diff -r 0d41a1b4c4fe Lib/test/test_codecs.py
--- a/Lib/test/test_codecs.py	Mon Oct 08 07:46:11 2012 +0200
+++ b/Lib/test/test_codecs.py	Mon Oct 08 18:26:34 2012 +0300
@@ -330,7 +330,7 @@
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "", # first byte of BOM read
                 "", # second byte of BOM read
@@ -352,6 +352,10 @@
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -386,7 +390,7 @@
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "",
                 "",
@@ -404,6 +408,10 @@
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -426,7 +434,7 @@
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "",
                 "",
@@ -444,6 +452,10 @@
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -494,7 +506,7 @@
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "", # first byte of BOM read
                 "", # second byte of BOM read => byteorder known
@@ -506,6 +518,10 @@
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -543,7 +559,7 @@
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "",
                 "\x00",
@@ -553,6 +569,10 @@
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -582,7 +602,7 @@
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u0100\uffff",
+            "\x00\xff\u0100\uffff\U00010000",
             [
                 "",
                 "\x00",
@@ -592,6 +612,10 @@
                 "\x00\xff\u0100",
                 "\x00\xff\u0100",
                 "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff",
+                "\x00\xff\u0100\uffff\U00010000",
             ]
         )
 
@@ -621,7 +645,7 @@
 
     def test_partial(self):
         self.check_partial(
-            "\x00\xff\u07ff\u0800\uffff",
+            "\x00\xff\u07ff\u0800\uffff\U00010000",
             [
                 "\x00",
                 "\x00",
@@ -634,6 +658,10 @@
                 "\x00\xff\u07ff\u0800",
                 "\x00\xff\u07ff\u0800",
                 "\x00\xff\u07ff\u0800\uffff",
+                "\x00\xff\u07ff\u0800\uffff",
+                "\x00\xff\u07ff\u0800\uffff",
+                "\x00\xff\u07ff\u0800\uffff",
+                "\x00\xff\u07ff\u0800\uffff\U00010000",
             ]
         )
 
@@ -812,7 +840,7 @@
 
     def test_partial(self):
         self.check_partial(
-            "\ufeff\x00\xff\u07ff\u0800\uffff",
+            "\ufeff\x00\xff\u07ff\u0800\uffff\U00010000",
             [
                 "",
                 "",
@@ -831,6 +859,10 @@
                 "\ufeff\x00\xff\u07ff\u0800",
                 "\ufeff\x00\xff\u07ff\u0800",
                 "\ufeff\x00\xff\u07ff\u0800\uffff",
+                "\ufeff\x00\xff\u07ff\u0800\uffff",
+                "\ufeff\x00\xff\u07ff\u0800\uffff",
+                "\ufeff\x00\xff\u07ff\u0800\uffff",
+                "\ufeff\x00\xff\u07ff\u0800\uffff\U00010000",
             ]
         )
 
diff -r 0d41a1b4c4fe Objects/unicodeobject.c
--- a/Objects/unicodeobject.c	Mon Oct 08 07:46:11 2012 +0200
+++ b/Objects/unicodeobject.c	Mon Oct 08 18:26:34 2012 +0300
@@ -5143,8 +5143,11 @@
             /* The remaining input chars are ignored if the callback
                chooses to skip the input */
         case 1:
+            q -= 2;
+            if (consumed)
+                goto End;
             errmsg = "unexpected end of data";
-            startinpos = ((const char *)q) - 2 - starts;
+            startinpos = ((const char *)q) - starts;
             endinpos = ((const char *)e) - starts;
             break;
         case 2: