[Emacs-diffs] emacs-25 8121757: Fix regex abort when it tries to reenter itself

[Paul Eggert]

branch: emacs-25
commit 8121757b3ae6cbb7a7bac12b11ac32f243657aee
Author: Paul Eggert <address@hidden>
Commit: Paul Eggert <address@hidden>

    Fix regex abort when it tries to reenter itself
    
    Problem reported by Ken Raeburn.
    Solution suggested by Stefan Monnier (Bug#21688).
    * src/regex.c (re_match_2_internal):
    Use new _FAST functions to avoid regex code reentering itself.
    * src/syntax.c (update_syntax_table_forward): New arg PROPERTIZE.
    All callers changed.
    * src/syntax.h (UPDATE_SYNTAX_TABLE_FORWARD_FAST)
    (UPDATE_SYNTAX_TABLE_FAST): New inline functions.
---
 src/regex.c  |   18 +++++++++---------
 src/syntax.c |   15 ++++++++-------
 src/syntax.h |   19 +++++++++++++++++--
 3 files changed, 34 insertions(+), 18 deletions(-)

diff --git a/src/regex.c b/src/regex.c
index dcf2864..4e00fd1 100644
--- a/src/regex.c
+++ b/src/regex.c
@@ -5945,12 +5945,12 @@ re_match_2_internal (struct re_pattern_buffer *bufp, 
const_re_char *string1,
 #ifdef emacs
                ssize_t offset = PTR_TO_OFFSET (d - 1);
                ssize_t charpos = SYNTAX_TABLE_BYTE_TO_CHAR (offset);
-               UPDATE_SYNTAX_TABLE (charpos);
+               UPDATE_SYNTAX_TABLE_FAST (charpos);
 #endif
                GET_CHAR_BEFORE_2 (c1, d, string1, end1, string2, end2);
                s1 = SYNTAX (c1);
 #ifdef emacs
-               UPDATE_SYNTAX_TABLE_FORWARD (charpos + 1);
+               UPDATE_SYNTAX_TABLE_FORWARD_FAST (charpos + 1);
 #endif
                PREFETCH_NOLIMIT ();
                GET_CHAR_AFTER (c2, d, dummy);
@@ -5987,7 +5987,7 @@ re_match_2_internal (struct re_pattern_buffer *bufp, 
const_re_char *string1,
 #ifdef emacs
              ssize_t offset = PTR_TO_OFFSET (d);
              ssize_t charpos = SYNTAX_TABLE_BYTE_TO_CHAR (offset);
-             UPDATE_SYNTAX_TABLE (charpos);
+             UPDATE_SYNTAX_TABLE_FAST (charpos);
 #endif
              PREFETCH ();
              GET_CHAR_AFTER (c2, d, dummy);
@@ -6032,7 +6032,7 @@ re_match_2_internal (struct re_pattern_buffer *bufp, 
const_re_char *string1,
 #ifdef emacs
              ssize_t offset = PTR_TO_OFFSET (d) - 1;
              ssize_t charpos = SYNTAX_TABLE_BYTE_TO_CHAR (offset);
-             UPDATE_SYNTAX_TABLE (charpos);
+             UPDATE_SYNTAX_TABLE_FAST (charpos);
 #endif
              GET_CHAR_BEFORE_2 (c1, d, string1, end1, string2, end2);
              s1 = SYNTAX (c1);
@@ -6047,7 +6047,7 @@ re_match_2_internal (struct re_pattern_buffer *bufp, 
const_re_char *string1,
                  PREFETCH_NOLIMIT ();
                  GET_CHAR_AFTER (c2, d, dummy);
 #ifdef emacs
-                 UPDATE_SYNTAX_TABLE_FORWARD (charpos);
+                 UPDATE_SYNTAX_TABLE_FORWARD_FAST (charpos);
 #endif
                  s2 = SYNTAX (c2);
 
@@ -6076,7 +6076,7 @@ re_match_2_internal (struct re_pattern_buffer *bufp, 
const_re_char *string1,
 #ifdef emacs
              ssize_t offset = PTR_TO_OFFSET (d);
              ssize_t charpos = SYNTAX_TABLE_BYTE_TO_CHAR (offset);
-             UPDATE_SYNTAX_TABLE (charpos);
+             UPDATE_SYNTAX_TABLE_FAST (charpos);
 #endif
              PREFETCH ();
              c2 = RE_STRING_CHAR (d, target_multibyte);
@@ -6119,7 +6119,7 @@ re_match_2_internal (struct re_pattern_buffer *bufp, 
const_re_char *string1,
 #ifdef emacs
              ssize_t offset = PTR_TO_OFFSET (d) - 1;
              ssize_t charpos = SYNTAX_TABLE_BYTE_TO_CHAR (offset);
-             UPDATE_SYNTAX_TABLE (charpos);
+             UPDATE_SYNTAX_TABLE_FAST (charpos);
 #endif
              GET_CHAR_BEFORE_2 (c1, d, string1, end1, string2, end2);
              s1 = SYNTAX (c1);
@@ -6134,7 +6134,7 @@ re_match_2_internal (struct re_pattern_buffer *bufp, 
const_re_char *string1,
                  PREFETCH_NOLIMIT ();
                  c2 = RE_STRING_CHAR (d, target_multibyte);
 #ifdef emacs
-                 UPDATE_SYNTAX_TABLE_FORWARD (charpos + 1);
+                 UPDATE_SYNTAX_TABLE_FORWARD_FAST (charpos + 1);
 #endif
                  s2 = SYNTAX (c2);
 
