Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Recommendations for Auditable Electronic Codex32 Implementations #57

Open
BenWestgate opened this issue Jul 28, 2023 · 107 comments · May be fixed by #59
Open

Recommendations for Auditable Electronic Codex32 Implementations #57

BenWestgate opened this issue Jul 28, 2023 · 107 comments · May be fixed by #59

Comments

@BenWestgate
Copy link
Contributor

BenWestgate commented Jul 28, 2023

While intended for paper computers, some benefits of codex32 can be maximized through electronic implementations. To enhance this aspect, I propose recommendations for creating auditable codex32 share sets when using electronics.

Inspired by the concept of auditable bitcoin wallets, I have developed an implementation of For a fresh master seed, which ensures full determinism in generating the codex32 share set.

I believe that deterministically choosing every aspect of the codex32 share set enhances protection against potential malicious tampering of shares by detecting any deviation from the determinism.

In issue #54, we discussed that the hash160(public_masterkey) from descriptors serves as a useful deterministic data to use as the default identifier. However, we need to address another aspect that could potentially leak information – the share indices and payloads of the k - 1 independent shares.

Currently, the BIP recommends selecting the next available letter from the bech32 alphabet, in alphabetical order, as share indices:

Take the next available letter from the bech32 alphabet, in alphabetical order, as a, c, d, ..., to be the share index

This approach may inadvertently reduce privacy, as finding share f, g, or h would reveal a backup of at least 5, 6, or 7 shares, respectively. To improve this, I propose an update to the BIP where, during hand creation, dice are rolled to select each new share index. This method is akin to how indexes are created in software like libgfshare-bin.

For electronic implementations, share indices should be determined by the codex32 secret alone. Below is an example of how I chose the share indices in Python:

import random
import hashlib
import hmac
salt=threshold+identifier+str(len(master_seed))
random.seed(a=hmac.digest(master_seed, salt, 'sha512'))
share_index = random.sample(indices_free, n)

The master_seed, threshold, identifier, and seed_length are the only variables in a codex32 secret, so I HMAC these together to seed the selection of share indices. Although a hash of the codex32 secret string would also work.

I think this is the right track but I'm looking to standardize recommendations for electronic implementations. If there's a better way please say so.

Below is an example of how I generated the share payloads in Python:

share_entropy = hashlib.scrypt(password=master_seed, salt=salt, n=2**20, r=8, p=1, maxmem=1025**3, dklen=(int(k)-1)*(seed_length+1))

While this a start, it may not be suitable as a standard recommendation due to scrypt parameters that won't run on hardware wallets. To address this, I am open to exploring alternative variable length secure hash functions. Your input on this matter would be appreciated.

Next, this stretched entropy is divided evenly into k - 1 share payloads. Additionally, an extra byte is allocated for filling the padding bits on the 26th or 52th characters.

Subsequently, the remaining n + 1 - k shares are derived according to the BIP:

    for x in range(int(k) - 1):
        data = list(share_entropy[x*(seed_length+1):x*(seed_length+1)+seed_length+1])
        share_list += [encode("ms", seed_length, k, identifier, share_index[x], data)]
    for x in range(int(k) - 1, n):
        derived_share_list += [derive_new_share(share_list,share_index[x])]

Lastly, if a new set of shares is created for an existing master_seed with the same threshold, I propose incrementing the last character of the identifier by 1, which will cause a completely new set of indicies and share payloads to be chosen.

I welcome any feedback or suggestions for improving the approach. Let's establish standardized and auditable recommendations for electronic codex32 implementations. Thank you for your collaboration.

@apoelstra
Copy link
Contributor

To improve this, I propose an update to the BIP where, during hand creation, dice are rolled to select each new share index.

This would actually complicate things a fair bit -- generating shares in order allows us to have a fixed set of tables that users can use to derive additional shares. If you were generating shares in random order you would need to do something akin to the recovery process -- and for each share index we'd need a "recovery wheel" for that index.

