linux-nvdimm.lists.01.org archive mirror
 help / color / mirror / Atom feed
From: Mimi Zohar <zohar@linux.ibm.com>
To: Dan Williams <dan.j.williams@intel.com>
Cc: Mimi Zohar <zohar@linux.vnet.ibm.com>,
	keyrings@vger.kernel.org,
	linux-nvdimm <linux-nvdimm@lists.01.org>
Subject: Re: [PATCH 01/11] keys-encrypted: add nvdimm key format type to encrypted keys
Date: Tue, 27 Nov 2018 14:35:02 -0500	[thread overview]
Message-ID: <1543347302.3902.202.camel@linux.ibm.com> (raw)
In-Reply-To: <CAPcyv4i5=tAx6=W44Vva7=0GO9A=j0w5vEEYMMv3NnC9_fMz6g@mail.gmail.com>

On Tue, 2018-11-27 at 11:10 -0800, Dan Williams wrote:
> On Tue, Nov 27, 2018 at 10:24 AM Mimi Zohar <zohar@linux.ibm.com> wrote:
> >
> > On Tue, 2018-11-27 at 09:20 -0700, Dave Jiang wrote:
> > >
> > > On 11/27/18 12:20 AM, Dan Williams wrote:
> > > > On Fri, Nov 9, 2018 at 2:13 PM Dave Jiang <dave.jiang@intel.com> wrote:
> > > >>
> > > >> Adding nvdimm key format type to encrypted keys in order to limit the size
> > > >
> > > > s/Adding/Add an/
> > > >
> > > >> of the key to 32-bytes.
> > > >>
> > > >> Signed-off-by: Dave Jiang <dave.jiang@intel.com>
> > > >> ---
> > > >>  security/keys/encrypted-keys/encrypted.c |   29 ++++++++++++++++++++---------
> > > >>  1 file changed, 20 insertions(+), 9 deletions(-)
> > > >>
> > > >> diff --git a/security/keys/encrypted-keys/encrypted.c b/security/keys/encrypted-keys/encrypted.c
> > > >> index d92cbf9687c3..182b4f136bdf 100644
> > > >> --- a/security/keys/encrypted-keys/encrypted.c
> > > >> +++ b/security/keys/encrypted-keys/encrypted.c
> > > >> @@ -45,6 +45,7 @@ static const char hmac_alg[] = "hmac(sha256)";
> > > >>  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_nvdimm[] = "nvdimm";
> > > >>  static unsigned int ivsize;
> > > >>  static int blksize;
> > > >>
> > > >> @@ -54,6 +55,7 @@ static int blksize;
> > > >>  #define HASH_SIZE SHA256_DIGEST_SIZE
> > > >>  #define MAX_DATA_SIZE 4096
> > > >>  #define MIN_DATA_SIZE  20
> > > >> +#define KEY_NVDIMM_PAYLOAD_LEN 32
> > > >>
> > > >>  static struct crypto_shash *hash_tfm;
> > > >>
> > > >> @@ -62,12 +64,13 @@ enum {
> > > >>  };
> > > >>
> > > >>  enum {
> > > >> -       Opt_error = -1, Opt_default, Opt_ecryptfs
> > > >> +       Opt_error = -1, Opt_default, Opt_ecryptfs, Opt_nvdimm
> > > >>  };
> > > >>
> > > >>  static const match_table_t key_format_tokens = {
> > > >>         {Opt_default, "default"},
> > > >>         {Opt_ecryptfs, "ecryptfs"},
> > > >> +       {Opt_nvdimm, "nvdimm"},
> > > >>         {Opt_error, NULL}
> > > >>  };
> > > >>
> > > >> @@ -195,6 +198,7 @@ static int datablob_parse(char *datablob, const char **format,
> > > >>         key_format = match_token(p, key_format_tokens, args);
> > > >>         switch (key_format) {
> > > >>         case Opt_ecryptfs:
> > > >> +       case Opt_nvdimm:
> > > >>         case Opt_default:
> > > >>                 *format = p;
> > > >>                 *master_desc = strsep(&datablob, " \t");
> > > >> @@ -625,15 +629,22 @@ static struct encrypted_key_payload *encrypted_key_alloc(struct key *key,
> > > >>         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_nvdimm)) {
> > > >> +                       if (decrypted_datalen != KEY_NVDIMM_PAYLOAD_LEN) {
> > > >> +                               pr_err("encrypted_key: nvdimm key payload incorrect length: %d\n",
> > > >> +                                               decrypted_datalen);
> > > >> +                               return ERR_PTR(-EINVAL);
> > > >> +                       }
> > > >
> > > > I suspect this may not be the last key type that gets added, but I
> > > > wonder if we should instead create key-types based on the dlen size.
> > > > I.e. create a generic 32-byte key-type "enc32"? That way if another
> > > > 32-byte requirement key comes along we don't need to come touch this
> > > > routine again.
> > > >
> > >
> > > I'm ok with that if Mimi is.
> >
> > If the usage (eg. format: ecryptfs, nvdimm) limits the dlen size(s),
> > how will defining generic key-types help?  If there are no usage size
> > limitations, then there would be no usage specific definition.
> >
> > I must be missing something.
> 
> ...or I did a poor job of describing the problem. I'm just looking
> ahead to another potential encrypted-key use case, but instead of
> nvdimms this would be to lock / unlock persistent memory "namespace"
> objects. If it turns out the namespace key is just another 32-byte
> encrypted key should the encrypted-keys core grow support for a new
> "namespace" key type, should it reuse the "nvdimm" key type, or should
> we define a generic 32-byte payload-size-requirement key type?

