Merged revisions 64114 via svnmerge from

svn+ssh://pythondev@svn.python.org/python/trunk

........
  r64114 | gregory.p.smith | 2008-06-11 09:41:16 +0200 (mer., 11 juin 2008) | 6 lines

  Merge in release25-maint r60793:

   Added checks for integer overflows, contributed by Google. Some are
   only available if asserts are left in the code, in cases where they
   can't be triggered from Python code.
........

6 files changed, 172 insertions, 26 deletions

diff --git a/Modules/_csv.c b/Modules/_csv.c
index c654712..430ccdc 100644
--- a/Modules/_csv.c
+++ b/Modules/_csv.c
@@ -533,6 +533,10 @@ parse_grow_buff(ReaderObj *self)
 		self->field = PyMem_New(Py_UNICODE, self->field_size);
 	}
 	else {
+		if (self->field_size > INT_MAX / 2) {
+			PyErr_NoMemory();
+			return 0;
+		} 
 		self->field_size *= 2;
 		self->field = PyMem_Resize(self->field, Py_UNICODE,
 					   self->field_size);
@@ -1038,6 +1042,12 @@ join_append_data(WriterObj *self, Py_UNICODE *field, int quote_empty,
 static int
 join_check_rec_size(WriterObj *self, int rec_len)
 {
+
+	if (rec_len < 0 || rec_len > INT_MAX - MEM_INCR) {
+		PyErr_NoMemory();
+		return 0;
+	}
+
 	if (rec_len > self->rec_size) {
 		if (self->rec_size == 0) {
 			self->rec_size = (rec_len / MEM_INCR + 1) * MEM_INCR;
diff --git a/Modules/arraymodule.c b/Modules/arraymodule.c
index 3e26369..d2ae367 100644
--- a/Modules/arraymodule.c
+++ b/Modules/arraymodule.c
@@ -642,6 +642,9 @@ array_concat(arrayobject *a, PyObject *bb)
 		PyErr_BadArgument();
 		return NULL;
 	}
+	if (Py_SIZE(a) > PY_SSIZE_T_MAX - Py_SIZE(b)) {
+		return PyErr_NoMemory();
+	}
 	size = Py_SIZE(a) + Py_SIZE(b);
 	np = (arrayobject *) newarrayobject(&Arraytype, size, a->ob_descr);
 	if (np == NULL) {
@@ -664,6 +667,9 @@ array_repeat(arrayobject *a, Py_ssize_t n)
 	Py_ssize_t nbytes;
 	if (n < 0)
 		n = 0;
+	if ((Py_SIZE(a) != 0) && (n > PY_SSIZE_T_MAX / Py_SIZE(a))) {
+		return PyErr_NoMemory();
+	}
 	size = Py_SIZE(a) * n;
 	np = (arrayobject *) newarrayobject(&Arraytype, size, a->ob_descr);
 	if (np == NULL)
@@ -853,6 +859,10 @@ array_inplace_repeat(arrayobject *self, Py_ssize_t n)
 		if (n < 0)
 			n = 0;
 		items = self->ob_item;
+		if ((self->ob_descr->itemsize != 0) && 
+			(Py_SIZE(self) > PY_SSIZE_T_MAX / self->ob_descr->itemsize)) {
+			return PyErr_NoMemory();
+		}
 		size = Py_SIZE(self) * self->ob_descr->itemsize;
 		if (n == 0) {
 			PyMem_FREE(items);
@@ -861,6 +871,9 @@ array_inplace_repeat(arrayobject *self, Py_ssize_t n)
 			self->allocated = 0;
 		}
 		else {
+			if (size > PY_SSIZE_T_MAX / n) {
+				return PyErr_NoMemory();
+			}
 			PyMem_Resize(items, char, n * size);
 			if (items == NULL)
 				return PyErr_NoMemory();
@@ -1142,6 +1155,10 @@ array_reduce(arrayobject *array)
 		Py_INCREF(dict);
 	}
 	if (Py_SIZE(array) > 0) {
+		if (array->ob_descr->itemsize 
+				> PY_SSIZE_T_MAX / Py_SIZE(array)) {
+			return PyErr_NoMemory();
+		}
 		result = Py_BuildValue("O(cy#)O", 
 			Py_TYPE(array), 
 			array->ob_descr->typecode,
@@ -1315,6 +1332,9 @@ array_fromlist(arrayobject *self, PyObject *list)
 			if ((*self->ob_descr->setitem)(self,
 					Py_SIZE(self) - n + i, v) != 0) {
 				Py_SIZE(self) -= n;
+				if (itemsize && (Py_SIZE(self) > PY_SSIZE_T_MAX / itemsize)) {
+					return PyErr_NoMemory();
+				}
 				PyMem_RESIZE(item, char,
 					          Py_SIZE(self) * itemsize);
 				self->ob_item = item;
@@ -1373,6 +1393,10 @@ array_fromstring(arrayobject *self, PyObject *args)
 	n = n / itemsize;
 	if (n > 0) {
 		char *item = self->ob_item;
+		if ((n > PY_SSIZE_T_MAX - Py_SIZE(self)) ||
+			((Py_SIZE(self) + n) > PY_SSIZE_T_MAX / itemsize)) {
+				return PyErr_NoMemory();
+		}
 		PyMem_RESIZE(item, char, (Py_SIZE(self) + n) * itemsize);
 		if (item == NULL) {
 			PyErr_NoMemory();
@@ -1398,8 +1422,12 @@ values, as if it had been read from a file using the fromfile() method).");
 static PyObject *
 array_tostring(arrayobject *self, PyObject *unused)
 {
-	return PyBytes_FromStringAndSize(self->ob_item,
-                                         Py_SIZE(self) * self->ob_descr->itemsize);
+	if (Py_SIZE(self) <= PY_SSIZE_T_MAX / self->ob_descr->itemsize) {
+		return PyBytes_FromStringAndSize(self->ob_item,
+				    Py_SIZE(self) * self->ob_descr->itemsize);
+	} else {
+		return PyErr_NoMemory();
+	}
 }
 