I wouldn't mind releasing a set of "recovery wheels" and suggesting people do this as an additional privacy measure (we've wanted to do this anyway as a "recover the remaining shares from an incomplete set" process), but I think it's too onerous to suggest doing as the default. Especially as the k/n parameters are not something that we think of as secret data. k is even encoded in the share header.

But I agree that this generation mode be the default for electronic implementations.

Below is an example of how I generated the share payloads in Python:

Because you're starting with the master seed, I don't think there's any value in hardening this. If the seed itself is of low-quality randomness then the user is screwed anyway. So I would actually suggest going in the other direction, toward chacha20 rather than scrypt, in the hope of coming up with something computationally simple enough that it could be (theoretically) audited by hand.

Similarly, with the share indices I don't even think we need cryptographic randomness -- the goal is simply to break the implication "share G implies that shares B,C,D,E,F exist". Though of course, if you are using the master seed as entropy, you want to run that through a cryptographic RNG to make sure that no information leaks into the indices themselves. I also don't think the share indices need to be auditable, though there's no reason for them not to be so we might as well.

BTW, I've been holding off on PR'ing to the BIP repo because it seems like you (and us) are still iterating on this, but I definitely do intend to incorporate your suggestions once we've nailed everything down. Thanks a ton for digging into this!

@apoelstra
Copy link
Contributor

@BenWestgate also, if you are on IRC, #volvelle-wizards is our discussion channel for this stuff.

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Jul 28, 2023

Thank you for your response I will make changes to my implementation to make it closer to a standard recommendation.

I will get rid of the KDF and use ChaCha20 for filling the secret payload padding, the share payloads and shuffling the indicies so auditing electronic codex32 share sets by hand is possible.

Share indices MUST be derived by a cryptographically secure randomness if not fixed, and we agree they should shouldn't be fixed by default in electronic implementations for the small "n" privacy gain.

By my math if the user made 31 shares and numbered them 1 to 31 it can leak a surprising amount of data: 87 bits.
So even with encrypted payloads, the seed would be effectively leaked by a malicious non-deterministic implementation with enough indices known. Although maybe 2nd level SSS encryption should encrypt the plaintext share's threshold, identifier, and indices for reasons like this. #16

Without encryption the max threshold is 9, so the maximum data leak we're concerned with is from 8 indices
log2(31*30*29*28*27*26*25*24) = 39 -bits leaked

Leaving only 89-bits of security for the master_seed, which is probably guessable by TLAs in our lifetime if not already.

Regarding the algorithm for the share index selection, using python random's random.seed() and then random.sample() is probably not hand auditable due to the Mersenne Twister.

Instead I could get at least 87-bits from CHACHA20, convert to bech32, and take the first unique character to be the next share index.

But how to know how much data to produce from the hash so I don't run out? Is there a better way to deterministically "shuffle" a list by hand computations?

  1. Use Chacha20 with a secret key (master_seed) to generate a keystream.
  2. Take the first byte of the keystream (or any desired number of bits, depending on the number of elements in the list) and use it as an index to select the first symbol in the shuffled list, unless the value is "s" or a previously selected character.
    3.Use the next byte of the keystream (or any desired number of bits) to use as an index to select the next symbol for the shuffled list
  3. Increment the counter for Chacha20 by 1 if out of bytes.
  4. Repeat steps 2 to 4 until all symbols are selected, creating the shuffled list.

Perhaps the chacha20 hash mod 31, mod 30, ... etc
What is the right balance between length of secure hash output needed and math operations to derive each index indices that's most efficient for hand auditing and also doesn't introduce modulo bias?

The share payloads are much easier in this regard as they allow repeating characters and use the full 5 bits per character so the chacha20 output stream can be directly used.

I joined the IRC channel as benwestgate.

@BenWestgate
Copy link
Contributor Author

Because you're starting with the master seed, I don't think there's any value in hardening this. If the seed itself is of low-quality randomness then the user is screwed anyway. So I would actually suggest going in the other direction, toward chacha20 rather than scrypt, in the hope of coming up with something computationally simple enough that it could be (theoretically) audited by hand.

from Cryptodome.Cipher import ChaCha20

padding _length = 32 - len(master_seed)

key = master_seed + bytes(padding_length)

nonce = bytes(hrp+k+identifier+"s", 'utf')

cipher = ChaCha20.new(key=key, nonce=nonce)

cipher_stream = cipher.encrypt(bytes(8))

Does this look good as the RNG stream for setting the padding bits in the last symbol of the codex32 secret payload, the share indicies and k-1 share payloads?

