Merge tag 'libnvdimm-for-4.21' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm

Pull libnvdimm updates from Dan Williams:
"The vast bulk of this update is the new support for the security
capabilities of some nvdimms.

The userspace tooling for this capability is still a work in progress,
but the changes survive the existing libnvdimm unit tests. The changes
also pass manual checkout on hardware and the new nfit_test emulation
of the security capability.

The touches of the security/keys/ files have received the necessary
acks from Mimi and David. Those changes were necessary to allow for a
new generic encrypted-key type, and allow the nvdimm sub-system to
lookup key material referenced by the libnvdimm-sysfs interface.

Summary:

- Add support for the security features of nvdimm devices that
implement a security model similar to ATA hard drive security. The
security model supports locking access to the media at
device-power-loss, to be unlocked with a passphrase, and
secure-erase (crypto-scramble).

Unlike the ATA security case where the kernel expects device
security to be managed in a pre-OS environment, the libnvdimm
security implementation allows key provisioning and key-operations
at OS runtime. Keys are managed with the kernel's encrypted-keys
facility to provide data-at-rest security for the libnvdimm key
material. The usage model mirrors fscrypt key management, but is
driven via libnvdimm sysfs.

- Miscellaneous updates for api usage and comment fixes"

* tag 'libnvdimm-for-4.21' of git://git.kernel.org/pub/scm/linux/kernel/git/nvdimm/nvdimm: (21 commits)
libnvdimm/security: Quiet security operations
libnvdimm/security: Add documentation for nvdimm security support
tools/testing/nvdimm: add Intel DSM 1.8 support for nfit_test
tools/testing/nvdimm: Add overwrite support for nfit_test
tools/testing/nvdimm: Add test support for Intel nvdimm security DSMs
acpi/nfit, libnvdimm/security: add Intel DSM 1.8 master passphrase support
acpi/nfit, libnvdimm/security: Add security DSM overwrite support
acpi/nfit, libnvdimm: Add support for issue secure erase DSM to Intel nvdimm
acpi/nfit, libnvdimm: Add enable/update passphrase support for Intel nvdimms
acpi/nfit, libnvdimm: Add disable passphrase support to Intel nvdimm.
acpi/nfit, libnvdimm: Add unlock of nvdimm support for Intel DIMMs
acpi/nfit, libnvdimm: Add freeze security support to Intel nvdimm
acpi/nfit, libnvdimm: Introduce nvdimm_security_ops
keys-encrypted: add nvdimm key format type to encrypted keys
keys: Export lookup_user_key to external users
acpi/nfit, libnvdimm: Store dimm id as a member to struct nvdimm
libnvdimm, namespace: Replace kmemdup() with kstrndup()
libnvdimm, label: Switch to bitmap_zalloc()
ACPI/nfit: Adjust annotation for why return 0 if fail to find NFIT at start
libnvdimm, bus: Check id immediately following ida_simple_get
...

Documentation/nvdimm/security.txt

···
+NVDIMM SECURITY
+===============
+
+1. Introduction
+---------------
+
+With the introduction of Intel Device Specific Methods (DSM) v1.8
+specification [1], security DSMs are introduced. The spec added the following
+security DSMs: "get security state", "set passphrase", "disable passphrase",
+"unlock unit", "freeze lock", "secure erase", and "overwrite". A security_ops
+data structure has been added to struct dimm in order to support the security
+operations and generic APIs are exposed to allow vendor neutral operations.
+
+2. Sysfs Interface
+------------------
+The "security" sysfs attribute is provided in the nvdimm sysfs directory. For
+example:
+/sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0012:00/ndbus0/nmem0/security
+
+The "show" attribute of that attribute will display the security state for
+that DIMM. The following states are available: disabled, unlocked, locked,
+frozen, and overwrite. If security is not supported, the sysfs attribute
+will not be visible.
+
+The "store" attribute takes several commands when it is being written to
+in order to support some of the security functionalities:
+update <old_keyid> <new_keyid> - enable or update passphrase.
+disable <keyid> - disable enabled security and remove key.
+freeze - freeze changing of security states.
+erase <keyid> - delete existing user encryption key.
+overwrite <keyid> - wipe the entire nvdimm.
+master_update <keyid> <new_keyid> - enable or update master passphrase.
+master_erase <keyid> - delete existing user encryption key.
+
+3. Key Management
+-----------------
+
+The key is associated to the payload by the DIMM id. For example:
+# cat /sys/devices/LNXSYSTM:00/LNXSYBUS:00/ACPI0012:00/ndbus0/nmem0/nfit/id
+8089-a2-1740-00000133
+The DIMM id would be provided along with the key payload (passphrase) to
+the kernel.
+
+The security keys are managed on the basis of a single key per DIMM. The
+key "passphrase" is expected to be 32bytes long. This is similar to the ATA
+security specification [2]. A key is initially acquired via the request_key()
+kernel API call during nvdimm unlock. It is up to the user to make sure that
+all the keys are in the kernel user keyring for unlock.
+
+A nvdimm encrypted-key of format enc32 has the description format of:
+nvdimm:<bus-provider-specific-unique-id>
+
+See file ``Documentation/security/keys/trusted-encrypted.rst`` for creating
+encrypted-keys of enc32 format. TPM usage with a master trusted key is
+preferred for sealing the encrypted-keys.
+
+4. Unlocking
+------------
+When the DIMMs are being enumerated by the kernel, the kernel will attempt to
+retrieve the key from the kernel user keyring. This is the only time
+a locked DIMM can be unlocked. Once unlocked, the DIMM will remain unlocked
+until reboot. Typically an entity (i.e. shell script) will inject all the
+relevant encrypted-keys into the kernel user keyring during the initramfs phase.
+This provides the unlock function access to all the related keys that contain
+the passphrase for the respective nvdimms.  It is also recommended that the
+keys are injected before libnvdimm is loaded by modprobe.
+
+5. Update
+---------
+When doing an update, it is expected that the existing key is removed from
+the kernel user keyring and reinjected as different (old) key. It's irrelevant
+what the key description is for the old key since we are only interested in the
+keyid when doing the update operation. It is also expected that the new key
+is injected with the description format described from earlier in this
+document.  The update command written to the sysfs attribute will be with
+the format:
+update <old keyid> <new keyid>
+
+If there is no old keyid due to a security enabling, then a 0 should be
+passed in.
+
+6. Freeze
+---------
+The freeze operation does not require any keys. The security config can be
+frozen by a user with root privelege.
+
+7. Disable
+----------
+The security disable command format is:
+disable <keyid>
+
+An key with the current passphrase payload that is tied to the nvdimm should be
+in the kernel user keyring.
+
+8. Secure Erase
+---------------
+The command format for doing a secure erase is:
+erase <keyid>
+
+An key with the current passphrase payload that is tied to the nvdimm should be
+in the kernel user keyring.
+
+9. Overwrite
+------------
+The command format for doing an overwrite is:
+overwrite <keyid>
+
+Overwrite can be done without a key if security is not enabled. A key serial
+of 0 can be passed in to indicate no key.
+
+The sysfs attribute "security" can be polled to wait on overwrite completion.
+Overwrite can last tens of minutes or more depending on nvdimm size.
+
+An encrypted-key with the current user passphrase that is tied to the nvdimm
+should be injected and its keyid should be passed in via sysfs.
+
+10. Master Update
+-----------------
+The command format for doing a master update is:
+update <old keyid> <new keyid>
+
+The operating mechanism for master update is identical to update except the
+master passphrase key is passed to the kernel. The master passphrase key
+is just another encrypted-key.
+
+This command is only available when security is disabled.
+
+11. Master Erase
+----------------
+The command format for doing a master erase is:
+master_erase <current keyid>
+
+This command has the same operating mechanism as erase except the master
+passphrase key is passed to the kernel. The master passphrase key is just
+another encrypted-key.
+
+This command is only available when the master security is enabled, indicated
+by the extended security status.
+
+[1]: http://pmem.io/documents/NVDIMM_DSM_Interface-V1.8.pdf
+[2]: http://www.t13.org/documents/UploadedDocuments/docs2006/e05179r4-ACS-SecurityClarifications.pdf

Documentation/security/keys/trusted-encrypted.rst

···
 
 Where::
 
-	format:= 'default | ecryptfs'
+	format:= 'default | ecryptfs | enc32'
 	key-type:= 'trusted' | 'user'
 
 
···
 in order to use encrypted keys to mount an eCryptfs filesystem.  More details
 about the usage can be found in the file
 ``Documentation/security/keys/ecryptfs.rst``.
+
+Another new format 'enc32' has been defined in order to support encrypted keys
+with payload size of 32 bytes. This will initially be used for nvdimm security
+but may expand to other usages that require 32 bytes payload.

drivers/acpi/nfit/Kconfig

···
 
 	  To compile this driver as a module, choose M here:
 	  the module will be called nfit.
+
+config NFIT_SECURITY_DEBUG
+	bool "Enable debug for NVDIMM security commands"
+	depends on ACPI_NFIT
+	help
+	  Some NVDIMM devices and controllers support encryption and
+	  other security features. The payloads for the commands that
+	  enable those features may contain sensitive clear-text
+	  security material. Disable debug of those command payloads
+	  by default. If you are a kernel developer actively working
+	  on NVDIMM security enabling say Y, otherwise say N.

drivers/acpi/nfit/Makefile

···
 obj-$(CONFIG_ACPI_NFIT) := nfit.o
 nfit-y := core.o
+nfit-y += intel.o
 nfit-$(CONFIG_X86_MCE) += mce.o

drivers/acpi/nfit/core.c

···
 #include <linux/nd.h>
 #include <asm/cacheflush.h>
 #include <acpi/nfit.h>
+#include "intel.h"
 #include "nfit.h"
 #include "intel.h"
 
···
 			[NVDIMM_INTEL_QUERY_FWUPDATE] = 2,
 			[NVDIMM_INTEL_SET_THRESHOLD] = 2,
 			[NVDIMM_INTEL_INJECT_ERROR] = 2,
+			[NVDIMM_INTEL_GET_SECURITY_STATE] = 2,
+			[NVDIMM_INTEL_SET_PASSPHRASE] = 2,
+			[NVDIMM_INTEL_DISABLE_PASSPHRASE] = 2,
+			[NVDIMM_INTEL_UNLOCK_UNIT] = 2,
+			[NVDIMM_INTEL_FREEZE_LOCK] = 2,
+			[NVDIMM_INTEL_SECURE_ERASE] = 2,
+			[NVDIMM_INTEL_OVERWRITE] = 2,
+			[NVDIMM_INTEL_QUERY_OVERWRITE] = 2,
+			[NVDIMM_INTEL_SET_MASTER_PASSPHRASE] = 2,
+			[NVDIMM_INTEL_MASTER_SECURE_ERASE] = 2,
 		},
 	};
 	u8 id;
···
 	if (id == 0)
 		return 1; /* default */
 	return id;
+}
+
+static bool payload_dumpable(struct nvdimm *nvdimm, unsigned int func)
+{
+	struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm);
+
+	if (nfit_mem && nfit_mem->family == NVDIMM_FAMILY_INTEL
+			&& func >= NVDIMM_INTEL_GET_SECURITY_STATE
+			&& func <= NVDIMM_INTEL_MASTER_SECURE_ERASE)
+		return IS_ENABLED(CONFIG_NFIT_SECURITY_DEBUG);
+	return true;
 }
 
 int acpi_nfit_ctl(struct nvdimm_bus_descriptor *nd_desc, struct nvdimm *nvdimm,
···
 
 	dev_dbg(dev, "%s cmd: %d: func: %d input length: %d\n",
 		dimm_name, cmd, func, in_buf.buffer.length);
-	print_hex_dump_debug("nvdimm in  ", DUMP_PREFIX_OFFSET, 4, 4,
-			in_buf.buffer.pointer,
-			min_t(u32, 256, in_buf.buffer.length), true);
+	if (payload_dumpable(nvdimm, func))
+		print_hex_dump_debug("nvdimm in  ", DUMP_PREFIX_OFFSET, 4, 4,
+				in_buf.buffer.pointer,
+				min_t(u32, 256, in_buf.buffer.length), true);
 
 	/* call the BIOS, prefer the named methods over _DSM if available */
 	if (nvdimm && cmd == ND_CMD_GET_CONFIG_SIZE
···
 static ssize_t id_show(struct device *dev,
 		struct device_attribute *attr, char *buf)
 {
-	struct acpi_nfit_control_region *dcr = to_nfit_dcr(dev);
+	struct nvdimm *nvdimm = to_nvdimm(dev);
+	struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm);
 
-	if (dcr->valid_fields & ACPI_NFIT_CONTROL_MFG_INFO_VALID)
-		return sprintf(buf, "%04x-%02x-%04x-%08x\n",
-				be16_to_cpu(dcr->vendor_id),
-				dcr->manufacturing_location,
-				be16_to_cpu(dcr->manufacturing_date),
-				be32_to_cpu(dcr->serial_number));
-	else
-		return sprintf(buf, "%04x-%08x\n",
-				be16_to_cpu(dcr->vendor_id),
-				be32_to_cpu(dcr->serial_number));
+	return sprintf(buf, "%s\n", nfit_mem->id);
 }
 static DEVICE_ATTR_RO(id);
 
