mlx4_core: For 64-bit systems, vmap() kernel queue buffers

Since kernel virtual memory is not a problem on 64-bit systems, there
is no reason to use our own 2-layer page mapping scheme for large
kernel queue buffers on such systems.  Instead, map the page list to a
single virtually contiguous buffer with vmap(), so that can we access
buffer memory via direct indexing.

Signed-off-by: Michael S. Tsirkin <mst@dev.mellanox.co.il>
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Roland Dreier <rolandd@cisco.com>
diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h
index a0afa75..6316077 100644
--- a/include/linux/mlx4/device.h
+++ b/include/linux/mlx4/device.h
@@ -189,7 +189,7 @@
 };
 
 struct mlx4_buf {
-	union {
+	struct {
 		struct mlx4_buf_list	direct;
 		struct mlx4_buf_list   *page_list;
 	} u;
@@ -310,7 +310,7 @@
 void mlx4_buf_free(struct mlx4_dev *dev, int size, struct mlx4_buf *buf);
 static inline void *mlx4_buf_offset(struct mlx4_buf *buf, int offset)
 {
-	if (buf->nbufs == 1)
+	if (BITS_PER_LONG == 64 || buf->nbufs == 1)
 		return buf->u.direct.buf + offset;
 	else
 		return buf->u.page_list[offset >> PAGE_SHIFT].buf +