dox4/ruby
1
0
Fork 0
mirror of https://github.com/ruby/ruby.git synced 2026-11-08 01:20:26 +08:00
Code Issues Packages Projects Releases Wiki Activity

Index the regexp match cache by OnigPosition

Browse Source 
The match cache index, the fail count and the threshold to enable the cache grow with the string length times the number of cache points, but were long, which is 32-bit on mswin. There the following raised RegexpError (failed to allocate memory) instead of returning false:

    /\A#{'.*' * 16}\z/.match?("a" * 2**27 + "\n")

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
Hiroshi SHIBATA 2026-10-12 12:06:54 +09:00
parent c5c692c818
commit f81c212132
Notes: git 2026-10-13 00:46:42 +00:00 
Merged: https://github.com/ruby/ruby/pull/19303
2 changed files with 11 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
regexec.c
View File

@ -1523,7 +1523,7 @@ stack_double(OnigStackType** arg_stk_base, OnigStackType** arg_stk_end,
#endif
#ifdef ONIG_DEBUG_MATCH_CACHE
# define MATCH_CACHE_DEBUG_MEMOIZE(stkp) fprintf(stderr, "MATCH CACHE: memoize (index=%ld mask=%d)\n", stkp->u.match_cache_point.index, stkp->u.match_cache_point.mask);
# define MATCH_CACHE_DEBUG_MEMOIZE(stkp) fprintf(stderr, "MATCH CACHE: memoize (index=%td mask=%d)\n", stkp->u.match_cache_point.index, stkp->u.match_cache_point.mask);
#else
# define MATCH_CACHE_DEBUG_MEMOIZE(stkp) ((void) 0)
#endif
@ -2281,7 +2281,7 @@ find_cache_point(regex_t* reg, const OnigCacheOpcode* cache_opcodes, long num_ca
}
static int
check_extended_match_cache_point(uint8_t *match_cache_buf, long match_cache_point_index, uint8_t match_cache_point_mask)
check_extended_match_cache_point(uint8_t *match_cache_buf, OnigPosition match_cache_point_index, uint8_t match_cache_point_mask)
{
if (match_cache_point_mask & 0x80) {
return (match_cache_buf[match_cache_point_index + 1] & 0x01) > 0;
@ -2292,7 +2292,7 @@ check_extended_match_cache_point(uint8_t *match_cache_buf, long match_cache_poin
}
static void
memoize_extended_match_cache_point(uint8_t *match_cache_buf, long match_cache_point_index, uint8_t match_cache_point_mask)
memoize_extended_match_cache_point(uint8_t *match_cache_buf, OnigPosition match_cache_point_index, uint8_t match_cache_point_mask)
{
match_cache_buf[match_cache_point_index] |= match_cache_point_mask;
if (match_cache_point_mask & 0x80) {
@ -2614,7 +2614,7 @@ match_at(regex_t* reg, const UChar* str, const UChar* end,
#ifdef USE_MATCH_CACHE
#ifdef ONIG_DEBUG_MATCH_CACHE
#define MATCH_CACHE_DEBUG fprintf(stderr, "MATCH CACHE: cache %ld (p=%p index=%ld mask=%d)\n", match_cache_point, pbegin, match_cache_point_index, match_cache_point_mask)
#define MATCH_CACHE_DEBUG fprintf(stderr, "MATCH CACHE: cache %td (p=%p index=%td mask=%d)\n", match_cache_point, pbegin, match_cache_point_index, match_cache_point_mask)
#define MATCH_CACHE_DEBUG_HIT fprintf(stderr, "MATCH CACHE: cache hit\n")
#else
#define MATCH_CACHE_DEBUG ((void) 0)
@ -2628,8 +2628,8 @@ match_at(regex_t* reg, const UChar* str, const UChar* end,
const OnigCacheOpcode *cache_opcode;\
long cache_point = find_cache_point(reg, msa->cache_opcodes, msa->num_cache_opcodes, pbegin, stk_base, repeat_stk, &cache_opcode);\
if (cache_point >= 0) {\
long match_cache_point = msa->num_cache_points * (long)(s - str) + cache_point;\
long match_cache_point_index = match_cache_point >> 3;\
OnigPosition match_cache_point = msa->num_cache_points * (OnigPosition)(s - str) + cache_point;\
OnigPosition match_cache_point_index = match_cache_point >> 3;\
uint8_t match_cache_point_mask = 1 << (match_cache_point & 7);\
MATCH_CACHE_DEBUG;\
if (msa->match_cache_buf[match_cache_point_index] & match_cache_point_mask) {\
@ -4130,7 +4130,7 @@ match_at(regex_t* reg, const UChar* str, const UChar* end,
#ifdef USE_MATCH_CACHE
if (
msa->match_cache_status != MATCH_CACHE_STATUS_DISABLED &&
++msa->num_fails >= (long)(end - str) * msa->num_cache_opcodes
++msa->num_fails >= (OnigPosition)(end - str) * msa->num_cache_opcodes
) {
if (msa->match_cache_status == MATCH_CACHE_STATUS_UNINIT) {
msa->match_cache_status = MATCH_CACHE_STATUS_INIT;
@ -4141,7 +4141,7 @@ match_at(regex_t* reg, const UChar* str, const UChar* end,
msa->match_cache_status = MATCH_CACHE_STATUS_DISABLED;
goto fail_match_cache;
}
if (msa->num_fails < (long)(end - str) * msa->num_cache_opcodes) {
if (msa->num_fails < (OnigPosition)(end - str) * msa->num_cache_opcodes) {
goto fail_match_cache;
}
if (msa->cache_opcodes == NULL) {
@ -4175,7 +4175,7 @@ match_at(regex_t* reg, const UChar* str, const UChar* end,
if (num_match_cache_points / length != (size_t)msa->num_cache_points) {
return ONIGERR_MEMORY;
}
if (num_match_cache_points >= LONG_MAX_LIMIT) {
if (num_match_cache_points >= (size_t)PTRDIFF_MAX) {
return ONIGERR_MEMORY;
}
size_t match_cache_buf_length = (num_match_cache_points >> 3) + (num_match_cache_points & 7 ? 1 : 0) + 1;

5
regint.h
View File

@ -364,7 +364,6 @@ typedef unsigned int BitStatusType;
#define INT_MAX_LIMIT ((1UL << (SIZEOF_INT * 8 - 1)) - 1)
#define LONG_MAX_LIMIT ((1UL << (SIZEOF_LONG * 8 - 1)) - 1)
/* Program offsets are RelAddrType, and the compile_length_*() functions sum
into int, so a larger program would wrap both. The quarter leaves room
@ -854,7 +853,7 @@ typedef struct _OnigStackType {
} absent_pos;
#ifdef USE_MATCH_CACHE
struct {
long index; /* index of the match cache buffer */
OnigPosition index; /* index of the match cache buffer */
uint8_t mask; /* bit-mask for the match cache buffer */
} match_cache_point;
#endif
@ -897,7 +896,7 @@ typedef struct {
#endif
#ifdef USE_MATCH_CACHE
int match_cache_status;
long num_fails;
OnigPosition num_fails;
long num_cache_opcodes;
OnigCacheOpcode* cache_opcodes;
long num_cache_points;