The nonce can be incremented by 1 whenever a 512-bit block is exhausted.

Do you prefer zero padding the decoded master_seed bytes as shown above or directly utf-8 encoding the 32 bech32 characters after "s", ie key = bytes(codex32_secret[9:41], 'utf-8')?

Unsure which is easier by hand.

Sad to see double_sha256 go as the padding as 256-bit master_seed's were direct truncation of WIF serialization. But that novelty is a liability for auditing by hand.

Likewise, it feels wrong to set the secret payload padding bits with a RNG stream that depends on the header, causing the payload for identical master_seed's to vary if the header changes.

Yet it seems smart to conserve labor for hand auditors by avoiding computing a fixed nonce ChaCha20 block only to set 2-4 padding bits then discard. These bits should be the first data extracted from the stream as a fast first check for malicious tampering. The next should be the share payloads, as these are also recoverable error detecting and auditable data, the last should be the indices as the order the software provided them is meant to be lost during recovery and so they cannot detect tampering.

@apoelstra
Copy link
Contributor

Share indices MUST be derived by a cryptographically secure randomness if not fixed, and we agree they should shouldn't be fixed by default in electronic implementations for the small "n" privacy gain.

By my math if the user made 31 shares and numbered them 1 to 31 it can leak a surprising amount of data: 87 bits.

Heh, yeah, this seems reasonable. It seems pretty implausible that the user would number their shares 1 to 31 like that but why have a leak when we don't need to. If users want to do this by hand they can use dice rolls, which are cryptographically secure.

Regarding the algorithm for the share index selection, using python random's random.seed() and then random.sample() is probably not hand auditable due to the Mersenne Twister.

Hmm, true. And I see now the value of auditability, since in theory each share index could be leaking 5 bits (to an attacker who knows the order that the shares was generated anyway).

But how to know how much data to produce from the hash so I don't run out? Is there a better way to deterministically "shuffle" a list by hand computations?

Knuth's "Algorithm P" can do this. (The question is about finding a uniform subset, but conveniently this algorithm finds a subset in a uniform order. If you want a sorted order, which is just as good from a security POV, you can use "Algorithm S" or just sort manually.) Howevery, algorithm P assumes you can uniformly sample from the range [i, 31] for each i, which is hard to do without a theoretically indefinite stream of randomness. Though if you use rejection sampling, the probabilities are at least much easier to compute.

Let me think about this.

Does this look good as the RNG stream for setting the padding bits in the last symbol of the codex32 secret payload, the share indicies and k-1 share payloads?

The RNG initialization looks good to me, although

  • I guess identifier comes out of a different RNG?.
  • I also don't know what encrypt() does; I assume it xors the input with the random stream.
  • It's a cool idea to set the nonce uniquely per seed; I'd have probably just left it at all-zeroes...see below comments.

Also, I wouldn't bother setting the padding bits randomly. You can just leave them as 0. They aren't used and do not feed into BIP32 derivations.

The nonce can be incremented by 1 whenever a 512-bit block is exhausted.

Yep, though this is a tiny bit subtle. If you were to produce 256 blocks then you'd be incrementing the identifier part of the nonce, and would produce the same output as you would've if you'd simply started with that identifier.

But since we're only producing 31 shares, each of which only takes half a block, this is fine.

Do you prefer zero padding the decoded master_seed bytes as shown above or directly utf-8 encoding the 32 bech32 characters after "s", ie key = bytes(codex32_secret[9:41], 'utf-8')?

Unsure which is easier by hand.

Directly using the bytes is definitely easier, and I think is also more elegant.

Sad to see double_sha256 go as the padding as 256-bit master_seed's were direct truncation of WIF serialization. But that novelty is a liability for auditing by hand.

