bpf, sockmap: Add sk_rmem_alloc check for sockmap

hulk inclusion
category: feature
bugzilla: 186640, https://gitee.com/openeuler/kernel/issues/I545NW



--------------------------------

A tcp socket in a sockmap. If the packets transmission rate is very fast
and the packets receiving rate is very slow, a large number of packets
are stacked in the ingress queue on the packets receiving side. As a
result the memory is exhausted and the system ooms.

To fix, we add sk_rmem_alloc while sk_msg queued in the ingress queue
and subtract sk_rmem_alloc while sk_msg dequeued from the ingress queue
and check sk_rmem_alloc at the beginning of bpf_tcp_ingress().

Signed-off-by: Wang Yufen <wangyufen@huawei.com>
Reviewed-by: Liu Jian <liujian56@huawei.com>
Reviewed-by: Wei Yongjun <weiyongjun1@huawei.com>
Reviewed-by: Yue Haibing <yuehaibing@huawei.com>
Signed-off-by: Zheng Zengkai <zhengzengkai@huawei.com>