IB/{hfi1, qib, rdmavt}: Schedule multi RC/UC packets instead of posting

The post_send() path determines if it should post directly or, schedule
the post for later.  The current logic is:

  if the swqe ring is empty or (for hfi1) wqe->length <= piothreshold
    post the send
  else
    schedule

This can allow large requests to call the send engine directly.  Large
requests can potentially produce a large number of packets prior to
returning to the caller, blocking the caller from posting more requests,
and allowing better parallel processing.

Allow the driver(s) more say in this logic (pass call_send to the driver,
rather than examining a return value).

Update hfi1/qib logic to schedule the send engine if an RC or UC message
is larger than the QP MTU size.

Reviewed-by: Mike Marciniszyn <mike.marciniszyn@intel.com>
Reviewed-by: Ira Weiny <ira.weiny@intel.com>
Signed-off-by: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
This commit is contained in:
Michael J. Ruhl
2018-09-10 09:49:27 -07:00
committed by Jason Gunthorpe
parent 3e5d60bcc8
commit 0b79b27748
6 changed files with 32 additions and 27 deletions

View File

@@ -214,8 +214,14 @@ struct rvt_driver_provided {
void (*schedule_send)(struct rvt_qp *qp);
void (*schedule_send_no_lock)(struct rvt_qp *qp);
/* Driver specific work request checking */
int (*check_send_wqe)(struct rvt_qp *qp, struct rvt_swqe *wqe);
/*
* Validate the wqe. This needs to be done prior to inserting the
* wqe into the ring, but after the wqe has been set up. Allow for
* driver specific work request checking by providing a callback.
* call_send indicates if the wqe should be posted or scheduled.
*/
int (*check_send_wqe)(struct rvt_qp *qp, struct rvt_swqe *wqe,
bool *call_send);
/*
* Sometimes rdmavt needs to kick the driver's send progress. That is