Author: Armin Rigo <[email protected]>
Branch:
Changeset: r949:0ee2e057cb83
Date: 2012-09-19 16:42 +0200
http://bitbucket.org/cffi/cffi/changeset/0ee2e057cb83/
Log: Write down places that we would need to fix to add a debugging mode
to check for accesses-after-freeing-memory.
diff --git a/c/_cffi_backend.c b/c/_cffi_backend.c
--- a/c/_cffi_backend.c
+++ b/c/_cffi_backend.c
@@ -511,6 +511,7 @@
static PY_LONG_LONG
read_raw_signed_data(char *target, int size)
{
+ /*READ(target, size)*/
if (size == sizeof(signed char))
return *((signed char*)target);
else if (size == sizeof(short))
@@ -530,6 +531,7 @@
static unsigned PY_LONG_LONG
read_raw_unsigned_data(char *target, int size)
{
+ /*READ(target, size)*/
if (size == sizeof(unsigned char))
return *((unsigned char*)target);
else if (size == sizeof(unsigned short))
@@ -549,6 +551,7 @@
static void
write_raw_integer_data(char *target, unsigned PY_LONG_LONG source, int size)
{
+ /*WRITE(target, size)*/
if (size == sizeof(unsigned char))
*((unsigned char*)target) = (unsigned char)source;
else if (size == sizeof(unsigned short))
@@ -566,6 +569,7 @@
static double
read_raw_float_data(char *target, int size)
{
+ /*READ(target, size)*/
if (size == sizeof(float))
return *((float*)target);
else if (size == sizeof(double))
@@ -579,12 +583,14 @@
static long double
read_raw_longdouble_data(char *target)
{
+ /*READ(target, sizeof(long double))*/
return *((long double*)target);
}
static void
write_raw_float_data(char *target, double source, int size)
{
+ /*WRITE(target, size)*/
if (size == sizeof(float))
*((float*)target) = (float)source;
else if (size == sizeof(double))
@@ -596,6 +602,7 @@
static void
write_raw_longdouble_data(char *target, long double source)
{
+ /*WRITE(target, sizeof(long double))*/
*((long double*)target) = source;
}
@@ -648,6 +655,7 @@
/* non-primitive types (check done just for performance) */
if (ct->ct_flags & (CT_POINTER|CT_FUNCTIONPTR)) {
char *ptrdata = *(char **)data;
+ /*READ(data, sizeof(char *))*/
return new_simple_cdata(ptrdata, ct);
}
else if (ct->ct_flags & CT_IS_OPAQUE) {
@@ -711,11 +719,15 @@
}
}
else if (ct->ct_flags & CT_PRIMITIVE_CHAR) {
- if (ct->ct_size == sizeof(char))
+ if (ct->ct_size == sizeof(char)) {
+ /*READ(data, 1)*/
return PyBytes_FromStringAndSize(data, 1);
+ }
#ifdef HAVE_WCHAR_H
- else
+ else {
+ /*READ(data, sizeof(wchar_t))*/
return _my_PyUnicode_FromWideChar((wchar_t *)data, 1);
+ }
#endif
}
@@ -780,7 +792,9 @@
if (CData_Check(init) &&
(((CDataObject *)init)->c_type->ct_flags & CT_PRIMITIVE_CHAR) &&
(((CDataObject *)init)->c_type->ct_size == sizeof(char))) {
- return *(unsigned char *)((CDataObject *)init)->c_data;
+ char *data = ((CDataObject *)init)->c_data;
+ /*READ(data, 1)*/
+ return *(unsigned char *)data;
}
PyErr_Format(PyExc_TypeError,
"initializer for ctype 'char' must be a "STR_OR_BYTES
@@ -799,7 +813,9 @@
if (CData_Check(init) &&
(((CDataObject *)init)->c_type->ct_flags & CT_PRIMITIVE_CHAR) &&
(((CDataObject *)init)->c_type->ct_size == sizeof(wchar_t))) {
- return *(wchar_t *)((CDataObject *)init)->c_data;
+ char *data = ((CDataObject *)init)->c_data;
+ /*READ(data, sizeof(wchar_t))*/
+ return *(wchar_t *)data;
}
PyErr_Format(PyExc_TypeError,
"initializer for ctype 'wchar_t' must be a unicode string "
@@ -923,6 +939,9 @@
const char *expected;
char buf[sizeof(PY_LONG_LONG)];
+ /*if (ct->ct_size >= 0)*/
+ /*WRITE(data, ct->ct_size)*/
+
if (ct->ct_flags & CT_ARRAY) {
return convert_array_from_object(data, ct, init);
}
@@ -1329,11 +1348,15 @@
return convert_to_object(cd->c_data, cd->c_type);
}
else if (cd->c_type->ct_flags & CT_PRIMITIVE_CHAR) {
- if (cd->c_type->ct_size == sizeof(char))
+ if (cd->c_type->ct_size == sizeof(char)) {
+ /*READ(cd->c_data, 1)*/
return PyInt_FromLong((unsigned char)cd->c_data[0]);
+ }
#ifdef HAVE_WCHAR_H
- else
+ else {
+ /*READ(cd->c_data, sizeof(wchar_t))*/
return PyInt_FromLong((long)*(wchar_t *)cd->c_data);
+ }
#endif
}
else if (cd->c_type->ct_flags & CT_PRIMITIVE_FLOAT) {
@@ -1871,6 +1894,7 @@
}
resultdata = buffer + cif_descr->exchange_offset_arg[0];
+ /*READ(cd->c_data, sizeof(void(*)(void)))*/
Py_BEGIN_ALLOW_THREADS
restore_errno();
@@ -4117,6 +4141,7 @@
}
if (cd->c_type->ct_itemdescr->ct_size == sizeof(char)) {
const char *start = cd->c_data;
+ /*READ(start, ?)*/
if (length < 0)
length = strlen(start);
else {
@@ -4130,6 +4155,7 @@
#ifdef HAVE_WCHAR_H
else if (cd->c_type->ct_itemdescr->ct_flags & CT_PRIMITIVE_CHAR) {
const wchar_t *start = (wchar_t *)cd->c_data;
+ /*READ(start, ?)*/
assert(cd->c_type->ct_itemdescr->ct_size == sizeof(wchar_t));
if (length < 0) {
length = 0;
@@ -4156,11 +4182,13 @@
CT_PRIMITIVE_SIGNED |
CT_PRIMITIVE_UNSIGNED)) {
if (cd->c_type->ct_size == sizeof(char)) {
+ /*READ(cd->c_data, 1)*/
return PyBytes_FromStringAndSize(cd->c_data, 1);
}
#ifdef HAVE_WCHAR_H
else if (cd->c_type->ct_flags & CT_PRIMITIVE_CHAR) {
assert(cd->c_type->ct_size == sizeof(wchar_t));
+ /*READ(cd->c_data, sizeof(wchar_t))*/
return _my_PyUnicode_FromWideChar((wchar_t *)cd->c_data, 1);
}
#endif
@@ -4198,6 +4226,7 @@
cd->c_type->ct_name);
return NULL;
}
+ /*WRITE(cd->c_data, size)*/
#if PY_MAJOR_VERSION < 3
return PyBuffer_FromReadWriteMemory(cd->c_data, size);
#else
_______________________________________________
pypy-commit mailing list
[email protected]
http://mail.python.org/mailman/listinfo/pypy-commit
