Merge tag 'nvme-6.14-2025-02-20' of git://git.infradead.org/nvme into…

… block-6.14

Pull NVMe fixes from Keith:

"nvme fixes for Linux 6.14

 - FC controller state check fixes (Daniel)
 - PCI Endpoint fixes (Damien)
 - TCP connection failure fixe (Caleb)
 - TCP handling C2HTermReq PDU (Maurizio)
 - RDMA queue state check (Ruozhu)
 - Apple controller fixes (Hector)
 - Target crash on disbaled namespace (Hannes)"

* tag 'nvme-6.14-2025-02-20' of git://git.infradead.org/nvme:
  nvme: only allow entering LIVE from CONNECTING state
  nvme-fc: rely on state transitions to handle connectivity loss
  apple-nvme: Support coprocessors left idle
  apple-nvme: Release power domains when probe fails
  nvmet: Use enum definitions instead of hardcoded values
  nvme: Cleanup the definition of the controller config register fields
  nvme/ioctl: add missing space in err message
  nvme-tcp: fix connect failure on receiving partial ICResp PDU
  nvme: tcp: Fix compilation warning with W=1
  nvmet: pci-epf: Avoid RCU stalls under heavy workload
  nvmet: pci-epf: Do not uselessly write the CSTS register
  nvmet: pci-epf: Correctly initialize CSTS when enabling the controller
  nvmet-rdma: recheck queue state is LIVE in state lock in recv done
  nvmet: Fix crash when a namespace is disabled
  nvme-tcp: add basic support for the C2HTermReq PDU
  nvme-pci: quirk Acer FA100 for non-uniqueue identifiers

drivers/nvme/host/apple.c

@@ -1011,25 +1011,37 @@ static void apple_nvme_reset_work(struct work_struct *work)
 		ret = apple_rtkit_shutdown(anv->rtk);
 		if (ret)
 			goto out;
+
+		writel(0, anv->mmio_coproc + APPLE_ANS_COPROC_CPU_CONTROL);
 	}
 
-	writel(0, anv->mmio_coproc + APPLE_ANS_COPROC_CPU_CONTROL);
+	/*
+	 * Only do the soft-reset if the CPU is not running, which means either we
+	 * or the previous stage shut it down cleanly.
+	 */
+	if (!(readl(anv->mmio_coproc + APPLE_ANS_COPROC_CPU_CONTROL) &
+		APPLE_ANS_COPROC_CPU_CONTROL_RUN)) {
 
-	ret = reset_control_assert(anv->reset);
-	if (ret)
-		goto out;
+		ret = reset_control_assert(anv->reset);
+		if (ret)
+			goto out;
 
-	ret = apple_rtkit_reinit(anv->rtk);
-	if (ret)
-		goto out;
+		ret = apple_rtkit_reinit(anv->rtk);
+		if (ret)
+			goto out;
 
-	ret = reset_control_deassert(anv->reset);
-	if (ret)
-		goto out;
+		ret = reset_control_deassert(anv->reset);
+		if (ret)
+			goto out;
+
+		writel(APPLE_ANS_COPROC_CPU_CONTROL_RUN,
+		       anv->mmio_coproc + APPLE_ANS_COPROC_CPU_CONTROL);
+
+		ret = apple_rtkit_boot(anv->rtk);
+	} else {
+		ret = apple_rtkit_wake(anv->rtk);
+	}
 
-	writel(APPLE_ANS_COPROC_CPU_CONTROL_RUN,
-	       anv->mmio_coproc + APPLE_ANS_COPROC_CPU_CONTROL);
-	ret = apple_rtkit_boot(anv->rtk);
 	if (ret) {
 		dev_err(anv->dev, "ANS did not boot");
 		goto out;
@@ -1516,6 +1528,7 @@ static struct apple_nvme *apple_nvme_alloc(struct platform_device *pdev)
 
 	return anv;
 put_dev:
+	apple_nvme_detach_genpd(anv);
 	put_device(anv->dev);
 	return ERR_PTR(ret);
 }
@@ -1549,6 +1562,7 @@ static int apple_nvme_probe(struct platform_device *pdev)
 	nvme_uninit_ctrl(&anv->ctrl);
 out_put_ctrl:
 	nvme_put_ctrl(&anv->ctrl);
+	apple_nvme_detach_genpd(anv);
 	return ret;
 }
 
