line_table->highest_line, col,
"backslash and newline separated by space");
if (buffer->next_line > buffer->rlimit)
{
cpp_error_with_line (pfile, CPP_DL_PEDWARN, pfile->line_table->highest_line, col,
"backslash-newline at end of file");
/* Prevent "no newline at end of file" warning. */
buffer->next_line = buffer->rlimit;
}
buffer->line_base = note->pos;
CPP_INCREMENT_LINE (pfile, 0);
}
else if (_cpp_trigraph_map[note->type])
{
if (CPP_OPTION (pfile, warn_trigraphs)
&& (!in_comment || warn_in_comment (pfile, note)))
{
if (CPP_OPTION (pfile, trigraphs))
cpp_warning_with_line (pfile, CPP_W_TRIGRAPHS,
pfile->line_table->highest_line, col,
"trigraph ??%c converted to %c",
note->type,
(int) _cpp_trigraph_map[note->type]);
else
{
cpp_warning_with_line
(pfile, CPP_W_TRIGRAPHS,
pfile->line_table->highest_line, col,
"trigraph ??%c ignored, use -trigraphs to enable",
note->type);
}
}
}
else if (note->type == 0)
/* Already processed in lex_raw_string. */;
else
abort ();
}
}
/* Skip a C-style block comment. We find the end of the comment by
seeing if an asterisk is before every '/' we encounter. Returns
nonzero if comment terminated by EOF, zero otherwise.
Buffer->cur points to the initial asterisk of the comment. */
bool
_cpp_skip_block_comment (cpp_reader *pfile)
{
cpp_buffer *buffer = pfile->buffer;
const uchar *cur = buffer->cur;
uchar c;
cur++;
if (*cur == '/')
cur++;
for (;;)
{
/* People like decorating comments with '*', so check for '/'
instead for efficiency. */
c = *cur++;
if (c == '/')
{
if (cur[-2] == '*')
break;
/* Warn about potential nested comments, but not if the '/'
comes immediately before the true comment delimiter.
Don't bother to get it right across escaped newlines. */
if (CPP_OPTION (pfile, warn_comments)
&& cur[0] == '*' && cur[1] != '/')
{
buffer->cur = cur;
cpp_warning_with_line (pfile, CPP_W_COMMENTS,
pfile->line_table->highest_line,
CPP_BUF_COL (buffer),
"\"/*\" within comment");
}
}
else if (c == '\n')
{
unsigned int cols;
buffer->cur = cur - 1;
_cpp_process_line_notes (pfile, true);
if (buffer->next_line >= buffer->rlimit)
return true;
_cpp_clean_line (pfile);
cols = buffer->next_line - buffer->line_base;
CPP_INCREMENT_LINE (pfile, cols);
cur = buffer->cur;
}
}
buffer->cur = cur;
_cpp_process_line_notes (pfile, true);
return false;
}
/* Skip a C++ line comment, leaving buffer->cur pointing to the
terminating newline. Handles escaped newlines. Returns nonzero
if a multiline comment. */
static int
skip_line_comment (cpp_reader *pfile)
{
cpp_buffer *buffer = pfile->buffer;
location_t orig_line = pfile->line_table->highest_line;
while (*buffer->cur != '\n')
buffer->cur++;
_cpp_process_line_notes (pfile, true);
return orig_line != pfile->line_table->highest_line;
}
/* Skips whitespace, saving the next non-whitespace character. */
static void
skip_whitespace (cpp_reader *pfile, cppchar_t c)
{
cpp_buffer *buffer = pfile->buffer;
bool saw_NUL = false;
do
{
/* Horizontal space always OK. */
if (c == ' ' || c == '\t')
;
/* Just \f \v or \0 left. */
else if (c == '\0')
saw_NUL = true;
else if (pfile->state.in_directive && CPP_PEDANTIC (pfile))
cpp_error_with_line (pfile, CPP_DL_PEDWARN, pfile->line_table->highest_line,
CPP_BUF_COL (buffer),
"%s in preprocessing directive",
c == '\f' ? "form feed" : "vertical tab");
c = *buffer->cur++;
}
/* We only want non-vertical space, i.e. ' ' \t \f \v \0. */
while (is_nvspace (c));
if (saw_NUL)
cpp_error (pfile, CPP_DL_WARNING, "null character(s) ignored");
buffer->cur--;
}
/* See if the characters of a number token are valid in a name (no
'.', '+' or '-'). */
static int
name_p (cpp_reader *pfile, const cpp_string *string)
{
unsigned int i;
for (i = 0; i < string->len; i++)
if (!is_idchar (string->text[i]))
return 0;
return 1;
}
/* After parsing an identifier or other sequence, produce a warning about
sequences not in NFC/NFKC. */
static void
warn_about_normalization (cpp_reader *pfile,
const cpp_token *token,
const struct normalize_state *s)
{
if (CPP_OPTION (pfile, warn_normalize) < NORMALIZE_STATE_RESULT (s)
&& !pfile->state.skipping)
{
/* Make sure that the token is printed using UCNs, even
if we'd otherwise happily print UTF-8. */
unsigned char *buf = XNEWVEC (unsigned char, cpp_token_len (token));
size_t sz;
sz = cpp_spell_token (pfile, token, buf, false) - buf;
if (NORMALIZE_STATE_RESULT (s) == normalized_C)
cpp_warning_with_line (pfile, CPP_W_NORMALIZE, token->src_loc, 0,
"`%.*s' is not in NFKC", (int) sz, buf);
else if (CPP_OPTION (pfile, cxx23_identifiers))
cpp_pedwarning_with_line (pfile, CPP_W_NORMALIZE, token->src_loc, 0,
"`%.*s' is not in NFC", (int) sz, buf);
else
