diff options
author | Jakub Jelinek <jakub@redhat.com> | 2004-12-10 06:50:32 +0000 |
---|---|---|
committer | Jakub Jelinek <jakub@redhat.com> | 2004-12-10 06:50:32 +0000 |
commit | f2b016bfc98c9ced9f9572d23580d31442a3519c (patch) | |
tree | d9f2e28f30b6d4aca132126bb24de7039805fca8 | |
parent | 0e7f328d0849e8e663eb14ade5074b4e451fe276 (diff) | |
download | glibc-f2b016bfc98c9ced9f9572d23580d31442a3519c.tar.gz glibc-f2b016bfc98c9ced9f9572d23580d31442a3519c.tar.xz glibc-f2b016bfc98c9ced9f9572d23580d31442a3519c.zip |
Updated to fedora-glibc-20041210T0634
-rw-r--r-- | ChangeLog | 77 | ||||
-rw-r--r-- | elf/ldd.bash.in | 26 | ||||
-rw-r--r-- | fedora/branch.mk | 4 | ||||
-rw-r--r-- | fedora/glibc.spec.in | 8 | ||||
-rw-r--r-- | linuxthreads/ChangeLog | 13 | ||||
-rw-r--r-- | malloc/malloc.c | 15 | ||||
-rw-r--r-- | nptl/ChangeLog | 37 | ||||
-rw-r--r-- | nptl/init.c | 3 | ||||
-rw-r--r-- | posix/regex_internal.c | 70 | ||||
-rw-r--r-- | posix/regex_internal.h | 22 | ||||
-rw-r--r-- | posix/regexec.c | 138 |
11 files changed, 259 insertions, 154 deletions
diff --git a/ChangeLog b/ChangeLog index 75a4f15a4d..4b5c3dca03 100644 --- a/ChangeLog +++ b/ChangeLog @@ -1,9 +1,58 @@ +2004-11-29 Jakub Jelinek <jakub@redhat.com> + + * stdlib/strtod_l.c (INTERNAL (__STRTOF)): If densize > 2 + and numsize < densize, always shift num up by empty + 1 limbs. + +2004-12-07 Paolo Bonzini <bonzini@gnu.org> + + * posix/regexec.c (proceed_next_node): Simplify treatment of epsilon + nodes. Pass the pushed node to push_fail_stack. + (push_fail_stack): Accept a single node rather than an array + of two epsilon destinations. + (build_sifted_states): Only walk non-epsilon nodes. + (check_arrival): Don't pass epsilon nodes to + check_arrival_add_next_nodes. + (check_arrival_add_next_nodes) [DEBUG]: Abort if an epsilon node is + found. + (check_node_accept): Do expensive checks later. + (add_epsilon_src_nodes): Cache result of merging the inveclosures. + * posix/regex_internal.h (re_dfastate_t): Add non_eps_nodes and + inveclosure. + (re_string_elem_size_at, re_string_char_size_at, re_string_wchar_at, + re_string_context_at, re_string_peek_byte_case, + re_string_fetch_byte_case, re_node_set_compare, re_node_set_contains): + Declare as pure. + * posix/regex_internal.c (create_newstate_common): Remove. + (register_state): Move part of it here. Initialize non_eps_nodes. + (free_state): Free inveclosure and non_eps_nodes. + (create_cd_newstate, create_ci_newstate): Allocate the new + re_dfastate_t here. + +2004-12-09 Ulrich Drepper <drepper@redhat.com> + + * malloc/malloc.c (public_rEALLOc): Add parameter checks. + (_int_free): Provide better error message for invalid pointers. + +2004-12-01 Jakub Jelinek <jakub@redhat.com> + + * posix/tst-regex.c: Use defined _POSIX_CPUTIME && _POSIX_CPUTIME >= 0 + conditionals instead of defined _POSIX_CPUTIME. + (main): If _POSIX_CPUTIME == 0, call sysconf to see if CPUTIME + option is available. + * posix/tst-regex.c2: Use defined _POSIX_CPUTIME && _POSIX_CPUTIME >= 0 + conditionals instead of defined _POSIX_CPUTIME. + (do_test): If _POSIX_CPUTIME == 0, call sysconf to see if CPUTIME + option is available. + * sysdeps/posix/sysconf.c (__sysconf): If _POSIX_CPUTIME resp. + _POSIX_THREAD_CPUTIME is defined to 0, return -1 for the corresponding + _SC_ argument. + 2004-12-08 Jakub Jelinek <jakub@redhat.com> - * elf/ldd.bash.in: For -u, set bind_now=yes. - If eval $add_env '"$file"' exits with exitcode 5, retry with - eval $add_env \${RTLD} '"$file"'. - Remove | cat usage, it breaks exit code propagation. + * elf/ldd.bash.in: When set -o pipefail is available, use that for + piping to cat; when not, don't use the pipe at all. + Pipe to cat in all cases of running the executable. + When direct running exits with code 5, retry running via ${RTLD}. * elf/rtld.c (process_envvars): If __libc_enable_secure and mode != normal, exit with exitcode 5. @@ -130,20 +179,6 @@ * sysdeps/unix/sysv/linux/arm/sysdep.h: Likewise. * sysdeps/unix/sysv/linux/arm/vfork.S: Likewise. -2004-12-01 Jakub Jelinek <jakub@redhat.com> - - * posix/tst-regex.c: Use defined _POSIX_CPUTIME && _POSIX_CPUTIME >= 0 - conditionals instead of defined _POSIX_CPUTIME. - (main): If _POSIX_CPUTIME == 0, call sysconf to see if CPUTIME - option is available. - * posix/tst-regex.c2: Use defined _POSIX_CPUTIME && _POSIX_CPUTIME >= 0 - conditionals instead of defined _POSIX_CPUTIME. - (do_test): If _POSIX_CPUTIME == 0, call sysconf to see if CPUTIME - option is available. - * sysdeps/posix/sysconf.c (__sysconf): If _POSIX_CPUTIME resp. - _POSIX_THREAD_CPUTIME is defined to 0, return -1 for the corresponding - _SC_ argument. - 2004-12-02 Roland McGrath <roland@redhat.com> * extra-lib.mk (object-suffixes-$(lib)): Add .oS when @@ -172,12 +207,6 @@ * sysdeps/gnu/Makefile ($(objpfx)errlist-compat.c): Do $(make-target-directory). -2004-11-29 Jakub Jelinek <jakub@redhat.com> - - * stdlib/strtod_l.c (INTERNAL (__STRTOF)): If densize > 2 - and numsize < densize, always shift num up by empty + 1 - limbs. - 2004-11-29 Roland McGrath <roland@redhat.com> * posix/confstr.c: Avoid punctuation in #error text. diff --git a/elf/ldd.bash.in b/elf/ldd.bash.in index c91886a1e6..5dbf88cdc8 100644 --- a/elf/ldd.bash.in +++ b/elf/ldd.bash.in @@ -106,6 +106,21 @@ add_env="$add_env LD_VERBOSE=$verbose" if test "$unused" = yes; then add_env="$add_env LD_DEBUG=\"$LD_DEBUG${LD_DEBUG:+,}unused\"" fi + +# The following use of cat is needed to make ldd work in SELinux +# environments where the executed program might not have permissions +# to write to the console/tty. But only bash 3.x supports the pipefail +# option, and we don't bother to handle the case for older bash versions. +if set -o pipefail 2> /dev/null; then + try_trace() { + eval $add_env '"$@"' | cat + } +else + try_trace() { + eval $add_env '"$@"' + } +fi + case $# in 0) echo >&2 'ldd:' $"missing file arguments" @@ -157,8 +172,13 @@ warning: you do not have execution permission for" "\`$file'" >&2 # If the program exits with exit code 5, it means the process has been # invoked with __libc_enable_secure. Fall back to running it through # the dynamic linker. - ( eval $add_env '"$file"'; ret=$?; [ $ret != 5 ] && exit $ret; - eval $add_env \${RTLD} '"$file"'; ) || result=1 + try_trace "$file" + rc=$? + if [ $rc = 5 ]; then + try_trace "$RTLD" "$file" + rc=$? + fi + [ $rc = 0 ] || result=1 ;; 1) # This can be a non-ELF binary or no binary at all. @@ -168,7 +188,7 @@ warning: you do not have execution permission for" "\`$file'" >&2 } ;; 2) - eval $add_env \${RTLD} '"$file"' || result=1 + try_trace "$RTLD" "$file" || result=1 ;; *) echo 'ldd:' ${RTLD} $"exited with unknown exit code" "($ret)" >&2 diff --git a/fedora/branch.mk