Yeah, that is a cool property. Though TBH I don't think users frequently see WIF serialization of xprivs so it's something of a curiosity.

Likewise, it feels wrong to set the secret payload padding bits with a RNG stream that depends on the header, causing the payload for identical master_seed's to vary if the header changes.

Yeah, hmm. There's two ways you could think of this:

  • Changing the threshold or identifier is a "reset the world" change and we should put this stuff into the RNG. This prevents users from re-sharing the same secret in different contexts while producing the same shares, which is good because the different share sets would interact algebraically in possibly surprising ways.
  • Changing the threshold or identifier is a "relabelling" which we ought to support. I actually considered producing a worksheet for this because in the early days I found myself relabelling stuff a lot (to get encrypted shares with distinct but human-readable headers).

The threshold value should definitely go into the RNG. Under no circumstances should you "relabel" the threshold, unless you are doing something dangerous and way off-script in which case the spec doesn't matter anyway. After thinking about it for a bit, I also think we should do this with the identifier. It creates a nice workflow where "if you want to re-split the secret, you have to change the identifier" and conversely "if the identifiers are the same (and the they come from the same seed), they're compatible shares".

Re-sharing secrets is a safe thing to do as long as the initial shares are independently random, and there are lots of reasons that users might want to do it.

@BenWestgate
Copy link
Contributor Author

I guess identifier comes out of a different RNG?.

#54 (comment)

identifier is either selected by the user or defaults to the hash160(master_pubkey), making it possible to tell which share set recovers spending a watch only wallet, as it is prepended to child xpubs in descriptors. Of various options this one seems most useful for these 4 characters.

I also don't know what encrypt() does; I assume it xors the input with the random stream.

That's exactly what it does, the stream is being XOR'd with zeros in order to print the cipher stream directly.

It's a cool idea to set the nonce uniquely per seed; I'd have probably just left it at all-zeroes...see below comments.

It is an essential feature.

Having the nonce be a function of the header allows for creating new backups for the same master_seed that don't reuse the original randomness. For example if a share or two was compromised, the old set could be destroyed and replaced with a share set using a new header (new randomness) to restore security without sweeping funds. (Whether it's advisable to "rotate" shares given an attacker may reach a threshold before you've Completely Destroyed the partially-compromised share set is another question.)

If the nonce was fixed to bytes(8), you'd get the same independent shares for the same master_seed every time, even if you changed the threshold or identifier. This even helps with access control, if you want one group to be able to recover at threshold 3 and another group to recover at threshold 2, make two share sets with each threshold and one set will NOT leak payload data of the other set.

@apoelstra
Copy link
Contributor

It is an essential feature.

This is a great argument. Agreed on all counts.

@BenWestgate
Copy link
Contributor Author

Also, I wouldn't bother setting the padding bits randomly. You can just leave them as 0. They aren't used and do not feed into BIP32 derivations.

I dislike wasted space, especially in handwriting. But I agree padding should not be set randomly. Let's see how to make those bits do some work:

  1. These bits are as secret as the secret, it's a waste to set them with a non-linear secure hash function. double_sha256(seed) etc with no efficient erasure correction ability.
  2. These bits are as secret as the secret, it's a waste to set them to some public data like the hash160(master_pubkey) as you can't see them until you can already compute the pubkey, this is just another form of 1 with extra wasted work.
  3. But they could be parity bits or linear binary error coding. These algorithms are pretty fast to do on the master_seed bytes by hand.

Do you have any suggestions?

Naive idea is upper half and lower half parity, which has nice property of keeping the same 64-bit chunk length for 128/256-bit seeds. But falls apart at 512-bit seeds which have 3-bits of padding (170 bit chunks?)
This is not the most efficient error coding. Proper ECC should correct contiguous bit erasures anywhere in the master_seed that are the length of the checksum and is what should be used. This offers a little extra help in the absolute worst case scenario, where you have threshold shares but some are invalid beyond correction by the codex32 BCH code but are less than a symbol away from getting the right correction, potentially making brute forcing your valid codex32 secret 4 to 16 times cheaper.

