From: Jason Xing <kernelxing@xxxxxxxxxxx>
Quoting from the commit 7c80b038d23e ("net: fix sk_wmem_schedule()
and sk_rmem_schedule() errors"):
"If sk->sk_forward_alloc is 150000, and we need to schedule 150001 bytes,
we want to allocate 1 byte more (rounded up to one page),
instead of 150001"
After applied this patch, we could avoid receive path scheduling
extra amount of memory.
Fixes: f970bd9e3a06 ("udp: implement memory accounting helpers")
Signed-off-by: Jason Xing <kernelxing@xxxxxxxxxxx>
---
net/ipv4/udp.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/net/ipv4/udp.c b/net/ipv4/udp.c
index 9592fe3e444a..a13f622cfa36 100644
--- a/net/ipv4/udp.c
+++ b/net/ipv4/udp.c
@@ -1567,16 +1567,16 @@ int __udp_enqueue_schedule_skb(struct sock *sk, struct sk_buff *skb)
goto uncharge_drop;
spin_lock(&list->lock);
- if (size >= sk->sk_forward_alloc) {
- amt = sk_mem_pages(size);
- delta = amt << PAGE_SHIFT;
+ if (size > sk->sk_forward_alloc) {
+ delta = size - sk->sk_forward_alloc;
+ amt = sk_mem_pages(delta);
if (!__sk_mem_raise_allocated(sk, delta, amt, SK_MEM_RECV)) {
err = -ENOBUFS;
spin_unlock(&list->lock);
goto uncharge_drop;
}
- sk->sk_forward_alloc += delta;
+ sk->sk_forward_alloc += amt << PAGE_SHIFT;
}
sk->sk_forward_alloc -= size;
--
2.37.3
