SWORD25: Fix compilation on 64bit arches by casting a pointer to ssize_t instead of int.

This should be no permanent solution. I added a small fixme comment near the
cast so it won't be forgotten.

svn-id: r53593

1 files changed, 6 insertions, 1 deletions

diff --git a/engines/sword25/util/pluto/pluto.cpp b/engines/sword25/util/pluto/pluto.cpp
index 9e7d0c804d..2a39e98719 100644
--- a/engines/sword25/util/pluto/pluto.cpp
+++ b/engines/sword25/util/pluto/pluto.cpp
@@ -624,7 +624,12 @@ static void persist(PersistInfo *pi)
 	if(!lua_isnil(pi->L, -1)) {
 					/* perms reftbl ... obj ref */
 		int zero = 0;
-		int ref = (int)lua_touserdata(pi->L, -1);
+		// FIXME: Casting a pointer to an integer data type is a bad idea we
+		// should really get rid of this by fixing the design of this code.
+		// For now casting to ssize_t should silence most (all?) compilers,
+		// since ssize_t is supposedly the same size as a pointer on most
+		// (modern) architectures.
+		int ref = (ssize_t)lua_touserdata(pi->L, -1);
 		pi->writer(pi->L, &zero, sizeof(int), pi->ud);
 		pi->writer(pi->L, &ref, sizeof(int), pi->ud);
 		lua_pop(pi->L, 1);