bpo-30061: Check if PyObject_Size()/PySequence_Size()/PyMapping_Size() (#1096)

Lib/test/test_io.py

@@ -543,6 +543,22 @@ def test_readline(self):
         with self.open(support.TESTFN, "r") as f:
             self.assertRaises(TypeError, f.readline, 5.3)
 
+    def test_readline_nonsizeable(self):
+        # Issue #30061
+        # Crash when readline() returns an object without __len__
+        class R(self.IOBase):
+            def readline(self):
+                return None
+        self.assertRaises((TypeError, StopIteration), next, R())
+
+    def test_next_nonsizeable(self):
+        # Issue #30061
+        # Crash when __next__() returns an object without __len__
+        class R(self.IOBase):
+            def __next__(self):
+                return None
+        self.assertRaises(TypeError, R().readlines, 1)
+
     def test_raw_bytes_io(self):
         f = self.BytesIO()
         self.write_ops(f)

Misc/NEWS

@@ -32,6 +32,11 @@ Core and Builtins
 Library
 -------
 
+- bpo-30061: Fixed crashes in IOBase methods __next__() and readlines() when
+  readline() or __next__() respectively return non-sizeable object.
+  Fixed possible other errors caused by not checking results of PyObject_Size(),
+  PySequence_Size(), or PyMapping_Size().
+
 - bpo-30017: Allowed calling the close() method of the zip entry writer object
   multiple times.  Writing to a closed writer now always produces a ValueError.
 

Modules/_io/iobase.c

@@ -625,7 +625,8 @@ iobase_iternext(PyObject *self)
     if (line == NULL)
         return NULL;
 
-    if (PyObject_Size(line) == 0) {
+    if (PyObject_Size(line) <= 0) {
+        /* Error or empty */
         Py_DECREF(line);
         return NULL;
     }
@@ -676,6 +677,7 @@ _io__IOBase_readlines_impl(PyObject *self, Py_ssize_t hint)
     }
 
     while (1) {
+        Py_ssize_t line_length;
         PyObject *line = PyIter_Next(it);
         if (line == NULL) {
             if (PyErr_Occurred()) {
@@ -689,11 +691,14 @@ _io__IOBase_readlines_impl(PyObject *self, Py_ssize_t hint)
             Py_DECREF(line);
             goto error;
         }
-        length += PyObject_Size(line);
+        line_length = PyObject_Size(line);
         Py_DECREF(line);
-
-        if (length > hint)
+        if (line_length < 0) {
+            goto error;
+        }
+        if (line_length > hint - length)
             break;
+        length += line_length;
     }
 
     Py_DECREF(it);

Modules/_winapi.c

@@ -722,17 +722,22 @@ getenvironment(PyObject* environment)
         return NULL;
     }
 
-    envsize = PyMapping_Length(environment);
-
     keys = PyMapping_Keys(environment);
     values = PyMapping_Values(environment);
     if (!keys || !values)
         goto error;
 
+    envsize = PySequence_Fast_GET_SIZE(keys);
+    if (PySequence_Fast_GET_SIZE(values) != envsize) {
+        PyErr_SetString(PyExc_RuntimeError,
+            "environment changed size during iteration");
+        goto error;
+    }
+
     totalsize = 1; /* trailing null character */
     for (i = 0; i < envsize; i++) {
-        PyObject* key = PyList_GET_ITEM(keys, i);
-        PyObject* value = PyList_GET_ITEM(values, i);
+        PyObject* key = PySequence_Fast_GET_ITEM(keys, i);
+        PyObject* value = PySequence_Fast_GET_ITEM(values, i);
 
         if (! PyUnicode_Check(key) || ! PyUnicode_Check(value)) {
             PyErr_SetString(PyExc_TypeError,
@@ -760,8 +765,8 @@ getenvironment(PyObject* environment)
     end = buffer + totalsize;
 
     for (i = 0; i < envsize; i++) {
-        PyObject* key = PyList_GET_ITEM(keys, i);
-        PyObject* value = PyList_GET_ITEM(values, i);
+        PyObject* key = PySequence_Fast_GET_ITEM(keys, i);
+        PyObject* value = PySequence_Fast_GET_ITEM(values, i);
         if (!PyUnicode_AsUCS4(key, p, end - p, 0))
             goto error;
         p += PyUnicode_GET_LENGTH(key);

Modules/cjkcodecs/multibytecodec.c

@@ -1670,6 +1670,9 @@ _multibytecodec_MultibyteStreamWriter_writelines(MultibyteStreamWriterObject *se
         if (r == -1)
             return NULL;
     }
+    /* PySequence_Length() can fail */
+    if (PyErr_Occurred())
+        return NULL;
 
     Py_RETURN_NONE;
 }

Modules/posixmodule.c

@@ -6650,14 +6650,17 @@ static PyObject *
 os_setgroups(PyObject *module, PyObject *groups)
 /*[clinic end generated code: output=3fcb32aad58c5ecd input=fa742ca3daf85a7e]*/
 {
-    int i, len;
+    Py_ssize_t i, len;
     gid_t grouplist[MAX_GROUPS];
 
     if (!PySequence_Check(groups)) {
         PyErr_SetString(PyExc_TypeError, "setgroups argument must be a sequence");
         return NULL;
     }
     len = PySequence_Size(groups);
+    if (len < 0) {
+        return NULL;
+    }
     if (len > MAX_GROUPS) {
         PyErr_SetString(PyExc_ValueError, "too many groups");
         return NULL;
@@ -7886,9 +7889,9 @@ os_read_impl(PyObject *module, int fd, Py_ssize_t length)
 #if (defined(HAVE_SENDFILE) && (defined(__FreeBSD__) || defined(__DragonFly__) \
     || defined(__APPLE__))) || defined(HAVE_READV) || defined(HAVE_WRITEV)
 static Py_ssize_t
-iov_setup(struct iovec **iov, Py_buffer **buf, PyObject *seq, int cnt, int type)
+iov_setup(struct iovec **iov, Py_buffer **buf, PyObject *seq, Py_ssize_t cnt, int type)
 {
-    int i, j;
+    Py_ssize_t i, j;
     Py_ssize_t blen, total = 0;
 
