Re: [PATCH v12 04/12] nfit/libnvdimm: add unlock of nvdimm support for Intel DIMMs

From: Dave Jiang <dave.jiang@intel.com>
To: Dan Williams <dan.j.williams@intel.com>
Cc: "Schofield, Alison" <alison.schofield@intel.com>,
	Kees Cook <keescook@chromium.org>,
	linux-nvdimm <linux-nvdimm@lists.01.org>,
	Eric Biggers <ebiggers3@gmail.com>,
	David Howells <dhowells@redhat.com>,
	keyrings@vger.kernel.org
Subject: Re: [PATCH v12 04/12] nfit/libnvdimm: add unlock of nvdimm support for Intel DIMMs
Date: Tue, 9 Oct 2018 12:45:34 -0700	[thread overview]
Message-ID: <0853af1b-e0db-88b3-5166-5ed076bc2cfe@intel.com> (raw)
In-Reply-To: <CAPcyv4iBZfqTV2nGtA4sTvJtyzVOSonUYMqwmHeJOaZem=NAxg@mail.gmail.com>

On 10/09/2018 12:07 PM, Dan Williams wrote:
> On Mon, Oct 8, 2018 at 5:01 PM Dave Jiang <dave.jiang@intel.com> wrote:
>>
>> Add support to allow query the security status of the Intel nvdimms and
>> also unlock the dimm via the kernel key management APIs. The passphrase is
>> expected to be pulled from userspace through keyutils. Moving the Intel
>> related bits to its own source file as well.
>>
>> Signed-off-by: Dave Jiang <dave.jiang@intel.com>
>> Reviewed-by: Dan Williams <dan.j.williams@intel.com>
>> Signed-off-by: Dan Williams <dan.j.williams@intel.com>
>> ---
>>  drivers/acpi/nfit/Makefile          |    1
>>  drivers/acpi/nfit/core.c            |    3 -
>>  drivers/acpi/nfit/intel.c           |  152 +++++++++++++++++++++++++++++++++++
>>  drivers/acpi/nfit/intel.h           |   15 +++
>>  drivers/nvdimm/Kconfig              |    5 +
>>  drivers/nvdimm/Makefile             |    1
>>  drivers/nvdimm/dimm.c               |    7 ++
>>  drivers/nvdimm/dimm_devs.c          |   12 +++
>>  drivers/nvdimm/dimm_devs_security.c |  129 ++++++++++++++++++++++++++++++
>>  drivers/nvdimm/nd-core.h            |    5 +
>>  drivers/nvdimm/nd.h                 |   21 +++++
>>  include/linux/libnvdimm.h           |   27 ++++++
>>  include/uapi/linux/ndctl.h          |    6 +
>>  tools/testing/nvdimm/Kbuild         |    2
>>  14 files changed, 383 insertions(+), 3 deletions(-)
>>  create mode 100644 drivers/acpi/nfit/intel.c
>>  create mode 100644 drivers/nvdimm/dimm_devs_security.c
>>
>> diff --git a/drivers/acpi/nfit/Makefile b/drivers/acpi/nfit/Makefile
>> index a407e769f103..443c7ef4e6a6 100644
>> --- a/drivers/acpi/nfit/Makefile
>> +++ b/drivers/acpi/nfit/Makefile
>> @@ -1,3 +1,4 @@
>>  obj-$(CONFIG_ACPI_NFIT) := nfit.o
>>  nfit-y := core.o
>> +nfit-$(CONFIG_X86) += intel.o
>>  nfit-$(CONFIG_X86_MCE) += mce.o
>> diff --git a/drivers/acpi/nfit/core.c b/drivers/acpi/nfit/core.c
>> index 94755a4b6327..e5f7c664a7c8 100644