@@ -1563,9 +1577,12 @@ static void apple_nvme_remove(struct platform_device *pdev)
 	apple_nvme_disable(anv, true);
 	nvme_uninit_ctrl(&anv->ctrl);
 
-	if (apple_rtkit_is_running(anv->rtk))
+	if (apple_rtkit_is_running(anv->rtk)) {
 		apple_rtkit_shutdown(anv->rtk);
 
+		writel(0, anv->mmio_coproc + APPLE_ANS_COPROC_CPU_CONTROL);
+	}
+
 	apple_nvme_detach_genpd(anv);
 }
 
@@ -1574,8 +1591,11 @@ static void apple_nvme_shutdown(struct platform_device *pdev)
 	struct apple_nvme *anv = platform_get_drvdata(pdev);
 
 	apple_nvme_disable(anv, true);
-	if (apple_rtkit_is_running(anv->rtk))
+	if (apple_rtkit_is_running(anv->rtk)) {
 		apple_rtkit_shutdown(anv->rtk);
+
+		writel(0, anv->mmio_coproc + APPLE_ANS_COPROC_CPU_CONTROL);
+	}
 }
 
 static int apple_nvme_resume(struct device *dev)
@@ -1592,10 +1612,11 @@ static int apple_nvme_suspend(struct device *dev)
 
 	apple_nvme_disable(anv, true);
 
-	if (apple_rtkit_is_running(anv->rtk))
+	if (apple_rtkit_is_running(anv->rtk)) {
 		ret = apple_rtkit_shutdown(anv->rtk);
 
-	writel(0, anv->mmio_coproc + APPLE_ANS_COPROC_CPU_CONTROL);
+		writel(0, anv->mmio_coproc + APPLE_ANS_COPROC_CPU_CONTROL);
+	}
 
 	return ret;
 }

drivers/nvme/host/core.c

@@ -564,8 +564,6 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl,
 	switch (new_state) {
 	case NVME_CTRL_LIVE:
 		switch (old_state) {
-		case NVME_CTRL_NEW:
-		case NVME_CTRL_RESETTING:
 		case NVME_CTRL_CONNECTING:
 			changed = true;
 			fallthrough;

drivers/nvme/host/fc.c

@@ -781,61 +781,12 @@ nvme_fc_abort_lsops(struct nvme_fc_rport *rport)
 static void
 nvme_fc_ctrl_connectivity_loss(struct nvme_fc_ctrl *ctrl)
 {
-	enum nvme_ctrl_state state;
-	unsigned long flags;
-
 	dev_info(ctrl->ctrl.device,
 		"NVME-FC{%d}: controller connectivity lost. Awaiting "
 		"Reconnect", ctrl->cnum);
 