cpp_warning_with_line (pfile, CPP_W_NORMALIZE, token->src_loc, 0,
"`%.*s' is not in NFC", (int) sz, buf);
free (buf);
}
}
static const cppchar_t utf8_signifier = 0xC0;
/* Returns TRUE if the sequence starting at buffer->cur is valid in
an identifier. FIRST is TRUE if this starts an identifier. */
static bool
forms_identifier_p (cpp_reader *pfile, int first,
struct normalize_state *state)
{
cpp_buffer *buffer = pfile->buffer;
if (*buffer->cur == '$')
{
if (!CPP_OPTION (pfile, dollars_in_ident))
return false;
buffer->cur++;
if (CPP_OPTION (pfile, warn_dollars) && !pfile->state.skipping)
{
CPP_OPTION (pfile, warn_dollars) = 0;
cpp_error (pfile, CPP_DL_PEDWARN, "'$' in identifier or number");
}
return true;
}
/* Is this a syntactically valid UCN or a valid UTF-8 char? */
if (CPP_OPTION (pfile, extended_identifiers))
{
cppchar_t s;
if (*buffer->cur >= utf8_signifier)
{
if (_cpp_valid_utf8 (pfile, &buffer->cur, buffer->rlimit, 1 + !first,
state, &s))
return true;
}
else if (*buffer->cur == '\\'
&& (buffer->cur[1] == 'u' || buffer->cur[1] == 'U'))
{
buffer->cur += 2;
if (_cpp_valid_ucn (pfile, &buffer->cur, buffer->rlimit, 1 + !first,
state, &s, NULL, NULL))
return true;
buffer->cur -= 2;
}
}
return false;
}
/* Helper function to issue error about improper __VA_OPT__ use. */
static void
maybe_va_opt_error (cpp_reader *pfile)
{
if (CPP_PEDANTIC (pfile) && !CPP_OPTION (pfile, va_opt))
{
/* __VA_OPT__ should not be accepted at all, but allow it in
system headers. */
if (!_cpp_in_system_header (pfile))
cpp_error (pfile, CPP_DL_PEDWARN,
"__VA_OPT__ is not available until C++20");
}
else if (!pfile->state.va_args_ok)
{
/* __VA_OPT__ should only appear in the replacement list of a
variadic macro. */
cpp_error (pfile, CPP_DL_PEDWARN,
"__VA_OPT__ can only appear in the expansion"
" of a C++20 variadic macro");
}
}
/* Helper function to get the cpp_hashnode of the identifier BASE. */
static cpp_hashnode *
lex_identifier_intern (cpp_reader *pfile, const uchar *base)
{
cpp_hashnode *result;
const uchar *cur;
unsigned int len;
unsigned int hash = HT_HASHSTEP (0, *base);
cur = base + 1;
while (ISIDNUM (*cur))
{
hash = HT_HASHSTEP (hash, *cur);
cur++;
}
len = cur - base;
hash = HT_HASHFINISH (hash, len);
result = CPP_HASHNODE (ht_lookup_with_hash (pfile->hash_table,
base, len, hash, HT_ALLOC));
/* Rarely, identifiers require diagnostics when lexed. */
if (__builtin_expect ((result->flags & NODE_DIAGNOSTIC)
&& !pfile->state.skipping, 0))
{
/* It is allowed to poison the same identifier twice. */
if ((result->flags & NODE_POISONED) && !pfile->state.poisoned_ok)
cpp_error (pfile, CPP_DL_ERROR, "attempt to use poisoned \"%s\"",
NODE_NAME (result));
/* Constraint 6.10.3.5: __VA_ARGS__ should only appear in the
replacement list of a variadic macro. */
if (result == pfile->spec_nodes.n__VA_ARGS__
&& !pfile->state.va_args_ok)
{
if (CPP_OPTION (pfile, cplusplus))
cpp_error (pfile, CPP_DL_PEDWARN,
"__VA_ARGS__ can only appear in the expansion"
" of a C++11 variadic macro");
else
cpp_error (pfile, CPP_DL_PEDWARN,
"__VA_ARGS__ can only appear in the expansion"
" of a C99 variadic macro");
}
if (result == pfile->spec_nodes.n__VA_OPT__)
maybe_va_opt_error (pfile);
/* For -Wc++-compat, warn about use of C++ named operators. */
if (result->flags & NODE_WARN_OPERATOR)
cpp_warning (pfile, CPP_W_CXX_OPERATOR_NAMES,
"identifier \"%s\" is a special operator name in C++",
NODE_NAME (result));
}
return result;
}
/* Get the cpp_hashnode of an identifier specified by NAME in
the current cpp_reader object. If none is found, NULL is returned. */
cpp_hashnode *
_cpp_lex_identifier (cpp_reader *pfile, const char *name)
{
cpp_hashnode *result;
result = lex_identifier_intern (pfile, (uchar *) name);
return result;
}
/* Lex an identifier starting at BUFFER->CUR - 1. */
static cpp_hashnode *
lex_identifier (cpp_reader *pfile, const uchar *base, bool starts_ucn,
struct normalize_state *nst, cpp_hashnode **spelling)
{
cpp_hashnode *result;
const uchar *cur;
unsigned int len;
unsigned int hash = HT_HASHSTEP (0, *base);
cur = pfile->buffer->cur;
if (! starts_ucn)
{
while (ISIDNUM (*cur))
{
hash = HT_HASHSTEP (hash, *cur);
cur++;
}
NORMALIZE_STATE_UPDATE_IDNUM (nst, *(cur - 1));
}
pfile->buffer->cur = cur;
if (starts_ucn || forms_identifier_p (pfile, false, nst))
{
/* Slower version for identifiers containing UCNs
or extended chars (including $). */
do {
while (ISIDNUM (*pfile->buffer->cur))
{
NORMALIZE_STATE_UPDATE_IDNUM (nst, *pfile->buffer->cur);
pfile->buffer->cur++;
}
} while (forms_identifier_p (pfile, false, nst));
result = _cpp_interpret_identifier (pfile, base,
pfile->buffer->cur - base);
