[PATCH] nvmet: support reservation feature

Sagi Grimberg sagi at grimberg.me
Wed Jan 10 07:47:21 PST 2024



On 1/9/24 14:10, Guixin Liu wrote:
> This patch implements the reservation feature, includes:
> 1. reservation register(register, unregister and replace).
> 2. reservation acquire(acquire, preempt, preempt and abort).
> 3. reservation release(release and clear).
> 4. reservation report.
> 
> And also make reservation configurable, one can set ns to support
> reservation before enable ns. The default of resv_enable is false.
> 
> Signed-off-by: Guixin Liu <kanie at linux.alibaba.com>
> ---
> Hi guys,
>      I've implemented the NVMe reservation feature. Please review it, all
> comments are welcome.
>      In addtion, I didn't implement event reporting because I didn't see
> any handling of these events on the host side. If these events are mandatory
> to report, please let me know so that I can implement them.
> 
>   drivers/nvme/target/Makefile    |   2 +-
>   drivers/nvme/target/admin-cmd.c |  14 +-
>   drivers/nvme/target/configfs.c  |  27 ++
>   drivers/nvme/target/core.c      |  37 +-
>   drivers/nvme/target/nvmet.h     |  26 ++
>   drivers/nvme/target/pr.c        | 806 ++++++++++++++++++++++++++++++++
>   include/linux/nvme.h            |  30 ++
>   7 files changed, 939 insertions(+), 3 deletions(-)
>   create mode 100644 drivers/nvme/target/pr.c
> 
> diff --git a/drivers/nvme/target/Makefile b/drivers/nvme/target/Makefile
> index c66820102493..f9bfc904a5b3 100644
> --- a/drivers/nvme/target/Makefile
> +++ b/drivers/nvme/target/Makefile
> @@ -10,7 +10,7 @@ obj-$(CONFIG_NVME_TARGET_FCLOOP)	+= nvme-fcloop.o
>   obj-$(CONFIG_NVME_TARGET_TCP)		+= nvmet-tcp.o
>   
>   nvmet-y		+= core.o configfs.o admin-cmd.o fabrics-cmd.o \
> -			discovery.o io-cmd-file.o io-cmd-bdev.o
> +			discovery.o io-cmd-file.o io-cmd-bdev.o pr.o
>   nvmet-$(CONFIG_NVME_TARGET_PASSTHRU)	+= passthru.o
>   nvmet-$(CONFIG_BLK_DEV_ZONED)		+= zns.o
>   nvmet-$(CONFIG_NVME_TARGET_AUTH)	+= fabrics-cmd-auth.o auth.o
> diff --git a/drivers/nvme/target/admin-cmd.c b/drivers/nvme/target/admin-cmd.c
> index 39cb570f833d..7da6f3085a4c 100644
> --- a/drivers/nvme/target/admin-cmd.c
> +++ b/drivers/nvme/target/admin-cmd.c
> @@ -550,7 +550,13 @@ static void nvmet_execute_identify_ns(struct nvmet_req *req)
>   	 */
>   	id->nmic = NVME_NS_NMIC_SHARED;
>   	id->anagrpid = cpu_to_le32(req->ns->anagrpid);
> -
> +	id->rescap = NVME_PR_SUPPORT_WRITE_EXCLUSIVE |
> +		     NVME_PR_SUPPORT_EXCLUSIVE_ACCESS |
> +		     NVME_PR_SUPPORT_WRITE_EXCLUSIVE_REG_ONLY |
> +		     NVME_PR_SUPPORT_EXCLUSIVE_ACCESS_REG_ONLY |
> +		     NVME_PR_SUPPORT_WRITE_EXCLUSIVE_ALL_REGS |
> +		     NVME_PR_SUPPORT_EXCLUSIVE_ACCESS_ALL_REGS |
> +		     NVME_PR_SUPPORT_IEKEY_DEF_LATER_VER_1_3;

NVME_PR_SUPPORT_IEKEY_VER_1_3_DEF

>   	memcpy(&id->nguid, &req->ns->nguid, sizeof(id->nguid));
>   
>   	id->lbaf[0].ds = req->ns->blksize_shift;
> @@ -1017,6 +1023,12 @@ u16 nvmet_parse_admin_cmd(struct nvmet_req *req)
>   	if (nvmet_is_passthru_req(req))
>   		return nvmet_parse_passthru_admin_cmd(req);
>   
> +	ret = nvmet_pr_check_cmd_access(req);
> +	if (unlikely(ret)) {
> +		req->error_loc = offsetof(struct nvme_common_command, opcode);
> +		return ret;
> +	}

I think some admin commands are defined to conflict and some are
defined to allowed, hence this check needs to be per (supported)
command.

Also, it should come after req->ns is populated and lose the lookup
in nvmet_pr_check_cmd_access.