-	spin_lock_irqsave(&ctrl->lock, flags);
 	set_bit(ASSOC_FAILED, &ctrl->flags);
-	state = nvme_ctrl_state(&ctrl->ctrl);
-	spin_unlock_irqrestore(&ctrl->lock, flags);
-
-	switch (state) {
-	case NVME_CTRL_NEW:
-	case NVME_CTRL_LIVE:
-		/*
-		 * Schedule a controller reset. The reset will terminate the
-		 * association and schedule the reconnect timer.  Reconnects
-		 * will be attempted until either the ctlr_loss_tmo
-		 * (max_retries * connect_delay) expires or the remoteport's
-		 * dev_loss_tmo expires.
-		 */
-		if (nvme_reset_ctrl(&ctrl->ctrl)) {
-			dev_warn(ctrl->ctrl.device,
-				"NVME-FC{%d}: Couldn't schedule reset.\n",
-				ctrl->cnum);
-			nvme_delete_ctrl(&ctrl->ctrl);
-		}
-		break;
-
-	case NVME_CTRL_CONNECTING:
-		/*
-		 * The association has already been terminated and the
-		 * controller is attempting reconnects.  No need to do anything
-		 * futher.  Reconnects will be attempted until either the
-		 * ctlr_loss_tmo (max_retries * connect_delay) expires or the
-		 * remoteport's dev_loss_tmo expires.
-		 */
-		break;
-
-	case NVME_CTRL_RESETTING:
-		/*
-		 * Controller is already in the process of terminating the
-		 * association.  No need to do anything further. The reconnect
-		 * step will kick in naturally after the association is
-		 * terminated.
-		 */
-		break;
-
-	case NVME_CTRL_DELETING:
-	case NVME_CTRL_DELETING_NOIO:
-	default:
-		/* no action to take - let it delete */
-		break;
-	}
+	nvme_reset_ctrl(&ctrl->ctrl);
 }
 
 /**
@@ -3071,7 +3022,6 @@ nvme_fc_create_association(struct nvme_fc_ctrl *ctrl)
 	struct nvmefc_ls_rcv_op *disls = NULL;
 	unsigned long flags;
 	int ret;
-	bool changed;
 
 	++ctrl->ctrl.nr_reconnects;
 
@@ -3177,23 +3127,18 @@ nvme_fc_create_association(struct nvme_fc_ctrl *ctrl)
 		else
 			ret = nvme_fc_recreate_io_queues(ctrl);
 	}
+	if (!ret && test_bit(ASSOC_FAILED, &ctrl->flags))
+		ret = -EIO;
 	if (ret)
 		goto out_term_aen_ops;
 
-	spin_lock_irqsave(&ctrl->lock, flags);
-	if (!test_bit(ASSOC_FAILED, &ctrl->flags))
-		changed = nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_LIVE);
-	else
+	if (!nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_LIVE)) {
 		ret = -EIO;
-	spin_unlock_irqrestore(&ctrl->lock, flags);
-
-	if (ret)
 		goto out_term_aen_ops;
+	}
 
 	ctrl->ctrl.nr_reconnects = 0;
-
-	if (changed)
-		nvme_start_ctrl(&ctrl->ctrl);
+	nvme_start_ctrl(&ctrl->ctrl);
 
 	return 0;	/* Success */
 

drivers/nvme/host/ioctl.c

@@ -283,8 +283,7 @@ static bool nvme_validate_passthru_nsid(struct nvme_ctrl *ctrl,
 {
 	if (ns && nsid != ns->head->ns_id) {
 		dev_err(ctrl->device,
-			"%s: nsid (%u) in cmd does not match nsid (%u)"
-			"of namespace\n",
+			"%s: nsid (%u) in cmd does not match nsid (%u) of namespace\n",
 			current->comm, nsid, ns->head->ns_id);
 		return false;
 	}

drivers/nvme/host/pci.c

@@ -3706,6 +3706,8 @@ static const struct pci_device_id nvme_id_table[] = {
 		.driver_data = NVME_QUIRK_BOGUS_NID, },
 	{ PCI_DEVICE(0x1cc1, 0x5350),   /* ADATA XPG GAMMIX S50 */
 		.driver_data = NVME_QUIRK_BOGUS_NID, },
+	{ PCI_DEVICE(0x1dbe, 0x5216),   /* Acer/INNOGRIT FA100/5216 NVMe SSD */
+		.driver_data = NVME_QUIRK_BOGUS_NID, },
 	{ PCI_DEVICE(0x1dbe, 0x5236),   /* ADATA XPG GAMMIX S70 */
 		.driver_data = NVME_QUIRK_BOGUS_NID, },
 	{ PCI_DEVICE(0x1e49, 0x0021),   /* ZHITAI TiPro5000 NVMe SSD */

drivers/nvme/host/tcp.c

@@ -763,6 +763,40 @@ static int nvme_tcp_handle_r2t(struct nvme_tcp_queue *queue,
 	return 0;
 }
 