*spelling = cpp_lookup (pfile, base, pfile->buffer->cur - base);
}
else
{
len = cur - base;
hash = HT_HASHFINISH (hash, len);
result = CPP_HASHNODE (ht_lookup_with_hash (pfile->hash_table,
base, len, hash, HT_ALLOC));
*spelling = result;
}
/* Rarely, identifiers require diagnostics when lexed. */
if (__builtin_expect ((result->flags & NODE_DIAGNOSTIC)
&& !pfile->state.skipping, 0))
{
/* It is allowed to poison the same identifier twice. */
if ((result->flags & NODE_POISONED) && !pfile->state.poisoned_ok)
cpp_error (pfile, CPP_DL_ERROR, "attempt to use poisoned \"%s\"",
NODE_NAME (result));
/* Constraint 6.10.3.5: __VA_ARGS__ should only appear in the
replacement list of a variadic macro. */
if (result == pfile->spec_nodes.n__VA_ARGS__
&& !pfile->state.va_args_ok)
{
if (CPP_OPTION (pfile, cplusplus))
cpp_error (pfile, CPP_DL_PEDWARN,
"__VA_ARGS__ can only appear in the expansion"
" of a C++11 variadic macro");
else
cpp_error (pfile, CPP_DL_PEDWARN,
"__VA_ARGS__ can only appear in the expansion"
" of a C99 variadic macro");
}
/* __VA_OPT__ should only appear in the replacement list of a
variadic macro. */
if (result == pfile->spec_nodes.n__VA_OPT__)
maybe_va_opt_error (pfile);
/* For -Wc++-compat, warn about use of C++ named operators. */
if (result->flags & NODE_WARN_OPERATOR)
cpp_warning (pfile, CPP_W_CXX_OPERATOR_NAMES,
"identifier \"%s\" is a special operator name in C++",
NODE_NAME (result));
}
return result;
}
/* Lex a number to NUMBER starting at BUFFER->CUR - 1. */
static void
lex_number (cpp_reader *pfile, cpp_string *number,
struct normalize_state *nst)
{
const uchar *cur;
const uchar *base;
uchar *dest;
base = pfile->buffer->cur - 1;
do
{
const uchar *adj_digit_sep = NULL;
cur = pfile->buffer->cur;
/* N.B. ISIDNUM does not include $. */
while (ISIDNUM (*cur)
|| (*cur == '.' && !DIGIT_SEP (cur[-1]))
|| DIGIT_SEP (*cur)
|| (VALID_SIGN (*cur, cur[-1]) && !DIGIT_SEP (cur[-2])))
{
NORMALIZE_STATE_UPDATE_IDNUM (nst, *cur);
/* Adjacent digit separators do not form part of the pp-number syntax.
However, they can safely be diagnosed here as an error, since '' is
not a valid preprocessing token. */
if (DIGIT_SEP (*cur) && DIGIT_SEP (cur[-1]) && !adj_digit_sep)
adj_digit_sep = cur;
cur++;
}
/* A number can't end with a digit separator. */
while (cur > pfile->buffer->cur && DIGIT_SEP (cur[-1]))
--cur;
if (adj_digit_sep && adj_digit_sep < cur)
cpp_error (pfile, CPP_DL_ERROR, "adjacent digit separators");
pfile->buffer->cur = cur;
}
while (forms_identifier_p (pfile, false, nst));
number->len = cur - base;
dest = _cpp_unaligned_alloc (pfile, number->len + 1);
memcpy (dest, base, number->len);
dest[number->len] = '\0';
number->text = dest;
}
/* Create a token of type TYPE with a literal spelling. */
static void
create_literal (cpp_reader *pfile, cpp_token *token, const uchar *base,
unsigned int len, enum cpp_ttype type)
{
token->type = type;
token->val.str.len = len;
token->val.str.text = cpp_alloc_token_string (pfile, base, len);
}
const uchar *
cpp_alloc_token_string (cpp_reader *pfile,
const unsigned char *ptr, unsigned len)
{
uchar *dest = _cpp_unaligned_alloc (pfile, len + 1);
dest[len] = 0;
memcpy (dest, ptr, len);
return dest;
}
/* A pair of raw buffer pointers. The currently open one is [1], the
first one is [0]. Used for string literal lexing. */
struct lit_accum {
_cpp_buff *first;
_cpp_buff *last;
const uchar *rpos;
size_t accum;
lit_accum ()
: first (NULL), last (NULL), rpos (0), accum (0)
{
}
void append (cpp_reader *, const uchar *, size_t);
void read_begin (cpp_reader *);
bool reading_p () const
{
return rpos != NULL;
}
char read_char ()
{
char c = *rpos++;
if (rpos == BUFF_FRONT (last))
rpos = NULL;
return c;
}
};
/* Subroutine of lex_raw_string: Append LEN chars from BASE to the buffer
sequence from *FIRST_BUFF_P to LAST_BUFF_P. */
void
lit_accum::append (cpp_reader *pfile, const uchar *base, size_t len)
{
if (!last)
/* Starting. */
first = last = _cpp_get_buff (pfile, len);
else if (len > BUFF_ROOM (last))
{
/* There is insufficient room in the buffer. Copy what we can,
and then either extend or create a new one. */
size_t room = BUFF_ROOM (last);
memcpy (BUFF_FRONT (last), base, room);
BUFF_FRONT (last) += room;
base += room;
len -= room;
accum += room;
gcc_checking_assert (!rpos);
last = _cpp_append_extend_buff (pfile, last, len);
}
memcpy (BUFF_FRONT (last), base, len);
BUFF_FRONT (last) += len;
accum += len;
}
void
lit_accum::read_begin (cpp_reader *pfile)
{
/* We never accumulate more than 4 chars to read. */
if (BUFF_ROOM (last) < 4)
last = _cpp_append_extend_buff (pfile, last, 4);
rpos = BUFF_FRONT (last);
}
/* Returns true if a macro has been defined.