What is the best ECC to put here? Will it wastefully repeat the same info as the codex32 secret checksum or can it supplement it?

Agreed, regarding re-sharing the same secret:

The BIP already says only shares with a matching identifier can be combined. Allowing "relabeling" is not very useful because it changes the checksum, even if share payloads do not change. And then you'd create a situation that the BIP says is not supported: two shares of different identifiers can be combined, but only if you ignore their checksum and identifier and just combine payloads. It's neat to just increment the LSB of the identifier to re-share an existing master_seed, then it still retains 3 symbols of the master_pubkey fingerprint, while allowing 32 re-shares. And setting a custom identifier to reshare should also be supported as long as it's unique from all previously used identifiers at that threshold.

On this topic of identifier uniqueness for re-sharing, it's not clear a user would know all previously used identifiers besides the one they have just combined with to do a reshare. Especially if they are not destroying old sets, they may combine with an older share set, and if we naively increment that, we will re-use the same nonce! So I think the default for "re-shares" needs to generate some independent randomness when users don't know all outstanding identifiers for this seed and threshold. Then we have a birthday problem and just pray for no collisions... The alternative is to not support useful cases:

Can you think of any reason to have 2 outstanding share sets of the same threshold but different identifiers? I did: prevent one set from combining with another set to defend against custodian collusion.

I might make a 2-of-3 with my dad and executor and a 2-of-3 with my mom and executor, each with unique identifier of same master_seed, and now recovering the master_seed requires me or my executor and 1 parent, but both my parents cannot combine shares to recover without me or my executor. Making them less attractive targets for a physical attack than if this was a 2-of-4 between my parents, myself and my executor. Useful.

How much randomness do we have to inject into the identifier if we don't know all the old identifiers to be pretty safe we don't reuse one? Can we still keep some bech32 encoding of the public key fingerprint, to identify the underlying wallet master_seed and connection between unique share sets of it or must inject all the entropy we can to be safe?

@apoelstra
Copy link
Contributor

I dislike wasted space, especially in handwriting. But I agree padding should not be set randomly. Let's see how to make those bits do some work:

This is an interesting direction. We can definitely stick a checksum in here and we can definitely make it "independent" of the existing checksum in the sense that it will meaningfully add error correction. But it will be a bit-level error correcting code which has some unintuitive properties when applied to data that's normally encoded as 5-bit characters.

In particular I think a 2-bit code will only be able to detect one error (I guess, two, if they're consecutive), so it would mean we could detect certain character errors but not others. This is not zero value but it's fairly limited. I think we should keep thinking whether there's a better use of 2 bits before committing to this.

Allowing "relabeling" is not very useful because it changes the checksum, even if share payloads do not change.

It does, but you can compute what the checksum will change to, far more quickly than recomputing the whole thing. You can even use a computer to do most of the work without exposing any secret data to the computer.

On this topic of identifier uniqueness for re-sharing, it's not clear a user would know all previously used identifiers besides the one they have just combined with to do a reshare. Especially if they are not destroying old sets, they may combine with an older share set, and if we naively increment that, we will re-use the same nonce! So I think the default for "re-shares" needs to generate some independent randomness when users don't know all outstanding identifiers for this seed and threshold. Then we have a birthday problem and just pray for no collisions... The alternative is to not support useful cases:

Agreed. Unfortunately I think we just need to live with the birthday problem. In practice I think it'd be okay, both because the chance of collisions is fairly low (you need close to 1024 re-shares before the probabilites get high, and like, what are you doing with your secrets then..) and because the failure mode is that you'll generate exactly the same share set as before. Which is not much worse than having the original set still out there.

Can you think of any reason to have 2 outstanding share sets of the same threshold but different identifiers? I did: prevent one set from combining with another set to defend against custodian collusion.

Yeah, this is sorta what I'm imagining. That, or having "backup shares" that might have the same/overlapping custodian sets, but nonetheless shouldn't be mixable. Unfortunately these are the exact cases where collisions would undermine your security model.