···
 	const guid_t *guid;
 	int i;
 	int family = -1;
+	struct acpi_nfit_control_region *dcr = nfit_mem->dcr;
 
 	/* nfit test assumes 1:1 relationship between commands and dsms */
 	nfit_mem->dsm_mask = acpi_desc->dimm_cmd_force_en;
 	nfit_mem->family = NVDIMM_FAMILY_INTEL;
+
+	if (dcr->valid_fields & ACPI_NFIT_CONTROL_MFG_INFO_VALID)
+		sprintf(nfit_mem->id, "%04x-%02x-%04x-%08x",
+				be16_to_cpu(dcr->vendor_id),
+				dcr->manufacturing_location,
+				be16_to_cpu(dcr->manufacturing_date),
+				be32_to_cpu(dcr->serial_number));
+	else
+		sprintf(nfit_mem->id, "%04x-%08x",
+				be16_to_cpu(dcr->vendor_id),
+				be32_to_cpu(dcr->serial_number));
+
 	adev = to_acpi_dev(acpi_desc);
 	if (!adev) {
 		/* unit test case */
···
 	mutex_unlock(&acpi_desc->init_mutex);
 }
 
+static const struct nvdimm_security_ops *acpi_nfit_get_security_ops(int family)
+{
+	switch (family) {
+	case NVDIMM_FAMILY_INTEL:
+		return intel_security_ops;
+	default:
+		return NULL;
+	}
+}
+
 static int acpi_nfit_register_dimms(struct acpi_nfit_desc *acpi_desc)
 {
 	struct nfit_mem *nfit_mem;
···
 
 		flush = nfit_mem->nfit_flush ? nfit_mem->nfit_flush->flush
 			: NULL;
-		nvdimm = nvdimm_create(acpi_desc->nvdimm_bus, nfit_mem,
+		nvdimm = __nvdimm_create(acpi_desc->nvdimm_bus, nfit_mem,
 				acpi_nfit_dimm_attribute_groups,
 				flags, cmd_mask, flush ? flush->hint_count : 0,
-				nfit_mem->flush_wpq);
+				nfit_mem->flush_wpq, &nfit_mem->id[0],
+				acpi_nfit_get_security_ops(nfit_mem->family));
 		if (!nvdimm)
 			return -ENOMEM;
 
···
 		nvdimm = nfit_mem->nvdimm;
 		if (!nvdimm)
 			continue;
+
+		rc = nvdimm_security_setup_events(nvdimm);
+		if (rc < 0)
+			dev_warn(acpi_desc->dev,
+				"security event setup failed: %d\n", rc);
 
 		nfit_kernfs = sysfs_get_dirent(nvdimm_kobj(nvdimm)->sd, "nfit");
 		if (nfit_kernfs)
···
 	return 0;
 }
 
-static int acpi_nfit_clear_to_send(struct nvdimm_bus_descriptor *nd_desc,
+static int __acpi_nfit_clear_to_send(struct nvdimm_bus_descriptor *nd_desc,
 		struct nvdimm *nvdimm, unsigned int cmd)
 {
 	struct acpi_nfit_desc *acpi_desc = to_acpi_nfit_desc(nd_desc);
···
 		return -EBUSY;
 
 	return 0;
+}
+
+/* prevent security commands from being issued via ioctl */
+static int acpi_nfit_clear_to_send(struct nvdimm_bus_descriptor *nd_desc,
+		struct nvdimm *nvdimm, unsigned int cmd, void *buf)
+{
+	struct nd_cmd_pkg *call_pkg = buf;
+	unsigned int func;
+
+	if (nvdimm && cmd == ND_CMD_CALL &&
+			call_pkg->nd_family == NVDIMM_FAMILY_INTEL) {
+		func = call_pkg->nd_command;
+		if ((1 << func) & NVDIMM_INTEL_SECURITY_CMDMASK)
+			return -EOPNOTSUPP;
+	}
+
+	return __acpi_nfit_clear_to_send(nd_desc, nvdimm, cmd);
 }
 
 int acpi_nfit_ars_rescan(struct acpi_nfit_desc *acpi_desc,
···
 
 	status = acpi_get_table(ACPI_SIG_NFIT, 0, &tbl);
 	if (ACPI_FAILURE(status)) {
-		/* This is ok, we could have an nvdimm hotplugged later */
+		/* The NVDIMM root device allows OS to trigger enumeration of
+		 * NVDIMMs through NFIT at boot time and re-enumeration at
+		 * root level via the _FIT method during runtime.
+		 * This is ok to return 0 here, we could have an nvdimm
+		 * hotplugged later and evaluate _FIT method which returns
+		 * data in the format of a series of NFIT Structures.
+		 */
 		dev_dbg(dev, "failed to find NFIT at startup\n");
 		return 0;
 	}

drivers/acpi/nfit/intel.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright(c) 2018 Intel Corporation. All rights reserved. */
+#include <linux/libnvdimm.h>
+#include <linux/ndctl.h>
+#include <linux/acpi.h>
+#include <asm/smp.h>
+#include "intel.h"
+#include "nfit.h"
+
+static enum nvdimm_security_state intel_security_state(struct nvdimm *nvdimm,
+		enum nvdimm_passphrase_type ptype)
+{
+	struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm);
+	struct {
+		struct nd_cmd_pkg pkg;
+		struct nd_intel_get_security_state cmd;
+	} nd_cmd = {
+		.pkg = {
+			.nd_command = NVDIMM_INTEL_GET_SECURITY_STATE,
+			.nd_family = NVDIMM_FAMILY_INTEL,
+			.nd_size_out =
+				sizeof(struct nd_intel_get_security_state),
+			.nd_fw_size =
+				sizeof(struct nd_intel_get_security_state),
+		},
+	};
+	int rc;
+
+	if (!test_bit(NVDIMM_INTEL_GET_SECURITY_STATE, &nfit_mem->dsm_mask))
+		return -ENXIO;
+
+	/*
+	 * Short circuit the state retrieval while we are doing overwrite.
+	 * The DSM spec states that the security state is indeterminate
+	 * until the overwrite DSM completes.
+	 */
+	if (nvdimm_in_overwrite(nvdimm) && ptype == NVDIMM_USER)
+		return NVDIMM_SECURITY_OVERWRITE;
+
+	rc = nvdimm_ctl(nvdimm, ND_CMD_CALL, &nd_cmd, sizeof(nd_cmd), NULL);
+	if (rc < 0)
+		return rc;
+	if (nd_cmd.cmd.status)
+		return -EIO;
+
+	/* check and see if security is enabled and locked */
+	if (ptype == NVDIMM_MASTER) {
+		if (nd_cmd.cmd.extended_state & ND_INTEL_SEC_ESTATE_ENABLED)
+			return NVDIMM_SECURITY_UNLOCKED;
+		else if (nd_cmd.cmd.extended_state &
+				ND_INTEL_SEC_ESTATE_PLIMIT)
+			return NVDIMM_SECURITY_FROZEN;
+	} else {
+		if (nd_cmd.cmd.state & ND_INTEL_SEC_STATE_UNSUPPORTED)
+			return -ENXIO;
+		else if (nd_cmd.cmd.state & ND_INTEL_SEC_STATE_ENABLED) {
+			if (nd_cmd.cmd.state & ND_INTEL_SEC_STATE_LOCKED)
+				return NVDIMM_SECURITY_LOCKED;
+			else if (nd_cmd.cmd.state & ND_INTEL_SEC_STATE_FROZEN
+					|| nd_cmd.cmd.state &
+					ND_INTEL_SEC_STATE_PLIMIT)
+				return NVDIMM_SECURITY_FROZEN;
+			else
+				return NVDIMM_SECURITY_UNLOCKED;
+		}
+	}
+
+	/* this should cover master security disabled as well */
+	return NVDIMM_SECURITY_DISABLED;
+}
+
+static int intel_security_freeze(struct nvdimm *nvdimm)
+{
+	struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm);
+	struct {
+		struct nd_cmd_pkg pkg;
+		struct nd_intel_freeze_lock cmd;
+	} nd_cmd = {
+		.pkg = {
+			.nd_command = NVDIMM_INTEL_FREEZE_LOCK,
+			.nd_family = NVDIMM_FAMILY_INTEL,
+			.nd_size_out = ND_INTEL_STATUS_SIZE,
+			.nd_fw_size = ND_INTEL_STATUS_SIZE,
+		},
+	};
+	int rc;
+
+	if (!test_bit(NVDIMM_INTEL_FREEZE_LOCK, &nfit_mem->dsm_mask))
+		return -ENOTTY;
+
+	rc = nvdimm_ctl(nvdimm, ND_CMD_CALL, &nd_cmd, sizeof(nd_cmd), NULL);
+	if (rc < 0)
+		return rc;
+	if (nd_cmd.cmd.status)
+		return -EIO;
+	return 0;
+}
+
+static int intel_security_change_key(struct nvdimm *nvdimm,
+		const struct nvdimm_key_data *old_data,
+		const struct nvdimm_key_data *new_data,
+		enum nvdimm_passphrase_type ptype)
+{
+	struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm);
+	unsigned int cmd = ptype == NVDIMM_MASTER ?
+		NVDIMM_INTEL_SET_MASTER_PASSPHRASE :
+		NVDIMM_INTEL_SET_PASSPHRASE;
+	struct {
+		struct nd_cmd_pkg pkg;
+		struct nd_intel_set_passphrase cmd;
+	} nd_cmd = {
+		.pkg = {
+			.nd_family = NVDIMM_FAMILY_INTEL,
+			.nd_size_in = ND_INTEL_PASSPHRASE_SIZE * 2,
+			.nd_size_out = ND_INTEL_STATUS_SIZE,
+			.nd_fw_size = ND_INTEL_STATUS_SIZE,
+			.nd_command = cmd,
+		},
+	};
+	int rc;
+
+	if (!test_bit(cmd, &nfit_mem->dsm_mask))
+		return -ENOTTY;
+
+	if (old_data)
+		memcpy(nd_cmd.cmd.old_pass, old_data->data,
+				sizeof(nd_cmd.cmd.old_pass));
+	memcpy(nd_cmd.cmd.new_pass, new_data->data,
+			sizeof(nd_cmd.cmd.new_pass));
+	rc = nvdimm_ctl(nvdimm, ND_CMD_CALL, &nd_cmd, sizeof(nd_cmd), NULL);
+	if (rc < 0)
+		return rc;
+
+	switch (nd_cmd.cmd.status) {
+	case 0:
+		return 0;
+	case ND_INTEL_STATUS_INVALID_PASS:
+		return -EINVAL;
+	case ND_INTEL_STATUS_NOT_SUPPORTED:
+		return -EOPNOTSUPP;
+	case ND_INTEL_STATUS_INVALID_STATE:
+	default:
+		return -EIO;
+	}
+}
+
+static void nvdimm_invalidate_cache(void);
+
+static int intel_security_unlock(struct nvdimm *nvdimm,
+		const struct nvdimm_key_data *key_data)
+{
+	struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm);
+	struct {
+		struct nd_cmd_pkg pkg;
+		struct nd_intel_unlock_unit cmd;
+	} nd_cmd = {
+		.pkg = {
+			.nd_command = NVDIMM_INTEL_UNLOCK_UNIT,
+			.nd_family = NVDIMM_FAMILY_INTEL,
+			.nd_size_in = ND_INTEL_PASSPHRASE_SIZE,
+			.nd_size_out = ND_INTEL_STATUS_SIZE,
+			.nd_fw_size = ND_INTEL_STATUS_SIZE,
+		},
+	};
+	int rc;
+
+	if (!test_bit(NVDIMM_INTEL_UNLOCK_UNIT, &nfit_mem->dsm_mask))
+		return -ENOTTY;
+
+	memcpy(nd_cmd.cmd.passphrase, key_data->data,
+			sizeof(nd_cmd.cmd.passphrase));
+	rc = nvdimm_ctl(nvdimm, ND_CMD_CALL, &nd_cmd, sizeof(nd_cmd), NULL);
+	if (rc < 0)
+		return rc;
+	switch (nd_cmd.cmd.status) {
+	case 0:
+		break;
+	case ND_INTEL_STATUS_INVALID_PASS:
+		return -EINVAL;
+	default:
+		return -EIO;
+	}
+
+	/* DIMM unlocked, invalidate all CPU caches before we read it */
+	nvdimm_invalidate_cache();
+
+	return 0;
+}
+
+static int intel_security_disable(struct nvdimm *nvdimm,
+		const struct nvdimm_key_data *key_data)
+{
+	int rc;
+	struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm);
+	struct {
+		struct nd_cmd_pkg pkg;
+		struct nd_intel_disable_passphrase cmd;
+	} nd_cmd = {
+		.pkg = {
+			.nd_command = NVDIMM_INTEL_DISABLE_PASSPHRASE,
+			.nd_family = NVDIMM_FAMILY_INTEL,
+			.nd_size_in = ND_INTEL_PASSPHRASE_SIZE,
+			.nd_size_out = ND_INTEL_STATUS_SIZE,
+			.nd_fw_size = ND_INTEL_STATUS_SIZE,
+		},
+	};
+
+	if (!test_bit(NVDIMM_INTEL_DISABLE_PASSPHRASE, &nfit_mem->dsm_mask))
+		return -ENOTTY;
+
+	memcpy(nd_cmd.cmd.passphrase, key_data->data,
+			sizeof(nd_cmd.cmd.passphrase));
+	rc = nvdimm_ctl(nvdimm, ND_CMD_CALL, &nd_cmd, sizeof(nd_cmd), NULL);
+	if (rc < 0)
+		return rc;
+
+	switch (nd_cmd.cmd.status) {
+	case 0:
+		break;
+	case ND_INTEL_STATUS_INVALID_PASS:
+		return -EINVAL;
+	case ND_INTEL_STATUS_INVALID_STATE:
+	default:
+		return -ENXIO;
+	}
+
+	return 0;
+}
+
+static int intel_security_erase(struct nvdimm *nvdimm,
+		const struct nvdimm_key_data *key,
+		enum nvdimm_passphrase_type ptype)
+{
+	int rc;
+	struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm);
+	unsigned int cmd = ptype == NVDIMM_MASTER ?
+		NVDIMM_INTEL_MASTER_SECURE_ERASE : NVDIMM_INTEL_SECURE_ERASE;
+	struct {
+		struct nd_cmd_pkg pkg;
+		struct nd_intel_secure_erase cmd;
+	} nd_cmd = {
+		.pkg = {
+			.nd_family = NVDIMM_FAMILY_INTEL,
+			.nd_size_in = ND_INTEL_PASSPHRASE_SIZE,
+			.nd_size_out = ND_INTEL_STATUS_SIZE,
+			.nd_fw_size = ND_INTEL_STATUS_SIZE,
+			.nd_command = cmd,
+		},
+	};
+
+	if (!test_bit(cmd, &nfit_mem->dsm_mask))
+		return -ENOTTY;
+
+	/* flush all cache before we erase DIMM */
+	nvdimm_invalidate_cache();
+	memcpy(nd_cmd.cmd.passphrase, key->data,
+			sizeof(nd_cmd.cmd.passphrase));
+	rc = nvdimm_ctl(nvdimm, ND_CMD_CALL, &nd_cmd, sizeof(nd_cmd), NULL);
+	if (rc < 0)
+		return rc;
+
+	switch (nd_cmd.cmd.status) {
+	case 0:
+		break;
+	case ND_INTEL_STATUS_NOT_SUPPORTED:
+		return -EOPNOTSUPP;
+	case ND_INTEL_STATUS_INVALID_PASS:
+		return -EINVAL;
+	case ND_INTEL_STATUS_INVALID_STATE:
+	default:
+		return -ENXIO;
+	}
+
+	/* DIMM erased, invalidate all CPU caches before we read it */
+	nvdimm_invalidate_cache();
+	return 0;
+}
+
+static int intel_security_query_overwrite(struct nvdimm *nvdimm)
+{
+	int rc;
+	struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm);
+	struct {
+		struct nd_cmd_pkg pkg;
+		struct nd_intel_query_overwrite cmd;
+	} nd_cmd = {
+		.pkg = {
+			.nd_command = NVDIMM_INTEL_QUERY_OVERWRITE,
+			.nd_family = NVDIMM_FAMILY_INTEL,
+			.nd_size_out = ND_INTEL_STATUS_SIZE,
+			.nd_fw_size = ND_INTEL_STATUS_SIZE,
+		},
+	};
+
+	if (!test_bit(NVDIMM_INTEL_QUERY_OVERWRITE, &nfit_mem->dsm_mask))
+		return -ENOTTY;
+
+	rc = nvdimm_ctl(nvdimm, ND_CMD_CALL, &nd_cmd, sizeof(nd_cmd), NULL);
+	if (rc < 0)
+		return rc;
+
+	switch (nd_cmd.cmd.status) {
+	case 0:
+		break;
+	case ND_INTEL_STATUS_OQUERY_INPROGRESS:
+		return -EBUSY;
+	default:
+		return -ENXIO;
+	}
+
+	/* flush all cache before we make the nvdimms available */
+	nvdimm_invalidate_cache();
+	return 0;
+}
+
+static int intel_security_overwrite(struct nvdimm *nvdimm,
+		const struct nvdimm_key_data *nkey)
+{
+	int rc;
+	struct nfit_mem *nfit_mem = nvdimm_provider_data(nvdimm);
+	struct {
+		struct nd_cmd_pkg pkg;
+		struct nd_intel_overwrite cmd;
+	} nd_cmd = {
+		.pkg = {
+			.nd_command = NVDIMM_INTEL_OVERWRITE,
+			.nd_family = NVDIMM_FAMILY_INTEL,
+			.nd_size_in = ND_INTEL_PASSPHRASE_SIZE,
+			.nd_size_out = ND_INTEL_STATUS_SIZE,
+			.nd_fw_size = ND_INTEL_STATUS_SIZE,
+		},
+	};
+
+	if (!test_bit(NVDIMM_INTEL_OVERWRITE, &nfit_mem->dsm_mask))
+		return -ENOTTY;
+
+	/* flush all cache before we erase DIMM */
+	nvdimm_invalidate_cache();
+	if (nkey)
+		memcpy(nd_cmd.cmd.passphrase, nkey->data,
+				sizeof(nd_cmd.cmd.passphrase));
+	rc = nvdimm_ctl(nvdimm, ND_CMD_CALL, &nd_cmd, sizeof(nd_cmd), NULL);
+	if (rc < 0)
+		return rc;
+
+	switch (nd_cmd.cmd.status) {
+	case 0:
+		return 0;
+	case ND_INTEL_STATUS_OVERWRITE_UNSUPPORTED:
+		return -ENOTSUPP;
+	case ND_INTEL_STATUS_INVALID_PASS:
+		return -EINVAL;
+	case ND_INTEL_STATUS_INVALID_STATE:
+	default:
+		return -ENXIO;
+	}
+}
+
+/*
+ * TODO: define a cross arch wbinvd equivalent when/if
+ * NVDIMM_FAMILY_INTEL command support arrives on another arch.
+ */
+#ifdef CONFIG_X86
+static void nvdimm_invalidate_cache(void)
+{
+	wbinvd_on_all_cpus();
+}
+#else
+static void nvdimm_invalidate_cache(void)
+{
+	WARN_ON_ONCE("cache invalidation required after unlock\n");
+}
+#endif
+
+static const struct nvdimm_security_ops __intel_security_ops = {
+	.state = intel_security_state,
+	.freeze = intel_security_freeze,
+	.change_key = intel_security_change_key,
+	.disable = intel_security_disable,
+#ifdef CONFIG_X86
+	.unlock = intel_security_unlock,
+	.erase = intel_security_erase,
+	.overwrite = intel_security_overwrite,
+	.query_overwrite = intel_security_query_overwrite,
+#endif
+};
+
+const struct nvdimm_security_ops *intel_security_ops = &__intel_security_ops;

