-----Shiraz Saleem <shiraz.saleem@xxxxxxxxx> wrote: -----
>To: Bernard Metzler <bmt@xxxxxxxxxxxxxx>
>From: Shiraz Saleem <shiraz.saleem@xxxxxxxxx>
>Date: 10/20/2017 02:58PM
>Cc: linux-rdma@xxxxxxxxxxxxxxx
>Subject: Re: [PATCH v2 11/13] SoftiWarp Completion Queue methods
>
>On Fri, Oct 06, 2017 at 08:28:51AM -0400, Bernard Metzler wrote:
>> Signed-off-by: Bernard Metzler <bmt@xxxxxxxxxxxxxx>
>> ---
>> drivers/infiniband/sw/siw/siw_cq.c | 164
>+++++++++++++++++++++++++++++++++++++
>> 1 file changed, 164 insertions(+)
>> create mode 100644 drivers/infiniband/sw/siw/siw_cq.c
>>
>> diff --git a/drivers/infiniband/sw/siw/siw_cq.c
>b/drivers/infiniband/sw/siw/siw_cq.c
>> new file mode 100644
>> index 000000000000..0cd6e010e7a7
>> --- /dev/null
>> +++ b/drivers/infiniband/sw/siw/siw_cq.c
>
>> +/*
>> + * Reap one CQE from the CQ.
>> + *
>> + * Caller must hold qp read lock
>
>Why is this a mandate?
>
>
Good point - it is not. Will remove that stupid comment.
>> + * TODO: Provide routine which can read more than one CQE
>> + */
>> +int siw_reap_cqe(struct siw_cq *cq, struct ib_wc *ofa_wc)
>> +{
>> + struct siw_cqe *cqe;
>> + unsigned long flags;
>> +
>> + spin_lock_irqsave(&cq->lock, flags);
>> +
>> + cqe = &cq->queue[cq->cq_get % cq->num_cqe];
>> + if (cqe->flags & SIW_WQE_VALID) {
>> + siw_wc_siw2ofa(cqe, ofa_wc);
>> +
>> + if (cq->kernel_verbs) {
>> + dprint(DBG_WR,
>> + " QP%d, CQ%d: Reap WQE type: %d at idx %d\n",
>> + QP_ID((struct siw_qp *)cqe->qp), OBJ_ID(cq),
>> + cqe->opcode, cq->cq_get % cq->num_cqe);
>> + siw_qp_put(cqe->qp);
>
>Where do we do the siw_qp_get()?
>
>
We do it when we create the CQE, that is, in siw_sqe_complete(), or
siw_rqe_complete(). See siw_qp.c
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
