* cygheap.cc (cygheap_init): Set Cygwin default locale values.

* cygheap.h (struct cygheap_locale): New structure.
	(struct user_heap_info): Add cygheap_locale member locale.
	* dcrt0.cc (dll_crt0_1): Revert to calling _setlocale_r so that only
	the applications locale is reverted to "C".
	* environ.cc (environ_init): Remove unused got_lc variable.
	* fhandler.h (class dev_console): Remove now unsed locale variables.
	* fhandler_console.cc (fhandler_console::get_tty_stuff): Remove
	setting dev_console's locale members.
	(dev_console::con_to_str): Use internal locale settings.  Default to
	__ascii_wctomb if charset is "ASCII".
	(fhandler_console::write_normal): Ditto.
	* strfuncs.cc (__ascii_wctomb): Drop declaration.
	(__db_wctomb): Use fixed value 2 instead of not
	necessarily matching MB_CUR_MAX.
	(__eucjp_wctomb): Use 3 instead of MB_CUR_MAX.
	(sys_cp_wcstombs): Remove special case for "C" locale.
	(sys_wcstombs): Implement here.  Use internal locale data stored on
	cygheap.
	(sys_cp_mbstowcs): Remove special case for "C" locale.
	(sys_mbstowcs): Implement here.  Use internal locale data stored on
	cygheap.
	* syscalls.cc (internal_setlocale): New function to set cygheap locale
	data and to reset CWD posix path.
	(setlocale): Just call internal_setlocale from here if necessary.
	* wchar.h (__ascii_wctomb): Declare.
	(sys_wcstombs): Don't define inline, just declare.
	(sys_mbstowcs): Ditto.

winsup/cygwin/ChangeLog

@@ -1,3 +1,34 @@
+2009-05-14  Corinna Vinschen  <corinna@vinschen.de>
+
+	* cygheap.cc (cygheap_init): Set Cygwin default locale values.
+	* cygheap.h (struct cygheap_locale): New structure.
+	(struct user_heap_info): Add cygheap_locale member locale.
+	* dcrt0.cc (dll_crt0_1): Revert to calling _setlocale_r so that only
+	the applications locale is reverted to "C".
+	* environ.cc (environ_init): Remove unused got_lc variable.
+	* fhandler.h (class dev_console): Remove now unsed locale variables.
+	* fhandler_console.cc (fhandler_console::get_tty_stuff): Remove
+	setting dev_console's locale members.
+	(dev_console::con_to_str): Use internal locale settings.  Default to
+	__ascii_wctomb if charset is "ASCII".
+	(fhandler_console::write_normal): Ditto.
+	* strfuncs.cc (__ascii_wctomb): Drop declaration.
+	(__db_wctomb): Use fixed value 2 instead of not
+	necessarily matching MB_CUR_MAX.
+	(__eucjp_wctomb): Use 3 instead of MB_CUR_MAX.
+	(sys_cp_wcstombs): Remove special case for "C" locale.
+	(sys_wcstombs): Implement here.  Use internal locale data stored on
+	cygheap.
+	(sys_cp_mbstowcs): Remove special case for "C" locale.
+	(sys_mbstowcs): Implement here.  Use internal locale data stored on
+	cygheap.
+	* syscalls.cc (internal_setlocale): New function to set cygheap locale
+	data and to reset CWD posix path.
+	(setlocale): Just call internal_setlocale from here if necessary.
+	* wchar.h (__ascii_wctomb): Declare.
+	(sys_wcstombs): Don't define inline, just declare.
+	(sys_mbstowcs): Ditto.
+
 2009-05-14  Christopher Faylor  <me+cygwin@cgf.cx>
 