+static void nvme_tcp_handle_c2h_term(struct nvme_tcp_queue *queue,
+		struct nvme_tcp_term_pdu *pdu)
+{
+	u16 fes;
+	const char *msg;
+	u32 plen = le32_to_cpu(pdu->hdr.plen);
+
+	static const char * const msg_table[] = {
+		[NVME_TCP_FES_INVALID_PDU_HDR] = "Invalid PDU Header Field",
+		[NVME_TCP_FES_PDU_SEQ_ERR] = "PDU Sequence Error",
+		[NVME_TCP_FES_HDR_DIGEST_ERR] = "Header Digest Error",
+		[NVME_TCP_FES_DATA_OUT_OF_RANGE] = "Data Transfer Out Of Range",
+		[NVME_TCP_FES_R2T_LIMIT_EXCEEDED] = "R2T Limit Exceeded",
+		[NVME_TCP_FES_UNSUPPORTED_PARAM] = "Unsupported Parameter",
+	};
+
+	if (plen < NVME_TCP_MIN_C2HTERM_PLEN ||
+	    plen > NVME_TCP_MAX_C2HTERM_PLEN) {
+		dev_err(queue->ctrl->ctrl.device,
+			"Received a malformed C2HTermReq PDU (plen = %u)\n",
+			plen);
+		return;
+	}
+
+	fes = le16_to_cpu(pdu->fes);
+	if (fes && fes < ARRAY_SIZE(msg_table))
+		msg = msg_table[fes];
+	else
+		msg = "Unknown";
+
+	dev_err(queue->ctrl->ctrl.device,
+		"Received C2HTermReq (FES = %s)\n", msg);
+}
+
 static int nvme_tcp_recv_pdu(struct nvme_tcp_queue *queue, struct sk_buff *skb,
 		unsigned int *offset, size_t *len)
 {
@@ -784,6 +818,15 @@ static int nvme_tcp_recv_pdu(struct nvme_tcp_queue *queue, struct sk_buff *skb,
 		return 0;
 
 	hdr = queue->pdu;
+	if (unlikely(hdr->type == nvme_tcp_c2h_term)) {
+		/*
+		 * C2HTermReq never includes Header or Data digests.
+		 * Skip the checks.
+		 */
+		nvme_tcp_handle_c2h_term(queue, (void *)queue->pdu);
+		return -EINVAL;
+	}
+
 	if (queue->hdr_digest) {
 		ret = nvme_tcp_verify_hdgst(queue, queue->pdu, hdr->hlen);
 		if (unlikely(ret))
@@ -1449,11 +1492,14 @@ static int nvme_tcp_init_connection(struct nvme_tcp_queue *queue)
 		msg.msg_control = cbuf;
 		msg.msg_controllen = sizeof(cbuf);
 	}
+	msg.msg_flags = MSG_WAITALL;
 	ret = kernel_recvmsg(queue->sock, &msg, &iov, 1,
 			iov.iov_len, msg.msg_flags);
-	if (ret < 0) {
+	if (ret < sizeof(*icresp)) {
 		pr_warn("queue %d: failed to receive icresp, error %d\n",
 			nvme_tcp_queue_id(queue), ret);
+		if (ret >= 0)
+			ret = -ECONNRESET;
 		goto free_icresp;
 	}
 	ret = -ENOTCONN;
@@ -1565,7 +1611,7 @@ static bool nvme_tcp_poll_queue(struct nvme_tcp_queue *queue)
 			  ctrl->io_queues[HCTX_TYPE_POLL];
 }
 
