Fix native code compiler infinite loop and update type info for 're'

The introduction of filenames being unicode binaries revealed a problem
in the type analysis of the native code compiler which resulted in an
infinite loop when compiling the 'filename' module.

In addition, the hard-coded type information for the built-in functions
of the 're' module was out-of-date, which resulted in erroneous type
information for 'filelib' functions being stored in the PLT.

2 files changed, 39 insertions, 12 deletions

diff --git a/lib/hipe/cerl/erl_bif_types.erl b/lib/hipe/cerl/erl_bif_types.erl
index ed5bf03804..eba95dc6af 100644
--- a/lib/hipe/cerl/erl_bif_types.erl
+++ b/lib/hipe/cerl/erl_bif_types.erl
@@ -48,6 +48,7 @@
 		    t_boolean/0,
 		    t_byte/0,
 		    t_char/0,
+		    t_charlist/0,
 		    t_cons/0,
 		    t_cons/2,
 		    t_cons_hd/1,
@@ -4517,11 +4518,11 @@ arg_types(os, timestamp, 0) ->
 arg_types(re, compile, 1) ->
   [t_iodata()];
 arg_types(re, compile, 2) ->
-  [t_iodata(), t_list(t_re_compile_option())];
+  [t_sup(t_iodata(), t_charlist()), t_list(t_re_compile_option())];
 arg_types(re, run, 2) ->
-  [t_iodata(), t_re_RE()];
+  [t_sup(t_iodata(), t_charlist()), t_re_RE()];
 arg_types(re, run, 3) ->
-  [t_iodata(), t_re_RE(), t_list(t_re_run_option())];
+  [t_sup(t_iodata(), t_charlist()), t_re_RE(), t_list(t_re_run_option())];
 %%------- string --------------------------------------------------------------
 arg_types(string, chars, 2) ->
   [t_char(), t_non_neg_integer()];
@@ -4978,8 +4979,7 @@ t_ets_info_items() ->
 %% =====================================================================
 
 t_prim_file_name() ->
-   t_sup([t_string(),
-	  t_binary()]).
+   t_sup(t_string(), t_binary()).
 
 %% =====================================================================
 %% These are used for the built-in functions of 'gen_tcp'
@@ -5136,13 +5136,14 @@ t_re_MP() ->  %% it's supposed to be an opaque data type
   t_tuple([t_atom('re_pattern'), t_integer(), t_integer(), t_binary()]).
 
 t_re_RE() ->
-  t_sup(t_re_MP(), t_iodata()).
+  t_sup([t_re_MP(), t_iodata(), t_charlist()]).
 
 t_re_compile_option() ->
-  t_sup([t_atoms(['anchored', 'caseless', 'dollar_endonly', 'dotall',
-		  'extended', 'firstline', 'multiline', 'no_auto_capture',
-		  'dupnames', 'ungreedy']),
-	 t_tuple([t_atom('newline'), t_re_NLSpec()])]).
+  t_sup([t_atoms(['unicode', 'anchored', 'caseless', 'dollar_endonly',
+		  'dotall', 'extended', 'firstline', 'multiline',
+		  'no_auto_capture', 'dupnames', 'ungreedy']),
+	 t_tuple([t_atom('newline'), t_re_NLSpec()]),
+	 t_atoms(['bsr_anycrlf', 'bsr_unicode'])]).
 
 t_re_run_option() ->
   t_sup([t_atoms(['anchored', 'global', 'notbol', 'noteol', 'notempty']),
@@ -5159,7 +5160,7 @@ t_re_Type() ->
   t_atoms(['index', 'list', 'binary']).
 
 t_re_NLSpec() ->
-  t_atoms(['cr', 'crlf', 'lf', 'anycrlf']).
+  t_atoms(['cr', 'crlf', 'lf', 'anycrlf', 'any']).
 
 t_re_ValueSpec() ->
   t_sup(t_atoms(['all', 'all_but_first', 'first', 'none']), t_re_ValueList()).
diff --git a/lib/hipe/cerl/erl_types.erl b/lib/hipe/cerl/erl_types.erl
index 1ed85af172..c8dc162457 100644
--- a/lib/hipe/cerl/erl_types.erl
+++ b/lib/hipe/cerl/erl_types.erl
@@ -62,6 +62,7 @@
 	 t_boolean/0,
 	 t_byte/0,
 	 t_char/0,
+	 t_charlist/0,
 	 t_collect_vars/1,
 	 t_cons/0,
 	 t_cons/2,
@@ -1455,6 +1456,21 @@ t_is_tuple(_) -> false.
 %% Non-primitive types, including some handy syntactic sugar types
 %%
 
+-spec t_charlist() -> erl_type().
+
+t_charlist() ->
+  t_charlist(1).
+
+-spec t_charlist(non_neg_integer()) -> erl_type().
+
+t_charlist(N) when N > 0 ->
+  t_maybe_improper_list(t_sup([t_unicode_char(),
+			       t_unicode_binary(),
+			       t_charlist(N-1)]),
+		        t_sup(t_unicode_binary(), t_nil()));
+t_charlist(0) ->
+  t_maybe_improper_list(t_any(), t_sup(t_unicode_binary(), t_nil())).
+
 -spec t_constant() -> erl_type().
 
 t_constant() ->
@@ -1549,6 +1565,16 @@ t_parameterized_module() ->
 t_timeout() ->
   t_sup(t_non_neg_integer(), t_atom('infinity')).
 
+-spec t_unicode_binary() -> erl_type().
+
+t_unicode_binary() ->
+  t_binary().  % with characters encoded in UTF-8 coding standard
+
+-spec t_unicode_char() -> erl_type().
+
+t_unicode_char() ->
+  t_integer(). % representing a valid unicode codepoint
+
 %%-----------------------------------------------------------------------------
 %% Some built-in opaque types
 %%
@@ -2825,7 +2851,7 @@ t_subtract(?list(Contents1, Termination1, Size1) = T,
 	true ->
 	  case {Size1, Size2} of
 	    {?nonempty_qual, ?unknown_qual} -> ?none;
-	    {?unknown_qual, ?nonempty_qual} -> Termination1;
+	    {?unknown_qual, ?nonempty_qual} -> ?nil;
 	    {S, S} -> ?none
 	  end;
 	false ->