How much randomness do we have to inject into the identifier if we don't know all the old identifiers to be pretty safe we don't reuse one? Can we still keep some bech32 encoding of the public key fingerprint, to identify the underlying wallet master_seed and connection between unique share sets of it or must inject all the entropy we can to be safe?

The birthday paradox shows up at around sqrt(N). So if you used the entire 20-bit identifier you'd expect collisions around the time you had 10 outstanding bits (1024). The curve is pretty steep so you have very low probabilities for quite a while, so it's fine to use 1000 as a rough estimate. If you used only half the identifier, 10 bits, you'd expect collisions after around 30 ... and have nontrivial probability of collisions after only a few.

My feeling is that you'd need to use the entire identifier for this.

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 1, 2023

I think using the entire identifier is fine. This is a sub-case of the support custom identifier case, where we MUST randomize it when the user and software can't recall all previously generated share set identifiers for this master_seed.

Otherwise implementations can default to suggesting to increment it, preserving most of the hash160(master_pubkey) fingerprint or as always the user can pick anything unique: "faml" "moma" "dada" "lawr" "acct" "aunt" etc. Wallets should store the identifiers they have created to prevent the user from repetition or at least warn in event it is deliberate to repair or extend an existing share set.

Since this "I forgot" case requires new randomness that cannot be derived from master_seed the following procedure from auditable bitcoin wallets is needed so that the new random identifier can't leak seed data or deliberately repeat an identifier to break security:

identifier = convertbits(hashlib.scrypt(password=user_data, salt=app_entropy, n=2**20, r=8, p=1, maxmem=1025**3, dklen=3), 8, 5)[:4]
  • These parameters will need to be reduced to support HWWs and hand auditing (see below)

Where the user provides randomness and the machine generates 20-bits and unconditionally displays it and they are mixed with a one-way compression function.

Process of creating the new identifier should be functionally equivalent to the following steps:

  1. Wallet generates a pseudo-random number (20 bits or more) that we call app entropy.
  2. Wallet unconditionally displays app entropy to the user (in bech32, hex etc).
  3. User is then asked for some data to be entered that allows to improve security in case app entropy generator has a backdoor (see Dual EC DRBG). This could be a one-time text entered by the user only for this purpose, or a passphrase containing enough entropy that is used later for other purposes.
  4. Wallet uses computationally expensive KDF to deterministically mix app entropy and user-entered text to produce the final identifier.

Important considerations:

  • App entropy must be displayed unconditionally so that wallet cannot know if user is going to audit it or not.
  • User-entered data should be visible and reproducible by the user to be able to arrive to the same data in an auditing software. Good: text or pattern of button presses. Bad: accelerometer data, mouse movements (hard to record and reproduce).
  • User-entered data should contain enough entropy and should be unique. If the wallet has a backdoored RNG which does not generate unique numbers, we're relying on user-entered data to avoid nonce reuse and if the wallet is trying to leak the master_seed we're relying on user entropy for secure key mixing.
  • Unless necessary, wallet should not rely solely on user input to generate identifier because they may inadvertently give a non-unique input. In such case, honest wallet with correct RNG implementation would with strong probability generate a unique nonce.
  • Process of mixing app entropy and user input must use computationally expensive KDF to compensate for non-uniqueness OR low-entropy in user input.

On second thought, it may not need to be an expensive KDF, but it may need to be a one way compression function. HMAC-SHA256 or our ChaCha20 hash. With the user input as key and app_entropy as nonce. I believe simple XOR won't provide protection from master_seed leakage when users enter less than 20-bits of entropy, which is likely when limited to 4 bech32 characters.

The worst case is the attacker predicts the user's input they can recover 20-bits of leaked master_seed. My KDF above costs $0.0002 to check all possible 20-bits with known user data. So there's no way to secure this other than that the user should be able to type more than 4 characters to maximize their chances of a unique string w/ 20-bits entropy.

from Cryptodome.Cipher import ChaCha20
from Cryptodome.Random import get_random_bytes
from segwit_addr import convertbits

# user data is limited to 32 bytes
padding _length = 32 - len(user_data)
key = user_data + bytes(padding_length)

nonce = get_random_bytes(8)
cipher = ChaCha20.new(key=key, nonce=nonce)

identifier = convertbits(cipher.encrypt(bytes(3)),8,5)[:4]

For auditing by hand, provide 8 de-biased dice rolls to select the custom identifier avoids doing a chacha20 hash just for this.

