[PATCH v7] nvme-fabrics: reject I/O to offline device

[Victor Gladkov]

On 2020/8/11 23:56, Sagi Grimberg wrote:
> 
> I think we still have a question to why is this fabrics specific, and if this is
> needed in fabrics, why is it not needed in pci as well.
> 
> Keith? Personally speaking, I also share Chirstoph's opinion that if it's not
> clearly fabrics specific, we should try to make pci and fabrics unified.
> 
> Your thoughts on this?

PCI doesn't define 'timeout'/'number of retries' for reconnect of the controller.
And the controller shuts down immediately if controller times out while starting.
NVME_CTRL_CONNECTING state used as transition state in the
nvme_reset_work() procedure to save similarities with fabric transports.
See below quotes from pci.c

1. 
----------------------------------------------------
Quote from nvme_timeout() function (pci.h)
	/*
	 * Shutdown immediately if controller times out while starting. The
	 * reset work will see the pci device disabled when it gets the forced
	 * cancellation error. All outstanding requests are completed on
	 * shutdown, so we return BLK_EH_DONE.
	 */
	switch (dev->ctrl.state) {
	case NVME_CTRL_CONNECTING:
		nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_DELETING);
		/* fall through */
----------------------------------------------------
2.
----------------------------------------------------
Quote from nvme_reset_work() function (pci.h)
	/*
	 * Introduce CONNECTING state from nvme-fc/rdma transports to mark the
	 * initializing procedure here.
	 */
	if (!nvme_change_ctrl_state(&dev->ctrl, NVME_CTRL_CONNECTING)) {
----------------------------------------------------