MatchData: Avoid large stack allocations in MatchData (GH-15872)

2 files changed, 31 insertions, 2 deletions

diff --git a/re.c b/re.c
index b2c1909c15..82e9407a0a 100644
--- a/re.c
+++ b/re.c
@@ -1014,6 +1014,7 @@ update_char_offset(VALUE match)
     char *s, *p, *q;
     rb_encoding *enc;
     pair_t *pairs;
+    VALUE pairs_obj = Qnil;
 
     if (rm->char_offset_num_allocated)
         return;
@@ -1035,7 +1036,7 @@ update_char_offset(VALUE match)
         return;
     }
 
-    pairs = ALLOCA_N(pair_t, num_regs*2);
+    pairs = RB_ALLOCV_N(pair_t, pairs_obj, num_regs * 2);
     num_pos = 0;
     for (i = 0; i < num_regs; i++) {
         if (BEG(i) < 0)
@@ -1070,6 +1071,8 @@ update_char_offset(VALUE match)
         found = bsearch(&key, pairs, num_pos, sizeof(pair_t), pair_byte_cmp);
         rm->char_offset[i].end = found->char_pos;
     }
+
+    RB_ALLOCV_END(pairs_obj);
 }
 
 static VALUE
@@ -2614,6 +2617,7 @@ match_inspect(VALUE match)
     struct re_registers *regs = RMATCH_REGS(match);
     int num_regs = regs->num_regs;
     struct backref_name_tag *names;
+    VALUE names_obj = Qnil;
     VALUE regexp = RMATCH(match)->regexp;
 
     if (regexp == 0) {
@@ -2624,7 +2628,7 @@ match_inspect(VALUE match)
                           cname, rb_reg_nth_match(0, match));
     }
 
-    names = ALLOCA_N(struct backref_name_tag, num_regs);
+    names = RB_ALLOCV_N(struct backref_name_tag, names_obj, num_regs);
     MEMZERO(names, struct backref_name_tag, num_regs);
 
     onig_foreach_name(RREGEXP_PTR(regexp),
@@ -2652,6 +2656,7 @@ match_inspect(VALUE match)
     }
     rb_str_buf_cat2(str, ">");
 
+    RB_ALLOCV_END(names_obj);
     return str;
 }
 
diff --git a/test/ruby/test_regexp.rb b/test/ruby/test_regexp.rb
index 2d7a67dd54..9feababa53 100644
--- a/test/ruby/test_regexp.rb
+++ b/test/ruby/test_regexp.rb
@@ -1669,6 +1669,30 @@ class TestRegexp < Test::Unit::TestCase
     assert_equal("hoge fuga", h["body"])
   end
 
+  def test_matchdata_large_capture_groups_stack
+    env = {"RUBY_THREAD_MACHINE_STACK_SIZE" => (256 * 1024).to_s}
+    assert_separately([env], <<~'RUBY')
+      n = 20000
+      require "rbconfig/sizeof"
+      stack = RubyVM::DEFAULT_PARAMS[:thread_machine_stack_size]
+      size = RbConfig::SIZEOF["long"]
+      required = (n + 1) * 4 * size
+      if !stack || stack == 0 || stack >= required
+        omit "thread machine stack size not reduced (#{stack}:#{required})"
+      end
+
+      inspect = Thread.new do
+        str = "\u{3042}" * n
+        m = Regexp.new("(.)" * n).match(str)
+        assert_not_nil(m)
+        assert_equal([n - 1, n], m.offset(n))
+        m.inspect
+      end.value
+
+      assert_include(inspect, "MatchData")
+    RUBY
+  end
+
   def test_regexp_popped
     EnvUtil.suppress_warning do
       assert_nothing_raised { eval("a = 1; /\#{ a }/; a") }