>> --- a/drivers/acpi/nfit/core.c
>> +++ b/drivers/acpi/nfit/core.c
>> @@ -1904,7 +1904,8 @@ static int acpi_nfit_register_dimms(struct acpi_nfit_desc *acpi_desc)
>>                 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->id[0]);
>> +                               nfit_mem->flush_wpq, &nfit_mem->id[0],
>> +                               acpi_nfit_get_security_ops(nfit_mem->family));
>>                 if (!nvdimm)
>>                         return -ENOMEM;
>>
>> diff --git a/drivers/acpi/nfit/intel.c b/drivers/acpi/nfit/intel.c
>> new file mode 100644
>> index 000000000000..4bfc1c1da339
>> --- /dev/null
>> +++ b/drivers/acpi/nfit/intel.c
>> @@ -0,0 +1,152 @@
>> +// SPDX-License-Identifier: GPL-2.0
>> +/* Copyright(c) 2018 Intel Corporation. All rights reserved. */
>> +/*
>> + * Intel specific NFIT ops
>> + */
>> +#include <linux/libnvdimm.h>
>> +#include <linux/module.h>
>> +#include <linux/mutex.h>
>> +#include <linux/ndctl.h>
>> +#include <linux/sysfs.h>
>> +#include <linux/delay.h>
>> +#include <linux/acpi.h>
>> +#include <linux/io.h>
>> +#include <linux/nd.h>
>> +#include <asm/cacheflush.h>
>> +#include <asm/smp.h>
>> +#include <acpi/nfit.h>
>> +#include "intel.h"
>> +#include "nfit.h"
>> +
>> +static int intel_dimm_security_unlock(struct nvdimm_bus *nvdimm_bus,
>> +               struct nvdimm *nvdimm, const struct nvdimm_key_data *nkey)
>> +{
>> +       struct nvdimm_bus_descriptor *nd_desc = to_nd_desc(nvdimm_bus);
>> +       int cmd_rc, rc = 0;
>> +       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,
>> +               },
>> +               .cmd = {
>> +                       .status = 0,
>> +               },
>> +       };
>> +
>> +       if (!test_bit(NVDIMM_INTEL_UNLOCK_UNIT, &nfit_mem->dsm_mask))
>> +               return -ENOTTY;
>> +
>> +       memcpy(nd_cmd.cmd.passphrase, nkey->data,
>> +                       sizeof(nd_cmd.cmd.passphrase));
>> +       rc = nd_desc->ndctl(nd_desc, nvdimm, ND_CMD_CALL, &nd_cmd,
>> +                       sizeof(nd_cmd), &cmd_rc);
>> +       if (rc < 0)
>> +               goto out;
>> +       if (cmd_rc < 0) {
>> +               rc = cmd_rc;
>> +               goto out;
>> +       }
>> +
>> +       switch (nd_cmd.cmd.status) {
>> +       case 0:
>> +               break;
>> +       case ND_INTEL_STATUS_INVALID_PASS:
>> +               rc = -EINVAL;
>> +               goto out;
>> +       case ND_INTEL_STATUS_INVALID_STATE:
>> +       default:
>> +               rc = -ENXIO;
>> +               goto out;
>> +       }
>> +
>> +       /*
>> +        * TODO: define a cross arch wbinvd when/if NVDIMM_FAMILY_INTEL
>> +        * support arrives on another arch.