@apoelstra
Copy link
Contributor

Heh, your comment inspired me to actually read dual-EC-DBRG. It's quite elegant, other than the obvious backdoor, and also because it's obscenely slow. My read of it is that backdoor allows an attacker who has seen 30 (aligned) bytes of output can learn the internal state of the RNG, with a little bit (16 bits) of grinding. Once they know the internal state they can predict every future random number. However they cannot learn previous states of the RNG, and in particular there is no way for them to learn the seed, assuming the discrete log problem is hard.

So for this particular usecase, dual-ec-drbg would actually be perfectly fine :P.

But regarding scrypt parameters, let's back up a bit. I spoke on the phone with @roconnor-blockstream and we briefly touched on this. One observation he made is that any considerations applied to the identifier also applies to the shares (we should assume our attacker has at least one share; arguably we should assume they have k-1 shares). So I think we should do the same thing in both cases, and in this case the 20-bit ID is the least of our problem next to the key-length-many-bits shares.

Our attack scenario is: an attacker has access to the 20-bit ID, a share, multiple shares, whatever. And they are trying to access the actual master secret. The scenarios are:

  • If the data was generated maliciously then the seed could just be leaked outright.
  • If the data was generated correctly but with a fast RNG, then an attacker can grind out the secret far faster (by a large constant factor) by re-generating the data than by deriving addresses.
  • If the data is generated by a slow RNG/KDF, then an attacker with the data has no advantage over one without.

Having auditability forces us into the middle choice. If we don't worry about auditability then we'll be in either the "all good, no attacker advantage to having share data" case or the "all bad, the seed was totally leaked" case. And we can bias toward "all good" with the usual techniques. Open source software, independent audits, etc.

Given this, my feeling is that if somebody cares enough to audit their random data by hand, they care enough to just generate all their data by hand (which would actually be much faster than auditing, no matter what RNG we chose). And if they don't care enough, they'd be better off if we used scrypt everywhere rather than chacha everywhere.

@BenWestgate
Copy link
Contributor Author

Auditing a wallet always requires a computer to verify the public keys. So hand auditable share sets are not very important considering as you say the share set could be generated by hand faster for the paranoid.

That said, it makes sense for a given codex32 secret to always produce the same shares and share index order on electronics, but it is fine to require electronics to compute the shares with an expensive KDF.

I initially used Scrypt because it was built in hashlib, but the current state-of-the-art for memory hard KDFs is Argon2id. Its main advantage is being able to set the time cost and memory cost independently, while with Scrypt the time cost depends on memory cost.

@apoelstra
Copy link
Contributor

Agreed. I don't have much of an opinion between scrypt and argon2id. I'd be tempted to use scrypt because it's more widely supported and has been tested pretty heavily by the various cryptocurrencies using it a PoW. I'm also not sure there's a ton of value, for us, in being able to set time and memory separately. (And we could emulate that somewhat by just iterating scrypt.)

@BenWestgate
Copy link
Contributor Author

  • If the data was generated correctly but with a fast RNG, then an attacker can grind out the secret far faster (by a large constant factor) by re-generating the [share] data than by deriving addresses.

  • If the data is generated by a slow RNG/KDF, then an attacker with the data has no advantage over one without.

This means the KDF hardness parameters only need to cost more than deriving an address? The cheapest "address" is the master pubkey hash we suggest to use as default identifier.

Cost to beat is: HMAC-SHA512, produce a pubkey, then ripemd160(sha256(master_pubkey))

That has 1 in 2^20 chance of a false positive. (Or it's faked) In that case, a known address must be derived.