     *iov = PyMem_New(struct iovec, cnt);
@@ -7965,8 +7968,7 @@ static Py_ssize_t
 os_readv_impl(PyObject *module, int fd, PyObject *buffers)
 /*[clinic end generated code: output=792da062d3fcebdb input=e679eb5dbfa0357d]*/
 {
-    int cnt;
-    Py_ssize_t n;
+    Py_ssize_t cnt, n;
     int async_err = 0;
     struct iovec *iov;
     Py_buffer *buf;
@@ -7978,6 +7980,8 @@ os_readv_impl(PyObject *module, int fd, PyObject *buffers)
     }
 
     cnt = PySequence_Size(buffers);
+    if (cnt < 0)
+        return -1;
 
     if (iov_setup(&iov, &buf, buffers, cnt, PyBUF_WRITABLE) < 0)
         return -1;
@@ -8116,15 +8120,24 @@ posix_sendfile(PyObject *self, PyObject *args, PyObject *kwdict)
                 "sendfile() headers must be a sequence");
             return NULL;
         } else {
-            Py_ssize_t i = 0; /* Avoid uninitialized warning */
-            sf.hdr_cnt = PySequence_Size(headers);
-            if (sf.hdr_cnt > 0 &&
-                (i = iov_setup(&(sf.headers), &hbuf,
-                                headers, sf.hdr_cnt, PyBUF_SIMPLE)) < 0)
+            Py_ssize_t i = PySequence_Size(headers);
+            if (i < 0)
+                return NULL;
+            if (i > INT_MAX) {
+                PyErr_SetString(PyExc_OverflowError,
+                    "sendfile() header is too large");
                 return NULL;
+            }
+            if (i > 0) {
+                sf.hdr_cnt = (int)i;
+                i = iov_setup(&(sf.headers), &hbuf,
+                              headers, sf.hdr_cnt, PyBUF_SIMPLE);
+                if (i < 0)
+                    return NULL;
 #ifdef __APPLE__
-            sbytes += i;
+                sbytes += i;
 #endif
+            }
         }
     }
     if (trailers != NULL) {
@@ -8133,15 +8146,24 @@ posix_sendfile(PyObject *self, PyObject *args, PyObject *kwdict)
                 "sendfile() trailers must be a sequence");
             return NULL;
         } else {
-            Py_ssize_t i = 0; /* Avoid uninitialized warning */
-            sf.trl_cnt = PySequence_Size(trailers);
-            if (sf.trl_cnt > 0 &&
-                (i = iov_setup(&(sf.trailers), &tbuf,
-                                trailers, sf.trl_cnt, PyBUF_SIMPLE)) < 0)
+            Py_ssize_t i = PySequence_Size(trailers);
+            if (i < 0)
+                return NULL;
+            if (i > INT_MAX) {
+                PyErr_SetString(PyExc_OverflowError,
+                    "sendfile() trailer is too large");
                 return NULL;
+            }
+            if (i > 0) {
+                sf.trl_cnt = (int)i;
+                i = iov_setup(&(sf.trailers), &tbuf,
+                              trailers, sf.trl_cnt, PyBUF_SIMPLE);
+                if (i < 0)
+                    return NULL;
 #ifdef __APPLE__
-            sbytes += i;
+                sbytes += i;
 #endif
+            }
         }
     }
 
@@ -8411,7 +8433,7 @@ static Py_ssize_t
 os_writev_impl(PyObject *module, int fd, PyObject *buffers)
 /*[clinic end generated code: output=56565cfac3aac15b input=5b8d17fe4189d2fe]*/
 {
-    int cnt;
+    Py_ssize_t cnt;
     Py_ssize_t result;
     int async_err = 0;
     struct iovec *iov;
@@ -8423,6 +8445,8 @@ os_writev_impl(PyObject *module, int fd, PyObject *buffers)
         return -1;
     }
     cnt = PySequence_Size(buffers);
+    if (cnt < 0)
+        return -1;
 
     if (iov_setup(&iov, &buf, buffers, cnt, PyBUF_SIMPLE) < 0) {
         return -1;

Objects/setobject.c

@@ -1550,20 +1550,26 @@ set_difference(PySetObject *so, PyObject *other)
     PyObject *key;
     Py_hash_t hash;
     setentry *entry;
-    Py_ssize_t pos = 0;
+    Py_ssize_t pos = 0, other_size;
     int rv;
 
     if (PySet_GET_SIZE(so) == 0) {
         return set_copy(so);
     }
 
-    if (!PyAnySet_Check(other)  && !PyDict_CheckExact(other)) {
+    if (PyAnySet_Check(other)) {
+        other_size = PySet_GET_SIZE(other);
+    }
+    else if (PyDict_CheckExact(other)) {
+        other_size = PyDict_Size(other);
+    }
+    else {
         return set_copy_and_difference(so, other);
     }
 
     /* If len(so) much more than len(other), it's more efficient to simply copy
      * so and then iterate other looking for common elements. */
-    if ((PySet_GET_SIZE(so) >> 2) > PyObject_Size(other)) {
+    if ((PySet_GET_SIZE(so) >> 2) > other_size) {
         return set_copy_and_difference(so, other);
     }