This might not work if compile with -save-temps,
or preprocess separately from compilation. */
static bool
is_macro(cpp_reader *pfile, const uchar *base)
{
const uchar *cur = base;
if (! ISIDST (*cur))
return false;
unsigned int hash = HT_HASHSTEP (0, *cur);
++cur;
while (ISIDNUM (*cur))
{
hash = HT_HASHSTEP (hash, *cur);
++cur;
}
hash = HT_HASHFINISH (hash, cur - base);
cpp_hashnode *result = CPP_HASHNODE (ht_lookup_with_hash (pfile->hash_table,
base, cur - base, hash, HT_NO_INSERT));
return result && cpp_macro_p (result);
}
/* Returns true if a literal suffix does not have the expected form
and is defined as a macro. */
static bool
is_macro_not_literal_suffix(cpp_reader *pfile, const uchar *base)
{
/* User-defined literals outside of namespace std must start with a single
underscore, so assume anything of that form really is a UDL suffix.
We don't need to worry about UDLs defined inside namespace std because
their names are reserved, so cannot be used as macro names in valid
programs. */
if (base[0] == '_' && base[1] != '_')
return false;
return is_macro (pfile, base);
}
/* Lexes a raw string. The stored string contains the spelling,
including double quotes, delimiter string, '(' and ')', any leading
'L', 'u', 'U' or 'u8' and 'R' modifier. The created token contains
the type of the literal, or CPP_OTHER if it was not properly
terminated.
BASE is the start of the token. Updates pfile->buffer->cur to just
after the lexed string.
The spelling is NUL-terminated, but it is not guaranteed that this
is the first NUL since embedded NULs are preserved. */
static void
lex_raw_string (cpp_reader *pfile, cpp_token *token, const uchar *base)
{
const uchar *pos = base;
/* 'tis a pity this information isn't passed down from the lexer's
initial categorization of the token. */
enum cpp_ttype type = CPP_STRING;
if (*pos == 'L')
{
type = CPP_WSTRING;
pos++;
}
else if (*pos == 'U')
{
type = CPP_STRING32;
pos++;
}
else if (*pos == 'u')
{
if (pos[1] == '8')
{
type = CPP_UTF8STRING;
pos++;
}
else
type = CPP_STRING16;
pos++;
}
gcc_checking_assert (pos[0] == 'R' && pos[1] == '"');
pos += 2;
_cpp_line_note *note = &pfile->buffer->notes[pfile->buffer->cur_note];
/* Skip notes before the ". */
while (note->pos < pos)
++note;
lit_accum accum;
uchar prefix[17];
unsigned prefix_len = 0;
enum Phase
{
PHASE_PREFIX = -2,
PHASE_NONE = -1,
PHASE_SUFFIX = 0
} phase = PHASE_PREFIX;
for (;;)
{
gcc_checking_assert (note->pos >= pos);
/* Undo any escaped newlines and trigraphs. */
if (!accum.reading_p () && note->pos == pos)
switch (note->type)
{
case '\\':
case ' ':
/* Restore backslash followed by newline. */
accum.append (pfile, base, pos - base);
base = pos;
accum.read_begin (pfile);
accum.append (pfile, UC"\\", 1);
after_backslash:
if (note->type == ' ')
/* GNU backslash whitespace newline extension. FIXME
could be any sequence of non-vertical space. When we
can properly restore any such sequence, we should
mark this note as handled so _cpp_process_line_notes
doesn't warn. */
accum.append (pfile, UC" ", 1);
accum.append (pfile, UC"\n", 1);
note++;
break;
case '\n':
/* This can happen for ??/<NEWLINE> when trigraphs are not
being interpretted. */
gcc_checking_assert (!CPP_OPTION (pfile, trigraphs));
note->type = 0;
note++;
break;
default:
gcc_checking_assert (_cpp_trigraph_map[note->type]);
/* Don't warn about this trigraph in
_cpp_process_line_notes, since trigraphs show up as
trigraphs in raw strings. */
uchar type = note->type;
note->type = 0;
if (CPP_OPTION (pfile, trigraphs))
{
accum.append (pfile, base, pos - base);
base = pos;
accum.read_begin (pfile);
accum.append (pfile, UC"??", 2);
accum.append (pfile, &type, 1);
/* ??/ followed by newline gets two line notes, one for
the trigraph and one for the backslash/newline. */
if (type == '/' && note[1].pos == pos)
{
note++;
gcc_assert (note->type == '\\' || note->type == ' ');
goto after_backslash;
}
/* Skip the replacement character. */
base = ++pos;
}
note++;
break;
}
/* Now get a char to process. Either from an expanded note, or
from the line buffer. */
bool read_note = accum.reading_p ();
char c = read_note ? accum.read_char () : *pos++;
if (phase == PHASE_PREFIX)
{
if (c == '(')
{
/* Done. */
phase = PHASE_NONE;
prefix[prefix_len++] = '"';
}
else if (prefix_len < 16
/* Prefix chars are any of the basic character set,
[lex.charset] except for '
()\\\t\v\f\n'. Optimized for a contiguous
alphabet. */
/* Unlike a switch, this collapses down to one or
two shift and bitmask operations on an ASCII
system, with an outlier or two. */
&& (('Z' - 'A' == 25
? ((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z'))
: ISIDST (c))
|| (c >= '0' && c <= '9')
|| c == '_' || c == '{' || c == '}'
|| c == '[' || c == ']' || c == '#'
|| c == '<' || c == '>' || c == '%'
|| c == ':' || c == ';' || c == '.' || c == '?'
