!2517 RDMA/hns: Support flexible wqe buffer page size

Merge Pull Request from: @stinft 
 
Upload two patches to support flexible wqe buffer page size.
Currently, driver fixedly allocates 4K pages for user space WQE buffer even in a 64K system. This 
results in HW reading WQE with a granularity of 4K even in a 64K system. Considering that we support 
1024-byte inline,
in the scenario of using SQ inline, HW will switch pages every 4 WQEs. This will introduce a delay of about 400ns, which is an average delay of 100ns per packet.
In order to improve performance, we allow user-mode driver to use more flexible WQE buffer page size allocation strategies, which allowing user-mode driver to configure WQE buffer using pages between 4K to system PAGESIZE.
This feature needs to be used in conjunction with the user-mode driver. In order to ensure forward compatibility, if the user-mode driver does not support this feature, the kernel mode will continue to use a fixed 4K pagesize.

The first patch is cleanup. The patch of this feature needs to add code to the hns_roce_alloc_ucontext() function, so that the number of lines is too large. Therefore, before uploading the patch of this feature, reconstruct the function to reduce the number of lines of the function and improve the readability.
The second patch is used to support this feature.

 Chengchang Tang (1):
   RDMA/hns: Support flexible wqe buffer page size
 Juan Zhou (1):
   RDMA/hns: Refactor hns_roce_alloc_ucontext()

https://gitee.com/openeuler/kernel/issues/I87LTM 
 
Link:https://gitee.com/openeuler/kernel/pulls/2517

 

Reviewed-by: Chengchang Tang <tangchengchang@huawei.com>
Signed-off-by: Jialin Zhang <zhangjialin11@huawei.com>