POSIX-1.2008 per-thread locales, groundwork part 1

Introduce first cut of struct _thr_locale_t used for the locale_t definition.
Introduce global instance called __global_locale used by default.
Introduce internal inline functions __get_global_locale, __get_locale_r,
__get_current_locale.

Remove usage of global variables in favor of accessor functions pointing to
__global_locale for now.  Include all local headers in locale subdir from
setlocale.h to get single include for internal locale access.

Introduce __CTYPE_PTR macro to replace direct access to __ctype_ptr__
and use throughout in isxxx functions.

Signed-off by: Corinna Vinschen <corinna@vinschen.de>

4 files changed, 65 insertions, 37 deletions

diff --git a/winsup/cygwin/common.din b/winsup/cygwin/common.din
index d8df00e..acb3fab 100644
--- a/winsup/cygwin/common.din
+++ b/winsup/cygwin/common.din
@@ -68,6 +68,7 @@ __isinfd = isinf NOSIGFE
 __isinff = isinff NOSIGFE
 __isnand = isnan NOSIGFE
 __isnanf = isnanf NOSIGFE
+__locale_ctype_ptr NOSIGFE
 __locale_mb_cur_max NOSIGFE
 __main NOSIGFE
 __mempcpy = mempcpy NOSIGFE
diff --git a/winsup/cygwin/ctype.cc b/winsup/cygwin/ctype.cc
index fa8aad3..94de81e 100644
--- a/winsup/cygwin/ctype.cc
+++ b/winsup/cygwin/ctype.cc
@@ -1,6 +1,7 @@
 #include "winsup.h"
 extern "C" {
 #include <ctype.h>
+#include "../locale/setlocale.h"
 #include <stdlib.h>
 #include <wctype.h>
 
@@ -18,9 +19,10 @@ extern const char __ctype_cp[22][128 + 256];		/* Newlib */
 extern const char __ctype_iso[15][128 + 256];		/* Newlib */
 
 void
-__set_ctype (const char *charset)
+__set_ctype (struct _reent *reent, const char *charset)
 {
   int idx;
+  char *ctype_ptr = NULL;
 
   switch (*charset)
     {
@@ -36,8 +38,8 @@ __set_ctype (const char *charset)
 	  memcpy (_ctype_b, __ctype_iso[idx], 128);
 	  memcpy (_ctype_b + 256, __ctype_iso[idx] + 256, 128);
 	}
-      __ctype_ptr__ = (char *) (__ctype_iso[idx] + 127);
-      return;
+      ctype_ptr = (char *) __ctype_iso[idx];
+      break;
     case 'C':
       idx = __cp_index (charset + 2);
       if (idx < 0)
@@ -47,17 +49,24 @@ __set_ctype (const char *charset)
 	  memcpy (_ctype_b, __ctype_cp[idx], 128);
 	  memcpy (_ctype_b + 256, __ctype_cp[idx] + 256, 128);
 	}
-      __ctype_ptr__ = (char *) (__ctype_cp[idx] + 127);
-      return;
+      ctype_ptr = (char *) __ctype_cp[idx];
+      break;
     default:
       break;
     }
-  if (CYGWIN_VERSION_CHECK_FOR_OLD_CTYPE)
+  if (!ctype_ptr)
     {
-      memset (_ctype_b, 0, 128);
-      memset (_ctype_b + 256, 0, 128);
+      if (CYGWIN_VERSION_CHECK_FOR_OLD_CTYPE)
+	{
+	  memset (_ctype_b, 0, 128);
+	  memset (_ctype_b + 256, 0, 128);
+	}
+      ctype_ptr = (char *) _ctype_b;
     }
-  __ctype_ptr__ = (char *) _ctype_b + 127;
+  if (reent)
+    __get_locale_r (reent)->ctype_ptr = ctype_ptr + 127;
+  else
+    __ctype_ptr__ = ctype_ptr + 127;
 }
 
