On Sun, 2010-08-22 at 18:56 +0200, Bart Van Assche wrote:
> @@ -929,6 +1008,14 @@ static void srp_handle_recv(struct srp_target_port
> *target, struct ib_wc *wc)
> PFX "Got target logout request\n");
> break;
>
> + case SRP_CRED_REQ:
> + srp_handle_req(target, iu, &__srp_handle_cred_req);
> + break;
> +
> + case SRP_AER_REQ:
> + srp_handle_req(target, iu, &__srp_handle_aer_req);
> + break;
Would prefer the two handlers just call a common response routine after
handling their part, but you seem dead-set against that.
You don't need to take the address of a function; there have been a
number of patches recently to change &function to just function.
> + * - SRP_IU_REQ_NORMAL: SRP_NORMAL_REQ_SQ_SIZE, since the SCSI mid-layer
> + * never queues more than Scsi_Host.can_queue requests.
> + * - SRP_IU_REQ_TSK_MGMT: SRP_TSK_MGMT_SQ_SIZE.
> + * - SRP_IU_RSP: 1, since a conforming SRP target never sends more than
> + * one unanswered SRP request to an initiator.
If you're going to churn the names, these are better than the SRP_TX_IU*
variants. Better still would be SRP_IU_CMD, SRP_IU_MGMT, SRP_IU_RSP or
SRP_IU_RESP, or leave them alone.
> static struct srp_iu *__srp_get_tx_iu(struct srp_target_port *target,
> - enum srp_request_type req_type)
> + enum srp_tx_iu_type iu_type)
> {
> - s32 rsv = (req_type == SRP_REQ_TASK_MGMT) ? 0 : SRP_TSK_MGMT_SQ_SIZE;
> + s32 rsv = (iu_type == SRP_IU_REQ_TSK_MGMT) ? 0 : SRP_TSK_MGMT_SQ_SIZE;
You check for credits even if this is an SRP_IU_RSP, so we'll fail to
allocate an IU if we're low on credits. Not a common condition, but
possible within the spec.
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
