[sgen] Performance improvement for ms_block_size computations

This reduces regressions for switching from MS_BLOCK_SIZE compile time constant
to ms_block_size variable by 85% on the tests where this was showing up. In the
future we should implement a block allocator, which would, amongst other benefits,
enable us to remove the block-size/page-size dependency.

mono/sgen/sgen-marksweep.c

@@ -1409,10 +1409,10 @@ static inline void
 sweep_block_for_size (MSBlockInfo *block, int count, int obj_size)
 {
 	int obj_index;
+	void *obj = MS_BLOCK_OBJ_FOR_SIZE (block, 0, obj_size);
 
-	for (obj_index = 0; obj_index < count; ++obj_index) {
+	for (obj_index = 0; obj_index < count; ++obj_index, obj = (void*)((mword)obj + obj_size)) {
 		int word, bit;
-		void *obj = MS_BLOCK_OBJ_FOR_SIZE (block, obj_index, obj_size);
 
 		MS_CALC_MARK_BIT (word, bit, obj);
 		if (MS_MARK_BIT (block, word, bit)) {