drivers/acpi/nfit/intel.h

···
 	};
 } __packed;
 
+extern const struct nvdimm_security_ops *intel_security_ops;
+
+#define ND_INTEL_STATUS_SIZE		4
+#define ND_INTEL_PASSPHRASE_SIZE	32
+
+#define ND_INTEL_STATUS_NOT_SUPPORTED	1
+#define ND_INTEL_STATUS_RETRY		5
+#define ND_INTEL_STATUS_NOT_READY	9
+#define ND_INTEL_STATUS_INVALID_STATE	10
+#define ND_INTEL_STATUS_INVALID_PASS	11
+#define ND_INTEL_STATUS_OVERWRITE_UNSUPPORTED	0x10007
+#define ND_INTEL_STATUS_OQUERY_INPROGRESS	0x10007
+#define ND_INTEL_STATUS_OQUERY_SEQUENCE_ERR	0x20007
+
+#define ND_INTEL_SEC_STATE_ENABLED	0x02
+#define ND_INTEL_SEC_STATE_LOCKED	0x04
+#define ND_INTEL_SEC_STATE_FROZEN	0x08
+#define ND_INTEL_SEC_STATE_PLIMIT	0x10
+#define ND_INTEL_SEC_STATE_UNSUPPORTED	0x20
+#define ND_INTEL_SEC_STATE_OVERWRITE	0x40
+
+#define ND_INTEL_SEC_ESTATE_ENABLED	0x01
+#define ND_INTEL_SEC_ESTATE_PLIMIT	0x02
+
+struct nd_intel_get_security_state {
+	u32 status;
+	u8 extended_state;
+	u8 reserved[3];
+	u8 state;
+	u8 reserved1[3];
+} __packed;
+
+struct nd_intel_set_passphrase {
+	u8 old_pass[ND_INTEL_PASSPHRASE_SIZE];
+	u8 new_pass[ND_INTEL_PASSPHRASE_SIZE];
+	u32 status;
+} __packed;
+
+struct nd_intel_unlock_unit {
+	u8 passphrase[ND_INTEL_PASSPHRASE_SIZE];
+	u32 status;
+} __packed;
+
+struct nd_intel_disable_passphrase {
+	u8 passphrase[ND_INTEL_PASSPHRASE_SIZE];
+	u32 status;
+} __packed;
+
+struct nd_intel_freeze_lock {
+	u32 status;
+} __packed;
+
+struct nd_intel_secure_erase {
+	u8 passphrase[ND_INTEL_PASSPHRASE_SIZE];
+	u32 status;
+} __packed;
+
+struct nd_intel_overwrite {
+	u8 passphrase[ND_INTEL_PASSPHRASE_SIZE];
+	u32 status;
+} __packed;
+
+struct nd_intel_query_overwrite {
+	u32 status;
+} __packed;
+
+struct nd_intel_set_master_passphrase {
+	u8 old_pass[ND_INTEL_PASSPHRASE_SIZE];
+	u8 new_pass[ND_INTEL_PASSPHRASE_SIZE];
+	u32 status;
+} __packed;
+
+struct nd_intel_master_secure_erase {
+	u8 passphrase[ND_INTEL_PASSPHRASE_SIZE];
+	u32 status;
+} __packed;
 #endif

drivers/acpi/nfit/nfit.h

···
 	NVDIMM_INTEL_QUERY_FWUPDATE = 16,
 	NVDIMM_INTEL_SET_THRESHOLD = 17,
 	NVDIMM_INTEL_INJECT_ERROR = 18,
+	NVDIMM_INTEL_GET_SECURITY_STATE = 19,
+	NVDIMM_INTEL_SET_PASSPHRASE = 20,
+	NVDIMM_INTEL_DISABLE_PASSPHRASE = 21,
+	NVDIMM_INTEL_UNLOCK_UNIT = 22,
+	NVDIMM_INTEL_FREEZE_LOCK = 23,
+	NVDIMM_INTEL_SECURE_ERASE = 24,
+	NVDIMM_INTEL_OVERWRITE = 25,
+	NVDIMM_INTEL_QUERY_OVERWRITE = 26,
+	NVDIMM_INTEL_SET_MASTER_PASSPHRASE = 27,
+	NVDIMM_INTEL_MASTER_SECURE_ERASE = 28,
 };
+
+#define NVDIMM_INTEL_SECURITY_CMDMASK \
+(1 << NVDIMM_INTEL_GET_SECURITY_STATE | 1 << NVDIMM_INTEL_SET_PASSPHRASE \
+| 1 << NVDIMM_INTEL_DISABLE_PASSPHRASE | 1 << NVDIMM_INTEL_UNLOCK_UNIT \
+| 1 << NVDIMM_INTEL_FREEZE_LOCK | 1 << NVDIMM_INTEL_SECURE_ERASE \
+| 1 << NVDIMM_INTEL_OVERWRITE | 1 << NVDIMM_INTEL_QUERY_OVERWRITE \
+| 1 << NVDIMM_INTEL_SET_MASTER_PASSPHRASE \
+| 1 << NVDIMM_INTEL_MASTER_SECURE_ERASE)
 
 #define NVDIMM_INTEL_CMDMASK \
 (NVDIMM_STANDARD_CMDMASK | 1 << NVDIMM_INTEL_GET_MODES \
  | 1 << NVDIMM_INTEL_GET_FWINFO | 1 << NVDIMM_INTEL_START_FWUPDATE \
  | 1 << NVDIMM_INTEL_SEND_FWUPDATE | 1 << NVDIMM_INTEL_FINISH_FWUPDATE \
  | 1 << NVDIMM_INTEL_QUERY_FWUPDATE | 1 << NVDIMM_INTEL_SET_THRESHOLD \
- | 1 << NVDIMM_INTEL_INJECT_ERROR | 1 << NVDIMM_INTEL_LATCH_SHUTDOWN)
+ | 1 << NVDIMM_INTEL_INJECT_ERROR | 1 << NVDIMM_INTEL_LATCH_SHUTDOWN \
+ | NVDIMM_INTEL_SECURITY_CMDMASK)
 
 enum nfit_uuids {
 	/* for simplicity alias the uuid index with the family id */
···
 	NFIT_MEM_DIRTY_COUNT,
 };
 
+#define NFIT_DIMM_ID_LEN	22
+
 /* assembled tables for a given dimm/memory-device */
 struct nfit_mem {
 	struct nvdimm *nvdimm;
···
 	struct list_head list;
 	struct acpi_device *adev;
 	struct acpi_nfit_desc *acpi_desc;
+	char id[NFIT_DIMM_ID_LEN+1];
 	struct resource *flush_wpq;
 	unsigned long dsm_mask;
 	unsigned long flags;

drivers/nvdimm/Kconfig

···
 
 	  Select Y if unsure.
 