 	* mount.cc (mount_info::init): Remove MOUNT_CYGWIN_EXEC setting when

winsup/cygwin/cygheap.cc

@@ -156,6 +156,10 @@ cygheap_init ()
 					 _cygheap_mid - _cygheap_start);
       cygheap_max = cygheap;
       _csbrk (sizeof (*cygheap));
+      /* Default locale settings. */
+      cygheap->locale.mbtowc = __utf8_mbtowc;
+      cygheap->locale.wctomb = __utf8_wctomb;
+      strcpy (cygheap->locale.charset, "ASCII");
       /* Set umask to a sane default. */
       cygheap->umask = 022;
     }

winsup/cygwin/cygheap.h

@@ -258,6 +258,13 @@ struct cygheap_debug
 };
 #endif
 
+struct cygheap_locale
+{
+  mbtowc_p mbtowc;
+  wctomb_p wctomb;
+  char charset[ENCODING_LEN + 1];
+};
+
 struct user_heap_info
 {
   void *base;
@@ -281,6 +288,7 @@ struct init_cygheap
   char *buckets[32];
   cygheap_root root;
   cygheap_user user;
+  cygheap_locale locale;
   user_heap_info user_heap;
   mode_t umask;
   HANDLE console_h;

winsup/cygwin/dcrt0.cc

@@ -930,8 +930,8 @@ dll_crt0_1 (void *)
      do this for noncygwin case since the signal thread is blocked due to
      LoadLibrary serialization. */
   ld_preload ();
-  /* Reset current locale to "C" per POSIX */
+  /* Reset current application locale to "C" per POSIX */
-  setlocale (LC_CTYPE, "C");
+  _setlocale_r (_REENT, LC_CTYPE, "C");
   if (user_data->main)
     cygwin_exit (user_data->main (__argc, __argv, *user_data->envptr));
   __asm__ ("				\n\

winsup/cygwin/environ.cc

@@ -732,7 +732,6 @@ environ_init (char **envp, int envc)
   static char NO_COPY cygterm[] = "TERM=cygwin";
   myfault efault;
   tmp_pathbuf tp;
-  bool got_lc = false;
   static const char *lc_arr[] = { "LC_ALL", "LC_CTYPE", "LANG", NULL };
 
   if (efault.faulted ())
@@ -791,10 +790,7 @@ environ_init (char **envp, int envc)
 	  char *buf = (char *) alloca (i);
 	  GetEnvironmentVariableA (lc_arr[lc], buf, i);
 	  if (setlocale (LC_CTYPE, buf))
-	    {
+	    break;
-	      got_lc = true;
-	      break;
-	    }
 	}
     }
   /* We also need the CYGWIN variable early to know the value of the

winsup/cygwin/fhandler.h

@@ -894,10 +894,6 @@ class dev_console
   bool use_mouse;
   bool raw_win32_keyboard_mode;
 
-  mbtowc_p con_mbtowc;
-  wctomb_p con_wctomb;
-  char con_charset[ENCODING_LEN + 1];
-
   inline UINT get_console_cp ();
   DWORD con_to_str (char *d, int dlen, WCHAR w);
   DWORD str_to_con (mbtowc_p, char *, PWCHAR d, const char *s, DWORD sz);

winsup/cygwin/fhandler_console.cc

@@ -99,25 +99,6 @@ fhandler_console::get_tty_stuff (int flags = 0)
 	dev_state->meta_mask |= RIGHT_ALT_PRESSED;
       dev_state->set_default_attr ();
       shared_console_info->tty_min_state.sethwnd ((HWND) INVALID_HANDLE_VALUE);
-
-      /* Set the console charset and the mb<->wc conversion functions from
-	 the current locale the first time the shared console info is created.
-	 When this initialization is called, the current locale is the one
-	 used when reading the environment.  This way we get a console setting
-	 which matches the setting of LC_ALL/LC_CTYPE/LANG at the time the
-	 first Cygwin process in this console starts.
-
-	 This has an interesting effect.  If none of the above environment
-	 variables is set, the setting is equivalent to before when
-	 CYGWIN=codepage was not set:  The console charset will be the
-	 default ANSI codepage.  So it's sort of backward compatible.
-
-	 TODO: Find out if that's a feasible approach.  It might be better
-	 in the long run to have a distinct console charset environment
-	 variable. */
-      dev_state->con_mbtowc = __mbtowc;
-      dev_state->con_wctomb = __wctomb;
-      strcpy (dev_state->con_charset, __locale_charset ());
     }
 