 PyDoc_STRVAR(tostring_doc,
@@ -1428,6 +1456,9 @@ array_fromunicode(arrayobject *self, PyObject *args)
 	}
 	if (n > 0) {
 		Py_UNICODE *item = (Py_UNICODE *) self->ob_item;
+		if (Py_SIZE(self) > PY_SSIZE_T_MAX - n) {
+			return PyErr_NoMemory();
+		}
 		PyMem_RESIZE(item, Py_UNICODE, Py_SIZE(self) + n);
 		if (item == NULL) {
 			PyErr_NoMemory();
diff --git a/Modules/audioop.c b/Modules/audioop.c
index 57f25c6..c660501 100644
--- a/Modules/audioop.c
+++ b/Modules/audioop.c
@@ -829,7 +829,7 @@ static PyObject *
 audioop_tostereo(PyObject *self, PyObject *args)
 {
         signed char *cp, *ncp;
-        int len, size, val1, val2, val = 0;
+        int len, new_len, size, val1, val2, val = 0;
         double fac1, fac2, fval, maxval;
         PyObject *rv;
         int i;
@@ -846,7 +846,14 @@ audioop_tostereo(PyObject *self, PyObject *args)
                 return 0;
         }
     
-        rv = PyBytes_FromStringAndSize(NULL, len*2);
+        new_len = len*2;
+        if (new_len < 0) {
+                PyErr_SetString(PyExc_MemoryError,
+                                "not enough memory for output buffer");
+                return 0;
+        }
+
+        rv = PyBytes_FromStringAndSize(NULL, new_len);
         if ( rv == 0 )
                 return 0;
         ncp = (signed char *)PyBytes_AsString(rv);
@@ -1009,7 +1016,7 @@ audioop_lin2lin(PyObject *self, PyObject *args)
 {
         signed char *cp;
         unsigned char *ncp;
-        int len, size, size2, val = 0;
+        int len, new_len, size, size2, val = 0;
         PyObject *rv;
         int i, j;
 
@@ -1023,7 +1030,13 @@ audioop_lin2lin(PyObject *self, PyObject *args)
                 return 0;
         }
     