b/fedora/branch.mk index e493d062a6..653f1e4aa7 100644 --- a/fedora/branch.mk +++ b/fedora/branch.mk @@ -1,5 +1,5 @@ # This file is updated automatically by Makefile. glibc-branch := fedora glibc-base := HEAD -fedora-sync-date := 2004-12-08 10:24 UTC -fedora-sync-tag := fedora-glibc-20041208T1024 +fedora-sync-date := 2004-12-10 06:34 UTC +fedora-sync-tag := fedora-glibc-20041210T0634 diff --git a/fedora/glibc.spec.in b/fedora/glibc.spec.in index 5c3f31c675..2ea3705bba 100644 --- a/fedora/glibc.spec.in +++ b/fedora/glibc.spec.in @@ -1,4 +1,4 @@ -%define glibcrelease 89 +%define glibcrelease 90 %define auxarches i586 i686 athlon sparcv9 alphaev6 %define prelinkarches noarch %define nptlarches i386 i686 athlon x86_64 ia64 s390 s390x sparcv9 ppc ppc64 @@ -1265,6 +1265,12 @@ rm -f *.filelist* %endif %changelog +* Fri Dec 10 2004 Jakub Jelinek <jakub@redhat.com> 2.3.3-90 +- update from CVS + - regex speedups + - use | cat in ldd if running under bash3+ to allow running + it on binaries that are not through SELinux allowed to access + console or tty - add __NR_waitid defines for alpha and ia64 * Wed Dec 8 2004 Jakub Jelinek <jakub@redhat.com> 2.3.3-89 diff --git a/linuxthreads/ChangeLog b/linuxthreads/ChangeLog index 9e34a9768e..1849b8a630 100644 --- a/linuxthreads/ChangeLog +++ b/linuxthreads/ChangeLog @@ -1,9 +1,4 @@ -2004-11-18 Daniel Jacobowitz <dan@codesourcery.com> - - * sysdeps/unix/sysv/linux/arm/sysdep-cancel.h: Update RETINSTR use. - * sysdeps/unix/sysv/linux/arm/vfork.S: Likewise. - -2004-12-01 Jakub Jelinek <jakub@redhat.com> +2004-12-01 Jakub Jelinek <jakub@redhat.coM. * sysdeps/unix/sysv/linux/bits/posix_opt.h (_POSIX_CPUTIME, _POSIX_THREAD_CPUTIME): Define to 0. @@ -18,6 +13,12 @@ * sysdeps/unix/sysv/linux/i386/bits/posix_opt.h: Removed. * sysdeps/unix/sysv/linux/ia64/bits/posix_opt.h: Removed. * sysdeps/unix/sysv/linux/x86_64/bits/posix_opt.h: Removed. + * tst-clock1.c (do_test): Check for availability of CPU clock. + +2004-11-18 Daniel Jacobowitz <dan@codesourcery.com> + + * sysdeps/unix/sysv/linux/arm/sysdep-cancel.h: Update RETINSTR use. + * sysdeps/unix/sysv/linux/arm/vfork.S: Likewise. 2004-12-02 Roland McGrath <roland@redhat.com> diff --git a/malloc/malloc.c b/malloc/malloc.c index b62ffb57c0..cf1b935ffd 100644 --- a/malloc/malloc.c +++ b/malloc/malloc.c @@ -3434,6 +3434,17 @@ public_rEALLOc(Void_t* oldmem, size_t bytes) oldp = mem2chunk(oldmem); oldsize = chunksize(oldp); + /* Little security check which won't hurt performance: the + allocator never wrapps around at the end of the address space. + Therefore we can exclude some size values which might appear + here by accident or by "design" from some intruder. */ + if (__builtin_expect ((uintptr_t) oldp > (uintptr_t) -oldsize, 0) + || __builtin_expect ((uintptr_t) oldp & MALLOC_ALIGN_MASK, 0)) + { + malloc_printerr (check_action, "realloc(): invalid pointer", oldmem); + return NULL; + } + checked_request2size(bytes, nb); #if HAVE_MMAP @@ -4205,7 +4216,6 @@ _int_free(mstate av, Void_t* mem) mchunkptr bck; /* misc temp for linking */ mchunkptr fwd; /* misc temp for linking */ - const char *errstr = NULL; p = mem2chunk(mem); @@ -4215,7 +4225,8 @@ _int_free(mstate av, Void_t* mem) allocator never wrapps around at the end of the address space. Therefore we can exclude some size values which might appear here by accident or by "design" from some intruder. */ - if (__builtin_expect ((uintptr_t) p > (uintptr_t) -size, 0)) + if (__builtin_expect ((uintptr_t) p > (uintptr_t) -size, 0) + || __builtin_expect ((uintptr_t) p & MALLOC_ALIGN_MASK, 0)) { errstr = "free(): invalid pointer"; errout: diff --git a/nptl/ChangeLog b/nptl/ChangeLog index 8352c6b904..981781807b 100644 --- a/nptl/ChangeLog +++ b/nptl/ChangeLog @@ -1,3 +1,24 @@ +2004-12-09 Ulrich Drepper <drepper@redhat.com> + + * init.c (__pthread_initialize_minimal_internal): Also unblock + SIGSETXID. + +2004-12-01 Jakub Jelinek <jakub@redhat.com> + + * sysdeps/unix/sysv/linux/bits/posix_opt.h (_POSIX_CPUTIME, + _POSIX_THREAD_CPUTIME): Define to 0. + * sysdeps/pthread/timer_create.c (timer_create): Remove unused code + handling CLOCK_PROCESS_CPUTIME_ID and CLOCK_THREAD_CPUTIME_ID. + * sysdeps/pthread/timer_routines.c (__timer_signal_thread_pclk, + __timer_signal_thread_tclk): Remove. + (init_module): Remove their initialization. + (thread_cleanup): Remove their cleanup assertions. + * sysdeps/pthread/posix-timer.h (__timer_signal_thread_pclk, + __timer_signal_thread_tclk): Remove. + * sysdeps/unix/sysv/linux/i386/bits/posix_opt.h: Removed. + * sysdeps/unix/sysv/linux/ia64/bits/posix_opt.h: Removed. + * sysdeps/unix/sysv/linux/x86_64/bits/posix_opt.h: Removed. + 2004-12-07 Jakub Jelinek <jakub@redhat.com> * sysdeps/ia64/tcb-offsets.sym (TID): Add. @@ -19,22 +40,6 @@ * sysdeps/unix/sysv/linux/i386/clone.S: New file. * sysdeps/unix/sysv/linux/x86_64/clone.S: New file. -2004-12-01 Jakub Jelinek <jakub@redhat.com> - - * sysdeps/unix/sysv/linux/bits/posix_opt.h (_POSIX_CPUTIME, - _POSIX_THREAD_CPUTIME): Define to 0. - * sysdeps/pthread/timer_create.c (timer_create): Remove unused code - handling CLOCK_PROCESS_CPUTIME_ID and CLOCK_THREAD_CPUTIME_ID. - * sysdeps/pthread/timer_routines.c (__timer_signal_thread_pclk, - __timer_signal_thread_tclk): Remove. - (init_module): Remove their initialization. - (thread_cleanup): Remove their cleanup assertions. - * sysdeps/pthread/posix-timer.h (__timer_signal_thread_pclk, - __timer_signal_thread_tclk): Remove. - * sysdeps/unix/sysv/linux/i386/bits/posix_opt.h: Removed. - * sysdeps/unix/sysv/linux/ia64/bits/posix_opt.h: Removed. - * sysdeps/unix/sysv/linux/x86_64/bits/posix_opt.h: Removed. - 2004-12-02 Roland McGrath <roland@redhat.com> * Makefile (libpthread-nonshared): Variable removed. diff --git a/nptl/init.c b/nptl/init.c index a18078edb8..3751e6be77 100644 --- a/nptl/init.c +++ b/nptl/init.c @@ -272,10 +272,11 @@ __pthread_initialize_minimal_internal (void) (void) __libc_sigaction (SIGSETXID, &sa, NULL); - /* The parent process might have left the signal blocked. Just in + /* The parent process might have left the signals blocked. Just in case, unblock it. We reuse the signal mask in the sigaction structure. It is already cleared. */ __sigaddset (&sa.sa_mask, SIGCANCEL); + __sigaddset (&sa.sa_mask, SIGSETXID); (void) INTERNAL_SYSCALL (rt_sigprocmask, err, 4, SIG_UNBLOCK, &sa.sa_mask, NULL, _NSIG / 8); diff --git a/posix/regex_internal.c b/posix/regex_internal.c index cb439e5d7c..001b50b134 100644 --- a/posix/regex_internal.c +++ b/posix/regex_internal.c @@ -26,9 +26,6 @@ static void re_string_construct_common (const char *str, int len, static int re_string_skip_chars (re_string_t *pstr, int new_raw_idx, wint_t *last_wc) internal_function; #endif /* RE_ENABLE_I18N */ -static re_dfastate_t *create_newstate_common (re_dfa_t *dfa, - const re_node_set *nodes, - unsigned int hash) internal_function; static reg_errcode_t register_state (re_dfa_t *dfa, re_dfastate_t *newstate, unsigned int hash) internal_function; static re_dfastate_t *create_ci_newstate (re_dfa_t *dfa, @@ -1298,7 +1295,7 @@ re_node_set_contains (set, elem) const re_node_set *set; int elem; { - int idx, right, mid; + unsigned int idx, right, mid; if (set->nelem <= 0) return 0; @@ -1484,43 +1481,32 @@ re_acquire_state_context (err, dfa, nodes, context) } } -/* Allocate memory for DFA state and initialize common properties. - Return the new state if succeeded, otherwise return NULL. */ +/* Finish initialization of the new state NEWSTATE, and using its hash value + HASH put in the appropriate bucket of DFA's state table. Return value + indicates the error code if failed. */ -static re_dfastate_t * -create_newstate_common (dfa, nodes, hash) +static reg_errcode_t +register_state (dfa, newstate, hash) re_dfa_t *dfa; - const re_node_set *nodes; + re_dfastate_t *newstate; unsigned int hash; { - re_dfastate_t *newstate; + struct re_state_table_entry *spot; reg_errcode_t err; - newstate = (re_dfastate_t *) calloc (sizeof (re_dfastate_t), 1); - if (BE (newstate == NULL, 0)) - return NULL; - err = re_node_set_init_copy (&newstate->nodes, nodes); + int i; + + newstate->hash = hash; + err = re_node_set_alloc (&newstate->non_eps_nodes, newstate->nodes.nelem); if (BE (err != REG_NOERROR, 0)) + return REG_ESPACE; + for (i = 0; i < newstate->nodes.nelem; i++) { - re_free (newstate); - return NULL; + int elem = newstate->nodes.elems[i]; + if (!IS_EPSILON_NODE (dfa->nodes[elem].type)) + re_node_set_insert_last (&newstate->non_eps_nodes, elem); } - newstate->trtable = NULL; - newstate->hash = hash; - return newstate; -} -/* Store the new state NEWSTATE whose hash value is HASH in appropriate - position. Return value indicate the error code if failed. */ - -static reg_errcode_t -register_state (dfa, newstate, hash) - re_dfa_t *dfa; - re_dfastate_t *newstate; - unsigned int hash; -{ - struct re_state_table_entry *spot; spot = dfa->state_table + (hash & dfa->state_hash_mask); - if (BE (spot->alloc <= spot->num, 0)) { int new_alloc = 2 * spot->num + 2; @@ -1547,11 +1533,18 @@ create_ci_newstate (dfa, nodes, hash) int i; reg_errcode_t err; re_dfastate_t *newstate; - newstate = create_newstate_common (dfa, nodes, hash); + + newstate = (re_dfastate_t *) calloc (sizeof (re_dfastate_t), 1); if (BE (newstate == NULL, 0)) return NULL; - newstate->entrance_nodes = &newstate->nodes; + err = re_node_set_init_copy (&newstate->nodes, nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_free (newstate); + return NULL; + } + newstate->entrance_nodes = &newstate->nodes; for (i = 0 ; i < nodes->nelem ; i++) { re_token_t *node = dfa->nodes + nodes->elems[i]; @@ -1595,9 +1588,16 @@ create_cd_newstate (dfa, nodes, context, hash) reg_errcode_t err; re_dfastate_t *newstate; - newstate = create_newstate_common (dfa, nodes, hash); + newstate = (re_dfastate_t *) calloc (sizeof (re_dfastate_t), 1); if (BE (newstate == NULL, 0)) return NULL; + err = re_node_set_init_copy (&newstate->nodes, nodes); + if (BE (err != REG_NOERROR, 0)) + { + re_free (newstate); + return NULL; + } + newstate->context = context; newstate->entrance_nodes = &newstate->nodes; @@ -1660,6 +1660,8 @@ static void free_state (state) re_dfastate_t *state; { + re_node_set_free (&state->non_eps_nodes); + re_node_set_free (&state->inveclosure); if (state->entrance_nodes != &state->nodes) { re_node_set_free (state->entrance_nodes); diff --git a/posix/regex_internal.h b/posix/regex_internal.h index 1345067a89..0ccd8d3665 100644 --- a/posix/regex_internal.h +++ b/posix/regex_internal.h @@ -388,18 +388,20 @@ static void re_string_translate_buffer (re_string_t *pstr) internal_function; static void re_string_destruct (re_string_t *pstr) internal_function; # ifdef RE_ENABLE_I18N static int re_string_elem_size_at (const re_string_t *pstr, int idx) - internal_function; + internal_function __attribute ((pure)); static inline int re_string_char_size_at (const re_string_t *pstr, int idx) - internal_function; + internal_function __attribute ((pure)); static inline wint_t re_string_wchar_at (const re_string_t *pstr, int idx) - internal_function; + internal_function __attribute ((pure)); # endif /* RE_ENABLE_I18N */ static unsigned int re_string_context_at (const re_string_t *input, int idx, - int eflags) internal_function; + int eflags) + internal_function __attribute ((pure)); static unsigned char re_string_peek_byte_case (const re_string_t *pstr, - int idx) internal_function; + int idx) + internal_function __attribute ((pure)); static unsigned char re_string_fetch_byte_case (re_string_t *pstr) - internal_function; + internal_function __attribute ((pure)); #endif #define re_string_peek_byte(pstr, offset) \ ((pstr)->mbs[(pstr)->cur_idx + offset]) @@ -481,6 +483,8 @@ struct re_dfastate_t { unsigned int hash; re_node_set nodes; + re_node_set non_eps_nodes; + re_node_set inveclosure; re_node_set *entrance_nodes; struct re_dfastate_t **trtable; unsigned int context : 4; @@ -665,8 +669,10 @@ static int re_node_set_insert (re_node_set *set, int elem) internal_function; static int re_node_set_insert_last (re_node_set *set, int elem) internal_function; static int re_node_set_compare (const re_node_set *set1, - const re_node_set *set2) internal_function; -static int re_node_set_contains (const re_node_set *set, int elem) internal_function; + const re_node_set *set2) + internal_function __attribute ((pure)); +static int re_node_set_contains (const re_node_set *set, int elem) + internal_function __attribute ((pure)); static void re_node_set_remove_at (re_node_set *set, int idx) internal_function; #define re_node_set_remove(set,id) \ (re_node_set_remove_at (set, re_node_set_contains (set, id) - 1)) diff --git a/posix/regexec.c b/posix/regexec.c index 22b806439b..91b48dd4a2 100644 --- a/posix/regexec.c +++ b/posix/regexec.c @@ -73,7 +73,7 @@ static int proceed_next_node (const re_match_context_t *mctx, int *pidx, int node, re_node_set *eps_via_nodes, struct re_fail_stack_t *fs) internal_function; static reg_errcode_t push_fail_stack (struct re_fail_stack_t *fs, - int str_idx, int *dests, int nregs, + int str_idx, int dest_node, int nregs, regmatch_t *regs, re_node_set *eps_via_nodes) internal_function; static int pop_fail_stack (struct re_fail_stack_t *fs, int *pidx, int nregs, @@ -1223,30 +1223,38 @@ proceed_next_node (mctx, nregs, regs, pidx, node, eps_via_nodes, fs) if (IS_EPSILON_NODE (dfa->nodes[node].type)) { re_node_set *cur_nodes = &mctx->state_log[*pidx]->nodes; - int