summaryrefslogtreecommitdiffstats
path: root/drivers/nvme/host
diff options
context:
space:
mode:
Diffstat (limited to 'drivers/nvme/host')
-rw-r--r--drivers/nvme/host/Kconfig2
-rw-r--r--drivers/nvme/host/core.c11
-rw-r--r--drivers/nvme/host/fabrics.c23
-rw-r--r--drivers/nvme/host/fabrics.h2
-rw-r--r--drivers/nvme/host/pci.c20
-rw-r--r--drivers/nvme/host/rdma.c127
6 files changed, 105 insertions, 80 deletions
diff --git a/drivers/nvme/host/Kconfig b/drivers/nvme/host/Kconfig
index db39d53cdfb9..0c644f7bdf80 100644
--- a/drivers/nvme/host/Kconfig
+++ b/drivers/nvme/host/Kconfig
@@ -31,7 +31,7 @@ config NVME_FABRICS
config NVME_RDMA
tristate "NVM Express over Fabrics RDMA host driver"
depends on INFINIBAND
- depends on BLK_DEV_NVME
+ select NVME_CORE
select NVME_FABRICS
select SG_POOL
help
diff --git a/drivers/nvme/host/core.c b/drivers/nvme/host/core.c
index 7ff2e820bbf4..2feacc70bf61 100644
--- a/drivers/nvme/host/core.c
+++ b/drivers/nvme/host/core.c
@@ -81,10 +81,12 @@ EXPORT_SYMBOL_GPL(nvme_cancel_request);
bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl,
enum nvme_ctrl_state new_state)
{
- enum nvme_ctrl_state old_state = ctrl->state;
+ enum nvme_ctrl_state old_state;
bool changed = false;
spin_lock_irq(&ctrl->lock);
+
+ old_state = ctrl->state;
switch (new_state) {
case NVME_CTRL_LIVE:
switch (old_state) {
@@ -140,11 +142,12 @@ bool nvme_change_ctrl_state(struct nvme_ctrl *ctrl,
default:
break;
}
- spin_unlock_irq(&ctrl->lock);
if (changed)
ctrl->state = new_state;
+ spin_unlock_irq(&ctrl->lock);
+
return changed;
}
EXPORT_SYMBOL_GPL(nvme_change_ctrl_state);
@@ -608,7 +611,7 @@ int nvme_get_features(struct nvme_ctrl *dev, unsigned fid, unsigned nsid,
ret = __nvme_submit_sync_cmd(dev->admin_q, &c, &cqe, NULL, 0, 0,
NVME_QID_ANY, 0, 0);
- if (ret >= 0)
+ if (ret >= 0 && result)
*result = le32_to_cpu(cqe.result);
return ret;
}
@@ -628,7 +631,7 @@ int nvme_set_features(struct nvme_ctrl *dev, unsigned fid, unsigned dword11,
ret = __nvme_submit_sync_cmd(dev->admin_q, &c, &cqe, NULL, 0, 0,
NVME_QID_ANY, 0, 0);
- if (ret >= 0)
+ if (ret >= 0 && result)
*result = le32_to_cpu(cqe.result);
return ret;
}
diff --git a/drivers/nvme/host/fabrics.c b/drivers/nvme/host/fabrics.c
index dc996761042f..4eff49174466 100644
--- a/drivers/nvme/host/fabrics.c
+++ b/drivers/nvme/host/fabrics.c
@@ -47,8 +47,10 @@ static struct nvmf_host *nvmf_host_add(const char *hostnqn)
mutex_lock(&nvmf_hosts_mutex);
host = __nvmf_host_find(hostnqn);
- if (host)
+ if (host) {
+ kref_get(&host->ref);
goto out_unlock;
+ }
host = kmalloc(sizeof(*host), GFP_KERNEL);
if (!host)
@@ -56,7 +58,7 @@ static struct nvmf_host *nvmf_host_add(const char *hostnqn)
kref_init(&host->ref);