Even with a generic length, there will be format (eg. ecryptfs,
nvdimm, namespace) specific code.  Otherwise you wouldn't be defining
a new format type.

switch(dlen) {
	case enc32:
		if (ecryptfs) goto fail;
		...
		break;
	case enc64:
		if (nvdimm || ns) goto fail;
		...
		break;
	default:
}

Mimi


_______________________________________________
Linux-nvdimm mailing list
Linux-nvdimm@lists.01.org
https://lists.01.org/mailman/listinfo/linux-nvdimm

  reply	other threads:[~2018-11-27 19:35 UTC|newest]

Thread overview: 34+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-11-09 22:13 [PATCH 00/11] Additional patches for nvdimm security support Dave Jiang
2018-11-09 22:13 ` [PATCH 01/11] keys-encrypted: add nvdimm key format type to encrypted keys Dave Jiang
2018-11-27  7:20   ` Dan Williams
2018-11-27 16:20     ` Dave Jiang
2018-11-27 18:24       ` Mimi Zohar
2018-11-27 19:10         ` Dan Williams
2018-11-27 19:35           ` Mimi Zohar [this message]
2018-11-27 19:48             ` Dan Williams
2018-11-27 20:10               ` Mimi Zohar
2018-11-27 20:15                 ` Dave Jiang
2018-11-09 22:13 ` [PATCH 02/11] libnvdimm/security: change clear text nvdimm keys " Dave Jiang
2018-11-10  1:45   ` Dan Williams
2018-11-11 17:27   ` Mimi Zohar
2018-11-11 19:20     ` Dan Williams
2018-11-11 20:09       ` Mimi Zohar
2018-11-12 15:42         ` Dave Jiang
2018-11-12 18:49           ` Mimi Zohar
2018-11-12 20:13             ` Dave Jiang
2018-11-12 15:45     ` Dave Jiang
2018-11-12 19:04       ` Mimi Zohar
2018-11-09 22:14 ` [PATCH 03/11] libnvdimm/security: add override module param for key self verification Dave Jiang
2018-11-09 22:14 ` [PATCH 04/11] libnvdimm/security: introduce NDD_SECURITY_BUSY flag Dave Jiang
2018-11-09 22:14 ` [PATCH 05/11] acpi/nfit, libnvdimm/security: Add security DSM overwrite support Dave Jiang
2018-11-09 22:14 ` [PATCH 06/11] tools/testing/nvdimm: Add overwrite support for nfit_test Dave Jiang
2018-11-09 22:14 ` [PATCH 07/11] libnvdimm/security: add overwrite status notification Dave Jiang
2018-11-10  2:59   ` Elliott, Robert (Persistent Memory)
2018-11-12 20:26     ` Dave Jiang
2018-11-09 22:14 ` [PATCH 08/11] libnvdimm/security: add documentation for ovewrite Dave Jiang
2018-11-09 22:14 ` [PATCH 09/11] acpi/nfit, libnvdimm/security: add Intel DSM 1.8 master passphrase support Dave Jiang
2018-11-25 21:24   ` Dan Williams
2018-11-09 22:14 ` [PATCH 10/11] tools/testing/nvdimm: add Intel DSM 1.8 support for nfit_test Dave Jiang
2018-11-10  3:15   ` Elliott, Robert (Persistent Memory)
2018-11-12 20:27     ` Dave Jiang
2018-11-09 22:14 ` [PATCH 11/11] acpi/nfit: prevent indiscriminate DSM payload dumping for security DSMs Dave Jiang

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1543347302.3902.202.camel@linux.ibm.com \
    --to=zohar@linux.ibm.com \
    --cc=dan.j.williams@intel.com \
    --cc=keyrings@vger.kernel.org \
    --cc=linux-nvdimm@lists.01.org \
    --cc=zohar@linux.vnet.ibm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).