 } /* extern "C" */
diff --git a/winsup/cygwin/nlsfuncs.cc b/winsup/cygwin/nlsfuncs.cc
index 72e5f45..79ab3d1 100644
--- a/winsup/cygwin/nlsfuncs.cc
+++ b/winsup/cygwin/nlsfuncs.cc
@@ -17,11 +17,7 @@ details. */
 #include "cygheap.h"
 #include "tls_pbuf.h"
 /* Internal headers from newlib */
-#include "../locale/timelocal.h"
-#include "../locale/lctype.h"
-#include "../locale/lnumeric.h"
-#include "../locale/lmonetary.h"
-#include "../locale/lmessages.h"
+#include "../locale/setlocale.h"
 #include "lc_msg.h"
 #include "lc_era.h"
 
@@ -689,8 +685,6 @@ __set_lc_time_from_win (const char *name,
   if (tmp != new_lc_time_buf)
     rebase_locale_buf (_time_locale, _time_locale + 1, tmp,
 		       new_lc_time_buf, lc_time_ptr);
-  if (*lc_time_buf)
-    free (*lc_time_buf);
   *lc_time_buf = tmp;
   return 1;
 }
@@ -764,8 +758,6 @@ __set_lc_ctype_from_win (const char *name,
   if (tmp != new_lc_ctype_buf)
     rebase_locale_buf (_ctype_locale, _ctype_locale + 1, tmp,
 		       new_lc_ctype_buf, lc_ctype_ptr);
-  if (*lc_ctype_buf)
-    free (*lc_ctype_buf);
   *lc_ctype_buf = tmp;
   return 1;
 }
@@ -841,8 +833,6 @@ __set_lc_numeric_from_win (const char *name,
   if (tmp != new_lc_numeric_buf)
     rebase_locale_buf (_numeric_locale, _numeric_locale + 1, tmp,
 		       new_lc_numeric_buf, lc_numeric_ptr);
-  if (*lc_numeric_buf)
-    free (*lc_numeric_buf);
   *lc_numeric_buf = tmp;
   return 1;
 }
@@ -981,8 +971,6 @@ __set_lc_monetary_from_win (const char *name,
   if (tmp != new_lc_monetary_buf)
     rebase_locale_buf (_monetary_locale, _monetary_locale + 1, tmp,
 		       new_lc_monetary_buf, lc_monetary_ptr);
-  if (*lc_monetary_buf)
-    free (*lc_monetary_buf);
   *lc_monetary_buf = tmp;
   return 1;
 }
@@ -1083,36 +1071,59 @@ __set_lc_messages_from_win (const char *name,
       _messages_locale->wnostr = (const wchar_t *) wc;
       wcpcpy (wc, msg->nostr);
     }
-  /* Aftermath. */
-  if (*lc_messages_buf)
-    free (*lc_messages_buf);
   *lc_messages_buf = new_lc_messages_buf;
   return 1;
 }
 
-LCID collate_lcid = 0;
-static mbtowc_p collate_mbtowc = __ascii_mbtowc;
-char collate_charset[ENCODING_LEN + 1] = "ASCII";
+struct lc_collate_T
+{
+  LCID lcid;
+  mbtowc_p mbtowc;
+  char codeset[ENCODING_LEN + 1];
+};
 
 /* Called from newlib's setlocale() if category is LC_COLLATE.  Stores
    LC_COLLATE locale information.  This is subsequently accessed by the
    below functions strcoll, strxfrm, wcscoll, wcsxfrm. */
 extern "C" int
-__collate_load_locale (const char *name, mbtowc_p f_mbtowc, const char *charset)
+__collate_load_locale (struct _thr_locale_t *locale, const char *name,
+		       mbtowc_p f_mbtowc, const char *charset)
 {
   LCID lcid = __get_lcid_from_locale (name);
   if (lcid == (LCID) -1)
     return -1;
-  collate_lcid = lcid;
-  collate_mbtowc = f_mbtowc;
-  stpcpy (collate_charset, charset);
+  struct lc_collate_T *cop;
+  if (lcid)
+    {
+      cop = (struct lc_collate_T *) calloc (1, sizeof *cop);
+      if (!cop)
+	return -1;
+      cop->lcid = lcid;
+      cop->mbtowc = f_mbtowc;
+      stpcpy (cop->codeset, charset);
+    }
   return 0;
 }
 