memcpy(host->nqn, hostnqn, NVMF_NQN_SIZE);
- uuid_le_gen(&host->id);
+ uuid_be_gen(&host->id);
list_add_tail(&host->list, &nvmf_hosts);
out_unlock:
@@ -73,9 +75,9 @@ static struct nvmf_host *nvmf_host_default(void)
return NULL;
kref_init(&host->ref);
- uuid_le_gen(&host->id);
+ uuid_be_gen(&host->id);
snprintf(host->nqn, NVMF_NQN_SIZE,
- "nqn.2014-08.org.nvmexpress:NVMf:uuid:%pUl", &host->id);
+ "nqn.2014-08.org.nvmexpress:NVMf:uuid:%pUb", &host->id);
mutex_lock(&nvmf_hosts_mutex);
list_add_tail(&host->list, &nvmf_hosts);
@@ -363,7 +365,14 @@ int nvmf_connect_admin_queue(struct nvme_ctrl *ctrl)
cmd.connect.opcode = nvme_fabrics_command;
cmd.connect.fctype = nvme_fabrics_type_connect;
cmd.connect.qid = 0;
- cmd.connect.sqsize = cpu_to_le16(ctrl->sqsize);
+
+ /*
+ * fabrics spec sets a minimum of depth 32 for admin queue,
+ * so set the queue with this depth always until
+ * justification otherwise.
+ */
+ cmd.connect.sqsize = cpu_to_le16(NVMF_AQ_DEPTH - 1);
+
/*
* Set keep-alive timeout in seconds granularity (ms * 1000)
* and add a grace period for controller kato enforcement
@@ -375,7 +384,7 @@ int nvmf_connect_admin_queue(struct nvme_ctrl *ctrl)
if (!data)
return -ENOMEM;
- memcpy(&data->hostid, &ctrl->opts->host->id, sizeof(uuid_le));
+ memcpy(&data->hostid, &ctrl->opts->host->id, sizeof(uuid_be));
data->cntlid = cpu_to_le16(0xffff);
strncpy(data->subsysnqn, ctrl->opts->subsysnqn, NVMF_NQN_SIZE);
strncpy(data->hostnqn, ctrl->opts->host->nqn, NVMF_NQN_SIZE);
@@ -434,7 +443,7 @@ int nvmf_connect_io_queue(struct nvme_ctrl *ctrl, u16 qid)
if (!data)
return -ENOMEM;
- memcpy(&data->hostid, &ctrl->opts->host->id, sizeof(uuid_le));
+ memcpy(&data->hostid, &ctrl->opts->host->id, sizeof(uuid_be));
data->cntlid = cpu_to_le16(ctrl->cntlid);
strncpy(data->subsysnqn, ctrl->opts->subsysnqn, NVMF_NQN_SIZE);
strncpy(data->hostnqn, ctrl->opts->host->nqn, NVMF_NQN_SIZE);
diff --git a/drivers/nvme/host/fabrics.h b/drivers/nvme/host/fabrics.h
index 89df52c8be97..46e460aee52d 100644
--- a/drivers/nvme/host/fabrics.h
+++ b/drivers/nvme/host/fabrics.h
@@ -34,7 +34,7 @@ struct nvmf_host {
struct kref ref;
struct list_head list;
char nqn[NVMF_NQN_SIZE];
- uuid_le id;
+ uuid_be id;
};
/**
diff --git a/drivers/nvme/host/pci.c b/drivers/nvme/host/pci.c
index d7c33f9361aa..8dcf5a960951 100644
--- a/drivers/nvme/host/pci.c
+++ b/drivers/nvme/host/pci.c
@@ -1543,15 +1543,10 @@ static void nvme_disable_io_queues(struct nvme_dev *dev)
reinit_completion(&dev->ioq_wait);
retry:
timeout = ADMIN_TIMEOUT;
- for (; i > 0; i--) {
- struct nvme_queue *nvmeq = dev->queues[i];
-
- if (!pass)
- nvme_suspend_queue(nvmeq);
- if (nvme_delete_queue(nvmeq, opcode))
+ for (; i > 0; i--, sent++)
+ if (nvme_delete_queue(dev->queues[i], opcode))
