RDMA/hns: Use GFP_ATOMIC under spin lock

mainline inclusion
from mainline-v5.13-rc1
commit 2abb7431
category: bugfix
bugzilla: 174002
CVE:NA

Reference: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=2abb7431736be539b2e0926388c7c2602a68a178

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

A spin lock is taken here so we should use GFP_ATOMIC.

Fixes: f91696f2 ("RDMA/hns: Support congestion control type selection according to the FW")
Link: https://lore.kernel.org/r/20210407154900.3486268-1-weiyongjun1@huawei.com


Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wei Yongjun <weiyongjun1@huawei.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
Signed-off-by: wangsirong <wangsirong@huawei.com>
Reviewed-by: ChunZhi Hu <huchunzhi@huawei.com>
Signed-off-by: Zheng Zengkai <zhengzengkai@huawei.com>