|| c == '*' || c == '+' || c == '-' || c == '/'
|| c == '^' || c == '&' || c == '|' || c == '~'
|| c == '!' || c == '=' || c == ','
|| c == '"' || c == '\''))
prefix[prefix_len++] = c;
else
{
/* Something is wrong. */
int col = CPP_BUF_COLUMN (pfile->buffer, pos) + read_note;
if (prefix_len == 16)
cpp_error_with_line (pfile, CPP_DL_ERROR, token->src_loc,
col, "raw string delimiter longer "
"than 16 characters");
else if (c == '\n')
cpp_error_with_line (pfile, CPP_DL_ERROR, token->src_loc,
col, "invalid new-line in raw "
"string delimiter");
else
cpp_error_with_line (pfile, CPP_DL_ERROR, token->src_loc,
col, "invalid character '%c' in "
"raw string delimiter", c);
type = CPP_OTHER;
phase = PHASE_NONE;
/* Continue until we get a close quote, that's probably
the best failure mode. */
prefix_len = 0;
}
if (c != '\n')
continue;
}
if (phase != PHASE_NONE)
{
if (prefix[phase] != c)
phase = PHASE_NONE;
else if (unsigned (phase + 1) == prefix_len)
break;
else
{
phase = Phase (phase + 1);
continue;
}
}
if (!prefix_len && c == '"')
/* Failure mode lexing. */
goto out;
else if (prefix_len && c == ')')
phase = PHASE_SUFFIX;
else if (!read_note && c == '\n')
{
pos--;
pfile->buffer->cur = pos;
if (pfile->state.in_directive
|| (pfile->state.parsing_args
&& pfile->buffer->next_line >= pfile->buffer->rlimit))
{
cpp_error_with_line (pfile, CPP_DL_ERROR, token->src_loc, 0,
"unterminated raw string");
type = CPP_OTHER;
goto out;
}
accum.append (pfile, base, pos - base + 1);
_cpp_process_line_notes (pfile, false);
if (pfile->buffer->next_line < pfile->buffer->rlimit)
CPP_INCREMENT_LINE (pfile, 0);
pfile->buffer->need_line = true;
if (!_cpp_get_fresh_line (pfile))
{
/* We ran out of file and failed to get a line. */
location_t src_loc = token->src_loc;
token->type = CPP_EOF;
/* Tell the compiler the line number of the EOF token. */
token->src_loc = pfile->line_table->highest_line;
token->flags = BOL;
if (accum.first)
_cpp_release_buff (pfile, accum.first);
cpp_error_with_line (pfile, CPP_DL_ERROR, src_loc, 0,
"unterminated raw string");
/* Now pop the buffer that _cpp_get_fresh_line did not. */
_cpp_pop_buffer (pfile);
return;
}
pos = base = pfile->buffer->cur;
note = &pfile->buffer->notes[pfile->buffer->cur_note];
}
}
if (CPP_OPTION (pfile, user_literals))
{
/* If a string format macro, say from inttypes.h, is placed touching
a string literal it could be parsed as a C++11 user-defined string
literal thus breaking the program. */
if (is_macro_not_literal_suffix (pfile, pos))
{
/* Raise a warning, but do not consume subsequent tokens. */
if (CPP_OPTION (pfile, warn_literal_suffix) && !pfile->state.skipping)
cpp_warning_with_line (pfile, CPP_W_LITERAL_SUFFIX,
token->src_loc, 0,
"invalid suffix on literal; C++11 requires "
"a space between literal and string macro");
}
/* Grab user defined literal suffix. */
else if (ISIDST (*pos))
{
type = cpp_userdef_string_add_type (type);
++pos;
while (ISIDNUM (*pos))
++pos;
}
}
out:
pfile->buffer->cur = pos;
if (!accum.accum)
create_literal (pfile, token, base, pos - base, type);
else
{
size_t extra_len = pos - base;
uchar *dest = _cpp_unaligned_alloc (pfile, accum.accum + extra_len + 1);
token->type = type;
token->val.str.len = accum.accum + extra_len;
token->val.str.text = dest;
for (_cpp_buff *buf = accum.first; buf; buf = buf->next)
{
size_t len = BUFF_FRONT (buf) - buf->base;
memcpy (dest, buf->base, len);
dest += len;
}
_cpp_release_buff (pfile, accum.first);
memcpy (dest, base, extra_len);
dest[extra_len] = '\0';
}
}
/* Lexes a string, character constant, or angle-bracketed header file
name. The stored string contains the spelling, including opening
quote and any leading 'L', 'u', 'U' or 'u8' and optional
'R' modifier. It returns the type of the literal, or CPP_OTHER
if it was not properly terminated, or CPP_LESS for an unterminated
header name which must be relexed as normal tokens.