> +
>   	switch (cmd->common.opcode) {
>   	case nvme_admin_get_log_page:
>   		req->execute = nvmet_execute_get_log_page;
> diff --git a/drivers/nvme/target/configfs.c b/drivers/nvme/target/configfs.c
> index d937fe05129e..1ac4802ec818 100644
> --- a/drivers/nvme/target/configfs.c
> +++ b/drivers/nvme/target/configfs.c
> @@ -714,6 +714,32 @@ static ssize_t nvmet_ns_revalidate_size_store(struct config_item *item,
>   
>   CONFIGFS_ATTR_WO(nvmet_ns_, revalidate_size);
>   
> +static ssize_t nvmet_ns_resv_enable_show(struct config_item *item, char *page)
> +{
> +	return sprintf(page, "%d\n", to_nvmet_ns(item)->pr.enable);
> +}
> +
> +static ssize_t nvmet_ns_resv_enable_store(struct config_item *item,
> +					const char *page, size_t count)
> +{
> +	struct nvmet_ns *ns = to_nvmet_ns(item);
> +	bool val;
> +
> +	if (kstrtobool(page, &val))
> +		return -EINVAL;
> +
> +	mutex_lock(&ns->subsys->lock);
> +	if (ns->enabled) {
> +		pr_err("the ns:%d is already enabled.\n", ns->nsid);
> +		mutex_unlock(&ns->subsys->lock);
> +		return -EINVAL;
> +	}
> +	ns->pr.enable = val;
> +	mutex_unlock(&ns->subsys->lock);
> +	return count;
> +}
> +CONFIGFS_ATTR(nvmet_ns_, resv_enable);
> +
>   static struct configfs_attribute *nvmet_ns_attrs[] = {
>   	&nvmet_ns_attr_device_path,
>   	&nvmet_ns_attr_device_nguid,
> @@ -722,6 +748,7 @@ static struct configfs_attribute *nvmet_ns_attrs[] = {
>   	&nvmet_ns_attr_enable,
>   	&nvmet_ns_attr_buffered_io,
>   	&nvmet_ns_attr_revalidate_size,
> +	&nvmet_ns_attr_resv_enable,
>   #ifdef CONFIG_PCI_P2PDMA
>   	&nvmet_ns_attr_p2pmem,
>   #endif
> diff --git a/drivers/nvme/target/core.c b/drivers/nvme/target/core.c
> index 3935165048e7..8eab81804b14 100644
> --- a/drivers/nvme/target/core.c
> +++ b/drivers/nvme/target/core.c
> @@ -598,6 +598,7 @@ int nvmet_ns_enable(struct nvmet_ns *ns)
>   	subsys->nr_namespaces++;
>   
>   	nvmet_ns_changed(subsys, ns->nsid);
> +	nvmet_pr_init_ns(ns);
>   	ns->enabled = true;
>   	ret = 0;
>   out_unlock:
> @@ -651,6 +652,7 @@ void nvmet_ns_disable(struct nvmet_ns *ns)
>   
>   	subsys->nr_namespaces--;
>   	nvmet_ns_changed(subsys, ns->nsid);
> +	nvmet_pr_clean_all_registrants(&ns->pr);
>   	nvmet_ns_dev_disable(ns);
>   out_unlock:
>   	mutex_unlock(&subsys->lock);
> @@ -904,6 +906,16 @@ static u16 nvmet_parse_io_cmd(struct nvmet_req *req)
>   		return ret;
>   	}
>   
> +	ret = nvmet_pr_check_cmd_access(req);
> +	if (unlikely(ret)) {
> +		req->error_loc = offsetof(struct nvme_common_command, opcode);
> +		return ret;
> +	}
> +
> +	ret = nvmet_parse_pr_cmd(req);
> +	if (!ret)
> +		return ret;

I think both calls should be if pr is enabled on the ns. Not inside
the functions.

> +
>   	switch (req->ns->csi) {
>   	case NVME_CSI_NVM:
>   		if (req->ns->file)
> @@ -1231,6 +1243,21 @@ static void nvmet_init_cap(struct nvmet_ctrl *ctrl)
>   		nvmet_passthrough_override_cap(ctrl);
>   }
>   
> +bool nvmet_is_host_still_connected(uuid_t *hostid, struct nvmet_subsys *subsys)
> +{
> +	struct nvmet_ctrl *ctrl;
> +
> +	mutex_lock(&subsys->lock);
> +	list_for_each_entry(ctrl, &subsys->ctrls, subsys_entry) {
> +		if (uuid_equal(hostid, &ctrl->hostid)) {
> +			mutex_unlock(&subsys->lock);
> +			return true;
> +		}
> +	}
> +	mutex_unlock(&subsys->lock);
> +	return false;
> +}
> +
>   struct nvmet_ctrl *nvmet_ctrl_find_get(const char *subsysnqn,
>   				       const char *hostnqn, u16 cntlid,
>   				       struct nvmet_req *req)
> @@ -1488,6 +1515,7 @@ static void nvmet_ctrl_free(struct kref *ref)
>   	cancel_work_sync(&ctrl->fatal_err_work);
>   
>   	nvmet_destroy_auth(ctrl);
> +	nvmet_pr_unregister_ctrl_host(ctrl);
>   
>   	ida_free(&cntlid_ida, ctrl->cntlid);
>   
> @@ -1673,11 +1701,17 @@ static int __init nvmet_init(void)
>   	if (error)
>   		goto out_free_nvmet_work_queue;
>   
> -	error = nvmet_init_configfs();
> +	error = nvmet_pr_init();
>   	if (error)
>   		goto out_exit_discovery;
> +
> +	error = nvmet_init_configfs();
> +	if (error)
> +		goto out_exit_pr;
>   	return 0;
>   
> +out_exit_pr:
> +	nvmet_pr_exit();
>   out_exit_discovery:
>   	nvmet_exit_discovery();
>   out_free_nvmet_work_queue:
> @@ -1700,6 +1734,7 @@ static void __exit nvmet_exit(void)
>   	destroy_workqueue(buffered_io_wq);
>   	destroy_workqueue(zbd_wq);
>   	kmem_cache_destroy(nvmet_bvec_cache);
> +	nvmet_pr_exit();
>   
>   	BUILD_BUG_ON(sizeof(struct nvmf_disc_rsp_page_entry) != 1024);
>   	BUILD_BUG_ON(sizeof(struct nvmf_disc_rsp_page_hdr) != 1024);
> diff --git a/drivers/nvme/target/nvmet.h b/drivers/nvme/target/nvmet.h
> index 6c8acebe1a1a..7670bdc13f9c 100644
> --- a/drivers/nvme/target/nvmet.h
> +++ b/drivers/nvme/target/nvmet.h
> @@ -56,6 +56,22 @@
>   #define IPO_IATTR_CONNECT_SQE(x)	\
>   	(cpu_to_le32(offsetof(struct nvmf_connect_command, x)))
>   
> +struct nvmet_pr_registrant {
> +	bool			is_holder;
> +	u64			rkey;
> +	uuid_t			hostid;
> +	struct list_head	entry;
> +};
> +
> +struct nvmet_pr {
> +	bool			enable;
> +	enum nvme_pr_type	rtype;
> +	u32			generation;
> +	struct nvmet_pr_registrant *holder;
> +	rwlock_t		pr_lock;
> +	struct list_head	registrant_list;
> +};
> +
>   struct nvmet_ns {
>   	struct percpu_ref	ref;
>   	struct bdev_handle	*bdev_handle;
> @@ -85,6 +101,7 @@ struct nvmet_ns {
>   	int			pi_type;
>   	int			metadata_size;
>   	u8			csi;
> +	struct nvmet_pr		pr;
>   };
>   
>   static inline struct nvmet_ns *to_nvmet_ns(struct config_item *item)
> @@ -750,4 +767,13 @@ static inline bool nvmet_has_auth(struct nvmet_ctrl *ctrl)
>   static inline const char *nvmet_dhchap_dhgroup_name(u8 dhgid) { return NULL; }
>   #endif
>   
> +int nvmet_pr_init(void);
> +void nvmet_pr_exit(void);
> +void nvmet_pr_init_ns(struct nvmet_ns *ns);
> +u16 nvmet_parse_pr_cmd(struct nvmet_req *req);
> +u16 nvmet_pr_check_cmd_access(struct nvmet_req *req);
> +void nvmet_pr_unregister_ctrl_host(struct nvmet_ctrl *ctrl);
> +void nvmet_pr_clean_all_registrants(struct nvmet_pr *pr);
> +bool nvmet_is_host_still_connected(uuid_t *hostid, struct nvmet_subsys *subsys);
> +
>   #endif /* _NVMET_H */
> diff --git a/drivers/nvme/target/pr.c b/drivers/nvme/target/pr.c
> new file mode 100644
> index 000000000000..e58e295820cf
> --- /dev/null
> +++ b/drivers/nvme/target/pr.c
> @@ -0,0 +1,806 @@
> +// SPDX-License-Identifier: GPL-2.0
> +/*
> + * NVMe over Fabrics Persist Reservation.
> + * Copyright (c) 2024 Guixin Liu, Alibaba Cloud.
> + * All rights reserved.
> + */
> +#include "nvmet.h"
> +#include <linux/slab.h>
> +#include <asm/unaligned.h>
> +
> +struct nvmet_pr_register_data {
> +	__le64	crkey;
> +	__le64	nrkey;
> +};
> +
> +struct nvmet_pr_acquire_data {
> +	__le64	crkey;
> +	__le64	prkey;
> +};
> +
> +struct nvmet_pr_release_data {
> +	__le64	crkey;
> +};
> +
> +static struct kmem_cache *registrant_cache;