>> +        */
>> +       /* DIMM unlocked, invalidate all CPU caches before we read it */
>> +       wbinvd_on_all_cpus();
>> +
>> + out:
>> +       return rc;
>> +}
>> +
>> +static int intel_dimm_security_state(struct nvdimm_bus *nvdimm_bus,
>> +               struct nvdimm *nvdimm, enum nvdimm_security_state *state)
>> +{
>> +       struct nvdimm_bus_descriptor *nd_desc = to_nd_desc(nvdimm_bus);
>> +       int cmd_rc, rc = 0;
>> +       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_in = 0,
>> +                       .nd_size_out =
>> +                               sizeof(struct nd_intel_get_security_state),
>> +                       .nd_fw_size =
>> +                               sizeof(struct nd_intel_get_security_state),
>> +               },
>> +               .cmd = {
>> +                       .status = 0,
>> +                       .state = 0,
>> +               },
>> +       };
>> +
>> +       if (!test_bit(NVDIMM_INTEL_GET_SECURITY_STATE, &nfit_mem->dsm_mask)) {
>> +               *state = NVDIMM_SECURITY_UNSUPPORTED;
>> +               return 0;
>> +       }
>> +
>> +       *state = NVDIMM_SECURITY_DISABLED;
>> +       rc = nd_desc->ndctl(nd_desc, nvdimm, ND_CMD_CALL, &nd_cmd,
>> +                       sizeof(nd_cmd), &cmd_rc);
>> +       if (rc < 0)
>> +               goto out;
>> +       if (cmd_rc < 0) {
>> +               rc = cmd_rc;
>> +               goto out;
>> +       }
>> +
>> +       switch (nd_cmd.cmd.status) {
>> +       case 0:
>> +               break;
>> +       case ND_INTEL_STATUS_RETRY:
>> +               rc = -EAGAIN;
>> +               goto out;
>> +       case ND_INTEL_STATUS_NOT_READY:
>> +       default:
>> +               rc = -ENXIO;
>> +               goto out;
>> +       }
>> +
>> +       /* check and see if security is enabled and locked */
>> +       if (nd_cmd.cmd.state & ND_INTEL_SEC_STATE_UNSUPPORTED)
>> +               *state = NVDIMM_SECURITY_UNSUPPORTED;
>> +       else if (nd_cmd.cmd.state & ND_INTEL_SEC_STATE_ENABLED) {
>> +               if (nd_cmd.cmd.state & ND_INTEL_SEC_STATE_LOCKED)
>> +                       *state = NVDIMM_SECURITY_LOCKED;
>> +               else
>> +                       *state = NVDIMM_SECURITY_UNLOCKED;
>> +       } else
>> +               *state = NVDIMM_SECURITY_DISABLED;
>> +
>> + out:
>> +       if (rc < 0)
>> +               *state = NVDIMM_SECURITY_INVALID;
>> +       return rc;
>> +}
>> +
>> +const struct nvdimm_security_ops intel_security_ops = {
>> +       .state = intel_dimm_security_state,
>> +       .unlock = intel_dimm_security_unlock,
>> +};
>> diff --git a/drivers/acpi/nfit/intel.h b/drivers/acpi/nfit/intel.h
>> index a6f8f4bc8ebb..f9ac718776ca 100644
>> --- a/drivers/acpi/nfit/intel.h
>> +++ b/drivers/acpi/nfit/intel.h
>> @@ -8,6 +8,8 @@
>>
>>  #ifdef CONFIG_X86
>>
>> +extern const struct nvdimm_security_ops intel_security_ops;
>> +
>>  #define ND_INTEL_STATUS_SIZE           4
>>  #define ND_INTEL_PASSPHRASE_SIZE       32
>>
>> @@ -64,4 +66,17 @@ struct nd_intel_query_overwrite {
>>  } __packed;
>>  #endif /* CONFIG_X86 */
>>
>> +static inline const struct nvdimm_security_ops *
>> +acpi_nfit_get_security_ops(int family)
>> +{
>> +       switch (family) {
>> +#ifdef CONFIG_X86
>> +       case NVDIMM_FAMILY_INTEL:
>> +               return &intel_security_ops;
>> +#endif
>> +       default:
>> +               return NULL;
>> +       }
>> +}
>> +
>>  #endif
>> diff --git a/drivers/nvdimm/Kconfig b/drivers/nvdimm/Kconfig
>> index 9d36473dc2a2..09fa6bba975d 100644
>> --- a/drivers/nvdimm/Kconfig
>> +++ b/drivers/nvdimm/Kconfig
>> @@ -112,4 +112,9 @@ config OF_PMEM
>>
>>           Select Y if unsure.