The spelling is NUL-terminated, but it is not guaranteed that this
is the first NUL since embedded NULs are preserved. */
static void
lex_string (cpp_reader *pfile, cpp_token *token, const uchar *base)
{
bool saw_NUL = false;
const uchar *cur;
cppchar_t terminator;
enum cpp_ttype type;
cur = base;
terminator = *cur++;
if (terminator == 'L' || terminator == 'U')
terminator = *cur++;
else if (terminator == 'u')
{
terminator = *cur++;
if (terminator == '8')
terminator = *cur++;
}
if (terminator == 'R')
{
lex_raw_string (pfile, token, base);
return;
}
if (terminator == '"')
type = (*base == 'L' ? CPP_WSTRING :
*base == 'U' ? CPP_STRING32 :
*base == 'u' ? (base[1] == '8' ? CPP_UTF8STRING : CPP_STRING16)
: CPP_STRING);
else if (terminator == '\'')
type = (*base == 'L' ? CPP_WCHAR :
*base == 'U' ? CPP_CHAR32 :
*base == 'u' ? (base[1] == '8' ? CPP_UTF8CHAR : CPP_CHAR16)
: CPP_CHAR);
else
terminator = '>', type = CPP_HEADER_NAME;
for (;;)
{
cppchar_t c = *cur++;
/* In #include-style directives, terminators are not escapable. */
if (c == '\\' && !pfile->state.angled_headers && *cur != '\n')
cur++;
else if (c == terminator)
break;
else if (c == '\n')
{
cur--;
/* Unmatched quotes always yield undefined behavior, but
greedy lexing means that what appears to be an unterminated
header name may actually be a legitimate sequence of tokens. */
if (terminator == '>')
{
token->type = CPP_LESS;
return;
}
type = CPP_OTHER;
break;
}
else if (c == '\0')
saw_NUL = true;
}
if (saw_NUL && !pfile->state.skipping)
cpp_error (pfile, CPP_DL_WARNING,
"null character(s) preserved in literal");
if (type == CPP_OTHER && CPP_OPTION (pfile, lang) != CLK_ASM)
cpp_error (pfile, CPP_DL_PEDWARN, "missing terminating %c character",
(int) terminator);
if (CPP_OPTION (pfile, user_literals))
{
/* If a string format macro, say from inttypes.h, is placed touching
a string literal it could be parsed as a C++11 user-defined string
literal thus breaking the program. */
if (is_macro_not_literal_suffix (pfile, cur))
{
/* Raise a warning, but do not consume subsequent tokens. */
if (CPP_OPTION (pfile, warn_literal_suffix) && !pfile->state.skipping)
cpp_warning_with_line (pfile, CPP_W_LITERAL_SUFFIX,
token->src_loc, 0,
"invalid suffix on literal; C++11 requires "
"a space between literal and string macro");
}
/* Grab user defined literal suffix. */
else if (ISIDST (*cur))
{
type = cpp_userdef_char_add_type (type);
type = cpp_userdef_string_add_type (type);
++cur;
while (ISIDNUM (*cur))
++cur;
}
}
else if (CPP_OPTION (pfile, cpp_warn_cxx11_compat)
&& is_macro (pfile, cur)
&& !pfile->state.skipping)
cpp_warning_with_line (pfile, CPP_W_CXX11_COMPAT,
token->src_loc, 0, "C++11 requires a space "
"between string literal and macro");
pfile->buffer->cur = cur;
create_literal (pfile, token, base, cur - base, type);
}
/* Return the comment table. The client may not make any assumption
about the ordering of the table. */
cpp_comment_table *
cpp_get_comments (cpp_reader *pfile)
{
return &pfile->comments;
}
/* Append a comment to the end of the comment table. */
static void
store_comment (cpp_reader *pfile, cpp_token *token)
{
int len;
if (pfile->comments.allocated == 0)
{
pfile->comments.allocated = 256;
pfile->comments.entries = (cpp_comment *) xmalloc
(pfile->comments.allocated * sizeof (cpp_comment));
}
if (pfile->comments.count == pfile->comments.allocated)
{
pfile->comments.allocated *= 2;
pfile->comments.entries = (cpp_comment *) xrealloc
(pfile->comments.entries,
pfile->comments.allocated * sizeof (cpp_comment));
}
len = token->val.str.len;
/* Copy comment. Note, token may not be NULL terminated. */
pfile->comments.entries[pfile->comments.count].comment =
(char *) xmalloc (sizeof (char) * (len + 1));
memcpy (pfile->comments.entries[pfile->comments.count].comment,
token->val.str.text, len);
pfile->comments.entries[pfile->comments.count].comment[len] = '\0';
/* Set source location. */
pfile->comments.entries[pfile->comments.count].sloc = token->src_loc;
/* Increment the count of entries in the comment table. */
pfile->comments.count++;
}
/* The stored comment includes the comment start and any terminator. */
static void
save_comment (cpp_reader *pfile, cpp_token *token, const unsigned char *from,
cppchar_t type)
{
unsigned char *buffer;
unsigned int len, clen, i;
len = pfile->buffer->cur - from + 1; /* + 1 for the initial '/'. */
/* C++ comments probably (not definitely) have moved past a new
line, which we don't want to save in the comment. */
if (is_vspace (pfile->buffer->cur[-1]))
len--;
/* If we are currently in a directive or in argument parsing, then
we need to store all C++ comments as C comments internally, and
so we need to allocate a little extra space in that case.