Why is there a need for a kmem_cache here?

> +
> +static inline void nvmet_pr_clean_holder(struct nvmet_pr *pr)
> +{
> +	pr->holder = NULL;
> +	pr->rtype = 0;
> +}
> +
> +static u16 nvmet_pr_validate_and_set_rtype(struct nvmet_pr *pr, u8 rtype)
> +{
> +	if (rtype < NVME_PR_WRITE_EXCLUSIVE ||
> +	    rtype > NVME_PR_EXCLUSIVE_ACCESS_ALL_REGS)
> +		return NVME_SC_INVALID_FIELD | NVME_SC_DNR;
> +
> +	pr->rtype = rtype;
> +	return 0;
> +}
> +
> +static u16 nvmet_pr_set_rtype_and_holder(struct nvmet_pr *pr, u8 rtype,
> +					 struct nvmet_pr_registrant *reg)
> +{
> +	u16 ret;
> +
> +	ret = nvmet_pr_validate_and_set_rtype(pr, rtype);
> +	if (!ret)
> +		pr->holder = reg;
> +	return ret;
> +}
> +
> +static void nvmet_pr_inc_generation(struct nvmet_pr *pr)
> +{
> +	u32 gen = pr->generation;
> +
> +	gen++;
> +	if (gen == U32_MAX)
> +		gen = 0;
> +	pr->generation = gen;
> +}

Why isn't this just a generation++ ?

> +
> +static u16 nvmet_pr_register(struct nvmet_req *req,
> +			     struct nvmet_pr_register_data *d)

It'd be nice if the bdev could be used instead if it is supported
and nvmet would simply use pr_ops for it. Like Mike did for LIO.

> +{
> +	struct nvmet_ctrl *ctrl = req->sq->ctrl;
> +	struct nvmet_pr_registrant *reg;
> +	struct nvmet_pr *pr = &req->ns->pr;
> +	u16 ret = NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +
> +	read_lock(&pr->pr_lock);
> +	list_for_each_entry(reg, &pr->registrant_list, entry) {
> +		if (uuid_equal(&reg->hostid, &ctrl->hostid)) {
> +			if (reg->rkey == le64_to_cpu(d->nrkey))
> +				ret = 0;

NVME_SC_SUCCESS ?

> +			read_unlock(&pr->pr_lock);
> +			goto out;
> +		}
> +	}
> +	read_unlock(&pr->pr_lock);
> +
> +	reg = kmem_cache_alloc(registrant_cache, GFP_KERNEL);

Why are you using a dedicated kmem_cache? Are you looking to
optimize registration path?

> +	if (!reg) {
> +		ret = NVME_SC_INTERNAL;
> +		goto out;
> +	}
> +
> +	reg->rkey = le64_to_cpu(d->nrkey);
> +	uuid_copy(&reg->hostid, &ctrl->hostid);
> +	reg->is_holder = false;
> +
> +	write_lock(&pr->pr_lock);
> +	list_add_tail(&reg->entry, &pr->registrant_list);
> +	write_unlock(&pr->pr_lock);
> +	ret = 0;
> +out:
> +	return ret;
> +}
> +
> +static void __nvmet_pr_unregister_one(struct nvmet_pr *pr,
> +				      struct nvmet_pr_registrant *reg)
> +{
> +	list_del(&reg->entry);
> +	kmem_cache_free(registrant_cache, reg);
> +
> +	/*
> +	 * Temporarily, don't send notification, because host does not
> +	 * handle this event.
> +	 */

You should send a notification. Regardless of what the host does
or doesn't do.

> +	if (pr->holder && pr->holder == reg) {
> +		if (pr->rtype == NVME_PR_WRITE_EXCLUSIVE_ALL_REGS ||
> +		    pr->rtype == NVME_PR_EXCLUSIVE_ACCESS_ALL_REGS) {
> +			reg = list_first_entry(&pr->registrant_list,
> +					struct nvmet_pr_registrant, entry);
> +			pr->holder = reg;
> +			if (!reg)
> +				pr->rtype = 0;
> +		} else {
> +			nvmet_pr_clean_holder(pr);
> +		}

I'd do it the other way around, clean the holder and then assign
another holder for all registrants resvs.