>>
>> +config NVDIMM_SECURITY
>> +       bool
>> +       depends on KEYS
>> +       default LIBNVDIMM
>> +
>>  endif
>> diff --git a/drivers/nvdimm/Makefile b/drivers/nvdimm/Makefile
>> index e8847045dac0..c476547348b1 100644
>> --- a/drivers/nvdimm/Makefile
>> +++ b/drivers/nvdimm/Makefile
>> @@ -27,3 +27,4 @@ libnvdimm-$(CONFIG_ND_CLAIM) += claim.o
>>  libnvdimm-$(CONFIG_BTT) += btt_devs.o
>>  libnvdimm-$(CONFIG_NVDIMM_PFN) += pfn_devs.o
>>  libnvdimm-$(CONFIG_NVDIMM_DAX) += dax_devs.o
>> +libnvdimm-$(CONFIG_NVDIMM_SECURITY) += dimm_devs_security.o
>> diff --git a/drivers/nvdimm/dimm.c b/drivers/nvdimm/dimm.c
>> index 6c8fb7590838..b6381ddbd6c1 100644
>> --- a/drivers/nvdimm/dimm.c
>> +++ b/drivers/nvdimm/dimm.c
>> @@ -51,6 +51,13 @@ static int nvdimm_probe(struct device *dev)
>>         get_device(dev);
>>         kref_init(&ndd->kref);
>>
>> +       nvdimm_security_get_state(dev);
>> +
>> +       /* unlock DIMM here before touch label */
>> +       rc = nvdimm_security_unlock_dimm(dev);
>> +       if (rc < 0)
>> +               dev_warn(dev, "failed to unlock dimm %s\n", dev_name(dev));
>> +
>>         /*
>>          * EACCES failures reading the namespace label-area-properties
>>          * are interpreted as the DIMM capacity being locked but the
>> diff --git a/drivers/nvdimm/dimm_devs.c b/drivers/nvdimm/dimm_devs.c
>> index 3f02008d75e7..024b1c4dfe88 100644
>> --- a/drivers/nvdimm/dimm_devs.c
>> +++ b/drivers/nvdimm/dimm_devs.c
>> @@ -18,6 +18,7 @@
>>  #include <linux/io.h>
>>  #include <linux/fs.h>
>>  #include <linux/mm.h>
>> +#include <linux/key.h>
>>  #include "nd-core.h"
>>  #include "label.h"
>>  #include "pmem.h"
>> @@ -212,7 +213,13 @@ void nvdimm_clear_locked(struct device *dev)
>>  static void nvdimm_release(struct device *dev)
>>  {
>>         struct nvdimm *nvdimm = to_nvdimm(dev);
>> +       struct key *key;
>>
>> +       mutex_lock(&nvdimm->key_mutex);
>> +       key = nvdimm_get_key(dev);
>> +       if (key)
>> +               key_put(key);
>> +       mutex_unlock(&nvdimm->key_mutex);
> 
> If the assumption is that the key is valid as long as it is attached
> to the nvdimm, can't we do this instead:
> 
>     mutex_lock(&nvdimm->key_mutex);
>     key_put(nvdimm->key);
>     nvdimm->key = NULL;
>     mutex_unlock(&nvdimm->key_mutex);

Yes, since key_put() already checks. We could have an invalid key if
security is not on.

> 
> key_put() already checks for key being NULL, and don't we need to
> disconnect the key from anyone else that might be doing a nvdimm->key
> usage right after we drop the lock?
> 

I don't think so. Whomever is using has to hold the mutex. So either
it's being in use and the release block until we can set the key to NULL
or we have set the key NULL before the user tries to acquire the key.
_______________________________________________
Linux-nvdimm mailing list
Linux-nvdimm@lists.01.org
https://lists.01.org/mailman/listinfo/linux-nvdimm