+config NVDIMM_KEYS
+	def_bool y
+	depends on ENCRYPTED_KEYS
+	depends on (LIBNVDIMM=ENCRYPTED_KEYS) || LIBNVDIMM=m
+
 endif

drivers/nvdimm/Makefile

···
 libnvdimm-$(CONFIG_BTT) += btt_devs.o
 libnvdimm-$(CONFIG_NVDIMM_PFN) += pfn_devs.o
 libnvdimm-$(CONFIG_NVDIMM_DAX) += dax_devs.o
+libnvdimm-$(CONFIG_NVDIMM_KEYS) += security.o

drivers/nvdimm/bus.c

···
 }
 EXPORT_SYMBOL_GPL(to_nvdimm_bus);
 
+struct nvdimm_bus *nvdimm_to_bus(struct nvdimm *nvdimm)
+{
+	return to_nvdimm_bus(nvdimm->dev.parent);
+}
+EXPORT_SYMBOL_GPL(nvdimm_to_bus);
+
 struct nvdimm_bus *nvdimm_bus_register(struct device *parent,
 		struct nvdimm_bus_descriptor *nd_desc)
 {
···
 	INIT_LIST_HEAD(&nvdimm_bus->mapping_list);
 	init_waitqueue_head(&nvdimm_bus->probe_wait);
 	nvdimm_bus->id = ida_simple_get(&nd_ida, 0, 0, GFP_KERNEL);
-	mutex_init(&nvdimm_bus->reconfig_mutex);
-	badrange_init(&nvdimm_bus->badrange);
 	if (nvdimm_bus->id < 0) {
 		kfree(nvdimm_bus);
 		return NULL;
 	}
+	mutex_init(&nvdimm_bus->reconfig_mutex);
+	badrange_init(&nvdimm_bus->badrange);
 	nvdimm_bus->nd_desc = nd_desc;
 	nvdimm_bus->dev.parent = parent;
 	nvdimm_bus->dev.release = nvdimm_bus_release;
···
 	 * i.e. remove classless children
 	 */
 	if (dev->class)
-		/* pass */;
-	else
-		nd_device_unregister(dev, ND_SYNC);
+		return 0;
+
+	if (is_nvdimm(dev)) {
+		struct nvdimm *nvdimm = to_nvdimm(dev);
+		bool dev_put = false;
+
+		/* We are shutting down. Make state frozen artificially. */
+		nvdimm_bus_lock(dev);
+		nvdimm->sec.state = NVDIMM_SECURITY_FROZEN;
+		if (test_and_clear_bit(NDD_WORK_PENDING, &nvdimm->flags))
+			dev_put = true;
+		nvdimm_bus_unlock(dev);
+		cancel_delayed_work_sync(&nvdimm->dwork);
+		if (dev_put)
+			put_device(dev);
+	}
+	nd_device_unregister(dev, ND_SYNC);
+
 	return 0;
 }
 
···
 
 	/* ask the bus provider if it would like to block this request */
 	if (nd_desc->clear_to_send) {
-		int rc = nd_desc->clear_to_send(nd_desc, nvdimm, cmd);
+		int rc = nd_desc->clear_to_send(nd_desc, nvdimm, cmd, data);
 
 		if (rc)
 			return rc;

drivers/nvdimm/dimm.c

···
 		return rc;
 	}
 
-	/* reset locked, to be validated below... */
+	/*
+	 * The locked status bit reflects explicit status codes from the
+	 * label reading commands, revalidate it each time the driver is
+	 * activated and re-reads the label area.
+	 */
 	nvdimm_clear_locked(dev);
 
 	ndd = kzalloc(sizeof(*ndd), GFP_KERNEL);
···
 	ndd->dev = dev;
 	get_device(dev);
 	kref_init(&ndd->kref);
+
+	/*
+	 * Attempt to unlock, if the DIMM supports security commands,
+	 * otherwise the locked indication is determined by explicit
+	 * status codes from the label reading commands.
+	 */
+	rc = nvdimm_security_unlock(dev);
+	if (rc < 0)
+		dev_dbg(dev, "failed to unlock dimm: %d\n", rc);
+
 
 	/*
 	 * EACCES failures reading the namespace label-area-properties

drivers/nvdimm/dimm_devs.c

···
 }
 static DEVICE_ATTR_RO(available_slots);
 
+__weak ssize_t security_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct nvdimm *nvdimm = to_nvdimm(dev);
+
+	switch (nvdimm->sec.state) {
+	case NVDIMM_SECURITY_DISABLED:
+		return sprintf(buf, "disabled\n");
+	case NVDIMM_SECURITY_UNLOCKED:
+		return sprintf(buf, "unlocked\n");
+	case NVDIMM_SECURITY_LOCKED:
+		return sprintf(buf, "locked\n");
+	case NVDIMM_SECURITY_FROZEN:
+		return sprintf(buf, "frozen\n");
+	case NVDIMM_SECURITY_OVERWRITE:
+		return sprintf(buf, "overwrite\n");
+	default:
+		return -ENOTTY;
+	}
+
+	return -ENOTTY;
+}
+
+#define OPS							\
+	C( OP_FREEZE,		"freeze",		1),	\
+	C( OP_DISABLE,		"disable",		2),	\
+	C( OP_UPDATE,		"update",		3),	\
+	C( OP_ERASE,		"erase",		2),	\
+	C( OP_OVERWRITE,	"overwrite",		2),	\
+	C( OP_MASTER_UPDATE,	"master_update",	3),	\
+	C( OP_MASTER_ERASE,	"master_erase",		2)
+#undef C
+#define C(a, b, c) a
+enum nvdimmsec_op_ids { OPS };
+#undef C
+#define C(a, b, c) { b, c }
+static struct {
+	const char *name;
+	int args;
+} ops[] = { OPS };
+#undef C
+
+#define SEC_CMD_SIZE 32
+#define KEY_ID_SIZE 10
+
+static ssize_t __security_store(struct device *dev, const char *buf, size_t len)
+{
+	struct nvdimm *nvdimm = to_nvdimm(dev);
+	ssize_t rc;
+	char cmd[SEC_CMD_SIZE+1], keystr[KEY_ID_SIZE+1],
+		nkeystr[KEY_ID_SIZE+1];
+	unsigned int key, newkey;
+	int i;
+
+	if (atomic_read(&nvdimm->busy))
+		return -EBUSY;
+
+	rc = sscanf(buf, "%"__stringify(SEC_CMD_SIZE)"s"
+			" %"__stringify(KEY_ID_SIZE)"s"
+			" %"__stringify(KEY_ID_SIZE)"s",
+			cmd, keystr, nkeystr);
+	if (rc < 1)
+		return -EINVAL;
+	for (i = 0; i < ARRAY_SIZE(ops); i++)
+		if (sysfs_streq(cmd, ops[i].name))
+			break;
+	if (i >= ARRAY_SIZE(ops))
+		return -EINVAL;
+	if (ops[i].args > 1)
+		rc = kstrtouint(keystr, 0, &key);
+	if (rc >= 0 && ops[i].args > 2)
+		rc = kstrtouint(nkeystr, 0, &newkey);
+	if (rc < 0)
+		return rc;
+
+	if (i == OP_FREEZE) {
+		dev_dbg(dev, "freeze\n");
+		rc = nvdimm_security_freeze(nvdimm);
+	} else if (i == OP_DISABLE) {
+		dev_dbg(dev, "disable %u\n", key);
+		rc = nvdimm_security_disable(nvdimm, key);
+	} else if (i == OP_UPDATE) {
+		dev_dbg(dev, "update %u %u\n", key, newkey);
+		rc = nvdimm_security_update(nvdimm, key, newkey, NVDIMM_USER);
+	} else if (i == OP_ERASE) {
+		dev_dbg(dev, "erase %u\n", key);
+		rc = nvdimm_security_erase(nvdimm, key, NVDIMM_USER);
+	} else if (i == OP_OVERWRITE) {
+		dev_dbg(dev, "overwrite %u\n", key);
+		rc = nvdimm_security_overwrite(nvdimm, key);
+	} else if (i == OP_MASTER_UPDATE) {
+		dev_dbg(dev, "master_update %u %u\n", key, newkey);
+		rc = nvdimm_security_update(nvdimm, key, newkey,
+				NVDIMM_MASTER);
+	} else if (i == OP_MASTER_ERASE) {
+		dev_dbg(dev, "master_erase %u\n", key);
+		rc = nvdimm_security_erase(nvdimm, key,
+				NVDIMM_MASTER);
+	} else
+		return -EINVAL;
+
+	if (rc == 0)
+		rc = len;
+	return rc;
+}
+
+static ssize_t security_store(struct device *dev,
+		struct device_attribute *attr, const char *buf, size_t len)
+
+{
+	ssize_t rc;
+
+	/*
+	 * Require all userspace triggered security management to be
+	 * done while probing is idle and the DIMM is not in active use
+	 * in any region.
+	 */
+	device_lock(dev);
+	nvdimm_bus_lock(dev);
+	wait_nvdimm_bus_probe_idle(dev);
+	rc = __security_store(dev, buf, len);
+	nvdimm_bus_unlock(dev);
+	device_unlock(dev);
+
+	return rc;
+}
+static DEVICE_ATTR_RW(security);
+
 static struct attribute *nvdimm_attributes[] = {
 	&dev_attr_state.attr,
 	&dev_attr_flags.attr,
 	&dev_attr_commands.attr,
 	&dev_attr_available_slots.attr,
+	&dev_attr_security.attr,
 	NULL,
 };
 
+static umode_t nvdimm_visible(struct kobject *kobj, struct attribute *a, int n)
+{
+	struct device *dev = container_of(kobj, typeof(*dev), kobj);
+	struct nvdimm *nvdimm = to_nvdimm(dev);
+
+	if (a != &dev_attr_security.attr)
+		return a->mode;
+	if (nvdimm->sec.state < 0)
+		return 0;
+	/* Are there any state mutation ops? */
+	if (nvdimm->sec.ops->freeze || nvdimm->sec.ops->disable
+			|| nvdimm->sec.ops->change_key
+			|| nvdimm->sec.ops->erase
+			|| nvdimm->sec.ops->overwrite)
+		return a->mode;
+	return 0444;
+}
+
 struct attribute_group nvdimm_attribute_group = {
 	.attrs = nvdimm_attributes,
+	.is_visible = nvdimm_visible,
 };
 EXPORT_SYMBOL_GPL(nvdimm_attribute_group);
 