> +	}
> +}
> +
> +static u16 nvmet_pr_unregister(struct nvmet_req *req,
> +			       struct nvmet_pr_register_data *d,
> +			       bool ignore_key)
> +{
> +	struct nvmet_pr *pr = &req->ns->pr;
> +	struct nvmet_ctrl *ctrl = req->sq->ctrl;
> +	struct nvmet_pr_registrant *reg;
> +	struct nvmet_pr_registrant *tmp;
> +	u16 ret = NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +
> +	write_lock(&pr->pr_lock);
> +	list_for_each_entry_safe(reg, tmp, &pr->registrant_list, entry) {
> +		if (uuid_equal(&reg->hostid, &ctrl->hostid)) {
> +			if (ignore_key || reg->rkey == le64_to_cpu(d->crkey)) {
> +				ret = 0;

NVME_SC_SUCCESS ?

> +				__nvmet_pr_unregister_one(pr, reg);
> +			}
> +			break;
> +		}
> +	}
> +	write_unlock(&pr->pr_lock);
> +	return ret;
> +}
> +
> +static u16 nvmet_pr_replace(struct nvmet_req *req,
> +			    struct nvmet_pr_register_data *d,
> +			    bool ignore_key)
> +{
> +	struct nvmet_ctrl *ctrl = req->sq->ctrl;
> +	struct nvmet_pr_registrant *reg;
> +	struct nvmet_pr *pr = &req->ns->pr;
> +	u16 ret = NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +
> +	write_lock(&pr->pr_lock);
> +	list_for_each_entry(reg, &pr->registrant_list, entry) {
> +		if (uuid_equal(&reg->hostid, &ctrl->hostid)) {
> +			if (ignore_key || reg->rkey == le64_to_cpu(d->crkey)) {
> +				reg->rkey = le64_to_cpu(d->nrkey);
> +				ret = 0;
> +			}
> +			break;
> +		}
> +	}
> +	write_unlock(&pr->pr_lock);
> +	return ret;
> +}
> +
> +static void nvmet_execute_pr_register(struct nvmet_req *req)
> +{
> +	u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10);
> +	u8 reg_act = cdw10 & 0x07;
> +	bool ignore_key = (bool)((cdw10 >> 3) & 1);
> +	u16 status = 0;
> +	struct nvmet_pr_register_data *d;
> +
> +	d = kmalloc(sizeof(*d), GFP_KERNEL);
> +	if (!d) {
> +		status = NVME_SC_INTERNAL;
> +		goto out;
> +	}
> +
> +	status = nvmet_copy_from_sgl(req, 0, d, sizeof(*d));
> +	if (status)
> +		goto free_data;
> +
> +	switch (reg_act) {
> +	case NVME_PR_REGISTER_ACT_REG:
> +		status = nvmet_pr_register(req, d);
> +		break;
> +	case NVME_PR_REGISTER_ACT_UNREG:
> +		status = nvmet_pr_unregister(req, d, ignore_key);
> +		break;
> +	case NVME_PR_REGISTER_ACT_REPLACE:
> +		status = nvmet_pr_replace(req, d, ignore_key);
> +		break;
> +	default:
> +		req->error_loc = offsetof(struct nvme_common_command,
> +						cdw10);
> +		status = NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
> +		break;
> +	}
> +free_data:
> +	kfree(d);
> +out:
> +	if (!status)
> +		nvmet_pr_inc_generation(&req->ns->pr);
> +	nvmet_req_complete(req, status);
> +}
> +
> +static u16 nvmet_pr_acquire(struct nvmet_req *req,
> +			    struct nvmet_pr_registrant *reg,
> +			    u8 rtype)
> +{
> +	struct nvmet_pr *pr = &req->ns->pr;
> +	u16 ret = NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +
> +	if (pr->holder && reg != pr->holder)
> +		goto out;
> +	if (pr->holder && reg == pr->holder) {
> +		if (pr->rtype == rtype)
> +			ret = 0;
> +		goto out;
> +	}
> +
> +	ret = nvmet_pr_set_rtype_and_holder(pr, rtype, reg);
> +out:
> +	return ret;
> +}
> +
> +static u16 nvmet_pr_unregister_by_prkey(struct nvmet_pr *pr, u64 prkey)
> +{
> +	struct nvmet_pr_registrant *reg;
> +	struct nvmet_pr_registrant *tmp;
> +	u16 ret = NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +
> +	list_for_each_entry_safe(reg, tmp,
> +				 &pr->registrant_list, entry) {
> +		if (reg->rkey == prkey) {
> +			ret = 0;

NVME_SC_SUCCESS ?