break;
- ++sent;
- }
+
while (sent--) {
timeout = wait_for_completion_io_timeout(&dev->ioq_wait, timeout);
if (timeout == 0)
@@ -1693,11 +1688,12 @@ static void nvme_dev_disable(struct nvme_dev *dev, bool shutdown)
nvme_stop_queues(&dev->ctrl);
csts = readl(dev->bar + NVME_REG_CSTS);
}
+
+ for (i = dev->queue_count - 1; i > 0; i--)
+ nvme_suspend_queue(dev->queues[i]);
+
if (csts & NVME_CSTS_CFS || !(csts & NVME_CSTS_RDY)) {
- for (i = dev->queue_count - 1; i >= 0; i--) {
- struct nvme_queue *nvmeq = dev->queues[i];
- nvme_suspend_queue(nvmeq);
- }
+ nvme_suspend_queue(dev->queues[0]);
} else {
nvme_disable_io_queues(dev);
nvme_disable_admin_queue(dev, shutdown);
diff --git a/drivers/nvme/host/rdma.c b/drivers/nvme/host/rdma.c
index 3e3ce2b0424e..ab545fb347a0 100644
--- a/drivers/nvme/host/rdma.c
+++ b/drivers/nvme/host/rdma.c
@@ -12,13 +12,11 @@
* more details.
*/
#define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
-#include <linux/delay.h>
#include <linux/module.h>
#include <linux/init.h>
#include <linux/slab.h>
#include <linux/err.h>
#include <linux/string.h>
-#include <linux/jiffies.h>
#include <linux/atomic.h>
#include <linux/blk-mq.h>
#include <linux/types.h>
@@ -26,7 +24,6 @@
#include <linux/mutex.h>
#include <linux/scatterlist.h>
#include <linux/nvme.h>
-#include <linux/t10-pi.h>
#include <asm/unaligned.h>
#include <rdma/ib_verbs.h>
@@ -46,10 +43,6 @@
#define NVME_RDMA_MAX_INLINE_SEGMENTS 1
-#define NVME_RDMA_MAX_PAGES_PER_MR 512
-
-#define NVME_RDMA_DEF_RECONNECT_DELAY 20
-
/*
* We handle AEN commands ourselves and don't even let the
* block layer know about them.
@@ -80,7 +73,6 @@ struct nvme_rdma_request {
u32 num_sge;
int nents;
bool inline_data;
- bool need_inval;
struct ib_reg_wr reg_wr;
struct ib_cqe reg_cqe;
struct nvme_rdma_queue *queue;
@@ -169,7 +161,6 @@ MODULE_PARM_DESC(register_always,
static int nvme_rdma_cm_handler(struct rdma_cm_id *cm_id,
struct rdma_cm_event *event);
static void nvme_rdma_recv_done(struct ib_cq *cq, struct ib_wc *wc);
-static int __nvme_rdma_del_ctrl(struct nvme_rdma_ctrl *ctrl);
/* XXX: really should move to a generic header sooner or later.. */
static inline void put_unaligned_le24(u32 val, u8 *p)
@@ -290,7 +281,7 @@ static int nvme_rdma_reinit_request(void *data, struct request *rq)
struct nvme_rdma_request *req = blk_mq_rq_to_pdu(rq);
int ret = 0;
- if (!req->need_inval)
+ if (!req->mr->need_inval)
goto out;
ib_dereg_mr(req->mr);
@@ -302,7 +293,7 @@ static int nvme_rdma_reinit_request(void *data, struct request *rq)
req->mr = NULL;
}
- req->need_inval = false;
+ req->mr->need_inval = false;
out:
return ret;
@@ -649,7 +640,8 @@ static int nvme_rdma_init_io_queues(struct nvme_rdma_ctrl *ctrl)
