PyUnicode_EncodeUTF8(): tightened the memory asserts a bit, and at least

tried to catch some possible arithmetic overflows in the debug build.

diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c
index 596cb3830b481a6368bf71de253e4fac7979d126..8dbca6de1ee3b9aed3e5e3bd60b3d6e855d76b46 100644 (file)
--- a/Objects/unicodeobject.c
+++ b/Objects/unicodeobject.c
@@ -1185,15 +1185,17 @@ PyUnicode_EncodeUTF8(const Py_UNICODE *s,
 {
     PyObject *v;
     char *p;
-    int i = 0;
-    int overalloc = 2;
     int len;
-    
-    /* Short-cut for emtpy strings */
+    int i = 0;
+    long overalloc = 2;
+    int nallocated;  /* overalloc * size; PyString_ adds one more for \0 */
+
+    /* Short-cut for empty strings */
     if (size == 0)
        return PyString_FromStringAndSize(NULL, 0);
 
-    v = PyString_FromStringAndSize(NULL, overalloc * size);
+    nallocated = Py_SAFE_DOWNCAST(overalloc * size, long, int);
+    v = PyString_FromStringAndSize(NULL, nallocated);
     if (v == NULL)
         return NULL;
 
@@ -1211,7 +1213,7 @@ PyUnicode_EncodeUTF8(const Py_UNICODE *s,
             *p++ = (char)(0xc0 | (ch >> 6));
             *p++ = (char)(0x80 | (ch & 0x3f));
         }
-                        
+
         else {
            /* Encode UCS2 Unicode ordinals */
            if (ch < 0x10000) {
@@ -1230,9 +1232,11 @@ PyUnicode_EncodeUTF8(const Py_UNICODE *s,
                 }
 
                if (overalloc < 3) {
-                   len = (int)(p - PyString_AS_STRING(v));
+                   len = Py_SAFE_DOWNCAST(p-PyString_AS_STRING(v), long, int);
+                    assert(len <= nallocated);
                    overalloc = 3;
-                   if (_PyString_Resize(&v, overalloc * size))
+                    nallocated = Py_SAFE_DOWNCAST(overalloc * size, long, int);
+                   if (_PyString_Resize(&v, nallocated))
                        goto onError;
                    p = PyString_AS_STRING(v) + len;
                }
@@ -1245,9 +1249,11 @@ PyUnicode_EncodeUTF8(const Py_UNICODE *s,
            /* Encode UCS4 Unicode ordinals */
        encodeUCS4:
            if (overalloc < 4) {
-               len = (int)(p - PyString_AS_STRING(v));
+                len = Py_SAFE_DOWNCAST(p - PyString_AS_STRING(v), long, int);
+                assert(len <= nallocated);
                overalloc = 4;
-               if (_PyString_Resize(&v, overalloc * size))
+                nallocated = Py_SAFE_DOWNCAST(overalloc * size, long, int);
+               if (_PyString_Resize(&v, nallocated))
                    goto onError;
                p = PyString_AS_STRING(v) + len;
            }
@@ -1257,9 +1263,11 @@ PyUnicode_EncodeUTF8(const Py_UNICODE *s,
            *p++ = (char)(0x80 | (ch & 0x3f));
        }
     }
+
     *p = '\0';
-    assert((p - PyString_AS_STRING(v)) <= overalloc*size);
-    if (_PyString_Resize(&v, (int)(p - PyString_AS_STRING(v))))
+    len = Py_SAFE_DOWNCAST(p - PyString_AS_STRING(v), long, int);
+    assert(len <= nallocated);
+    if (_PyString_Resize(&v, len))
        goto onError;
     return v;