Revert "net/smc: don't req_notify until all CQEs drained"
authorDust Li <dust.li@linux.alibaba.com>
Fri, 4 Mar 2022 09:17:19 +0000 (17:17 +0800)
committerDavid S. Miller <davem@davemloft.net>
Sun, 6 Mar 2022 10:57:12 +0000 (10:57 +0000)
commit925a24213b5cc80fcef8858e6dc1f97ea2b17afb
treec971de4c85ffef91afceeb6f8f6af0c54d5517fd
parentd59e3cbaef707f0d3dc1e3b6735cb25060ca74c2
Revert "net/smc: don't req_notify until all CQEs drained"

This reverts commit a505cce6f7cfaf2aa2385aab7286063c96444526.

Leon says:
  We already discussed that. SMC should be changed to use
  RDMA CQ pool API
  drivers/infiniband/core/cq.c.
  ib_poll_handler() has much better implementation (tracing,
  IRQ rescheduling, proper error handling) than this SMC variant.

Since we will switch to ib_poll_handler() in the future,
revert this patch.

Link: https://lore.kernel.org/netdev/20220301105332.GA9417@linux.alibaba.com/
Suggested-by: Leon Romanovsky <leon@kernel.org>
Suggested-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Dust Li <dust.li@linux.alibaba.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/smc/smc_wr.c