   return &shared_console_info->tty_min_state;
@@ -146,7 +127,9 @@ tty_list::get_tty (int n)
 inline DWORD
 dev_console::con_to_str (char *d, int dlen, WCHAR w)
 {
-  return sys_cp_wcstombs (con_wctomb, con_charset, d, dlen, &w, 1);
+  return sys_cp_wcstombs (*cygheap->locale.charset == 'A'
+			  ? __ascii_wctomb : cygheap->locale.wctomb,
+			  cygheap->locale.charset, d, dlen, &w, 1);
 }
 
 inline UINT
@@ -1467,8 +1450,8 @@ fhandler_console::write_normal (const unsigned char *src,
     f_mbtowc = __set_charset_from_codepage (cp, charset = charsetbuf);
   else
     {
-      f_mbtowc = dev_state->con_mbtowc;
+      charset = cygheap->locale.charset;
-      charset = dev_state->con_charset;
+      f_mbtowc = (*charset == 'A') ? __ascii_mbtowc : cygheap->locale.mbtowc;
     }
 
   /* First check if we have cached lead bytes of a former try to write

winsup/cygwin/strfuncs.cc

@@ -50,7 +50,7 @@ __db_wctomb (struct _reent *r, char *s, wchar_t wchar, UINT cp)
 
   BOOL def_used = false;
   int ret = WideCharToMultiByte (cp, WC_NO_BEST_FIT_CHARS, &wchar, 1, s,
-				 MB_CUR_MAX, NULL, &def_used);
+				 2, NULL, &def_used);
   if (ret > 0 && !def_used)
     return ret;
 
@@ -65,8 +65,6 @@ __sjis_wctomb (struct _reent *r, char *s, wchar_t wchar, const char *charset,
   return __db_wctomb (r,s, wchar, 932);
 }
 
-extern "C" int __ascii_wctomb (struct _reent *, char *, wchar_t, const char *,
-			       mbstate_t *);
 extern "C" int
 __jis_wctomb (struct _reent *r, char *s, wchar_t wchar, const char *charset,
 	       mbstate_t *state)
@@ -101,7 +99,7 @@ __eucjp_wctomb (struct _reent *r, char *s, wchar_t wchar, const char *charset,
 
   BOOL def_used = false;
   int ret = WideCharToMultiByte (20932, WC_NO_BEST_FIT_CHARS, &wchar, 1, s,
-				 MB_CUR_MAX, NULL, &def_used);
+				 3, NULL, &def_used);
   if (ret > 0 && !def_used)
     {
       /* CP20932 representation of JIS-X-0212 character? */
@@ -418,8 +416,6 @@ sys_cp_wcstombs (wctomb_p f_wctomb, char *charset, char *dst, size_t len,
   mbstate_t ps;
   save_errno save;
 
-  if (f_wctomb == __ascii_wctomb)
-    f_wctomb = __utf8_wctomb;
   memset (&ps, 0, sizeof ps);
   if (dst == NULL)
     len = (size_t) -1;
@@ -479,6 +475,13 @@ sys_cp_wcstombs (wctomb_p f_wctomb, char *charset, char *dst, size_t len,
   return n;
 }
 
+size_t __stdcall
+sys_wcstombs (char *dst, size_t len, const wchar_t * src, size_t nwc)
+{
+  return sys_cp_wcstombs (cygheap->locale.wctomb, cygheap->locale.charset,
+			  dst, len, src, nwc);
+}
+
 /* Allocate a buffer big enough for the string, always including the
    terminating '\0'.  The buffer pointer is returned in *dst_p, the return
    value is the number of bytes written to the buffer, as usual.
@@ -527,8 +530,6 @@ sys_cp_mbstowcs (mbtowc_p f_mbtowc, char *charset, wchar_t *dst, size_t dlen,
   mbstate_t ps;
   save_errno save;
 
-  if (f_mbtowc == __ascii_mbtowc)
-    f_mbtowc = __utf8_mbtowc;
   memset (&ps, 0, sizeof ps);
   if (dst == NULL)
     len = (size_t)-1;
@@ -597,6 +598,13 @@ sys_cp_mbstowcs (mbtowc_p f_mbtowc, char *charset, wchar_t *dst, size_t dlen,
   return count;
 }
 