-        rv = PyBytes_FromStringAndSize(NULL, (len/size)*size2);
+        new_len = (len/size)*size2;
+        if (new_len < 0) {
+                PyErr_SetString(PyExc_MemoryError,
+                                "not enough memory for output buffer");
+                return 0;
+        }
+        rv = PyBytes_FromStringAndSize(NULL, new_len);
         if ( rv == 0 )
                 return 0;
         ncp = (unsigned char *)PyBytes_AsString(rv);
@@ -1059,6 +1072,7 @@ audioop_ratecv(PyObject *self, PyObject *args)
         int chan, d, *prev_i, *cur_i, cur_o;
         PyObject *state, *samps, *str, *rv = NULL;
         int bytes_per_frame;
+        size_t alloc_size;
 
         weightA = 1;
         weightB = 0;
@@ -1101,8 +1115,14 @@ audioop_ratecv(PyObject *self, PyObject *args)
         inrate /= d;
         outrate /= d;
 
-        prev_i = (int *) malloc(nchannels * sizeof(int));
-        cur_i = (int *) malloc(nchannels * sizeof(int));
+        alloc_size = sizeof(int) * (unsigned)nchannels;
+        if (alloc_size < nchannels) {
+                PyErr_SetString(PyExc_MemoryError,
+                                "not enough memory for output buffer");
+                return 0;
+        }
+        prev_i = (int *) malloc(alloc_size);
+        cur_i = (int *) malloc(alloc_size);
         if (prev_i == NULL || cur_i == NULL) {
                 (void) PyErr_NoMemory();
                 goto exit;
@@ -1275,7 +1295,7 @@ audioop_ulaw2lin(PyObject *self, PyObject *args)
         unsigned char *cp;
         unsigned char cval;
         signed char *ncp;
-        int len, size, val;
+        int len, new_len, size, val;
         PyObject *rv;
         int i;
 
@@ -1288,12 +1308,18 @@ audioop_ulaw2lin(PyObject *self, PyObject *args)
                 return 0;
         }
     
-        rv = PyBytes_FromStringAndSize(NULL, len*size);
+        new_len = len*size;
+        if (new_len < 0) {
+                PyErr_SetString(PyExc_MemoryError,
+                                "not enough memory for output buffer");
+                return 0;
+        }
+        rv = PyBytes_FromStringAndSize(NULL, new_len);
         if ( rv == 0 )
                 return 0;
         ncp = (signed char *)PyBytes_AsString(rv);
     
-        for ( i=0; i < len*size; i += size ) {
+        for ( i=0; i < new_len; i += size ) {
                 cval = *cp++;
                 val = st_ulaw2linear16(cval);
         
@@ -1343,7 +1369,7 @@ audioop_alaw2lin(PyObject *self, PyObject *args)
         unsigned char *cp;
         unsigned char cval;
         signed char *ncp;
-        int len, size, val;
+        int len, new_len, size, val;
         PyObject *rv;
         int i;
 
@@ -1356,12 +1382,18 @@ audioop_alaw2lin(PyObject *self, PyObject *args)
                 return 0;
         }
     
-        rv = PyBytes_FromStringAndSize(NULL, len*size);
+        new_len = len*size;
+        if (new_len < 0) {
+                PyErr_SetString(PyExc_MemoryError,
+                                "not enough memory for output buffer");
+                return 0;
+        }
+        rv = PyBytes_FromStringAndSize(NULL, new_len);
         if ( rv == 0 )
                 return 0;
         ncp = (signed char *)PyBytes_AsString(rv);
     
-        for ( i=0; i < len*size; i += size ) {
+        for ( i=0; i < new_len; i += size ) {
                 cval = *cp++;
                 val = st_alaw2linear16(cval);
         
@@ -1486,7 +1518,7 @@ audioop_adpcm2lin(PyObject *self, PyObject *args)
 {
         signed char *cp;
         signed char *ncp;
-        int len, size, valpred, step, delta, index, sign, vpdiff;
+        int len, new_len, size, valpred, step, delta, index, sign, vpdiff;
         PyObject *rv, *str, *state;
         int i, inputbuffer = 0, bufferstep;
 
@@ -1508,7 +1540,13 @@ audioop_adpcm2lin(PyObject *self, PyObject *args)
         } else if ( !PyArg_ParseTuple(state, "ii", &valpred, &index) )
                 return 0;
     
