[PATCH net-next 1/2] net: openvswitch: Use struct_size()

* [PATCH net-next 1/2] net: openvswitch: Use struct_size()
@ 2023-10-01 11:07 Christophe JAILLET
  2023-10-01 11:07 ` [PATCH net-next 2/2] net: openvswitch: Annotate struct mask_array with __counted_byUse struct_size() Christophe JAILLET
  0 siblings, 1 reply; 4+ messages in thread
From: Christophe JAILLET @ 2023-10-01 11:07 UTC (permalink / raw)
  To: keescook, Pravin B Shelar, David S. Miller, Eric Dumazet,
	Jakub Kicinski, Paolo Abeni
  Cc: linux-hardening, linux-kernel, kernel-janitors,
	Christophe JAILLET, netdev, dev

Use struct_size() instead of hand writing it.
This is less verbose and more robust.

Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
---
This is IMHO more readable, even if not perfect.

However (untested):
+	new = kzalloc(size_add(struct_size(new, masks, size),
			       size_mul(sizeof(u64), size)), GFP_KERNEL);

looks completely unreadable to me.
---
 net/openvswitch/flow_table.c | 7 ++-----
 1 file changed, 2 insertions(+), 5 deletions(-)

diff --git a/net/openvswitch/flow_table.c b/net/openvswitch/flow_table.c
index 4f3b1798e0b2..d108ae0bd0ee 100644
--- a/net/openvswitch/flow_table.c
+++ b/net/openvswitch/flow_table.c
@@ -220,16 +220,13 @@ static struct mask_array *tbl_mask_array_alloc(int size)
 	struct mask_array *new;
 
 	size = max(MASK_ARRAY_SIZE_MIN, size);
-	new = kzalloc(sizeof(struct mask_array) +
-		      sizeof(struct sw_flow_mask *) * size +
+	new = kzalloc(struct_size(new, masks, size) +
 		      sizeof(u64) * size, GFP_KERNEL);
 	if (!new)
 		return NULL;
 
 	new->masks_usage_zero_cntr = (u64 *)((u8 *)new +
-					     sizeof(struct mask_array) +
-					     sizeof(struct sw_flow_mask *) *
-					     size);
+					     struct_size(new, masks, size));
 
 	new->masks_usage_stats = __alloc_percpu(sizeof(struct mask_array_stats) +
 						sizeof(u64) * size,
-- 
2.34.1


^ permalink raw reply related	[flat|nested] 4+ messages in thread