@@ -6157,7 +6157,7 @@ re_match_2_internal (struct re_pattern_buffer *bufp, 
const_re_char *string1,
            {
              ssize_t offset = PTR_TO_OFFSET (d);
              ssize_t pos1 = SYNTAX_TABLE_BYTE_TO_CHAR (offset);
-             UPDATE_SYNTAX_TABLE (pos1);
+             UPDATE_SYNTAX_TABLE_FAST (pos1);
            }
 #endif
            {
diff --git a/src/syntax.c b/src/syntax.c
index 1dcb3a5..687bef7 100644
--- a/src/syntax.c
+++ b/src/syntax.c
@@ -246,7 +246,7 @@ SETUP_SYNTAX_TABLE (ptrdiff_t from, ptrdiff_t count)
   if (parse_sexp_lookup_properties)
     {
       if (count > 0)
-       update_syntax_table_forward (from, true, Qnil);
+       update_syntax_table_forward (from, true, true, Qnil);
       else if (from > BEGV)
        {
          update_syntax_table (from - 1, count, true, Qnil);
@@ -502,27 +502,28 @@ parse_sexp_propertize (ptrdiff_t charpos)
         e_property_truncated, so the e_property_truncated flag may
         occasionally be left raised spuriously.  This should be rare.  */
       gl_state.e_property_truncated = false;
-      update_syntax_table_forward (charpos, false, Qnil);
+      update_syntax_table_forward (charpos, false, true, Qnil);
     }
 }
 
 void
-update_syntax_table_forward (ptrdiff_t charpos, bool init,
+update_syntax_table_forward (ptrdiff_t charpos, bool init, bool propertize,
                             Lisp_Object object)
 {
   if (gl_state.e_property_truncated)
     {
       eassert (NILP (object));
       eassert (charpos >= gl_state.e_property);
-      parse_sexp_propertize (charpos);
     }
   else
     {
       update_syntax_table (charpos, 1, init, object);
-      if (gl_state.e_property > syntax_propertize__done
-         && NILP (object))
-       parse_sexp_propertize (charpos);
+      propertize &= (NILP (object)
+                    && gl_state.e_property > syntax_propertize__done);
     }
+
+  if (propertize)
+    parse_sexp_propertize (charpos);
 }
 
 /* Returns true if char at CHARPOS is quoted.
diff --git a/src/syntax.h b/src/syntax.h
index 06ce0ec..01c92ae 100644
--- a/src/syntax.h
+++ b/src/syntax.h
@@ -27,7 +27,7 @@ along with GNU Emacs.  If not, see 
<http://www.gnu.org/licenses/>.  */
 INLINE_HEADER_BEGIN
 
 extern void update_syntax_table (ptrdiff_t, EMACS_INT, bool, Lisp_Object);
-extern void update_syntax_table_forward (ptrdiff_t, bool, Lisp_Object);
+extern void update_syntax_table_forward (ptrdiff_t, bool, bool, Lisp_Object);
 
 /* The standard syntax table is stored where it will automatically
    be used in all new buffers.  */
@@ -183,7 +183,15 @@ UPDATE_SYNTAX_TABLE_FORWARD (ptrdiff_t charpos)
 { /* Performs just-in-time syntax-propertization.  */
   if (parse_sexp_lookup_properties && charpos >= gl_state.e_property)
     update_syntax_table_forward (charpos + gl_state.offset,
-                                false, gl_state.object);
+                                false, true, gl_state.object);
+}
+
+INLINE void
+UPDATE_SYNTAX_TABLE_FORWARD_FAST (ptrdiff_t charpos)
+{
+  if (parse_sexp_lookup_properties && charpos >= gl_state.e_property)
+    update_syntax_table_forward (charpos + gl_state.offset,
+                                false, false, gl_state.object);
 }
 
 /* Make syntax table state (gl_state) good for CHARPOS, assuming it is
@@ -205,6 +213,13 @@ UPDATE_SYNTAX_TABLE (ptrdiff_t charpos)
   UPDATE_SYNTAX_TABLE_FORWARD (charpos);
 }
 
+INLINE void
+UPDATE_SYNTAX_TABLE_FAST (ptrdiff_t charpos)
+{
+  UPDATE_SYNTAX_TABLE_BACKWARD (charpos);
+  UPDATE_SYNTAX_TABLE_FORWARD_FAST (charpos);
+}
+
 /* Set up the buffer-global syntax table.  */
 
 INLINE void