-        str = PyBytes_FromStringAndSize(NULL, len*size*2);
+        new_len = len*size*2;
+        if (new_len < 0) {
+                PyErr_SetString(PyExc_MemoryError,
+                                "not enough memory for output buffer");
+                return 0;
+        }
+        str = PyBytes_FromStringAndSize(NULL, new_len);
         if ( str == 0 )
                 return 0;
         ncp = (signed char *)PyBytes_AsString(str);
@@ -1516,7 +1554,7 @@ audioop_adpcm2lin(PyObject *self, PyObject *args)
         step = stepsizeTable[index];
         bufferstep = 0;
     
-        for ( i=0; i < len*size*2; i += size ) {
+        for ( i=0; i < new_len; i += size ) {
                 /* Step 1 - get the delta value and compute next index */
                 if ( bufferstep ) {
                         delta = inputbuffer & 0xf;
diff --git a/Modules/binascii.c b/Modules/binascii.c
index d3e8a51..ea34bcd 100644
--- a/Modules/binascii.c
+++ b/Modules/binascii.c
@@ -198,6 +198,8 @@ binascii_a2b_uu(PyObject *self, PyObject *args)
 	if ( !PyArg_ParseTuple(args, "t#:a2b_uu", &ascii_data, &ascii_len) )
 		return NULL;
 
+	assert(ascii_len >= 0);
+
 	/* First byte: binary data length (in bytes) */
 	bin_len = (*ascii_data++ - ' ') & 077;
 	ascii_len--;
@@ -355,6 +357,11 @@ binascii_a2b_base64(PyObject *self, PyObject *args)
 	if ( !PyArg_ParseTuple(args, "t#:a2b_base64", &ascii_data, &ascii_len) )
 		return NULL;
 
+	assert(ascii_len >= 0);
+
+	if (ascii_len > PY_SSIZE_T_MAX - 3)
+		return PyErr_NoMemory();
+
 	bin_len = ((ascii_len+3)/4)*3; /* Upper bound, corrected later */
 
 	/* Allocate the buffer */
@@ -448,6 +455,9 @@ binascii_b2a_base64(PyObject *self, PyObject *args)
 
 	if ( !PyArg_ParseTuple(args, "s#:b2a_base64", &bin_data, &bin_len) )
 		return NULL;
+
+	assert(bin_len >= 0);
+
 	if ( bin_len > BASE64_MAXBIN ) {
 		PyErr_SetString(Error, "Too much data for base64 line");
 		return NULL;
@@ -507,6 +517,11 @@ binascii_a2b_hqx(PyObject *self, PyObject *args)
 	if ( !PyArg_ParseTuple(args, "t#:a2b_hqx", &ascii_data, &len) )
 		return NULL;
 
+	assert(len >= 0);
+
+	if (len > PY_SSIZE_T_MAX - 2)
+		return PyErr_NoMemory();
+
 	/* Allocate a string that is too big (fixed later) 
 	   Add two to the initial length to prevent interning which
 	   would preclude subsequent resizing.  */
@@ -574,6 +589,11 @@ binascii_rlecode_hqx(PyObject *self, PyObject *args)
 	if ( !PyArg_ParseTuple(args, "s#:rlecode_hqx", &in_data, &len) )
 		return NULL;
 
+	assert(len >= 0);
+
+	if (len > PY_SSIZE_T_MAX / 2 - 2)
+		return PyErr_NoMemory();
+
 	/* Worst case: output is twice as big as input (fixed later) */
 	if ( (rv=PyBytes_FromStringAndSize(NULL, len*2+2)) == NULL )
 		return NULL;
@@ -627,6 +647,11 @@ binascii_b2a_hqx(PyObject *self, PyObject *args)
 	if ( !PyArg_ParseTuple(args, "s#:b2a_hqx", &bin_data, &len) )
 		return NULL;
 
+	assert(len >= 0);
+
+	if (len > PY_SSIZE_T_MAX / 2 - 2)
+		return PyErr_NoMemory();
+
 	/* Allocate a buffer that is at least large enough */
 	if ( (rv=PyBytes_FromStringAndSize(NULL, len*2+2)) == NULL )
 		return NULL;
@@ -669,9 +694,13 @@ binascii_rledecode_hqx(PyObject *self, PyObject *args)
 	if ( !PyArg_ParseTuple(args, "s#:rledecode_hqx", &in_data, &in_len) )
 		return NULL;
 