Note that the only time we encounter a directive here is
when we are saving comments in a "#define". */
clen = ((pfile->state.in_directive || pfile->state.parsing_args)
&& type == '/') ? len + 2 : len;
buffer = _cpp_unaligned_alloc (pfile, clen);
token->type = CPP_COMMENT;
token->val.str.len = clen;
token->val.str.text = buffer;
buffer[0] = '/';
memcpy (buffer + 1, from, len - 1);
/* Finish conversion to a C comment, if necessary. */
if ((pfile->state.in_directive || pfile->state.parsing_args) && type == '/')
{
buffer[1] = '*';
buffer[clen - 2] = '*';
buffer[clen - 1] = '/';
/* As there can be in a C++ comments illegal sequences for C comments
we need to filter them out. */
for (i = 2; i < (clen - 2); i++)
if (buffer[i] == '/' && (buffer[i - 1] == '*' || buffer[i + 1] == '*'))
buffer[i] = '|';
}
/* Finally store this comment for use by clients of libcpp. */
store_comment (pfile, token);
}
/* Returns true if comment at COMMENT_START is a recognized FALLTHROUGH
comment. */
static bool
fallthrough_comment_p (cpp_reader *pfile, const unsigned char *comment_start)
{
const unsigned char *from = comment_start + 1;
switch (CPP_OPTION (pfile, cpp_warn_implicit_fallthrough))
{
/* For both -Wimplicit-fallthrough=0 and -Wimplicit-fallthrough=5 we
don't recognize any comments. The latter only checks attributes,
the former doesn't warn. */
case 0:
default:
return false;
/* -Wimplicit-fallthrough=1 considers any comment, no matter what
content it has. */
case 1:
return true;
case 2:
/* -Wimplicit-fallthrough=2 looks for (case insensitive)
.*falls?[ \t-]*thr(u|ough).* regex. */
for (; (size_t) (pfile->buffer->cur - from) >= sizeof "fallthru" - 1;
from++)
{
/* Is there anything like strpbrk with upper boundary, or
memchr looking for 2 characters rather than just one? */
if (from[0] != 'f' && from[0] != 'F')
continue;
if (from[1] != 'a' && from[1] != 'A')
continue;
if (from[2] != 'l' && from[2] != 'L')
continue;
if (from[3] != 'l' && from[3] != 'L')
continue;
from += sizeof "fall" - 1;
if (from[0] == 's' || from[0] == 'S')
from++;
while (*from == ' ' || *from == '\t' || *from == '-')
from++;
if (from[0] != 't' && from[0] != 'T')
continue;
if (from[1] != 'h' && from[1] != 'H')
continue;
if (from[2] != 'r' && from[2] != 'R')
continue;
if (from[3] == 'u' || from[3] == 'U')
return true;
if (from[3] != 'o' && from[3] != 'O')
continue;
if (from[4] != 'u' && from[4] != 'U')
continue;
if (from[5] != 'g' && from[5] != 'G')
continue;
if (from[6] != 'h' && from[6] != 'H')
continue;
return true;
}
return false;
case 3:
case 4:
break;
}
/* Whole comment contents:
-fallthrough
@fallthrough@
*/
if (*from == '-' || *from == '@')
{
size_t len = sizeof "fallthrough" - 1;
if ((size_t) (pfile->buffer->cur - from - 1) < len)
return false;
if (memcmp (from + 1, "fallthrough", len))
return false;
if (*from == '@')
{
if (from[len + 1] != '@')
return false;
len++;
}
from += 1 + len;
}
/* Whole comment contents (regex):
lint -fallthrough[ \t]*
*/
else if (*from == 'l')
{
size_t len = sizeof "int -fallthrough" - 1;
if ((size_t) (pfile->buffer->cur - from - 1) < len)
return false;
if (memcmp (from + 1, "int -fallthrough", len))
return false;
from += 1 + len;
while (*from == ' ' || *from == '\t')
from++;
}
/* Whole comment contents (regex):
[ \t]*FALLTHR(U|OUGH)[ \t]*
*/
else if (CPP_OPTION (pfile, cpp_warn_implicit_fallthrough) == 4)
{
while (*from == ' ' || *from == '\t')
from++;
if ((size_t) (pfile->buffer->cur - from) < sizeof "FALLTHRU" - 1)
return false;
if (memcmp (from, "FALLTHR", sizeof "FALLTHR" - 1))
return false;
from += sizeof "FALLTHR" - 1;
if (*from == 'U')
from++;
else if ((size_t) (pfile->buffer->cur - from) < sizeof "OUGH" - 1)
return false;
else if (memcmp (from, "OUGH", sizeof "OUGH" - 1))
return false;
else
from += sizeof "OUGH" - 1;
while (*from == ' ' || *from == '\t')
from++;
}
/* Whole comment contents (regex):
[ \t.!]*(ELSE,? |INTENTIONAL(LY)? )?FALL(S | |-)?THR(OUGH|U)[ \t.!]*(-[^\n\r]*)?
[ \t.!]*(Else,? |Intentional(ly)? )?Fall((s | |-)[Tt]|t)hr(ough|u)[ \t.!]*(-[^\n\r]*)?
[ \t.!]*([Ee]lse,? |[Ii]ntentional(ly)? )?fall(s | |-)?thr(ough|u)[ \t.!]*(-[^\n\r]*)?
*/
else
{
while (*from == ' ' || *from == '\t' || *from == '.' || *from == '!')
from++;
unsigned char f = *from;
bool all_upper = false;
if (f == 'E' || f == 'e')
{
if ((size_t) (pfile->buffer->cur - from)
< sizeof "else fallthru" - 1)
return false;
if (f == 'E' && memcmp (from + 1, "LSE", sizeof "LSE" - 1) == 0)
all_upper = true;
else if (memcmp (from + 1, "lse", sizeof "lse" - 1))
return false;
from += sizeof "else" - 1;
if (*from == ',')
from++;
if (*from != ' ')
return false;
from++;
if (all_upper && *from == 'f')
return false;
if (f == 'e' && *from == 'F')
return false;
f = *from;
}
else if (f == 'I' || f == 'i')
{
if ((size_t) (pfile->buffer->cur - from)
< sizeof "intentional fallthru" - 1)
return false;
if (f == 'I' && memcmp (from + 1, "NTENTIONAL",
sizeof "NTENTIONAL" - 1) == 0)
all_upper = true;
else if (memcmp (from + 1, "ntentional",
sizeof "ntentional" - 1))
return false;
from += sizeof "intentional" - 1;
if (*from == ' ')
{
from++;
if (all_upper && *from == 'f')
return false;
}
else if (all_upper)
{
if (memcmp (from, "LY F", sizeof "LY F" - 1))
return false;
from += sizeof "LY " - 1;
}
else
{
if (memcmp (from, "ly ", sizeof "ly " - 1))
return false;
from += sizeof "ly " - 1;
}
if (f == 'i' && *from == 'F')
return false;
f = *from;
}
if (f != 'F' && f != 'f')
return false;
if ((size_t) (pfile->buffer->cur - from) < sizeof "fallthru" - 1)
return false;
if (f == 'F' && memcmp (from + 1, "ALL", sizeof "ALL" - 1) == 0)
all_upper = true;
else if (all_upper)
return false;
else if (memcmp (from + 1, "all", sizeof "all" - 1))
return false;
from += sizeof "fall" - 1;
if (*from == (all_upper ? 'S' : 's') && from[1] == ' ')
from += 2;
else if (*from == ' ' || *from == '-')
from++;
else if (*from != (all_upper ? 'T' : 't'))
return false;
if ((f == 'f' || *from != 'T') && (all_upper || *from != 't'))
return false;
if ((size_t) (pfile->buffer->cur - from) < sizeof "thru" - 1)
return false;
if (memcmp (from + 1, all_upper ? "HRU" : "hru", sizeof "hru" - 1))
{
if ((size_t) (pfile->buffer->cur - from) < sizeof "through" - 1)
return false;
if (memcmp (from + 1, all_upper ? "HROUGH" : "hrough",
sizeof "hrough" - 1))
return false;
from += sizeof "through" - 1;
}
else
from += sizeof "thru" - 1;
while (*from == ' ' || *from == '\t' || *from == '.' || *from == '!')