-struct nvdimm *nvdimm_create(struct nvdimm_bus *nvdimm_bus, void *provider_data,
-		const struct attribute_group **groups, unsigned long flags,
-		unsigned long cmd_mask, int num_flush,
-		struct resource *flush_wpq)
+struct nvdimm *__nvdimm_create(struct nvdimm_bus *nvdimm_bus,
+		void *provider_data, const struct attribute_group **groups,
+		unsigned long flags, unsigned long cmd_mask, int num_flush,
+		struct resource *flush_wpq, const char *dimm_id,
+		const struct nvdimm_security_ops *sec_ops)
 {
 	struct nvdimm *nvdimm = kzalloc(sizeof(*nvdimm), GFP_KERNEL);
 	struct device *dev;
···
 		kfree(nvdimm);
 		return NULL;
 	}
+
+	nvdimm->dimm_id = dimm_id;
 	nvdimm->provider_data = provider_data;
 	nvdimm->flags = flags;
 	nvdimm->cmd_mask = cmd_mask;
···
 	dev->type = &nvdimm_device_type;
 	dev->devt = MKDEV(nvdimm_major, nvdimm->id);
 	dev->groups = groups;
+	nvdimm->sec.ops = sec_ops;
+	nvdimm->sec.overwrite_tmo = 0;
+	INIT_DELAYED_WORK(&nvdimm->dwork, nvdimm_security_overwrite_query);
+	/*
+	 * Security state must be initialized before device_add() for
+	 * attribute visibility.
+	 */
+	/* get security state and extended (master) state */
+	nvdimm->sec.state = nvdimm_security_state(nvdimm, NVDIMM_USER);
+	nvdimm->sec.ext_state = nvdimm_security_state(nvdimm, NVDIMM_MASTER);
 	nd_device_register(dev);
 
 	return nvdimm;
 }
-EXPORT_SYMBOL_GPL(nvdimm_create);
+EXPORT_SYMBOL_GPL(__nvdimm_create);
+
+int nvdimm_security_setup_events(struct nvdimm *nvdimm)
+{
+	nvdimm->sec.overwrite_state = sysfs_get_dirent(nvdimm->dev.kobj.sd,
+			"security");
+	if (!nvdimm->sec.overwrite_state)
+		return -ENODEV;
+	return 0;
+}
+EXPORT_SYMBOL_GPL(nvdimm_security_setup_events);
+
+int nvdimm_in_overwrite(struct nvdimm *nvdimm)
+{
+	return test_bit(NDD_SECURITY_OVERWRITE, &nvdimm->flags);
+}
+EXPORT_SYMBOL_GPL(nvdimm_in_overwrite);
+
+int nvdimm_security_freeze(struct nvdimm *nvdimm)
+{
+	int rc;
+
+	WARN_ON_ONCE(!is_nvdimm_bus_locked(&nvdimm->dev));
+
+	if (!nvdimm->sec.ops || !nvdimm->sec.ops->freeze)
+		return -EOPNOTSUPP;
+
+	if (nvdimm->sec.state < 0)
+		return -EIO;
+
+	if (test_bit(NDD_SECURITY_OVERWRITE, &nvdimm->flags)) {
+		dev_warn(&nvdimm->dev, "Overwrite operation in progress.\n");
+		return -EBUSY;
+	}
+
+	rc = nvdimm->sec.ops->freeze(nvdimm);
+	nvdimm->sec.state = nvdimm_security_state(nvdimm, NVDIMM_USER);
+
+	return rc;
+}
 
 int alias_dpa_busy(struct device *dev, void *data)
 {

drivers/nvdimm/label.c

···
 	victims = 0;
 	if (old_num_resources) {
 		/* convert old local-label-map to dimm-slot victim-map */
-		victim_map = kcalloc(BITS_TO_LONGS(nslot), sizeof(long),
-				GFP_KERNEL);
+		victim_map = bitmap_zalloc(nslot, GFP_KERNEL);
 		if (!victim_map)
 			return -ENOMEM;
 
···
 	/* don't allow updates that consume the last label */
 	if (nfree - alloc < 0 || nfree - alloc + victims < 1) {
 		dev_info(&nsblk->common.dev, "insufficient label space\n");
-		kfree(victim_map);
+		bitmap_free(victim_map);
 		return -ENOSPC;
 	}
 	/* from here on we need to abort on error */
···
 
  out:
 	kfree(old_res_list);
-	kfree(victim_map);
+	bitmap_free(victim_map);
 	return rc;
 
  abort:

drivers/nvdimm/namespace_devs.c

···
 	if (dev->driver || to_ndns(dev)->claim)
 		return -EBUSY;
 
-	input = kmemdup(buf, len + 1, GFP_KERNEL);
+	input = kstrndup(buf, len, GFP_KERNEL);
 	if (!input)
 		return -ENOMEM;
 
-	input[len] = '\0';
 	pos = strim(input);
 	if (strlen(pos) + 1 > NSLABEL_NAME_LEN) {
 		rc = -EINVAL;

drivers/nvdimm/nd-core.h

···
 extern struct list_head nvdimm_bus_list;
 extern struct mutex nvdimm_bus_list_mutex;
 extern int nvdimm_major;
+extern struct workqueue_struct *nvdimm_wq;
 
 struct nvdimm_bus {
 	struct nvdimm_bus_descriptor *nd_desc;
···
 	atomic_t busy;
 	int id, num_flush;
 	struct resource *flush_wpq;
+	const char *dimm_id;
+	struct {
+		const struct nvdimm_security_ops *ops;
+		enum nvdimm_security_state state;
+		enum nvdimm_security_state ext_state;
+		unsigned int overwrite_tmo;
+		struct kernfs_node *overwrite_state;
+	} sec;
+	struct delayed_work dwork;
 };
+
+static inline enum nvdimm_security_state nvdimm_security_state(
+		struct nvdimm *nvdimm, bool master)
+{
+	if (!nvdimm->sec.ops)
+		return -ENXIO;
+
+	return nvdimm->sec.ops->state(nvdimm, master);
+}
+int nvdimm_security_freeze(struct nvdimm *nvdimm);
+#if IS_ENABLED(CONFIG_NVDIMM_KEYS)
+int nvdimm_security_disable(struct nvdimm *nvdimm, unsigned int keyid);
+int nvdimm_security_update(struct nvdimm *nvdimm, unsigned int keyid,
+		unsigned int new_keyid,
+		enum nvdimm_passphrase_type pass_type);
+int nvdimm_security_erase(struct nvdimm *nvdimm, unsigned int keyid,
+		enum nvdimm_passphrase_type pass_type);
+int nvdimm_security_overwrite(struct nvdimm *nvdimm, unsigned int keyid);
+void nvdimm_security_overwrite_query(struct work_struct *work);
+#else
+static inline int nvdimm_security_disable(struct nvdimm *nvdimm,
+		unsigned int keyid)
+{
+	return -EOPNOTSUPP;
+}
+static inline int nvdimm_security_update(struct nvdimm *nvdimm,
+		unsigned int keyid,
+		unsigned int new_keyid,
+		enum nvdimm_passphrase_type pass_type)
+{
+	return -EOPNOTSUPP;
+}
+static inline int nvdimm_security_erase(struct nvdimm *nvdimm,
+		unsigned int keyid,
+		enum nvdimm_passphrase_type pass_type)
+{
+	return -EOPNOTSUPP;
+}
+static inline int nvdimm_security_overwrite(struct nvdimm *nvdimm,
+		unsigned int keyid)
+{
+	return -EOPNOTSUPP;
+}
+static inline void nvdimm_security_overwrite_query(struct work_struct *work)
+{
+}
+#endif
 
 /**
  * struct blk_alloc_info - tracking info for BLK dpa scanning

drivers/nvdimm/nd.h

···
 void nvdimm_set_aliasing(struct device *dev);
 void nvdimm_set_locked(struct device *dev);
 void nvdimm_clear_locked(struct device *dev);
+#if IS_ENABLED(CONFIG_NVDIMM_KEYS)
+int nvdimm_security_unlock(struct device *dev);
+#else
+static inline int nvdimm_security_unlock(struct device *dev)
+{
+	return 0;
+}
+#endif
 struct nd_btt *to_nd_btt(struct device *dev);
 
 struct nd_gen_sb {

drivers/nvdimm/region_devs.c

···
 		struct nd_mapping *nd_mapping = &nd_region->mapping[i];
 		struct nvdimm *nvdimm = nd_mapping->nvdimm;
 
+		if (test_bit(NDD_SECURITY_OVERWRITE, &nvdimm->flags)) {
+			nvdimm_bus_unlock(&nd_region->dev);
+			return -EBUSY;
+		}
+
 		/* at least one null hint slot per-dimm for the "no-hint" case */
 		flush_data_size += sizeof(void *);
 		num_flush = min_not_zero(num_flush, nvdimm->num_flush);

