Check encoding name to replicate

https://hackerone.com/reports/954433

3 files changed, 37 insertions, 17 deletions

diff --git a/encoding.c b/encoding.c
index 7f0ea73ad1..7f798cd78d 100644
--- a/encoding.c
+++ b/encoding.c
@@ -272,6 +272,7 @@ int
 rb_to_encoding_index(VALUE enc)
 {
     int idx;
+    const char *name;
 
     idx = enc_check_encoding(enc);
     if (idx >= 0) {
@@ -283,20 +284,33 @@ rb_to_encoding_index(VALUE enc)
     if (!rb_enc_asciicompat(rb_enc_get(enc))) {
 	return -1;
     }
-    return rb_enc_find_index(StringValueCStr(enc));
+    if (!(name = rb_str_to_cstr(enc))) {
+	return -1;
+    }
+    return rb_enc_find_index(name);
+}
+
+static const char *
+name_for_encoding(volatile VALUE *enc)
+{
+    VALUE name = StringValue(*enc);
+    const char *n;
+
+    if (!rb_enc_asciicompat(rb_enc_get(name))) {
+	rb_raise(rb_eArgError, "invalid encoding name (non ASCII)");
+    }
+    if (!(n = rb_str_to_cstr(name))) {
+	rb_raise(rb_eArgError, "invalid encoding name (NUL byte)");
+    }
+    return n;
 }
 
 /* Returns encoding index or UNSPECIFIED_ENCODING */
 static int
 str_find_encindex(VALUE enc)
 {
-    int idx;
-
-    StringValue(enc);
-    if (!rb_enc_asciicompat(rb_enc_get(enc))) {
-	rb_raise(rb_eArgError, "invalid name encoding (non ASCII)");
-    }
-    idx = rb_enc_find_index(StringValueCStr(enc));
+    int idx = rb_enc_find_index(name_for_encoding(&enc));
+    RB_GC_GUARD(enc);
     return idx;
 }
 
@@ -385,9 +399,8 @@ enc_register(struct enc_table *enc_table, const char *name, rb_encoding *encodin
 {
     int index = enc_table->count;
 
-    if ((index = enc_table_expand(enc_table, index + 1)) < 0) return -1;
-    enc_table->count = index;
-    return enc_register_at(enc_table, index - 1, name, encoding);
+    enc_table->count = enc_table_expand(enc_table, index + 1);
+    return enc_register_at(enc_table, index, name, encoding);
 }
 
 static void set_encoding_const(const char *, rb_encoding *);
@@ -524,6 +537,7 @@ enc_replicate(struct enc_table *enc_table, const char *name, rb_encoding *encodi
 
     enc_check_duplication(enc_table, name);
     idx = enc_register(enc_table, name, encoding);
+    if (idx < 0) rb_raise(rb_eArgError, "invalid encoding name: %s", name);
     set_base_encoding(enc_table, idx, encoding);
     set_encoding_const(name, rb_enc_from_index(idx));
     return idx;
@@ -552,9 +566,9 @@ rb_enc_replicate(const char *name, rb_encoding *encoding)
 static VALUE
 enc_replicate_m(VALUE encoding, VALUE name)
 {
-    return rb_enc_from_encoding_index(
-	rb_enc_replicate(StringValueCStr(name),
-			 rb_to_encoding(encoding)));
+    int idx = rb_enc_replicate(name_for_encoding(&name), rb_to_encoding(encoding));
+    RB_GC_GUARD(name);
+    return rb_enc_from_encoding_index(idx);
 }
 
 static int
diff --git a/test/ruby/test_encoding.rb b/test/ruby/test_encoding.rb
index 6fc5c48179..2965c0bc7b 100644
--- a/test/ruby/test_encoding.rb
+++ b/test/ruby/test_encoding.rb
@@ -61,7 +61,7 @@ class TestEncoding < Test::Unit::TestCase
     assert_instance_of(Encoding, Encoding::ISO_2022_JP.replicate("ISO-2022-JP-ANOTHER#{Time.now.to_f}"))
     bug3127 = '[ruby-dev:40954]'
     assert_raise(TypeError, bug3127) {Encoding::UTF_8.replicate(0)}
-    assert_raise(ArgumentError, bug3127) {Encoding::UTF_8.replicate("\0")}
+    assert_raise_with_message(ArgumentError, /\bNUL\b/, bug3127) {Encoding::UTF_8.replicate("\0")}
     END;
   end
 
@@ -79,6 +79,12 @@ class TestEncoding < Test::Unit::TestCase
 
       assert_equal(e, (("x"*30).force_encoding(e)*1).encoding)
       GC.start
+
+      name = "A" * 64
+      Encoding.list.each do |enc|
+        assert_raise(ArgumentError) {enc.replicate(name)}
+        name.succ!
+      end
     end;
   end
 
diff --git a/test/ruby/test_io_m17n.rb b/test/ruby/test_io_m17n.rb
index 7b37893514..27b16a2a36 100644
--- a/test/ruby/test_io_m17n.rb
+++ b/test/ruby/test_io_m17n.rb
@@ -776,10 +776,10 @@ EOT
            assert_equal(eucjp, r.read)
          end)
 
-    assert_raise_with_message(ArgumentError, /invalid name encoding/) do
+    assert_raise_with_message(ArgumentError, /invalid encoding name/) do
       with_pipe("UTF-8", "UTF-8".encode("UTF-32BE")) {}
     end
-    assert_raise_with_message(ArgumentError, /invalid name encoding/) do
+    assert_raise_with_message(ArgumentError, /invalid encoding name/) do
       with_pipe("UTF-8".encode("UTF-32BE")) {}
     end