> +			__nvmet_pr_unregister_one(pr, reg);
> +		}
> +	}
> +	return ret;
> +}
> +
> +static void nvmet_pr_unregister_except_hostid(struct nvmet_pr *pr,
> +					      uuid_t *hostid)
> +{
> +	struct nvmet_pr_registrant *reg;
> +	struct nvmet_pr_registrant *tmp;
> +
> +	list_for_each_entry_safe(reg, tmp,
> +				 &pr->registrant_list, entry) {
> +		if (!uuid_equal(&reg->hostid, hostid))
> +			__nvmet_pr_unregister_one(pr, reg);
> +	}
> +}
> +
> +static u16 nvmet_pr_preempt(struct nvmet_req *req,
> +			    struct nvmet_pr_registrant *reg,
> +			    u8 rtype,
> +			    struct nvmet_pr_acquire_data *d)
> +{
> +	struct nvmet_ctrl *ctrl = req->sq->ctrl;
> +	struct nvmet_pr *pr = &req->ns->pr;
> +	u16 ret = 0;
> +
> +	if (!pr->holder)
> +		goto unregister_by_prkey;
> +
> +	if (pr->rtype == NVME_PR_WRITE_EXCLUSIVE_ALL_REGS ||
> +	    pr->rtype == NVME_PR_EXCLUSIVE_ACCESS_ALL_REGS) {
> +		if (!le64_to_cpu(d->prkey)) {
> +			nvmet_pr_unregister_except_hostid(pr, &ctrl->hostid);
> +			ret = nvmet_pr_set_rtype_and_holder(pr, rtype, reg);
> +			goto out;
> +		}
> +		goto unregister_by_prkey;
> +	}
> +
> +	if (pr->holder == reg) {
> +		nvmet_pr_validate_and_set_rtype(pr, rtype);
> +		goto out;
> +	}
> +
> +	if (le64_to_cpu(d->prkey) == pr->holder->rkey)
> +		goto new_reserve;
> +
> +	if (le64_to_cpu(d->prkey))
> +		goto unregister_by_prkey;
> +	ret = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
> +	goto out;
> +
> +new_reserve:
> +	ret = nvmet_pr_set_rtype_and_holder(pr, rtype, reg);
> +	if (ret)
> +		return ret;
> +unregister_by_prkey:
> +	ret = nvmet_pr_unregister_by_prkey(pr, le64_to_cpu(d->prkey));
> +out:
> +	return ret;
> +}
> +
> +static u16 nvmet_pr_preempt_and_abort(struct nvmet_req *req,
> +				      struct nvmet_pr_registrant *reg,
> +				      u8 rtype,
> +				      struct nvmet_pr_acquire_data *d)
> +{
> +	return nvmet_pr_preempt(req, reg, rtype, d);
> +}

What does this wrapper do?