+	assert(in_len >= 0);
+
 	/* Empty string is a special case */
 	if ( in_len == 0 )
 		return PyBytes_FromStringAndSize("", 0);
+	else if (in_len > PY_SSIZE_T_MAX / 2)
+		return PyErr_NoMemory();
 
 	/* Allocate a buffer of reasonable size. Resized when needed */
 	out_len = in_len*2;
@@ -697,6 +726,7 @@ binascii_rledecode_hqx(PyObject *self, PyObject *args)
 #define OUTBYTE(b) \
 	do { \
 		 if ( --out_len_left < 0 ) { \
+			  if ( out_len > PY_SSIZE_T_MAX / 2) return PyErr_NoMemory(); \
 			  if (_PyBytes_Resize(&rv, 2*out_len) < 0) \
 			    { Py_DECREF(rv); return NULL; } \
 			  out_data = (unsigned char *)PyBytes_AS_STRING(rv) \
@@ -769,7 +799,7 @@ binascii_crc_hqx(PyObject *self, PyObject *args)
 	if ( !PyArg_ParseTuple(args, "s#i:crc_hqx", &bin_data, &len, &crc) )
 		return NULL;
 
-	while(len--) {
+	while(len-- > 0) {
 		crc=((crc<<8)&0xff00)^crctab_hqx[((crc>>8)&0xff)^*bin_data++];
 	}
 
@@ -925,7 +955,7 @@ binascii_crc32(PyObject *self, PyObject *args)
 		return NULL;
 
 	crc = ~ crc;
-	while (len--) {
+	while (len-- > 0) {
 		crc = crc_32_tab[(crc ^ *bin_data++) & 0xff] ^ (crc >> 8);
 		/* Note:  (crc >> 8) MUST zero fill on left */
 	}
@@ -948,6 +978,10 @@ binascii_hexlify(PyObject *self, PyObject *args)
 	if (!PyArg_ParseTuple(args, "s#:b2a_hex", &argbuf, &arglen))
 		return NULL;
 
+	assert(arglen >= 0);
+	if (arglen > PY_SSIZE_T_MAX / 2)
+		return PyErr_NoMemory();
+
 	retval = PyBytes_FromStringAndSize(NULL, arglen*2);
 	if (!retval)
 		return NULL;
@@ -999,6 +1033,8 @@ binascii_unhexlify(PyObject *self, PyObject *args)
 	if (!PyArg_ParseTuple(args, "s#:a2b_hex", &argbuf, &arglen))
 		return NULL;
 
+	assert(arglen >= 0);
+
 	/* XXX What should we do about strings with an odd length?  Should
 	 * we add an implicit leading zero, or a trailing zero?  For now,
 	 * raise an exception.
diff --git a/Modules/cjkcodecs/multibytecodec.c b/Modules/cjkcodecs/multibytecodec.c
index 546f4e2..d6bafe7 100644
--- a/Modules/cjkcodecs/multibytecodec.c
+++ b/Modules/cjkcodecs/multibytecodec.c
@@ -172,13 +172,17 @@ static PyGetSetDef codecctx_getsets[] = {
 static int
 expand_encodebuffer(MultibyteEncodeBuffer *buf, Py_ssize_t esize)
 {
-	Py_ssize_t orgpos, orgsize;
+	Py_ssize_t orgpos, orgsize, incsize;
 
 	orgpos = (Py_ssize_t)((char *)buf->outbuf -
 				PyBytes_AS_STRING(buf->outobj));
 	orgsize = PyBytes_GET_SIZE(buf->outobj);
-	if (_PyBytes_Resize(&buf->outobj, orgsize + (
-	    esize < (orgsize >> 1) ? (orgsize >> 1) | 1 : esize)) == -1)
+	incsize = (esize < (orgsize >> 1) ? (orgsize >> 1) | 1 : esize);
+
+	if (orgsize > PY_SSIZE_T_MAX - incsize)
+		return -1;
+
+	if (_PyBytes_Resize(&buf->outobj, orgsize + incsize) == -1)
 		return -1;
 