-/**
+/*
  * Track the number of queues assigned to each cpu using a global per-cpu
  * counter and select the least used cpu from the mq_map. Our goal is to spread
  * different controllers I/O threads across different cpu cores.

drivers/nvme/target/core.c

@@ -606,13 +606,19 @@ int nvmet_ns_enable(struct nvmet_ns *ns)
 			goto out_dev_put;
 	}
 
+	if (percpu_ref_init(&ns->ref, nvmet_destroy_namespace, 0, GFP_KERNEL))
+		goto out_pr_exit;
+
 	nvmet_ns_changed(subsys, ns->nsid);
 	ns->enabled = true;
 	xa_set_mark(&subsys->namespaces, ns->nsid, NVMET_NS_ENABLED);
 	ret = 0;
 out_unlock:
 	mutex_unlock(&subsys->lock);
 	return ret;
+out_pr_exit:
+	if (ns->pr.enable)
+		nvmet_pr_exit_ns(ns);
 out_dev_put:
 	list_for_each_entry(ctrl, &subsys->ctrls, subsys_entry)
 		pci_dev_put(radix_tree_delete(&ctrl->p2p_ns_map, ns->nsid));
@@ -638,6 +644,19 @@ void nvmet_ns_disable(struct nvmet_ns *ns)
 
 	mutex_unlock(&subsys->lock);
 
+	/*
+	 * Now that we removed the namespaces from the lookup list, we
+	 * can kill the per_cpu ref and wait for any remaining references
+	 * to be dropped, as well as a RCU grace period for anyone only
+	 * using the namepace under rcu_read_lock().  Note that we can't
+	 * use call_rcu here as we need to ensure the namespaces have
+	 * been fully destroyed before unloading the module.
+	 */
+	percpu_ref_kill(&ns->ref);
+	synchronize_rcu();
+	wait_for_completion(&ns->disable_done);
+	percpu_ref_exit(&ns->ref);
+
 	if (ns->pr.enable)
 		nvmet_pr_exit_ns(ns);
 
@@ -660,22 +679,6 @@ void nvmet_ns_free(struct nvmet_ns *ns)
 	if (ns->nsid == subsys->max_nsid)
 		subsys->max_nsid = nvmet_max_nsid(subsys);
 
-	mutex_unlock(&subsys->lock);
-
-	/*
-	 * Now that we removed the namespaces from the lookup list, we
-	 * can kill the per_cpu ref and wait for any remaining references
-	 * to be dropped, as well as a RCU grace period for anyone only
-	 * using the namepace under rcu_read_lock().  Note that we can't
-	 * use call_rcu here as we need to ensure the namespaces have
-	 * been fully destroyed before unloading the module.
-	 */
-	percpu_ref_kill(&ns->ref);
-	synchronize_rcu();
-	wait_for_completion(&ns->disable_done);
-	percpu_ref_exit(&ns->ref);
-
-	mutex_lock(&subsys->lock);
 	subsys->nr_namespaces--;
 	mutex_unlock(&subsys->lock);
 
@@ -705,9 +708,6 @@ struct nvmet_ns *nvmet_ns_alloc(struct nvmet_subsys *subsys, u32 nsid)
 	ns->nsid = nsid;
 	ns->subsys = subsys;
 
-	if (percpu_ref_init(&ns->ref, nvmet_destroy_namespace, 0, GFP_KERNEL))
-		goto out_free;
-
 	if (ns->nsid > subsys->max_nsid)
 		subsys->max_nsid = nsid;
 
@@ -730,8 +730,6 @@ struct nvmet_ns *nvmet_ns_alloc(struct nvmet_subsys *subsys, u32 nsid)
 	return ns;
 out_exit:
 	subsys->max_nsid = nvmet_max_nsid(subsys);
-	percpu_ref_exit(&ns->ref);
-out_free:
 	kfree(ns);
 out_unlock:
 	mutex_unlock(&subsys->lock);