> +
> +static u16 __nvmet_execute_pr_acquire(struct nvmet_req *req,
> +				      struct nvmet_pr_registrant *reg,
> +				      u8 acquire_act,
> +				      u8 rtype,
> +				      struct nvmet_pr_acquire_data *d)
> +{
> +	u16 status;
> +
> +	switch (acquire_act) {
> +	case NVME_PR_ACQUIRE_ACT_ACQUIRE:
> +		status = nvmet_pr_acquire(req, reg, rtype);
> +		goto out;
> +	case NVME_PR_ACQUIRE_ACT_PREEMPT:
> +		status = nvmet_pr_preempt(req, reg, rtype, d);
> +		goto inc_gen;
> +	case NVME_PR_ACQUIRE_ACT_PREEMPT_AND_ABORT:
> +		status = nvmet_pr_preempt_and_abort(req, reg, rtype, d);
> +		goto inc_gen;
> +	default:
> +		req->error_loc = offsetof(struct nvme_common_command,
> +						cdw10);
> +		status = NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
> +		goto out;
> +	}
> +inc_gen:
> +	nvmet_pr_inc_generation(&req->ns->pr);
> +out:
> +	return status;
> +}
> +
> +static void nvmet_execute_pr_acquire(struct nvmet_req *req)
> +{
> +	u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10);
> +	u8 acquire_act = cdw10 & 0x07;
> +	bool ignore_key = (bool)((cdw10 >> 3) & 1);
> +	u8 rtype = (u8)((cdw10 >> 8) & 0xff);
> +	u16 status = 0;
> +	struct nvmet_pr_acquire_data *d = NULL;
> +	struct nvmet_pr *pr = &req->ns->pr;
> +	struct nvmet_pr_registrant *reg;
> +	struct nvmet_ctrl *ctrl = req->sq->ctrl;
> +
> +	if (ignore_key) {
> +		status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
> +		goto out;
> +	}
> +
> +	d = kmalloc(sizeof(*d), GFP_KERNEL);
> +	if (!d) {
> +		status = NVME_SC_INTERNAL;
> +		goto out;
> +	}
> +
> +	status = nvmet_copy_from_sgl(req, 0, d, sizeof(*d));
> +	if (status)
> +		goto free_data;
> +
> +	status = NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +	write_lock(&pr->pr_lock);
> +	list_for_each_entry(reg, &pr->registrant_list, entry) {
> +		if (uuid_equal(&reg->hostid, &ctrl->hostid) &&
> +		    reg->rkey == le64_to_cpu(d->crkey)) {
> +			status = __nvmet_execute_pr_acquire(req, reg,
> +							    acquire_act,
> +							    rtype, d);
> +			break;
> +		}
> +	}
> +	write_unlock(&pr->pr_lock);
> +
> +free_data:
> +	kfree(d);
> +out:
> +	nvmet_req_complete(req, status);
> +}
> +
> +static u16 nvmet_pr_release(struct nvmet_req *req,
> +			    struct nvmet_pr_registrant *reg,
> +			    u8 rtype)
> +{
> +	struct nvmet_pr *pr = &req->ns->pr;
> +
> +	if (reg != pr->holder || !pr->holder)
> +		return 0;
> +	if (pr->rtype != rtype)
> +		return NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +
> +	nvmet_pr_clean_holder(pr);
> +
> +	/*
> +	 * Temporarily, don't send notification, because host does not
> +	 * handle this event.
> +	 */
> +	return 0;
> +}
> +
> +static void __nvmet_pr_clean_all_registrants(struct nvmet_pr *pr)
> +{
> +	struct nvmet_pr_registrant *tmp_reg;
> +	struct nvmet_pr_registrant *tmp;
> +
> +	list_for_each_entry_safe(tmp_reg, tmp, &pr->registrant_list, entry)
> +		__nvmet_pr_unregister_one(pr, tmp_reg);
> +}
> +
> +static void nvmet_pr_clear(struct nvmet_req *req,
> +			   struct nvmet_pr_registrant *reg)
> +{
> +	struct nvmet_pr *pr = &req->ns->pr;
> +
> +	__nvmet_pr_clean_all_registrants(pr);
> +
> +	nvmet_pr_inc_generation(&req->ns->pr);
> +
> +	/*
> +	 * Temporarily, don't send notification, because host does not
> +	 * handle this event.
> +	 */
> +}
> +
> +static u16 __nvmet_execute_pr_release(struct nvmet_req *req,
> +				      struct nvmet_pr_registrant *reg,
> +				      u8 release_act, u8 rtype)
> +{
> +	switch (release_act) {
> +	case NVME_PR_RELEASE_ACT_RELEASE:
> +		return nvmet_pr_release(req, reg, rtype);
> +	case NVME_PR_RELEASE_ACT_CLEAR:
> +		nvmet_pr_clear(req, reg);
> +		return 0;
> +	default:
> +		req->error_loc = offsetof(struct nvme_common_command,
> +					  cdw10);
> +		return NVME_SC_INVALID_OPCODE | NVME_SC_DNR;
> +	}
> +}
> +
> +static void nvmet_execute_pr_release(struct nvmet_req *req)
> +{
> +	u32 cdw10 = le32_to_cpu(req->cmd->common.cdw10);
> +	u8 release_act = cdw10 & 0x07;
> +	bool ignore_key = (bool)((cdw10 >> 3) & 1);
> +	u8 rtype = (u8)((cdw10 >> 8) & 0xff);
> +	struct nvmet_pr_release_data *d;
> +	struct nvmet_pr *pr = &req->ns->pr;
> +	struct nvmet_pr_registrant *reg;
> +	struct nvmet_pr_registrant *tmp;
> +	struct nvmet_ctrl *ctrl = req->sq->ctrl;
> +	u16 status;
> +
> +	if (ignore_key) {
> +		status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
> +		goto out;
> +	}
> +
> +	d = kmalloc(sizeof(*d), GFP_KERNEL);
> +	if (!d) {
> +		status = NVME_SC_INTERNAL;
> +		goto out;
> +	}
> +
> +	status = nvmet_copy_from_sgl(req, 0, d, sizeof(*d));
> +	if (status)
> +		goto free_data;
> +
> +	status = NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +	write_lock(&pr->pr_lock);
> +	list_for_each_entry_safe(reg, tmp,
> +				 &pr->registrant_list, entry) {
> +		if (uuid_equal(&reg->hostid, &ctrl->hostid) &&
> +		    reg->rkey == le64_to_cpu(d->crkey)) {
> +			status = __nvmet_execute_pr_release(req, reg,
> +							    release_act, rtype);
> +			break;
> +		}
> +	}
> +	write_unlock(&pr->pr_lock);
> +free_data:
> +	kfree(d);
> +out:
> +	nvmet_req_complete(req, status);
> +}
> +
> +static struct nvmet_pr_registrant *nvmet_pr_find_registrant_by_hostid(
> +						struct nvmet_pr *pr,
> +						uuid_t *hostid)
> +{
> +	struct nvmet_pr_registrant *reg;
> +
> +	list_for_each_entry(reg, &pr->registrant_list, entry) {
> +		if (uuid_equal(&reg->hostid, hostid))
> +			return reg;
> +	}
> +	return NULL;
> +}
> +
> +static void nvmet_execute_pr_report(struct nvmet_req *req)
> +{
> +	u32 cdw11 = le32_to_cpu(req->cmd->common.cdw11);
> +	u8 eds = cdw11 & 1;
> +	u16 status = 0;
> +	u32 num_bytes = 4 * (le32_to_cpu(req->cmd->common.cdw10) + 1);
> +	struct nvmet_subsys *subsys = req->sq->ctrl->subsys;
> +	struct nvmet_ctrl *ctrl;
> +	struct nvmet_pr *pr = &req->ns->pr;
> +	struct nvme_reservation_status_ext *data;
> +	struct nvme_registered_ctrl_ext *ctrl_data;
> +	struct nvmet_pr_registrant *reg;
> +	u16 num_ctrls = 0;
> +
> +	/* nvmet hostid(uuid_t) is 128 bit. */
> +	if (!eds) {
> +		req->error_loc = offsetof(struct nvme_common_command,
> +					  cdw11);
> +		status = NVME_SC_HOST_ID_INCONSIST | NVME_SC_DNR;
> +		goto out;
> +	}
> +
> +	if (num_bytes < sizeof(struct nvme_reservation_status_ext)) {
> +		req->error_loc = offsetof(struct nvme_common_command,
> +					  cdw10);
> +		status = NVME_SC_INVALID_FIELD | NVME_SC_DNR;
> +		goto out;
> +	}
> +
> +	mutex_lock(&subsys->lock);
> +	read_lock(&pr->pr_lock);
> +	list_for_each_entry(ctrl, &subsys->ctrls, subsys_entry) {
> +		if (nvmet_pr_find_registrant_by_hostid(pr, &ctrl->hostid))
> +			num_ctrls++;
> +	}
> +
> +	num_bytes = min(num_bytes,
> +			sizeof(struct nvme_reservation_status_ext) +
> +			sizeof(struct nvme_registered_ctrl_ext) * num_ctrls);
> +
> +	data = kmalloc(num_bytes, GFP_KERNEL);
> +	if (!data) {
> +		status = NVME_SC_INTERNAL;
> +		goto out;
> +	}
> +	memset(data, 0, num_bytes);
> +	data->gen = cpu_to_le32(pr->generation);
> +	data->rtype = pr->rtype;
> +	put_unaligned_le16(num_ctrls, data->regctl);
> +	data->ptpls = 0;
> +	ctrl_data = data->regctl_eds;
> +
> +	list_for_each_entry(ctrl, &subsys->ctrls, subsys_entry) {
> +		if ((void *)ctrl_data >= (void *)(data + num_bytes))
> +			break;
> +		reg = nvmet_pr_find_registrant_by_hostid(pr, &ctrl->hostid);
> +		if (!reg)
> +			continue;
> +		ctrl_data->cntlid = cpu_to_le16(ctrl->cntlid);
> +		if (pr->rtype == NVME_PR_WRITE_EXCLUSIVE_ALL_REGS ||
> +		    pr->rtype == NVME_PR_EXCLUSIVE_ACCESS_ALL_REGS)
> +			ctrl_data->rcsts = 1;
> +		if (pr->holder &&
> +		    uuid_equal(&ctrl->hostid, &pr->holder->hostid))
> +			ctrl_data->rcsts = 1;
> +		uuid_copy((uuid_t *)&ctrl_data->hostid, &ctrl->hostid);
> +		ctrl_data->rkey = cpu_to_le64(reg->rkey);
> +		ctrl_data++;
> +	}
> +	status = nvmet_copy_to_sgl(req, 0, data, num_bytes);
> +out:
> +	read_unlock(&pr->pr_lock);

it is unclear what pr_lock protects when it is released here. Can you
explain why it is released here?