drivers/nvdimm/security.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright(c) 2018 Intel Corporation. All rights reserved. */
+
+#include <linux/module.h>
+#include <linux/device.h>
+#include <linux/ndctl.h>
+#include <linux/slab.h>
+#include <linux/io.h>
+#include <linux/mm.h>
+#include <linux/cred.h>
+#include <linux/key.h>
+#include <linux/key-type.h>
+#include <keys/user-type.h>
+#include <keys/encrypted-type.h>
+#include "nd-core.h"
+#include "nd.h"
+
+#define NVDIMM_BASE_KEY		0
+#define NVDIMM_NEW_KEY		1
+
+static bool key_revalidate = true;
+module_param(key_revalidate, bool, 0444);
+MODULE_PARM_DESC(key_revalidate, "Require key validation at init.");
+
+static void *key_data(struct key *key)
+{
+	struct encrypted_key_payload *epayload = dereference_key_locked(key);
+
+	lockdep_assert_held_read(&key->sem);
+
+	return epayload->decrypted_data;
+}
+
+static void nvdimm_put_key(struct key *key)
+{
+	if (!key)
+		return;
+
+	up_read(&key->sem);
+	key_put(key);
+}
+
+/*
+ * Retrieve kernel key for DIMM and request from user space if
+ * necessary. Returns a key held for read and must be put by
+ * nvdimm_put_key() before the usage goes out of scope.
+ */
+static struct key *nvdimm_request_key(struct nvdimm *nvdimm)
+{
+	struct key *key = NULL;
+	static const char NVDIMM_PREFIX[] = "nvdimm:";
+	char desc[NVDIMM_KEY_DESC_LEN + sizeof(NVDIMM_PREFIX)];
+	struct device *dev = &nvdimm->dev;
+
+	sprintf(desc, "%s%s", NVDIMM_PREFIX, nvdimm->dimm_id);
+	key = request_key(&key_type_encrypted, desc, "");
+	if (IS_ERR(key)) {
+		if (PTR_ERR(key) == -ENOKEY)
+			dev_dbg(dev, "request_key() found no key\n");
+		else
+			dev_dbg(dev, "request_key() upcall failed\n");
+		key = NULL;
+	} else {
+		struct encrypted_key_payload *epayload;
+
+		down_read(&key->sem);
+		epayload = dereference_key_locked(key);
+		if (epayload->decrypted_datalen != NVDIMM_PASSPHRASE_LEN) {
+			up_read(&key->sem);
+			key_put(key);
+			key = NULL;
+		}
+	}
+
+	return key;
+}
+
+static struct key *nvdimm_lookup_user_key(struct nvdimm *nvdimm,
+		key_serial_t id, int subclass)
+{
+	key_ref_t keyref;
+	struct key *key;
+	struct encrypted_key_payload *epayload;
+	struct device *dev = &nvdimm->dev;
+
+	keyref = lookup_user_key(id, 0, 0);
+	if (IS_ERR(keyref))
+		return NULL;
+
+	key = key_ref_to_ptr(keyref);
+	if (key->type != &key_type_encrypted) {
+		key_put(key);
+		return NULL;
+	}
+
+	dev_dbg(dev, "%s: key found: %#x\n", __func__, key_serial(key));
+
+	down_read_nested(&key->sem, subclass);
+	epayload = dereference_key_locked(key);
+	if (epayload->decrypted_datalen != NVDIMM_PASSPHRASE_LEN) {
+		up_read(&key->sem);
+		key_put(key);
+		key = NULL;
+	}
+	return key;
+}
+
+static struct key *nvdimm_key_revalidate(struct nvdimm *nvdimm)
+{
+	struct key *key;
+	int rc;
+
+	if (!nvdimm->sec.ops->change_key)
+		return NULL;
+
+	key = nvdimm_request_key(nvdimm);
+	if (!key)
+		return NULL;
+
+	/*
+	 * Send the same key to the hardware as new and old key to
+	 * verify that the key is good.
+	 */
+	rc = nvdimm->sec.ops->change_key(nvdimm, key_data(key),
+			key_data(key), NVDIMM_USER);
+	if (rc < 0) {
+		nvdimm_put_key(key);
+		key = NULL;
+	}
+	return key;
+}
+
+static int __nvdimm_security_unlock(struct nvdimm *nvdimm)
+{
+	struct device *dev = &nvdimm->dev;
+	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
+	struct key *key = NULL;
+	int rc;
+
+	/* The bus lock should be held at the top level of the call stack */
+	lockdep_assert_held(&nvdimm_bus->reconfig_mutex);
+
+	if (!nvdimm->sec.ops || !nvdimm->sec.ops->unlock
+			|| nvdimm->sec.state < 0)
+		return -EIO;
+
+	if (test_bit(NDD_SECURITY_OVERWRITE, &nvdimm->flags)) {
+		dev_dbg(dev, "Security operation in progress.\n");
+		return -EBUSY;
+	}
+
+	/*
+	 * If the pre-OS has unlocked the DIMM, attempt to send the key
+	 * from request_key() to the hardware for verification.  Failure
+	 * to revalidate the key against the hardware results in a
+	 * freeze of the security configuration. I.e. if the OS does not
+	 * have the key, security is being managed pre-OS.
+	 */
+	if (nvdimm->sec.state == NVDIMM_SECURITY_UNLOCKED) {
+		if (!key_revalidate)
+			return 0;
+
+		key = nvdimm_key_revalidate(nvdimm);
+		if (!key)
+			return nvdimm_security_freeze(nvdimm);
+	} else
+		key = nvdimm_request_key(nvdimm);
+
+	if (!key)
+		return -ENOKEY;
+
+	rc = nvdimm->sec.ops->unlock(nvdimm, key_data(key));
+	dev_dbg(dev, "key: %d unlock: %s\n", key_serial(key),
+			rc == 0 ? "success" : "fail");
+
+	nvdimm_put_key(key);
+	nvdimm->sec.state = nvdimm_security_state(nvdimm, NVDIMM_USER);
+	return rc;
+}
+
+int nvdimm_security_unlock(struct device *dev)
+{
+	struct nvdimm *nvdimm = to_nvdimm(dev);
+	int rc;
+
+	nvdimm_bus_lock(dev);
+	rc = __nvdimm_security_unlock(nvdimm);
+	nvdimm_bus_unlock(dev);
+	return rc;
+}
+
+int nvdimm_security_disable(struct nvdimm *nvdimm, unsigned int keyid)
+{
+	struct device *dev = &nvdimm->dev;
+	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
+	struct key *key;
+	int rc;
+
+	/* The bus lock should be held at the top level of the call stack */
+	lockdep_assert_held(&nvdimm_bus->reconfig_mutex);
+
+	if (!nvdimm->sec.ops || !nvdimm->sec.ops->disable
+			|| nvdimm->sec.state < 0)
+		return -EOPNOTSUPP;
+
+	if (nvdimm->sec.state >= NVDIMM_SECURITY_FROZEN) {
+		dev_dbg(dev, "Incorrect security state: %d\n",
+				nvdimm->sec.state);
+		return -EIO;
+	}
+
+	if (test_bit(NDD_SECURITY_OVERWRITE, &nvdimm->flags)) {
+		dev_dbg(dev, "Security operation in progress.\n");
+		return -EBUSY;
+	}
+
+	key = nvdimm_lookup_user_key(nvdimm, keyid, NVDIMM_BASE_KEY);
+	if (!key)
+		return -ENOKEY;
+
+	rc = nvdimm->sec.ops->disable(nvdimm, key_data(key));
+	dev_dbg(dev, "key: %d disable: %s\n", key_serial(key),
+			rc == 0 ? "success" : "fail");
+
+	nvdimm_put_key(key);
+	nvdimm->sec.state = nvdimm_security_state(nvdimm, NVDIMM_USER);
+	return rc;
+}
+
+int nvdimm_security_update(struct nvdimm *nvdimm, unsigned int keyid,
+		unsigned int new_keyid,
+		enum nvdimm_passphrase_type pass_type)
+{
+	struct device *dev = &nvdimm->dev;
+	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
+	struct key *key, *newkey;
+	int rc;
+
+	/* The bus lock should be held at the top level of the call stack */
+	lockdep_assert_held(&nvdimm_bus->reconfig_mutex);
+
+	if (!nvdimm->sec.ops || !nvdimm->sec.ops->change_key
+			|| nvdimm->sec.state < 0)
+		return -EOPNOTSUPP;
+
+	if (nvdimm->sec.state >= NVDIMM_SECURITY_FROZEN) {
+		dev_dbg(dev, "Incorrect security state: %d\n",
+				nvdimm->sec.state);
+		return -EIO;
+	}
+
+	if (keyid == 0)
+		key = NULL;
+	else {
+		key = nvdimm_lookup_user_key(nvdimm, keyid, NVDIMM_BASE_KEY);
+		if (!key)
+			return -ENOKEY;
+	}
+
+	newkey = nvdimm_lookup_user_key(nvdimm, new_keyid, NVDIMM_NEW_KEY);
+	if (!newkey) {
+		nvdimm_put_key(key);
+		return -ENOKEY;
+	}
+
+	rc = nvdimm->sec.ops->change_key(nvdimm, key ? key_data(key) : NULL,
+			key_data(newkey), pass_type);
+	dev_dbg(dev, "key: %d %d update%s: %s\n",
+			key_serial(key), key_serial(newkey),
+			pass_type == NVDIMM_MASTER ? "(master)" : "(user)",
+			rc == 0 ? "success" : "fail");
+
+	nvdimm_put_key(newkey);
+	nvdimm_put_key(key);
+	if (pass_type == NVDIMM_MASTER)
+		nvdimm->sec.ext_state = nvdimm_security_state(nvdimm,
+				NVDIMM_MASTER);
+	else
+		nvdimm->sec.state = nvdimm_security_state(nvdimm,
+				NVDIMM_USER);
+	return rc;
+}
+
+int nvdimm_security_erase(struct nvdimm *nvdimm, unsigned int keyid,
+		enum nvdimm_passphrase_type pass_type)
+{
+	struct device *dev = &nvdimm->dev;
+	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
+	struct key *key;
+	int rc;
+
+	/* The bus lock should be held at the top level of the call stack */
+	lockdep_assert_held(&nvdimm_bus->reconfig_mutex);
+
+	if (!nvdimm->sec.ops || !nvdimm->sec.ops->erase
+			|| nvdimm->sec.state < 0)
+		return -EOPNOTSUPP;
+
+	if (atomic_read(&nvdimm->busy)) {
+		dev_dbg(dev, "Unable to secure erase while DIMM active.\n");
+		return -EBUSY;
+	}
+
+	if (nvdimm->sec.state >= NVDIMM_SECURITY_FROZEN) {
+		dev_dbg(dev, "Incorrect security state: %d\n",
+				nvdimm->sec.state);
+		return -EIO;
+	}
+
+	if (test_bit(NDD_SECURITY_OVERWRITE, &nvdimm->flags)) {
+		dev_dbg(dev, "Security operation in progress.\n");
+		return -EBUSY;
+	}
+
+	if (nvdimm->sec.ext_state != NVDIMM_SECURITY_UNLOCKED
+			&& pass_type == NVDIMM_MASTER) {
+		dev_dbg(dev,
+			"Attempt to secure erase in wrong master state.\n");
+		return -EOPNOTSUPP;
+	}
+
+	key = nvdimm_lookup_user_key(nvdimm, keyid, NVDIMM_BASE_KEY);
+	if (!key)
+		return -ENOKEY;
+
+	rc = nvdimm->sec.ops->erase(nvdimm, key_data(key), pass_type);
+	dev_dbg(dev, "key: %d erase%s: %s\n", key_serial(key),
+			pass_type == NVDIMM_MASTER ? "(master)" : "(user)",
+			rc == 0 ? "success" : "fail");
+
+	nvdimm_put_key(key);
+	nvdimm->sec.state = nvdimm_security_state(nvdimm, NVDIMM_USER);
+	return rc;
+}
+
+int nvdimm_security_overwrite(struct nvdimm *nvdimm, unsigned int keyid)
+{
+	struct device *dev = &nvdimm->dev;
+	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(dev);
+	struct key *key;
+	int rc;
+
+	/* The bus lock should be held at the top level of the call stack */
+	lockdep_assert_held(&nvdimm_bus->reconfig_mutex);
+
+	if (!nvdimm->sec.ops || !nvdimm->sec.ops->overwrite
+			|| nvdimm->sec.state < 0)
+		return -EOPNOTSUPP;
+
+	if (atomic_read(&nvdimm->busy)) {
+		dev_dbg(dev, "Unable to overwrite while DIMM active.\n");
+		return -EBUSY;
+	}
+
+	if (dev->driver == NULL) {
+		dev_dbg(dev, "Unable to overwrite while DIMM active.\n");
+		return -EINVAL;
+	}
+
+	if (nvdimm->sec.state >= NVDIMM_SECURITY_FROZEN) {
+		dev_dbg(dev, "Incorrect security state: %d\n",
+				nvdimm->sec.state);
+		return -EIO;
+	}
+
+	if (test_bit(NDD_SECURITY_OVERWRITE, &nvdimm->flags)) {
+		dev_dbg(dev, "Security operation in progress.\n");
+		return -EBUSY;
+	}
+
+	if (keyid == 0)
+		key = NULL;
+	else {
+		key = nvdimm_lookup_user_key(nvdimm, keyid, NVDIMM_BASE_KEY);
+		if (!key)
+			return -ENOKEY;
+	}
+
+	rc = nvdimm->sec.ops->overwrite(nvdimm, key ? key_data(key) : NULL);
+	dev_dbg(dev, "key: %d overwrite submission: %s\n", key_serial(key),
+			rc == 0 ? "success" : "fail");
+
+	nvdimm_put_key(key);
+	if (rc == 0) {
+		set_bit(NDD_SECURITY_OVERWRITE, &nvdimm->flags);
+		set_bit(NDD_WORK_PENDING, &nvdimm->flags);
+		nvdimm->sec.state = NVDIMM_SECURITY_OVERWRITE;
+		/*
+		 * Make sure we don't lose device while doing overwrite
+		 * query.
+		 */
+		get_device(dev);
+		queue_delayed_work(system_wq, &nvdimm->dwork, 0);
+	}
+
+	return rc;
+}
+
+void __nvdimm_security_overwrite_query(struct nvdimm *nvdimm)
+{
+	struct nvdimm_bus *nvdimm_bus = walk_to_nvdimm_bus(&nvdimm->dev);
+	int rc;
+	unsigned int tmo;
+
+	/* The bus lock should be held at the top level of the call stack */
+	lockdep_assert_held(&nvdimm_bus->reconfig_mutex);
+
+	/*
+	 * Abort and release device if we no longer have the overwrite
+	 * flag set. It means the work has been canceled.
+	 */
+	if (!test_bit(NDD_WORK_PENDING, &nvdimm->flags))
+		return;
+
+	tmo = nvdimm->sec.overwrite_tmo;
+
+	if (!nvdimm->sec.ops || !nvdimm->sec.ops->query_overwrite
+			|| nvdimm->sec.state < 0)
+		return;
+
+	rc = nvdimm->sec.ops->query_overwrite(nvdimm);
+	if (rc == -EBUSY) {
+
+		/* setup delayed work again */
+		tmo += 10;
+		queue_delayed_work(system_wq, &nvdimm->dwork, tmo * HZ);
+		nvdimm->sec.overwrite_tmo = min(15U * 60U, tmo);
+		return;
+	}
+
+	if (rc < 0)
+		dev_dbg(&nvdimm->dev, "overwrite failed\n");
+	else
+		dev_dbg(&nvdimm->dev, "overwrite completed\n");
+
+	if (nvdimm->sec.overwrite_state)
+		sysfs_notify_dirent(nvdimm->sec.overwrite_state);
+	nvdimm->sec.overwrite_tmo = 0;
+	clear_bit(NDD_SECURITY_OVERWRITE, &nvdimm->flags);
+	clear_bit(NDD_WORK_PENDING, &nvdimm->flags);
+	put_device(&nvdimm->dev);
+	nvdimm->sec.state = nvdimm_security_state(nvdimm, NVDIMM_USER);
+	nvdimm->sec.ext_state = nvdimm_security_state(nvdimm, NVDIMM_MASTER);
+}
+
+void nvdimm_security_overwrite_query(struct work_struct *work)
+{
+	struct nvdimm *nvdimm =
+		container_of(work, typeof(*nvdimm), dwork.work);
+
+	nvdimm_bus_lock(&nvdimm->dev);
+	__nvdimm_security_overwrite_query(nvdimm);
+	nvdimm_bus_unlock(&nvdimm->dev);
+}

include/linux/key.h

···
 
 extern void key_set_timeout(struct key *, unsigned);
 
+extern key_ref_t lookup_user_key(key_serial_t id, unsigned long flags,
+				 key_perm_t perm);
+
 /*
  * The permissions required on a key that we're looking up.
  */

include/linux/libnvdimm.h

···
 	NDD_UNARMED = 1,
 	/* locked memory devices should not be accessed */
 	NDD_LOCKED = 2,
+	/* memory under security wipes should not be accessed */
+	NDD_SECURITY_OVERWRITE = 3,
+	/*  tracking whether or not there is a pending device reference */
+	NDD_WORK_PENDING = 4,
 