int i, ret;
for (i = 1; i < ctrl->queue_count; i++) {
- ret = nvme_rdma_init_queue(ctrl, i, ctrl->ctrl.sqsize);
+ ret = nvme_rdma_init_queue(ctrl, i,
+ ctrl->ctrl.opts->queue_size);
if (ret) {
dev_info(ctrl->ctrl.device,
"failed to initialize i/o queue: %d\n", ret);
@@ -687,11 +679,6 @@ static void nvme_rdma_free_ctrl(struct nvme_ctrl *nctrl)
list_del(&ctrl->list);
mutex_unlock(&nvme_rdma_ctrl_mutex);
- if (ctrl->ctrl.tagset) {
- blk_cleanup_queue(ctrl->ctrl.connect_q);
- blk_mq_free_tag_set(&ctrl->tag_set);
- nvme_rdma_dev_put(ctrl->device);
- }
kfree(ctrl->queues);
nvmf_free_options(nctrl->opts);
free_ctrl:
@@ -748,8 +735,11 @@ static void nvme_rdma_reconnect_ctrl_work(struct work_struct *work)
changed = nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_LIVE);
WARN_ON_ONCE(!changed);
- if (ctrl->queue_count > 1)
+ if (ctrl->queue_count > 1) {
nvme_start_queues(&ctrl->ctrl);
+ nvme_queue_scan(&ctrl->ctrl);
+ nvme_queue_async_events(&ctrl->ctrl);
+ }
dev_info(ctrl->ctrl.device, "Successfully reconnected\n");
@@ -855,7 +845,7 @@ static void nvme_rdma_unmap_data(struct nvme_rdma_queue *queue,
if (!blk_rq_bytes(rq))
return;
- if (req->need_inval) {
+ if (req->mr->need_inval) {
res = nvme_rdma_inv_rkey(queue, req);
if (res < 0) {
dev_err(ctrl->ctrl.device,
@@ -941,7 +931,7 @@ static int nvme_rdma_map_sg_fr(struct nvme_rdma_queue *queue,
IB_ACCESS_REMOTE_READ |
IB_ACCESS_REMOTE_WRITE;
- req->need_inval = true;
+ req->mr->need_inval = true;
sg->addr = cpu_to_le64(req->mr->iova);
put_unaligned_le24(req->mr->length, sg->length);
@@ -964,7 +954,7 @@ static int nvme_rdma_map_data(struct nvme_rdma_queue *queue,
req->num_sge = 1;
req->inline_data = false;
- req->need_inval = false;
+ req->mr->need_inval = false;
c->common.flags |= NVME_CMD_SGL_METABUF;
@@ -1151,7 +1141,7 @@ static int nvme_rdma_process_nvme_rsp(struct nvme_rdma_queue *queue,
if ((wc->wc_flags & IB_WC_WITH_INVALIDATE) &&
wc->ex.invalidate_rkey == req->mr->rkey)
- req->need_inval = false;
+ req->mr->need_inval = false;
blk_mq_complete_request(rq, status);
@@ -1269,7 +1259,7 @@ static int nvme_rdma_route_resolved(struct nvme_rdma_queue *queue)
{
struct nvme_rdma_ctrl *ctrl = queue->ctrl;
struct rdma_conn_param param = { };
- struct nvme_rdma_cm_req priv;
+ struct nvme_rdma_cm_req priv = { };
int ret;
param.qp_num = queue->qp->qp_num;
@@ -1284,8 +1274,22 @@ static int nvme_rdma_route_resolved(struct nvme_rdma_queue *queue)
priv.recfmt = cpu_to_le16(NVME_RDMA_CM_FMT_1_0);
priv.qid = cpu_to_le16(nvme_rdma_queue_idx(queue));
- priv.hrqsize = cpu_to_le16(queue->queue_size);
- priv.hsqsize = cpu_to_le16(queue->queue_size);
+ /*
+ * set the admin queue depth to the minimum size
+ * specified by the Fabrics standard.