> +	mutex_unlock(&subsys->lock);
> +	nvmet_req_complete(req, status);
> +}
> +
> +u16 nvmet_parse_pr_cmd(struct nvmet_req *req)
> +{
> +	struct nvme_command *cmd = req->cmd;
> +
> +	if (!req->ns->pr.enable)
> +		return 1;
> +
> +	switch (cmd->common.opcode) {
> +	case nvme_cmd_resv_register:
> +		req->execute = nvmet_execute_pr_register;
> +		break;
> +	case nvme_cmd_resv_acquire:
> +		req->execute = nvmet_execute_pr_acquire;
> +		break;
> +	case nvme_cmd_resv_release:
> +		req->execute = nvmet_execute_pr_release;
> +		break;
> +	case nvme_cmd_resv_report:
> +		req->execute = nvmet_execute_pr_report;
> +		break;
> +	default:
> +		return 1;
> +	}
> +	return 0;
> +}
> +
> +static bool nvmet_is_req_copy_cmd_group(struct nvmet_req *req)
> +{
> +	struct nvme_command *cmd = req->cmd;
> +	u8 opcode = cmd->common.opcode;
> +
> +	return req->sq->qid && opcode == nvme_cmd_copy;
> +}
> +
> +static bool nvmet_is_req_write_cmd_group(struct nvmet_req *req)
> +{
> +	struct nvme_command *cmd = req->cmd;
> +	u8 opcode = cmd->common.opcode;
> +
> +	if (req->sq->qid) {
> +		switch (opcode) {
> +		case nvme_cmd_flush:
> +		case nvme_cmd_write:
> +		case nvme_cmd_write_zeroes:
> +		case nvme_cmd_dsm:
> +		case nvme_cmd_write_uncor:
> +			return true;
> +		default:
> +			return false;
> +		}
> +	}
> +	switch (opcode) {
> +	case nvme_admin_cap_mgmt:
> +	case nvme_admin_format_nvm:
> +	case nvme_admin_ns_attach:
> +	case nvme_admin_ns_mgmt:
> +	case nvme_admin_sanitize_nvm:
> +	case nvme_admin_security_send:
> +		return true;
> +	default:
> +		return false;
> +	}
> +}
> +
> +static bool nvmet_is_req_read_cmd_group(struct nvmet_req *req)
> +{
> +	struct nvme_command *cmd = req->cmd;
> +	u8 opcode = cmd->common.opcode;
> +
> +	if (req->sq->qid) {
> +		switch (opcode) {
> +		case nvme_cmd_read:
> +		case nvme_cmd_compare:
> +		case nvme_cmd_verify:
> +			return true;
> +		default:
> +			return false;
> +		}
> +	}
> +	switch (opcode) {
> +	case nvme_admin_security_recv:
> +		return true;
> +	default:
> +		return false;
> +	}
> +}
> +
> +u16 nvmet_pr_check_cmd_access(struct nvmet_req *req)
> +{
> +	struct nvmet_ctrl *ctrl = req->sq->ctrl;
> +	struct nvmet_ns *ns = req->ns;
> +	struct nvmet_pr *pr;
> +	u16 ret = 0;
> +
> +	if (!ns) {
> +		ns = xa_load(&nvmet_req_subsys(req)->namespaces,
> +			     le32_to_cpu(req->cmd->common.nsid));
> +		if (!ns)
> +			return 0;
> +		percpu_ref_get(&ns->ref);
> +	}
> +
> +	pr = &ns->pr;
> +	if (!pr->enable)
> +		goto out;
> +
> +	read_lock(&pr->pr_lock);

I'm wandering if for the I/O path we can get away with a lighter-weight
lock than a rwlock ? Maybe an rcu for the I/O path?

> +	if (!pr->holder)
> +		goto unlock;
> +	if (uuid_equal(&ctrl->hostid, &pr->holder->hostid))
> +		goto unlock;
> +
> +	/*
> +	 * The Reservation command group is checked in executing,
> +	 * allow it here.
> +	 */
> +	switch ((u8)pr->rtype) {
> +	case NVME_PR_WRITE_EXCLUSIVE:
> +		if (nvmet_is_req_write_cmd_group(req) ||
> +		    nvmet_is_req_copy_cmd_group(req))

nvmet does not support copy, so it should fail before that. Once
nvmet adds copy-offload support, than this can be added (or maybe
it comes before that), but lets not add an unsupported opcode for
this redundant check.

> +			ret = NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +		break;
> +	case NVME_PR_EXCLUSIVE_ACCESS:
> +		if (nvmet_is_req_copy_cmd_group(req) ||
> +		    nvmet_is_req_read_cmd_group(req) ||
> +		    nvmet_is_req_write_cmd_group(req))
> +			ret = NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +		break;
> +	case NVME_PR_WRITE_EXCLUSIVE_REG_ONLY:
> +	case NVME_PR_WRITE_EXCLUSIVE_ALL_REGS:
> +		if ((nvmet_is_req_copy_cmd_group(req) ||
> +		    nvmet_is_req_write_cmd_group(req)) &&
> +		    !nvmet_pr_find_registrant_by_hostid(pr, &ctrl->hostid))
> +			ret = NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +		break;
> +	case NVME_PR_EXCLUSIVE_ACCESS_REG_ONLY:
> +	case NVME_PR_EXCLUSIVE_ACCESS_ALL_REGS:
> +		if ((nvmet_is_req_copy_cmd_group(req) ||
> +		    nvmet_is_req_read_cmd_group(req) ||
> +		    nvmet_is_req_write_cmd_group(req)) &&
> +		    !nvmet_pr_find_registrant_by_hostid(pr, &ctrl->hostid))
> +			ret = NVME_SC_RESERVATION_CONFLICT | NVME_SC_DNR;
> +		break;
> +	default:
> +		WARN_ON_ONCE(1);

WARN? perhaps just a pr_warn?