ndest, dest_nodes[2]; + re_node_set *edests = &dfa->edests[node]; + int dest_node; err = re_node_set_insert (eps_via_nodes, node); if (BE (err < 0, 0)) return -2; - /* Pick up valid destinations. */ - for (ndest = 0, i = 0; i < dfa->edests[node].nelem; ++i) + /* Pick up a valid destination, or return -1 if none is found. */ + for (dest_node = -1, i = 0; i < edests->nelem; ++i) { - int candidate = dfa->edests[node].elems[i]; + int candidate = edests->elems[i]; if (!re_node_set_contains (cur_nodes, candidate)) continue; - dest_nodes[0] = (ndest == 0) ? candidate : dest_nodes[0]; - dest_nodes[1] = (ndest == 1) ? candidate : dest_nodes[1]; - ++ndest; + if (dest_node == -1) + dest_node = candidate; + + else + { + /* In order to avoid infinite loop like "(a*)*", return the second + epsilon-transition if the first was already considered. */ + if (re_node_set_contains (eps_via_nodes, dest_node)) + return candidate; + + /* Otherwise, push the second epsilon-transition on the fail stack. */ + else if (fs != NULL + && push_fail_stack (fs, *pidx, candidate, nregs, regs, + eps_via_nodes)) + return -2; + + /* We know we are going to exit. */ + break; + } } - if (ndest <= 1) - return ndest == 0 ? -1 : (ndest == 1 ? dest_nodes[0] : 0); - /* In order to avoid infinite loop like "(a*)*". */ - if (re_node_set_contains (eps_via_nodes, dest_nodes[0])) - return dest_nodes[1]; - if (fs != NULL - && push_fail_stack (fs, *pidx, dest_nodes, nregs, regs, - eps_via_nodes)) - return -2; - return dest_nodes[0]; + return dest_node; } else { @@ -1304,9 +1312,9 @@ proceed_next_node (mctx, nregs, regs, pidx, node, eps_via_nodes, fs) } static reg_errcode_t -push_fail_stack (fs, str_idx, dests, nregs, regs, eps_via_nodes) +push_fail_stack (fs, str_idx, dest_node, nregs, regs, eps_via_nodes) struct re_fail_stack_t *fs; - int str_idx, *dests, nregs; + int str_idx, dest_node, nregs; regmatch_t *regs; re_node_set *eps_via_nodes; { @@ -1323,7 +1331,7 @@ push_fail_stack (fs, str_idx, dests, nregs, regs, eps_via_nodes) fs->stack = new_array; } fs->stack[num].idx = str_idx; - fs->stack[num].node = dests[1]; + fs->stack[num].node = dest_node; fs->stack[num].regs = re_malloc (regmatch_t, nregs); if (fs->stack[num].regs == NULL) return REG_ESPACE; @@ -1600,7 +1608,7 @@ build_sifted_states (mctx, sctx, str_idx, cur_dest) re_node_set *cur_dest; { re_dfa_t *const dfa = mctx->dfa; - re_node_set *cur_src = &mctx->state_log[str_idx]->nodes; + re_node_set *cur_src = &mctx->state_log[str_idx]->non_eps_nodes; int i; /* Then build the next sifted state. @@ -1608,16 +1616,20 @@ build_sifted_states (mctx, sctx, str_idx, cur_dest) `sifted_states[str_idx]' with `cur_dest'. Note: `cur_dest' is the sifted state from `state_log[str_idx + 1]'. - `cur_src' points the node_set of the old `state_log[str_idx]'. */ + `cur_src' points the node_set of the old `state_log[str_idx]' + (with the epsilon nodes pre-filtered out). */ for (i = 0; i < cur_src->nelem; i++) { int prev_node = cur_src->elems[i]; int naccepted = 0; - re_token_type_t type = dfa->nodes[prev_node].type; int ret; - if (IS_EPSILON_NODE (type)) - continue; +#if defined DEBUG || defined RE_ENABLE_I18N + re_token_type_t type = dfa->nodes[prev_node].type; +#endif +#ifdef DEBUG + assert (!IS_EPSILON_NODE (type)); +#endif #ifdef RE_ENABLE_I18N /* If the node may accept `multi byte'. */ if (ACCEPT_MB_NODE (type)) @@ -1764,26 +1776,24 @@ add_epsilon_src_nodes (dfa, dest_nodes, candidates) re_node_set *dest_nodes; const re_node_set *candidates; { - reg_errcode_t err; - int src_idx; - re_node_set src_copy; + reg_errcode_t err = REG_NOERROR; + int i; - err = re_node_set_init_copy (&src_copy, dest_nodes); + re_dfastate_t *state = re_acquire_state (&err, dfa, dest_nodes); if (BE (err != REG_NOERROR, 0)) return err; - for (src_idx = 0; src_idx < src_copy.nelem; ++src_idx) + + if (!state->inveclosure.alloc) { - err = re_node_set_add_intersect (dest_nodes, candidates, - dfa->inveclosures - + src_copy.elems[src_idx]); + err = re_node_set_alloc (&state->inveclosure, dest_nodes->nelem); if (BE (err != REG_NOERROR, 0)) - { - re_node_set_free (&src_copy); - return err; - } + return REG_ESPACE; + for (i = 0; i < dest_nodes->nelem; i++) + re_node_set_merge (&state->inveclosure, + dfa->inveclosures + dest_nodes->elems[i]); } - re_node_set_free (&src_copy); - return REG_NOERROR; + return re_node_set_add_intersect (dest_nodes, candidates, + &state->inveclosure); } static reg_errcode_t @@ -2935,7 +2945,7 @@ check_arrival (mctx, path, top_node, top_str, last_node, last_str, if (cur_state) { err = check_arrival_add_next_nodes (mctx, str_idx, - &cur_state->nodes, &next_nodes); + &cur_state->non_eps_nodes, &next_nodes); if (BE (err != REG_NOERROR, 0)) { re_node_set_free (&next_nodes); @@ -3009,9 +3019,12 @@ check_arrival_add_next_nodes (mctx, str_idx, cur_nodes, next_nodes) { int naccepted = 0; int cur_node = cur_nodes->elems[cur_idx]; +#if defined DEBUG || defined RE_ENABLE_I18N re_token_type_t type = dfa->nodes[cur_node].type; - if (IS_EPSILON_NODE (type)) - continue; +#endif +#ifdef DEBUG + assert (!IS_EPSILON_NODE (type)); +#endif #ifdef RE_ENABLE_I18N /* If the node may accept `multi byte'. */ if (ACCEPT_MB_NODE (type)) @@ -3988,24 +4001,20 @@ check_node_accept (mctx, node, idx) const re_token_t *node; int idx; { - re_dfa_t *const dfa = mctx->dfa; unsigned char ch; - if (node->constraint) - { - /* The node has constraints. Check whether the current context - satisfies the constraints. */ - unsigned int context = re_string_context_at (&mctx->input, idx, - mctx->eflags); - if (NOT_SATISFY_NEXT_CONSTRAINT (node->constraint, context)) - return 0; - } ch = re_string_byte_at (&mctx->input, idx); switch (node->type) { case CHARACTER: - return node->opr.c == ch; + if (node->opr.c != ch) + return 0; + break; + case SIMPLE_BRACKET: - return bitset_contain (node->opr.sbcset, ch); + if (!bitset_contain (node->opr.sbcset, ch)) + return 0; + break; + #ifdef RE_ENABLE_I18N case OP_UTF8_PERIOD: if (ch >= 0x80) @@ -4013,11 +4022,26 @@ check_node_accept (mctx, node, idx) /* FALLTHROUGH */ #endif case OP_PERIOD: - return !((ch == '\n' && !(dfa->syntax & RE_DOT_NEWLINE)) - || (ch == '\0' && (dfa->syntax & RE_DOT_NOT_NULL))); + if ((ch == '\n' && !(mctx->dfa->syntax & RE_DOT_NEWLINE)) + || (ch == '\0' && (mctx->dfa->syntax & RE_DOT_NOT_NULL))) + return 0; + break; + default: return 0; } + + if (node->constraint) + { + /* The node has constraints. Check whether the current context + satisfies the constraints. */ + unsigned int context = re_string_context_at (&mctx->input, idx, + mctx->eflags); + if (NOT_SATISFY_NEXT_CONSTRAINT (node->constraint, context)) + return 0; + } + + return 1; } /* Extend the buffers, if the buffers have run out. */ |