+ */
+ if (priv.qid == 0) {
+ priv.hrqsize = cpu_to_le16(NVMF_AQ_DEPTH);
+ priv.hsqsize = cpu_to_le16(NVMF_AQ_DEPTH - 1);
+ } else {
+ /*
+ * current interpretation of the fabrics spec
+ * is at minimum you make hrqsize sqsize+1, or a
+ * 1's based representation of sqsize.
+ */
+ priv.hrqsize = cpu_to_le16(queue->queue_size);
+ priv.hsqsize = cpu_to_le16(queue->ctrl->ctrl.sqsize);
+ }
ret = rdma_connect(queue->cm_id, &param);
if (ret) {
@@ -1318,37 +1322,39 @@ out_destroy_queue_ib:
* that caught the event. Since we hold the callout until the controller
* deletion is completed, we'll deadlock if the controller deletion will
* call rdma_destroy_id on this queue's cm_id. Thus, we claim ownership
- * of destroying this queue before-hand, destroy the queue resources
- * after the controller deletion completed with the exception of destroying
- * the cm_id implicitely by returning a non-zero rc to the callout.
+ * of destroying this queue before-hand, destroy the queue resources,
+ * then queue the controller deletion which won't destroy this queue and
+ * we destroy the cm_id implicitely by returning a non-zero rc to the callout.
*/
static int nvme_rdma_device_unplug(struct nvme_rdma_queue *queue)
{
struct nvme_rdma_ctrl *ctrl = queue->ctrl;
- int ret, ctrl_deleted = 0;
+ int ret = 0;
- /* First disable the queue so ctrl delete won't free it */
- if (!test_and_clear_bit(NVME_RDMA_Q_CONNECTED, &queue->flags))
- goto out;
+ /* Own the controller deletion */
+ if (!nvme_change_ctrl_state(&ctrl->ctrl, NVME_CTRL_DELETING))
+ return 0;
- /* delete the controller */
- ret = __nvme_rdma_del_ctrl(ctrl);
- if (!ret) {
- dev_warn(ctrl->ctrl.device,
- "Got rdma device removal event, deleting ctrl\n");
- flush_work(&ctrl->delete_work);
+ dev_warn(ctrl->ctrl.device,
+ "Got rdma device removal event, deleting ctrl\n");
- /* Return non-zero so the cm_id will destroy implicitly */
- ctrl_deleted = 1;
+ /* Get rid of reconnect work if its running */
+ cancel_delayed_work_sync(&ctrl->reconnect_work);
+ /* Disable the queue so ctrl delete won't free it */
+ if (test_and_clear_bit(NVME_RDMA_Q_CONNECTED, &queue->flags)) {
/* Free this queue ourselves */
- rdma_disconnect(queue->cm_id);
- ib_drain_qp(queue->qp);
+ nvme_rdma_stop_queue(queue);
nvme_rdma_destroy_queue_ib(queue);
+
+ /* Return non-zero so the cm_id will destroy implicitly */
+ ret = 1;
}
-out:
- return ctrl_deleted;
+ /* Queue controller deletion */
+ queue_work(nvme_rdma_wq, &ctrl->delete_work);
+ flush_work(&ctrl->delete_work);
+ return ret;
}
static int nvme_rdma_cm_handler(struct rdma_cm_id *cm_id,
@@ -1465,7 +1471,7 @@ static int nvme_rdma_queue_rq(struct blk_mq_hw_ctx *hctx,
if (rq->cmd_type == REQ_TYPE_FS && req_op(rq) == REQ_OP_FLUSH)
flush = true;
ret = nvme_rdma_post_send(queue, sqe, req->sge, req->num_sge,
