On 9 December 2019 18:03:11 GMT, James Bottomley <James.Bottomley@xxxxxxxxxxxxxxxxxxxxx> wrote:
>On Mon, 2019-12-09 at 10:18 +0000, David Woodhouse wrote:
>> On Sat, 2019-12-07 at 21:12 -0800, James Bottomley wrote:
>> > This commit adds the ability to specify a PCR lock policy to TPM2
>> > keys. There is a complexity in that the creator of the key must
>> > chose either to use a PCR lock policy or to use authentication. At
>> > the current time they can't use both due to a complexity with the
>> > way authentication works when policy registers are in use. The way
>> > to construct a pcrinfo statement for a key is simply to use the
>> > TPMS_PCR_SELECT structure to specify the PCRs and follow this by a
>> > hash of all their values in order of ascending PCR number.
>> >
>> > For simplicity, we require the policy name hash and the hash used
>> > for the PCRs to be the same. Thus to construct a policy around the
>> > value of the resettable PCR 16 using the sha1 bank, first reset the
>> > pcr to zero giving a hash of all zeros as:
>> >
>> > 6768033e216468247bd031a0a2d9876d79818f8f
>> >
>> > Then the TPMS_PCR_SELECT value for PCR 16 is
>> >
>> > 03000001
>> >
>> > So create a new 32 byte key with a policy policy locking the key to
>> > this value of PCR 16 with a parent key of 81000001 would be:
>> >
>> > keyctl new 32 keyhandle=0x81000001 hash=sha1
>> > pcrinfo=030000016768033e216468247bd031a0a2d9876d79818f8f" @u
>>
>> OK... but I've love to see a more formal definition of this binary
>> format, as part of the "standard" we allegedly have for the overall
>> ASN.1 representation.
>
>It's actually defined in the TPM2 command manual ... it's basically the
>policy commands you send to the TPM ordered so they can be directly
>hashed.
>
>However, I agree a standards definition would be good. This format
>doesn't support TPM2_PolicyOr directly (and the command manual is
>silent on how it should be supported), so that's going to have to be
>defined in the standard anyway.
>
>[...]
>> > +int tpm2_encode_policy(struct tpm2_policies *pols, u8 **data, u32
>> > *len)
>> > +{
>> > + u8 *buf = kmalloc(2 * PAGE_SIZE, GFP_KERNEL);
>> > + u8 *work = buf + PAGE_SIZE, *ptr;
>> > + int i;
>> > +
>> > + if (!buf)
>> > + return -ENOMEM;
>> > +
>> > + for (i = 0; i < pols->count; i++) {
>> > + u8 *seq_len, *tag_len;
>> > + u32 cmd = pols->code[i];
>> > + int l;
>> > +
>> > + /*
>> > + * cheat a bit here: we know a policy is < 128
>> > bytes,
>> > + * so the sequence and cons tags will only be two
>> > + * bytes long
>> > + */
>> > + *work++ = _tag(UNIV, CONS, SEQ);
>> > + seq_len = work++;
>> > + *work++ = _tagn(CONT, CONS, 0);
>> > + tag_len = work++;
>> > + asn1_encode_integer(&work, cmd);
>> > + *tag_len = work - tag_len - 1;
>> > + *work++ = _tagn(CONT, CONS, 1);
>> > + tag_len = work++;
>> > + asn1_encode_octet_string(&work, pols->policies[i],
>> > + pols->len[i]);
>> > + *tag_len = work - tag_len - 1;
>> > + l = work - seq_len - 1;
>> > + /* our assumption about policy length failed */
>> > + if (WARN(l > 127,
>> > + "policy is too long: %d but must be <
>> > 128", l)) {
>> > + kfree(buf);
>> > + return -EINVAL;
>> > + }
>> > + *seq_len = l;
>>
>>
>>
>> You're not even using your own sequence encoding here, because it
>> only works when you know the length in advance. How about setting
>> *seq_len to 0x80 to start with, for an indeterminate length.
>
>I already did that in the asn.1 patch, so I've updated this one to use
>it.
>
>> Then in the happy case where it is <128, just go back and fill it in
>> as you currently do. Otherwise append 0x00 0x00 as the end marker.
>
>That doesn't work ... the format of these octet strings is likely to
>have two zeros together, so they *have* to be definite length encoded.
The octet-strings sure, but we know the length of those. It was the sequence you have that <127 check and bail out for... wasn't it?
>> None of this has to be DER, does it?
>
>None of what? The policy? the DER format is already in use so we
>can't change it.
What we *output* doesn't need to be DER (mandatory definite length) and can be BER though, right?
>> <Insert more whining about PAGE_SIZE assumptions and buffer
>> overflows>
>
>OK, OK, I fixed that too.
>
>James
--
Sent from my Android device with K-9 Mail. Please excuse my brevity.