+size_t __stdcall
+sys_mbstowcs (wchar_t * dst, size_t dlen, const char *src, size_t nms)
+{
+  return sys_cp_mbstowcs (cygheap->locale.mbtowc, cygheap->locale.charset,
+			  dst, dlen, src, nms);
+}
+
 /* Same as sys_wcstombs_alloc, just backwards. */
 size_t __stdcall
 sys_mbstowcs_alloc (wchar_t **dst_p, int type, const char *src, size_t nms)

winsup/cygwin/syscalls.cc

@@ -4006,15 +4006,32 @@ unlinkat (int dirfd, const char *pathname, int flags)
   return (flags & AT_REMOVEDIR) ? rmdir (path) : unlink (path);
 }
 
-extern "C" char *
+static void
-setlocale (int category, const char *locale)
+internal_setlocale ()
 {
+  if (*cygheap->locale.charset == 'A')
+    {
+      cygheap->locale.mbtowc = __utf8_mbtowc;
+      cygheap->locale.wctomb = __utf8_wctomb;
+    }
+  else
+    {
+      cygheap->locale.mbtowc = __mbtowc;
+      cygheap->locale.wctomb = __wctomb;
+    }
+  strcpy (cygheap->locale.charset, __locale_charset ());
   /* Each setlocale potentially changes the multibyte representation
      of the CWD.  Therefore we have to rest the CWD's posix path and
      reevaluate the next time it's used. */
   /* FIXME: Other buffered paths might be affected as well. */
+  cygheap->cwd.reset_posix ();
+}
+
+extern "C" char *
+setlocale (int category, const char *locale)
+{
   char *ret = _setlocale_r (_REENT, category, locale);
-  if (ret)
+  if (ret && locale && (category == LC_ALL || category == LC_CTYPE))
-    cygheap->cwd.reset_posix ();
+    internal_setlocale ();
   return ret;
 }

winsup/cygwin/wchar.h

@@ -34,6 +34,7 @@ typedef int wctomb_f (struct _reent *, char *, wchar_t, const char *,
 typedef wctomb_f *wctomb_p;
 
 extern wctomb_p __wctomb;
+extern wctomb_f __ascii_wctomb;
 extern wctomb_f __utf8_wctomb;
 
 extern char *__locale_charset ();
@@ -48,12 +49,9 @@ extern mbtowc_p __set_charset_from_codepage (unsigned int cp, char *charset);
 size_t __stdcall sys_cp_wcstombs (wctomb_p, char *, char *, size_t,
 				  const wchar_t *, size_t = (size_t) -1)
        __attribute__ ((regparm(3)));
-inline size_t
+size_t __stdcall sys_wcstombs (char *dst, size_t len, const wchar_t * src,
-__stdcall sys_wcstombs (char *dst, size_t len, const wchar_t * src,
+			       size_t nwc = (size_t) -1)
-			size_t nwc = (size_t) -1)
+       __attribute__ ((regparm(3)));
-{
-  return sys_cp_wcstombs (__wctomb, __locale_charset (), dst, len, src, nwc);
-}
 size_t __stdcall sys_wcstombs_alloc (char **, int, const wchar_t *,
 				     size_t = (size_t) -1)
        __attribute__ ((regparm(3)));
@@ -61,12 +59,9 @@ size_t __stdcall sys_wcstombs_alloc (char **, int, const wchar_t *,
 size_t __stdcall sys_cp_mbstowcs (mbtowc_p, char *, wchar_t *, size_t,
 				  const char *, size_t = (size_t) -1)
        __attribute__ ((regparm(3)));
-inline size_t
+size_t sys_mbstowcs (wchar_t * dst, size_t dlen, const char *src,
-sys_mbstowcs (wchar_t * dst, size_t dlen, const char *src,
+		     size_t nms = (size_t) -1)
-	      size_t nms = (size_t) -1)
+       __attribute__ ((regparm(3)));
-{
-  return sys_cp_mbstowcs (__mbtowc, __locale_charset (), dst, dlen, src, nms);
-}
 size_t __stdcall sys_mbstowcs_alloc (wchar_t **, int, const char *,
 				     size_t = (size_t) -1)
        __attribute__ ((regparm(3)));