- req->need_inval ? &req->reg_wr.wr : NULL, flush);
+ req->mr->need_inval ? &req->reg_wr.wr : NULL, flush);
if (ret) {
nvme_rdma_unmap_data(queue, rq);
goto err;
@@ -1648,7 +1654,7 @@ static void nvme_rdma_shutdown_ctrl(struct nvme_rdma_ctrl *ctrl)
nvme_rdma_free_io_queues(ctrl);
}
- if (ctrl->ctrl.state == NVME_CTRL_LIVE)
+ if (test_bit(NVME_RDMA_Q_CONNECTED, &ctrl->queues[0].flags))
nvme_shutdown_ctrl(&ctrl->ctrl);
blk_mq_stop_hw_queues(ctrl->ctrl.admin_q);
@@ -1657,15 +1663,27 @@ static void nvme_rdma_shutdown_ctrl(struct nvme_rdma_ctrl *ctrl)
nvme_rdma_destroy_admin_queue(ctrl);
}
+static void __nvme_rdma_remove_ctrl(struct nvme_rdma_ctrl *ctrl, bool shutdown)
+{
+ nvme_uninit_ctrl(&ctrl->ctrl);
+ if (shutdown)
+ nvme_rdma_shutdown_ctrl(ctrl);
+
+ if (ctrl->ctrl.tagset) {
+ blk_cleanup_queue(ctrl->ctrl.connect_q);
+ blk_mq_free_tag_set(&ctrl->tag_set);
+ nvme_rdma_dev_put(ctrl->device);
+ }
+
+ nvme_put_ctrl(&ctrl->ctrl);
+}
+
static void nvme_rdma_del_ctrl_work(struct work_struct *work)
{
struct nvme_rdma_ctrl *ctrl = container_of(work,
struct nvme_rdma_ctrl, delete_work);
- nvme_remove_namespaces(&ctrl->ctrl);
- nvme_rdma_shutdown_ctrl(ctrl);
- nvme_uninit_ctrl(&ctrl->ctrl);
- nvme_put_ctrl(&ctrl->ctrl);
+ __nvme_rdma_remove_ctrl(ctrl, true);
}
static int __nvme_rdma_del_ctrl(struct nvme_rdma_ctrl *ctrl)
@@ -1698,9 +1716,7 @@ static void nvme_rdma_remove_ctrl_work(struct work_struct *work)
struct nvme_rdma_ctrl *ctrl = container_of(work,
struct nvme_rdma_ctrl, delete_work);
- nvme_remove_namespaces(&ctrl->ctrl);
- nvme_uninit_ctrl(&ctrl->ctrl);
- nvme_put_ctrl(&ctrl->ctrl);
+ __nvme_rdma_remove_ctrl(ctrl, false);
}
static void nvme_rdma_reset_ctrl_work(struct work_struct *work)
@@ -1739,6 +1755,7 @@ static void nvme_rdma_reset_ctrl_work(struct work_struct *work)
if (ctrl->queue_count > 1) {
nvme_start_queues(&ctrl->ctrl);
nvme_queue_scan(&ctrl->ctrl);
+ nvme_queue_async_events(&ctrl->ctrl);
}
return;
@@ -1809,7 +1826,7 @@ static int nvme_rdma_create_io_queues(struct nvme_rdma_ctrl *ctrl)
memset(&ctrl->tag_set, 0, sizeof(ctrl->tag_set));
ctrl->tag_set.ops = &nvme_rdma_mq_ops;
- ctrl->tag_set.queue_depth = ctrl->ctrl.sqsize;
+ ctrl->tag_set.queue_depth = ctrl->ctrl.opts->queue_size;
ctrl->tag_set.reserved_tags = 1; /* fabric connect */
ctrl->tag_set.numa_node = NUMA_NO_NODE;
ctrl->tag_set.flags = BLK_MQ_F_SHOULD_MERGE;
@@ -1907,7 +1924,7 @@ static struct nvme_ctrl *nvme_rdma_create_ctrl(struct device *dev,
spin_lock_init(&ctrl->lock);
ctrl->queue_count = opts->nr_io_queues + 1; /* +1 for admin queue */
- ctrl->ctrl.sqsize = opts->queue_size;
+ ctrl->ctrl.sqsize = opts->queue_size - 1;
ctrl->ctrl.kato = opts->kato;
ret = -ENOMEM;
generated by cgit v1.2.3 (git 2.39.1) at 2025-01-26 23:03:26 +0100