> +		break;
> +	}
> +
> +unlock:
> +	read_unlock(&pr->pr_lock);
> +out:
> +	if (!req->ns)
> +		nvmet_put_namespace(ns);
> +	return ret;
> +}
> +
> +void nvmet_pr_unregister_ctrl_host(struct nvmet_ctrl *ctrl)
> +{
> +	struct nvmet_ns *ns;
> +	unsigned long idx;
> +	struct nvmet_pr_registrant *reg;
> +	struct nvmet_pr_registrant *tmp;
> +
> +	if (nvmet_is_host_still_connected(&ctrl->hostid, ctrl->subsys))
> +		return;
> +
> +	xa_for_each(&ctrl->subsys->namespaces, idx, ns) {
> +		write_lock(&ns->pr.pr_lock);
> +		list_for_each_entry_safe(reg, tmp,
> +					 &ns->pr.registrant_list, entry) {
> +			if (uuid_equal(&reg->hostid, &ctrl->hostid)) {
> +				__nvmet_pr_unregister_one(&ns->pr, reg);
> +				break;
> +			}
> +		}
> +		write_unlock(&ns->pr.pr_lock);
> +	}
> +}
> +
> +void nvmet_pr_clean_all_registrants(struct nvmet_pr *pr)
> +{
> +	write_lock(&pr->pr_lock);
> +	__nvmet_pr_clean_all_registrants(pr);
> +	write_unlock(&pr->pr_lock);
> +}
> +
> +void nvmet_pr_init_ns(struct nvmet_ns *ns)
> +{
> +	ns->pr.holder = NULL;
> +	ns->pr.generation = 0;
> +	ns->pr.rtype = 0;
> +	rwlock_init(&ns->pr.pr_lock);
> +	INIT_LIST_HEAD(&ns->pr.registrant_list);
> +}
> +
> +static void nvmet_pr_registrant_init(void *addr)
> +{
> +	struct nvmet_pr_registrant *reg = (struct nvmet_pr_registrant *)addr;
> +
> +	memset(reg, 0, sizeof(*reg));
> +	INIT_LIST_HEAD(&reg->entry);
> +}
> +
> +int nvmet_pr_init(void)
> +{
> +	registrant_cache = kmem_cache_create("nvmet-registrant-cache",
> +					sizeof(struct nvmet_pr_registrant),
> +					sizeof(struct nvmet_pr_registrant),
> +					0,
> +					nvmet_pr_registrant_init);
> +	if (!registrant_cache)
> +		return -ENOMEM;
> +	return 0;
> +}
> +
> +void nvmet_pr_exit(void)
> +{
> +	kmem_cache_destroy(registrant_cache);
> +}
> diff --git a/include/linux/nvme.h b/include/linux/nvme.h
> index 44325c068b6a..adf956ea424c 100644
> --- a/include/linux/nvme.h
> +++ b/include/linux/nvme.h
> @@ -766,6 +766,17 @@ enum {
>   	NVME_LBART_ATTRIB_HIDE	= 1 << 1,
>   };
>   
> +enum nvme_pr_capabilities {
> +	NVME_PR_SUPPORT_PTPL				= 1,
> +	NVME_PR_SUPPORT_WRITE_EXCLUSIVE			= 1 << 1,
> +	NVME_PR_SUPPORT_EXCLUSIVE_ACCESS		= 1 << 2,
> +	NVME_PR_SUPPORT_WRITE_EXCLUSIVE_REG_ONLY	= 1 << 3,
> +	NVME_PR_SUPPORT_EXCLUSIVE_ACCESS_REG_ONLY	= 1 << 4,
> +	NVME_PR_SUPPORT_WRITE_EXCLUSIVE_ALL_REGS	= 1 << 5,
> +	NVME_PR_SUPPORT_EXCLUSIVE_ACCESS_ALL_REGS	= 1 << 6,
> +	NVME_PR_SUPPORT_IEKEY_DEF_LATER_VER_1_3		= 1 << 7,
> +};
> +
>   enum nvme_pr_type {
>   	NVME_PR_WRITE_EXCLUSIVE			= 1,
>   	NVME_PR_EXCLUSIVE_ACCESS		= 2,
> @@ -775,6 +786,23 @@ enum nvme_pr_type {
>   	NVME_PR_EXCLUSIVE_ACCESS_ALL_REGS	= 6,
>   };
>   
> +enum nvme_pr_register_action {
> +	NVME_PR_REGISTER_ACT_REG		= 0,
> +	NVME_PR_REGISTER_ACT_UNREG		= 1,
> +	NVME_PR_REGISTER_ACT_REPLACE		= 1 << 1,
> +};
> +
> +enum nvme_pr_acquire_action {
> +	NVME_PR_ACQUIRE_ACT_ACQUIRE		= 0,
> +	NVME_PR_ACQUIRE_ACT_PREEMPT		= 1,
> +	NVME_PR_ACQUIRE_ACT_PREEMPT_AND_ABORT	= 1 << 1,
> +};
> +
> +enum nvme_pr_release_action {
> +	NVME_PR_RELEASE_ACT_RELEASE		= 0,
> +	NVME_PR_RELEASE_ACT_CLEAR		= 1,
> +};
> +
>   enum nvme_eds {
>   	NVME_EXTENDED_DATA_STRUCT	= 0x1,
>   };
> @@ -838,6 +866,7 @@ enum nvme_opcode {
>   	nvme_cmd_resv_report	= 0x0e,
>   	nvme_cmd_resv_acquire	= 0x11,
>   	nvme_cmd_resv_release	= 0x15,
> +	nvme_cmd_copy		= 0x19,
>   	nvme_cmd_zone_mgmt_send	= 0x79,
>   	nvme_cmd_zone_mgmt_recv	= 0x7a,
>   	nvme_cmd_zone_append	= 0x7d,
> @@ -1162,6 +1191,7 @@ enum nvme_admin_opcode {
>   	nvme_admin_virtual_mgmt		= 0x1c,
>   	nvme_admin_nvme_mi_send		= 0x1d,
>   	nvme_admin_nvme_mi_recv		= 0x1e,
> +	nvme_admin_cap_mgmt		= 0x20,
>   	nvme_admin_dbbuf		= 0x7C,
>   	nvme_admin_format_nvm		= 0x80,
>   	nvme_admin_security_send	= 0x81,



More information about the Linux-nvme mailing list