[Python-checkins] r42429 - python/trunk/Modules/mmapmodule.c
tim.peters
python-checkins at python.org
Fri Feb 17 00:46:02 CET 2006
Author: tim.peters
Date: Fri Feb 17 00:46:01 2006
New Revision: 42429
Modified:
python/trunk/Modules/mmapmodule.c
Log:
new_mmap_object(), Windows flavor.
On a box where sizeof(size_t) == 4, C doesn't define
what happens when a size_t value is shifted right by
32 bits, and this caused test_mmap to fail on Windows
in a debug build. So use different code to break
the size apart depending on how large size_t actually
is.
This looks like an illusion, since lots of code in this
module still appears to assume sizes can't be more
than 32 bits (e.g., the internal _GetMapSize() still
returns an int), but at least test_mmap passes again.
Modified: python/trunk/Modules/mmapmodule.c
==============================================================================
--- python/trunk/Modules/mmapmodule.c (original)
+++ python/trunk/Modules/mmapmodule.c Fri Feb 17 00:46:01 2006
@@ -980,6 +980,8 @@
mmap_object *m_obj;
PyObject *map_size_obj = NULL;
int map_size;
+ DWORD size_hi; /* upper 32 bits of m_obj->size */
+ DWORD size_lo; /* lower 32 bits of m_obj->size */
char *tagname = "";
DWORD dwErr = 0;
int fileno;
@@ -1089,11 +1091,23 @@
m_obj->tagname = NULL;
m_obj->access = (access_mode)access;
+ /* DWORD is a 4-byte int. If we're on a box where size_t consumes
+ * more then 4 bytes, we need to break it apart. Else (size_t
+ * consumes 4 bytes), C doesn't define what happens if we shift
+ * right by 32, so we need different code.
+ */
+#if SIZEOF_SIZE_T > 4
+ size_hi = (DWORD)(m_obj->size >> 32);
+ size_lo = (DWORD)(m_obj->size & 0xFFFFFFFF);
+#else
+ size_hi = 0;
+ size_lo = (DWORD)m_obj->size;
+#endif
m_obj->map_handle = CreateFileMapping (m_obj->file_handle,
NULL,
flProtect,
- (DWORD)(m_obj->size >> 32),
- (DWORD)(m_obj->size & 0xFFFFFFFF),
+ size_hi,
+ size_lo,
m_obj->tagname);
if (m_obj->map_handle != NULL) {
m_obj->data = (char *) MapViewOfFile (m_obj->map_handle,
More information about the Python-checkins
mailing list