 	buf->outbuf = (unsigned char *)PyBytes_AS_STRING(buf->outobj) +orgpos;
@@ -481,6 +485,12 @@ multibytecodec_encode(MultibyteCodec *codec,
 	buf.excobj = NULL;
 	buf.inbuf = buf.inbuf_top = *data;
 	buf.inbuf_end = buf.inbuf_top + datalen;
+
+	if (datalen > (PY_SSIZE_T_MAX - 16) / 2) {
+		PyErr_NoMemory();
+		goto errorexit;
+	}
+
 	buf.outobj = PyBytes_FromStringAndSize(NULL, datalen * 2 + 16);
 	if (buf.outobj == NULL)
 		goto errorexit;
@@ -743,6 +753,11 @@ encoder_encode_stateful(MultibyteStatefulEncoderContext *ctx,
 	origpending = ctx->pendingsize;
 
 	if (origpending > 0) {
+		if (datalen > PY_SSIZE_T_MAX - ctx->pendingsize) {
+			PyErr_NoMemory();
+			/* inbuf_tmp == NULL */
+			goto errorexit;
+		}
 		inbuf_tmp = PyMem_New(Py_UNICODE, datalen + ctx->pendingsize);
 		if (inbuf_tmp == NULL)
 			goto errorexit;
@@ -805,9 +820,10 @@ decoder_append_pending(MultibyteStatefulDecoderContext *ctx,
 	Py_ssize_t npendings;
 
 	npendings = (Py_ssize_t)(buf->inbuf_end - buf->inbuf);
-	if (npendings + ctx->pendingsize > MAXDECPENDING) {
-		PyErr_SetString(PyExc_UnicodeError, "pending buffer overflow");
-		return -1;
+	if (npendings + ctx->pendingsize > MAXDECPENDING ||
+		npendings > PY_SSIZE_T_MAX - ctx->pendingsize) {
+			PyErr_SetString(PyExc_UnicodeError, "pending buffer overflow");
+			return -1;
 	}
 	memcpy(ctx->pending + ctx->pendingsize, buf->inbuf, npendings);
 	ctx->pendingsize += npendings;
@@ -1009,7 +1025,7 @@ mbidecoder_decode(MultibyteIncrementalDecoderObject *self,
 		  PyObject *args, PyObject *kwargs)
 {
 	MultibyteDecodeBuffer buf;
-	char *data, *wdata;
+	char *data, *wdata = NULL;
 	Py_ssize_t wsize, finalsize = 0, size, origpending;
 	int final = 0;
 
@@ -1025,6 +1041,10 @@ mbidecoder_decode(MultibyteIncrementalDecoderObject *self,
 		wdata = data;
 	}
 	else {
+		if (size > PY_SSIZE_T_MAX - self->pendingsize) {
+			PyErr_NoMemory();
+			goto errorexit;
+		}
 		wsize = size + self->pendingsize;
 		wdata = PyMem_Malloc(wsize);
 		if (wdata == NULL)
@@ -1244,6 +1264,10 @@ mbstreamreader_iread(MultibyteStreamReaderObject *self,
 			PyObject *ctr;
 			char *ctrdata;
 
+			if (PyBytes_GET_SIZE(cres) > PY_SSIZE_T_MAX - self->pendingsize) {
+				PyErr_NoMemory();
+				goto errorexit;
+            }
 			rsize = PyBytes_GET_SIZE(cres) + self->pendingsize;
 			ctr = PyBytes_FromStringAndSize(NULL, rsize);
 			if (ctr == NULL)
diff --git a/Modules/datetimemodule.c b/Modules/datetimemodule.c
index 6c2d5a0..4dd4fe31 100644
--- a/Modules/datetimemodule.c
+++ b/Modules/datetimemodule.c
@@ -1111,6 +1111,8 @@ format_utcoffset(char *buf, size_t buflen, const char *sep,
 	char sign;
 	int none;
 
+	assert(buflen >= 1);
+
 	offset = call_utcoffset(tzinfo, tzinfoarg, &none);
 	if (offset == -1 && PyErr_Occurred())
 		return -1;
@@ -1250,6 +1252,11 @@ wrap_strftime(PyObject *object, PyObject *format, PyObject *timetuple,
 	 * a new format.  Since computing the replacements for those codes
 	 * is expensive, don't unless they're actually used.
 	 */
+	if (flen > INT_MAX - 1) {
+		PyErr_NoMemory();
+		goto Done;
+	}
+
 	totalnew = flen + 1;	/* realistic if no %z/%Z */
 	newfmt = PyBytes_FromStringAndSize(NULL, totalnew);
 	if (newfmt == NULL) goto Done;