On Mon, 2019-03-18 at 23:16 +-0800, Ming Lei wrote: +AD4 I am not familiar with SRP, could you explain what SRP initiator driver +AD4 will do when the controller is in bad state? Especially about dealing with +AD4 in-flight IO requests under this situation. Hi Ming, Just like the NVMeOF initiator driver, the SRP initiator driver uses an RDMA RC connection for all of its communication over the network. If communication between initiator and target fails the target driver will close the connection or one of the work requests that was posted by the initiator driver will complete with an error status (wc-+AD4-status +ACEAPQ IB+AF8-WC+AF8-SUCCESS). In the latter case the function srp+AF8-handle+AF8-qp+AF8-err() will try to reestablish the connection between initiator and target after a certain delay: if (delay +AD4 0) queue+AF8-delayed+AF8-work(system+AF8-long+AF8-wq, +ACY-rport-+AD4-reconnect+AF8-work, 1UL +ACo delay +ACo HZ)+ADs SCSI timeouts may kick the SCSI error handler. That results in calls of the srp+AF8-reset+AF8-device() and/or srp+AF8-reset+AF8-host() functions. srp+AF8-reset+AF8-host() terminates all outstanding requests after having disconnected the RDMA RC connection. Disconnecting the RC connection first guarantees that there are no concurrent request completion calls from the regular completion path and from the error handler. Bart.