 	/* need to set a limit somewhere, but yes, this is likely overkill */
 	ND_IOCTL_MAX_BUFLEN = SZ_4M,
···
 	ndctl_fn ndctl;
 	int (*flush_probe)(struct nvdimm_bus_descriptor *nd_desc);
 	int (*clear_to_send)(struct nvdimm_bus_descriptor *nd_desc,
-			struct nvdimm *nvdimm, unsigned int cmd);
+			struct nvdimm *nvdimm, unsigned int cmd, void *data);
 };
 
 struct nd_cmd_desc {
···
 
 }
 
+enum nvdimm_security_state {
+	NVDIMM_SECURITY_DISABLED,
+	NVDIMM_SECURITY_UNLOCKED,
+	NVDIMM_SECURITY_LOCKED,
+	NVDIMM_SECURITY_FROZEN,
+	NVDIMM_SECURITY_OVERWRITE,
+};
+
+#define NVDIMM_PASSPHRASE_LEN		32
+#define NVDIMM_KEY_DESC_LEN		22
+
+struct nvdimm_key_data {
+	u8 data[NVDIMM_PASSPHRASE_LEN];
+};
+
+enum nvdimm_passphrase_type {
+	NVDIMM_USER,
+	NVDIMM_MASTER,
+};
+
+struct nvdimm_security_ops {
+	enum nvdimm_security_state (*state)(struct nvdimm *nvdimm,
+			enum nvdimm_passphrase_type pass_type);
+	int (*freeze)(struct nvdimm *nvdimm);
+	int (*change_key)(struct nvdimm *nvdimm,
+			const struct nvdimm_key_data *old_data,
+			const struct nvdimm_key_data *new_data,
+			enum nvdimm_passphrase_type pass_type);
+	int (*unlock)(struct nvdimm *nvdimm,
+			const struct nvdimm_key_data *key_data);
+	int (*disable)(struct nvdimm *nvdimm,
+			const struct nvdimm_key_data *key_data);
+	int (*erase)(struct nvdimm *nvdimm,
+			const struct nvdimm_key_data *key_data,
+			enum nvdimm_passphrase_type pass_type);
+	int (*overwrite)(struct nvdimm *nvdimm,
+			const struct nvdimm_key_data *key_data);
+	int (*query_overwrite)(struct nvdimm *nvdimm);
+};
+
 void badrange_init(struct badrange *badrange);
 int badrange_add(struct badrange *badrange, u64 addr, u64 length);
 void badrange_forget(struct badrange *badrange, phys_addr_t start,
···
 		struct nvdimm_bus_descriptor *nfit_desc);
 void nvdimm_bus_unregister(struct nvdimm_bus *nvdimm_bus);
 struct nvdimm_bus *to_nvdimm_bus(struct device *dev);
+struct nvdimm_bus *nvdimm_to_bus(struct nvdimm *nvdimm);
 struct nvdimm *to_nvdimm(struct device *dev);
 struct nd_region *to_nd_region(struct device *dev);
 struct device *nd_region_dev(struct nd_region *nd_region);
···
 struct kobject *nvdimm_kobj(struct nvdimm *nvdimm);
 unsigned long nvdimm_cmd_mask(struct nvdimm *nvdimm);
 void *nvdimm_provider_data(struct nvdimm *nvdimm);
-struct nvdimm *nvdimm_create(struct nvdimm_bus *nvdimm_bus, void *provider_data,
-		const struct attribute_group **groups, unsigned long flags,
-		unsigned long cmd_mask, int num_flush,
-		struct resource *flush_wpq);
+struct nvdimm *__nvdimm_create(struct nvdimm_bus *nvdimm_bus,
+		void *provider_data, const struct attribute_group **groups,
+		unsigned long flags, unsigned long cmd_mask, int num_flush,
+		struct resource *flush_wpq, const char *dimm_id,
+		const struct nvdimm_security_ops *sec_ops);
+static inline struct nvdimm *nvdimm_create(struct nvdimm_bus *nvdimm_bus,
+		void *provider_data, const struct attribute_group **groups,
+		unsigned long flags, unsigned long cmd_mask, int num_flush,
+		struct resource *flush_wpq)
+{
+	return __nvdimm_create(nvdimm_bus, provider_data, groups, flags,
+			cmd_mask, num_flush, flush_wpq, NULL, NULL);
+}
+
+int nvdimm_security_setup_events(struct nvdimm *nvdimm);
 const struct nd_cmd_desc *nd_cmd_dimm_desc(int cmd);
 const struct nd_cmd_desc *nd_cmd_bus_desc(int cmd);
 u32 nd_cmd_in_size(struct nvdimm *nvdimm, int cmd,
···
 void nvdimm_flush(struct nd_region *nd_region);
 int nvdimm_has_flush(struct nd_region *nd_region);
 int nvdimm_has_cache(struct nd_region *nd_region);
+int nvdimm_in_overwrite(struct nvdimm *nvdimm);
+
+static inline int nvdimm_ctl(struct nvdimm *nvdimm, unsigned int cmd, void *buf,
+		unsigned int buf_len, int *cmd_rc)
+{
+	struct nvdimm_bus *nvdimm_bus = nvdimm_to_bus(nvdimm);
+	struct nvdimm_bus_descriptor *nd_desc = to_nd_desc(nvdimm_bus);
+
+	return nd_desc->ndctl(nd_desc, nvdimm, cmd, buf, buf_len, cmd_rc);
+}
 
 #ifdef CONFIG_ARCH_HAS_PMEM_API
 #define ARCH_MEMREMAP_PMEM MEMREMAP_WB

security/keys/encrypted-keys/encrypted.c

···
 static const char blkcipher_alg[] = "cbc(aes)";
 static const char key_format_default[] = "default";
 static const char key_format_ecryptfs[] = "ecryptfs";
+static const char key_format_enc32[] = "enc32";
 static unsigned int ivsize;
 static int blksize;
 
···
 #define HASH_SIZE SHA256_DIGEST_SIZE
 #define MAX_DATA_SIZE 4096
 #define MIN_DATA_SIZE  20
+#define KEY_ENC32_PAYLOAD_LEN 32
 
 static struct crypto_shash *hash_tfm;
 
···
 };
 
 enum {
-	Opt_error = -1, Opt_default, Opt_ecryptfs
+	Opt_error = -1, Opt_default, Opt_ecryptfs, Opt_enc32
 };
 
 static const match_table_t key_format_tokens = {
 	{Opt_default, "default"},
 	{Opt_ecryptfs, "ecryptfs"},
+	{Opt_enc32, "enc32"},
 	{Opt_error, NULL}
 };
 
···
 	key_format = match_token(p, key_format_tokens, args);
 	switch (key_format) {
 	case Opt_ecryptfs:
+	case Opt_enc32:
 	case Opt_default:
 		*format = p;
 		*master_desc = strsep(&datablob, " \t");
···
 	format_len = (!format) ? strlen(key_format_default) : strlen(format);
 	decrypted_datalen = dlen;
 	payload_datalen = decrypted_datalen;
-	if (format && !strcmp(format, key_format_ecryptfs)) {
-		if (dlen != ECRYPTFS_MAX_KEY_BYTES) {
-			pr_err("encrypted_key: keylen for the ecryptfs format "
-			       "must be equal to %d bytes\n",
-			       ECRYPTFS_MAX_KEY_BYTES);
-			return ERR_PTR(-EINVAL);
+	if (format) {
+		if (!strcmp(format, key_format_ecryptfs)) {
+			if (dlen != ECRYPTFS_MAX_KEY_BYTES) {
+				pr_err("encrypted_key: keylen for the ecryptfs format must be equal to %d bytes\n",
+					ECRYPTFS_MAX_KEY_BYTES);
+				return ERR_PTR(-EINVAL);
+			}
+			decrypted_datalen = ECRYPTFS_MAX_KEY_BYTES;
+			payload_datalen = sizeof(struct ecryptfs_auth_tok);
+		} else if (!strcmp(format, key_format_enc32)) {
+			if (decrypted_datalen != KEY_ENC32_PAYLOAD_LEN) {
+				pr_err("encrypted_key: enc32 key payload incorrect length: %d\n",
+						decrypted_datalen);
+				return ERR_PTR(-EINVAL);
+			}
 		}
-		decrypted_datalen = ECRYPTFS_MAX_KEY_BYTES;
-		payload_datalen = sizeof(struct ecryptfs_auth_tok);
 	}
 
 	encrypted_datalen = roundup(decrypted_datalen, blksize);

security/keys/internal.h

···
 
 extern bool lookup_user_key_possessed(const struct key *key,
 				      const struct key_match_data *match_data);
-extern key_ref_t lookup_user_key(key_serial_t id, unsigned long flags,
-				 key_perm_t perm);
 #define KEY_LOOKUP_CREATE	0x01
 #define KEY_LOOKUP_PARTIAL	0x02
 #define KEY_LOOKUP_FOR_UNLINK	0x04

security/keys/process_keys.c

···
 	put_cred(ctx.cred);
 	goto try_again;
 }
+EXPORT_SYMBOL(lookup_user_key);
 
 /*
  * Join the named keyring as the session keyring if possible else attempt to

tools/testing/nvdimm/Kbuild

···
 obj-$(CONFIG_DEV_DAX_PMEM) += dax_pmem.o
 
 nfit-y := $(ACPI_SRC)/core.o
+nfit-y += $(ACPI_SRC)/intel.o
 nfit-$(CONFIG_X86_MCE) += $(ACPI_SRC)/mce.o
 nfit-y += acpi_nfit_test.o
 nfit-y += config_check.o
···
 libnvdimm-$(CONFIG_BTT) += $(NVDIMM_SRC)/btt_devs.o
 libnvdimm-$(CONFIG_NVDIMM_PFN) += $(NVDIMM_SRC)/pfn_devs.o
 libnvdimm-$(CONFIG_NVDIMM_DAX) += $(NVDIMM_SRC)/dax_devs.o
+libnvdimm-$(CONFIG_NVDIMM_KEYS) += $(NVDIMM_SRC)/security.o
+libnvdimm-y += dimm_devs.o
 libnvdimm-y += libnvdimm_test.o
 libnvdimm-y += config_check.o
 

tools/testing/nvdimm/dimm_devs.c

···
+// SPDX-License-Identifier: GPL-2.0
+/* Copyright Intel Corp. 2018 */
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/moduleparam.h>
+#include <linux/nd.h>
+#include "pmem.h"
+#include "pfn.h"
+#include "nd.h"
+#include "nd-core.h"
+
+ssize_t security_show(struct device *dev,
+		struct device_attribute *attr, char *buf)
+{
+	struct nvdimm *nvdimm = to_nvdimm(dev);
+
+	/*
+	 * For the test version we need to poll the "hardware" in order
+	 * to get the updated status for unlock testing.
+	 */
+	nvdimm->sec.state = nvdimm_security_state(nvdimm, false);
+	nvdimm->sec.ext_state = nvdimm_security_state(nvdimm, true);
+
+	switch (nvdimm->sec.state) {
+	case NVDIMM_SECURITY_DISABLED:
+		return sprintf(buf, "disabled\n");
+	case NVDIMM_SECURITY_UNLOCKED:
+		return sprintf(buf, "unlocked\n");
+	case NVDIMM_SECURITY_LOCKED:
+		return sprintf(buf, "locked\n");
+	case NVDIMM_SECURITY_FROZEN:
+		return sprintf(buf, "frozen\n");
+	case NVDIMM_SECURITY_OVERWRITE:
+		return sprintf(buf, "overwrite\n");
+	default:
+		return -ENOTTY;
+	}
+
+	return -ENOTTY;
+}
+

tools/testing/nvdimm/test/nfit.c

···
 
 static unsigned long dimm_fail_cmd_flags[ARRAY_SIZE(handle)];
 static int dimm_fail_cmd_code[ARRAY_SIZE(handle)];
+struct nfit_test_sec {
+	u8 state;
+	u8 ext_state;
+	u8 passphrase[32];
+	u8 master_passphrase[32];
+	u64 overwrite_end_time;
+} dimm_sec_info[NUM_DCR];
 
 static const struct nd_intel_smart smart_def = {
 	.flags = ND_INTEL_SMART_HEALTH_VALID
···
 	return rc;
 }
 