+extern "C" LCID
+__get_current_collate_lcid ()
+{
+  struct _thr_locale_t *cur_locale = __get_current_locale ();
+  return cur_locale->collate ? cur_locale->collate->lcid : 0;
+}
+
 extern "C" const char *
 __get_current_collate_codeset (void)
 {
-  return collate_charset;
+  struct _thr_locale_t *cur_locale = __get_current_locale ();
+  return cur_locale->collate ? cur_locale->collate->codeset : "ASCII";
+}
+
+static mbtowc_p
+__get_current_collate_mbtowc ()
+{
+  struct _thr_locale_t *cur_locale = __get_current_locale ();
+  return cur_locale->collate ? cur_locale->collate->mbtowc : __ascii_mbtowc;
 }
 
 /* We use the Windows functions for locale-specific string comparison and
@@ -1122,6 +1133,7 @@ extern "C" int
 wcscoll (const wchar_t *__restrict ws1, const wchar_t *__restrict ws2)
 {
   int ret;
+  LCID collate_lcid = __get_current_collate_lcid ();
 
   if (!collate_lcid)
     return wcscmp (ws1, ws2);
@@ -1138,11 +1150,14 @@ strcoll (const char *__restrict s1, const char *__restrict s2)
   wchar_t *ws1, *ws2;
   tmp_pathbuf tp;
   int ret;
+  LCID collate_lcid = __get_current_collate_lcid ();
 
   if (!collate_lcid)
     return strcmp (s1, s2);
   /* The ANSI version of CompareString uses the default charset of the lcid,
      so we must use the Unicode version. */
+  mbtowc_p collate_mbtowc = __get_current_collate_mbtowc ();
+  const char *collate_charset = __get_current_collate_codeset ();
   n1 = lc_mbstowcs (collate_mbtowc, collate_charset, NULL, s1, 0) + 1;
   ws1 = (n1 > NT_MAX_PATH ? (wchar_t *) malloc (n1 * sizeof (wchar_t))
 			  : tp.w_get ());
@@ -1176,6 +1191,7 @@ extern "C" size_t
 wcsxfrm (wchar_t *__restrict ws1, const wchar_t *__restrict ws2, size_t wsn)
 {
   size_t ret;
+  LCID collate_lcid = __get_current_collate_lcid ();
 
   if (!collate_lcid)
     return wcslcpy (ws1, ws2, wsn);
@@ -1211,11 +1227,14 @@ strxfrm (char *__restrict s1, const char *__restrict s2, size_t sn)
   size_t n2;
   wchar_t *ws2;
   tmp_pathbuf tp;
+  LCID collate_lcid = __get_current_collate_lcid ();
 
   if (!collate_lcid)
     return strlcpy (s1, s2, sn);
   /* The ANSI version of LCMapString uses the default charset of the lcid,
      so we must use the Unicode version. */
+  mbtowc_p collate_mbtowc = __get_current_collate_mbtowc ();
+  const char *collate_charset = __get_current_collate_codeset ();
   n2 = lc_mbstowcs (collate_mbtowc, collate_charset, NULL, s2, 0) + 1;
   ws2 = (n2 > NT_MAX_PATH ? (wchar_t *) malloc (n2 * sizeof (wchar_t))
 			  : tp.w_get ());
diff --git a/winsup/cygwin/regex/regcomp.c b/winsup/cygwin/regex/regcomp.c
index 554b43a..50d5dad 100644
--- a/winsup/cygwin/regex/regcomp.c
+++ b/winsup/cygwin/regex/regcomp.c
@@ -63,9 +63,8 @@ __FBSDID("$FreeBSD: src/lib/libc/regex/regcomp.c,v 1.36 2007/06/11 03:05:54 delp
 #include "cname.h"
 
 #ifdef __CYGWIN__
-/* These are defined in nlsfuncs.cc. */
-extern LCID collate_lcid;
-extern char collate_charset[];
+/* Defined in nlsfuncs.cc. */
+extern LCID __get_current_collate_lcid ();
 #endif
 
 /*
@@ -832,7 +831,7 @@ p_b_term(struct parse *p, cset *cs)
 			CHadd(p, cs, start);
 		else {
 #ifdef __CYGWIN__
-			if (!collate_lcid) {
+			if (!__get_current_collate_lcid ()) {
 #else
 			if (__collate_load_error) {
 #endif