from++;
if (*from == '-')
{
from++;
if (*comment_start == '*')
{
do
{
while (*from && *from != '*'
&& *from != '\n' && *from != '\r')
from++;
if (*from != '*' || from[1] == '/')
break;
from++;
}
while (1);
}
else
while (*from && *from != '\n' && *from != '\r')
from++;
}
}
/* C block comment. */
if (*comment_start == '*')
{
if (*from != '*' || from[1] != '/')
return false;
}
/* C++ line comment. */
else if (*from != '\n')
return false;
return true;
}
/* Allocate COUNT tokens for RUN. */
void
_cpp_init_tokenrun (tokenrun *run, unsigned int count)
{
run->base = XNEWVEC (cpp_token, count);
run->limit = run->base + count;
run->next = NULL;
}
/* Returns the next tokenrun, or creates one if there is none. */
static tokenrun *
next_tokenrun (tokenrun *run)
{
if (run->next == NULL)
{
run->next = XNEW (tokenrun);
run->next->prev = run;
_cpp_init_tokenrun (run->next, 250);
}
return run->next;
}
/* Return the number of not yet processed token in a given
context. */
int
_cpp_remaining_tokens_num_in_context (cpp_context *context)
{
if (context->tokens_kind == TOKENS_KIND_DIRECT)
return (LAST (context).token - FIRST (context).token);
else if (context->tokens_kind == TOKENS_KIND_INDIRECT
|| context->tokens_kind == TOKENS_KIND_EXTENDED)
return (LAST (context).ptoken - FIRST (context).ptoken);
else
abort ();
}
/* Returns the token present at index INDEX in a given context. If
INDEX is zero, the next token to be processed is returned. */
static const cpp_token*
_cpp_token_from_context_at (cpp_context *context, int index)
{
if (context->tokens_kind == TOKENS_KIND_DIRECT)
return &(FIRST (context).token[index]);
else if (context->tokens_kind == TOKENS_KIND_INDIRECT
|| context->tokens_kind == TOKENS_KIND_EXTENDED)
return FIRST (context).ptoken[index];
else
abort ();
}
/* Look ahead in the input stream. */
const cpp_token *
cpp_peek_token (cpp_reader *pfile, int index)
{
cpp_context *context = pfile->context;
const cpp_token *peektok;
int count;
/* First, scan through any pending cpp_context objects. */
while (context->prev)
{
ptrdiff_t sz = _cpp_remaining_tokens_num_in_context (context);
if (index < (int) sz)
return _cpp_token_from_context_at (context, index);
index -= (int) sz;
context = context->prev;
}
/* We will have to read some new tokens after all (and do so
without invalidating preceding tokens). */
count = index;
pfile->keep_tokens++;
/* For peeked tokens temporarily disable line_change reporting,
until the tokens are parsed for real. */
void (*line_change) (cpp_reader *, const cpp_token *, int)
= pfile->cb.line_change;
pfile->cb.line_change = NULL;
do
{
peektok = _cpp_lex_token (pfile);
if (peektok->type == CPP_EOF)
{
index--;
break;
}
else if (peektok->type == CPP_PRAGMA)
{
/* Don't peek past a pragma. */
if (peektok == &pfile->directive_result)
/* Save the pragma in the buffer. */
*pfile->cur_token++ = *peektok;
index--;
break;
}
}
while (index--);
_cpp_backup_tokens_direct (pfile, count - index);
pfile->keep_tokens--;
pfile->cb.line_change = line_change;
return peektok;
}
/* Allocate a single token that is invalidated at the same time as the
rest of the tokens on the line. Has its line and col set to the
same as the last lexed token, so that diagnostics appear in the
right place. */
cpp_token *
_cpp_temp_token (cpp_reader *pfile)
{
cpp_token *old, *result;
ptrdiff_t sz = pfile->cur_run->limit - pfile->cur_token;
ptrdiff_t la = (ptrdiff_t) pfile->lookaheads;
old = pfile->cur_token - 1;
/* Any pre-existing lookaheads must not be clobbered. */
if (la)
{
if (sz <= la)
{
tokenrun *next = next_tokenrun (pfile->cur_run);
if (sz < la)
memmove (next->base + 1, next->base,
(la - sz) * sizeof (cpp_token));
next->base[0] = pfile->cur_run->limit[-1];
}
if (sz > 1)
memmove (pfile->cur_token + 1, pfile->cur_token,
MIN (la, sz - 1) * sizeof (cpp_token));
}
if (!sz && pfile->cur_token == pfile->cur_run->limit)
{
pfile->cur_run = next_tokenrun (pfile->cur_run);
pfile->cur_token = pfile->cur_run->base;
}
result = pfile->cur_token++;
result->src_loc = old->src_loc;
return result;
}