notes - cedric... i think i may have found a problem here...

SVN revision: 34443
This commit is contained in:
Carsten Haitzler 2008-05-03 08:31:26 +00:00
parent 21bfa5b0d6
commit 0688234e2f
1 changed files with 11 additions and 6 deletions

View File

@ -153,16 +153,21 @@ _evas_common_rgba_image_surface_alloc(Image_Entry *ie, int w, int h)
else
siz = w * h * sizeof(DATA32);
if (im->image.no_free)
im->image.data = malloc(siz);
else
/* FIXME: this is wrong. we seem to realloc data here that ismaked "no_free"
* whihc is wrong, but when we fix it, we break all sorts of stuff
* whihc used to work. i suspect the cache changes in evas */
#if 1
im->image.data = realloc(im->image.data, siz);
if (im->image.data == NULL) return -1;
#else
if (!im->image.no_free)
{
// im->image.data = realloc(im->image.data, siz);
if (im->image.data) free(im->image.data);
im->image.data = malloc(siz);
if (im->image.data == NULL) return -1;
}
if (im->image.data == NULL) return -1;
#endif
#ifdef HAVE_VALGRIND
VALGRIND_MAKE_READABLE(im->image.data, siz);
#endif