+static int nd_intel_test_cmd_security_status(struct nfit_test *t,
+		struct nd_intel_get_security_state *nd_cmd,
+		unsigned int buf_len, int dimm)
+{
+	struct device *dev = &t->pdev.dev;
+	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
+
+	nd_cmd->status = 0;
+	nd_cmd->state = sec->state;
+	nd_cmd->extended_state = sec->ext_state;
+	dev_dbg(dev, "security state (%#x) returned\n", nd_cmd->state);
+
+	return 0;
+}
+
+static int nd_intel_test_cmd_unlock_unit(struct nfit_test *t,
+		struct nd_intel_unlock_unit *nd_cmd,
+		unsigned int buf_len, int dimm)
+{
+	struct device *dev = &t->pdev.dev;
+	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
+
+	if (!(sec->state & ND_INTEL_SEC_STATE_LOCKED) ||
+			(sec->state & ND_INTEL_SEC_STATE_FROZEN)) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_STATE;
+		dev_dbg(dev, "unlock unit: invalid state: %#x\n",
+				sec->state);
+	} else if (memcmp(nd_cmd->passphrase, sec->passphrase,
+				ND_INTEL_PASSPHRASE_SIZE) != 0) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_PASS;
+		dev_dbg(dev, "unlock unit: invalid passphrase\n");
+	} else {
+		nd_cmd->status = 0;
+		sec->state = ND_INTEL_SEC_STATE_ENABLED;
+		dev_dbg(dev, "Unit unlocked\n");
+	}
+
+	dev_dbg(dev, "unlocking status returned: %#x\n", nd_cmd->status);
+	return 0;
+}
+
+static int nd_intel_test_cmd_set_pass(struct nfit_test *t,
+		struct nd_intel_set_passphrase *nd_cmd,
+		unsigned int buf_len, int dimm)
+{
+	struct device *dev = &t->pdev.dev;
+	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
+
+	if (sec->state & ND_INTEL_SEC_STATE_FROZEN) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_STATE;
+		dev_dbg(dev, "set passphrase: wrong security state\n");
+	} else if (memcmp(nd_cmd->old_pass, sec->passphrase,
+				ND_INTEL_PASSPHRASE_SIZE) != 0) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_PASS;
+		dev_dbg(dev, "set passphrase: wrong passphrase\n");
+	} else {
+		memcpy(sec->passphrase, nd_cmd->new_pass,
+				ND_INTEL_PASSPHRASE_SIZE);
+		sec->state |= ND_INTEL_SEC_STATE_ENABLED;
+		nd_cmd->status = 0;
+		dev_dbg(dev, "passphrase updated\n");
+	}
+
+	return 0;
+}
+
+static int nd_intel_test_cmd_freeze_lock(struct nfit_test *t,
+		struct nd_intel_freeze_lock *nd_cmd,
+		unsigned int buf_len, int dimm)
+{
+	struct device *dev = &t->pdev.dev;
+	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
+
+	if (!(sec->state & ND_INTEL_SEC_STATE_ENABLED)) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_STATE;
+		dev_dbg(dev, "freeze lock: wrong security state\n");
+	} else {
+		sec->state |= ND_INTEL_SEC_STATE_FROZEN;
+		nd_cmd->status = 0;
+		dev_dbg(dev, "security frozen\n");
+	}
+
+	return 0;
+}
+
+static int nd_intel_test_cmd_disable_pass(struct nfit_test *t,
+		struct nd_intel_disable_passphrase *nd_cmd,
+		unsigned int buf_len, int dimm)
+{
+	struct device *dev = &t->pdev.dev;
+	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
+
+	if (!(sec->state & ND_INTEL_SEC_STATE_ENABLED) ||
+			(sec->state & ND_INTEL_SEC_STATE_FROZEN)) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_STATE;
+		dev_dbg(dev, "disable passphrase: wrong security state\n");
+	} else if (memcmp(nd_cmd->passphrase, sec->passphrase,
+				ND_INTEL_PASSPHRASE_SIZE) != 0) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_PASS;
+		dev_dbg(dev, "disable passphrase: wrong passphrase\n");
+	} else {
+		memset(sec->passphrase, 0, ND_INTEL_PASSPHRASE_SIZE);
+		sec->state = 0;
+		dev_dbg(dev, "disable passphrase: done\n");
+	}
+
+	return 0;
+}
+
+static int nd_intel_test_cmd_secure_erase(struct nfit_test *t,
+		struct nd_intel_secure_erase *nd_cmd,
+		unsigned int buf_len, int dimm)
+{
+	struct device *dev = &t->pdev.dev;
+	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
+
+	if (!(sec->state & ND_INTEL_SEC_STATE_ENABLED) ||
+			(sec->state & ND_INTEL_SEC_STATE_FROZEN)) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_STATE;
+		dev_dbg(dev, "secure erase: wrong security state\n");
+	} else if (memcmp(nd_cmd->passphrase, sec->passphrase,
+				ND_INTEL_PASSPHRASE_SIZE) != 0) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_PASS;
+		dev_dbg(dev, "secure erase: wrong passphrase\n");
+	} else {
+		memset(sec->passphrase, 0, ND_INTEL_PASSPHRASE_SIZE);
+		memset(sec->master_passphrase, 0, ND_INTEL_PASSPHRASE_SIZE);
+		sec->state = 0;
+		sec->ext_state = ND_INTEL_SEC_ESTATE_ENABLED;
+		dev_dbg(dev, "secure erase: done\n");
+	}
+
+	return 0;
+}
+
+static int nd_intel_test_cmd_overwrite(struct nfit_test *t,
+		struct nd_intel_overwrite *nd_cmd,
+		unsigned int buf_len, int dimm)
+{
+	struct device *dev = &t->pdev.dev;
+	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
+
+	if ((sec->state & ND_INTEL_SEC_STATE_ENABLED) &&
+			memcmp(nd_cmd->passphrase, sec->passphrase,
+				ND_INTEL_PASSPHRASE_SIZE) != 0) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_PASS;
+		dev_dbg(dev, "overwrite: wrong passphrase\n");
+		return 0;
+	}
+
+	memset(sec->passphrase, 0, ND_INTEL_PASSPHRASE_SIZE);
+	sec->state = ND_INTEL_SEC_STATE_OVERWRITE;
+	dev_dbg(dev, "overwrite progressing.\n");
+	sec->overwrite_end_time = get_jiffies_64() + 5 * HZ;
+
+	return 0;
+}
+
+static int nd_intel_test_cmd_query_overwrite(struct nfit_test *t,
+		struct nd_intel_query_overwrite *nd_cmd,
+		unsigned int buf_len, int dimm)
+{
+	struct device *dev = &t->pdev.dev;
+	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
+
+	if (!(sec->state & ND_INTEL_SEC_STATE_OVERWRITE)) {
+		nd_cmd->status = ND_INTEL_STATUS_OQUERY_SEQUENCE_ERR;
+		return 0;
+	}
+
+	if (time_is_before_jiffies64(sec->overwrite_end_time)) {
+		sec->overwrite_end_time = 0;
+		sec->state = 0;
+		sec->ext_state = ND_INTEL_SEC_ESTATE_ENABLED;
+		dev_dbg(dev, "overwrite is complete\n");
+	} else
+		nd_cmd->status = ND_INTEL_STATUS_OQUERY_INPROGRESS;
+	return 0;
+}
+
+static int nd_intel_test_cmd_master_set_pass(struct nfit_test *t,
+		struct nd_intel_set_master_passphrase *nd_cmd,
+		unsigned int buf_len, int dimm)
+{
+	struct device *dev = &t->pdev.dev;
+	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
+
+	if (!(sec->ext_state & ND_INTEL_SEC_ESTATE_ENABLED)) {
+		nd_cmd->status = ND_INTEL_STATUS_NOT_SUPPORTED;
+		dev_dbg(dev, "master set passphrase: in wrong state\n");
+	} else if (sec->ext_state & ND_INTEL_SEC_ESTATE_PLIMIT) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_STATE;
+		dev_dbg(dev, "master set passphrase: in wrong security state\n");
+	} else if (memcmp(nd_cmd->old_pass, sec->master_passphrase,
+				ND_INTEL_PASSPHRASE_SIZE) != 0) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_PASS;
+		dev_dbg(dev, "master set passphrase: wrong passphrase\n");
+	} else {
+		memcpy(sec->master_passphrase, nd_cmd->new_pass,
+				ND_INTEL_PASSPHRASE_SIZE);
+		sec->ext_state = ND_INTEL_SEC_ESTATE_ENABLED;
+		dev_dbg(dev, "master passphrase: updated\n");
+	}
+
+	return 0;
+}
+
+static int nd_intel_test_cmd_master_secure_erase(struct nfit_test *t,
+		struct nd_intel_master_secure_erase *nd_cmd,
+		unsigned int buf_len, int dimm)
+{
+	struct device *dev = &t->pdev.dev;
+	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
+
+	if (!(sec->ext_state & ND_INTEL_SEC_ESTATE_ENABLED)) {
+		nd_cmd->status = ND_INTEL_STATUS_NOT_SUPPORTED;
+		dev_dbg(dev, "master secure erase: in wrong state\n");
+	} else if (sec->ext_state & ND_INTEL_SEC_ESTATE_PLIMIT) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_STATE;
+		dev_dbg(dev, "master secure erase: in wrong security state\n");
+	} else if (memcmp(nd_cmd->passphrase, sec->master_passphrase,
+				ND_INTEL_PASSPHRASE_SIZE) != 0) {
+		nd_cmd->status = ND_INTEL_STATUS_INVALID_PASS;
+		dev_dbg(dev, "master secure erase: wrong passphrase\n");
+	} else {
+		/* we do not erase master state passphrase ever */
+		sec->ext_state = ND_INTEL_SEC_ESTATE_ENABLED;
+		memset(sec->passphrase, 0, ND_INTEL_PASSPHRASE_SIZE);
+		sec->state = 0;
+		dev_dbg(dev, "master secure erase: done\n");
+	}
+
+	return 0;
+}
+
+
 static int get_dimm(struct nfit_mem *nfit_mem, unsigned int func)
 {
 	int i;
···
 				return i;
 
 			switch (func) {
+			case NVDIMM_INTEL_GET_SECURITY_STATE:
+				rc = nd_intel_test_cmd_security_status(t,
+						buf, buf_len, i);
+				break;
+			case NVDIMM_INTEL_UNLOCK_UNIT:
+				rc = nd_intel_test_cmd_unlock_unit(t,
+						buf, buf_len, i);
+				break;
+			case NVDIMM_INTEL_SET_PASSPHRASE:
+				rc = nd_intel_test_cmd_set_pass(t,
+						buf, buf_len, i);
+				break;
+			case NVDIMM_INTEL_DISABLE_PASSPHRASE:
+				rc = nd_intel_test_cmd_disable_pass(t,
+						buf, buf_len, i);
+				break;
+			case NVDIMM_INTEL_FREEZE_LOCK:
+				rc = nd_intel_test_cmd_freeze_lock(t,
+						buf, buf_len, i);
+				break;
+			case NVDIMM_INTEL_SECURE_ERASE:
+				rc = nd_intel_test_cmd_secure_erase(t,
+						buf, buf_len, i);
+				break;
+			case NVDIMM_INTEL_OVERWRITE:
+				rc = nd_intel_test_cmd_overwrite(t,
+						buf, buf_len, i - t->dcr_idx);
+				break;
+			case NVDIMM_INTEL_QUERY_OVERWRITE:
+				rc = nd_intel_test_cmd_query_overwrite(t,
+						buf, buf_len, i - t->dcr_idx);
+				break;
+			case NVDIMM_INTEL_SET_MASTER_PASSPHRASE:
+				rc = nd_intel_test_cmd_master_set_pass(t,
+						buf, buf_len, i);
+				break;
+			case NVDIMM_INTEL_MASTER_SECURE_ERASE:
+				rc = nd_intel_test_cmd_master_secure_erase(t,
+						buf, buf_len, i);
+				break;
 			case ND_INTEL_ENABLE_LSS_STATUS:
 				rc = nd_intel_test_cmd_set_lss_status(t,
 						buf, buf_len);
···
 }
 static DEVICE_ATTR_RW(fail_cmd_code);
 
+static ssize_t lock_dimm_store(struct device *dev,
+		struct device_attribute *attr, const char *buf, size_t size)
+{
+	int dimm = dimm_name_to_id(dev);
+	struct nfit_test_sec *sec = &dimm_sec_info[dimm];
+
+	sec->state = ND_INTEL_SEC_STATE_ENABLED | ND_INTEL_SEC_STATE_LOCKED;
+	return size;
+}
+static DEVICE_ATTR_WO(lock_dimm);
+
 static struct attribute *nfit_test_dimm_attributes[] = {
 	&dev_attr_fail_cmd.attr,
 	&dev_attr_fail_cmd_code.attr,
 	&dev_attr_handle.attr,
+	&dev_attr_lock_dimm.attr,
 	NULL,
 };
 
···
 			return -ENOMEM;
 	}
 	return 0;
+}
+
+static void security_init(struct nfit_test *t)
+{
+	int i;
+
+	for (i = 0; i < t->num_dcr; i++) {
+		struct nfit_test_sec *sec = &dimm_sec_info[i];
+
+		sec->ext_state = ND_INTEL_SEC_ESTATE_ENABLED;
+	}
 }
 
 static void smart_init(struct nfit_test *t)
···
 	if (nfit_test_dimm_init(t))
 		return -ENOMEM;
 	smart_init(t);
+	security_init(t);
 	return ars_state_init(&t->pdev.dev, &t->ars_state);
 }
 
···
 	set_bit(ND_INTEL_FW_FINISH_UPDATE, &acpi_desc->dimm_cmd_force_en);
 	set_bit(ND_INTEL_FW_FINISH_QUERY, &acpi_desc->dimm_cmd_force_en);
 	set_bit(ND_INTEL_ENABLE_LSS_STATUS, &acpi_desc->dimm_cmd_force_en);
+	set_bit(NVDIMM_INTEL_GET_SECURITY_STATE,
+			&acpi_desc->dimm_cmd_force_en);
+	set_bit(NVDIMM_INTEL_SET_PASSPHRASE, &acpi_desc->dimm_cmd_force_en);
+	set_bit(NVDIMM_INTEL_DISABLE_PASSPHRASE,
+			&acpi_desc->dimm_cmd_force_en);
+	set_bit(NVDIMM_INTEL_UNLOCK_UNIT, &acpi_desc->dimm_cmd_force_en);
+	set_bit(NVDIMM_INTEL_FREEZE_LOCK, &acpi_desc->dimm_cmd_force_en);
+	set_bit(NVDIMM_INTEL_SECURE_ERASE, &acpi_desc->dimm_cmd_force_en);
+	set_bit(NVDIMM_INTEL_OVERWRITE, &acpi_desc->dimm_cmd_force_en);
+	set_bit(NVDIMM_INTEL_QUERY_OVERWRITE, &acpi_desc->dimm_cmd_force_en);
+	set_bit(NVDIMM_INTEL_SET_MASTER_PASSPHRASE,
+			&acpi_desc->dimm_cmd_force_en);
+	set_bit(NVDIMM_INTEL_MASTER_SECURE_ERASE,
+			&acpi_desc->dimm_cmd_force_en);
 }
 
 static void nfit_test1_setup(struct nfit_test *t)