Setting time cost and memory cost an order of magnitude (or more) higher than an address derivation should be safe and able to run on HWWs

@apoelstra
Copy link
Contributor

This means the KDF hardness parameters only need to cost more than deriving an address? The cheapest "address" is the master pubkey hash we suggest to use as default identifier.

Yeah, that's my thinking.

Setting time cost and memory cost an order of magnitude (or more) higher than an address derivation should be safe and able to run on HWWs

+1

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 2, 2023

I did some research with the chat robot and it claims HWWs are using "a few kB" of memory for address derivation from a BIP32 seed and most popular HWWs could use as much as 64-512kB for KDF memory cost parameters.

This makes memory hard algorithms on HWWs useless against GPUs: 8GB / 0.5MB > # GPU cores, but may still help somewhat against ASICs.

PBKDF2 in BIP39 is especially weak against GPUs (needs 8+ random word passphrases to be secure).

Given this, creating a secure auditable master_seed may want to use different parameters on mobile and desktop wallets (1+ GB) than hardware wallets (512kB). Here's an example from my implementation that costs 1GB and a few seconds:

def fresh_master_seed(seed_length,user_entropy,bitcoin_core_entropy):
    """Derive a fresh master seed with seed length bytes from user entropy and bitcoin core's entropy aka app entropy."""
    if 16 > seed_length > 64:
        return None
    master_seed = hashlib.scrypt(password=bytes(user_entropy+str(seed_length),"utf"), salt=bytes(bitcoin_core_entropy,"utf"), n=2**20, r=8, p=1, maxmem=1025**3, dklen=seed_length)
    return master_seed

A time_cost of 5-10 seconds may be recommended to reduce the disadvantages of weak hardware. Unlike authentication, where user must wait time_cost each login so is limited to 1-2 seconds or less, this is a one time wait at master_seed generation.

As long as the parameters were something unconditionally displayed like App Entropy anyone can verify they get the same codex32 secret by computing the KDF on a trusted device.

For share set generation, which only needs to cost more than an address derivation which I'm told takes "a few seconds or less" for a HWW then targeting something that takes 10 seconds for the worst hardware we want to support and uses 64kB or more on should be 2-10x as time costly and 10-100x as memory costly as address derivation.

Here all wallets should use the same parameters so codex32 secrets always derive identical codex32 share sets. Much like how 2048 rounds is part of bip39.

@apoelstra
Copy link
Contributor

Yeah, "a few KB" for address generation sounds about right. Memory-wise it's extremely cheap and therefore extremely parallelizable. There's also a time-memory tradeoff. I think I could probably do it with <100 bytes of RAM but it wouldn't be very fast.

Given this, creating a secure auditable master_seed may want to use different parameters on mobile and desktop wallets (1+ GB) than hardware wallets (512kB). Here's an example from my implementation that costs 1GB and a few seconds:

Is the assumption here that the user entropy or Core entropy might be weak? I think if together they exceed 128 bits of entropy then there isn't any need for hardening at all. Or am I confused?

For share set generation, which only needs to cost more than an address derivation which I'm told takes "a few seconds or less" for a HWW then targeting something that takes 10 seconds for the worst hardware we want to support and uses 64kB or more on should be 2-10x as time costly and 10-100x as memory costly as address derivation.

Yep, sounds good to me.

@BenWestgate
Copy link
Contributor Author

Is the assumption here that the user entropy or Core entropy might be weak? I think if together they exceed 128 bits of entropy then there isn't any need for hardening at all. Or am I confused?

It's to protect against RNG backdoors in "secure element" HWWs.

Neither the device or user should be trusted, so the hardening makes the most of what entropy the user is willing to provide to protect themselves. Increasing the cost tremendously vs SHA2 for the backdoor author to steal.

This is the "lite" effort version of the dice debiasing worksheet. KDF offers real protection even without achieving 128-bits user entropy, even half that is act of congress expensive:

Attack cost estimates of memory hard KDFs by entropy

65-bits would be on order of 10 billion USD with 1GB and 1 second argon2id parameters.

50-bits would be around $1 million. It's less profitable to backdoor the RNG if user can mash their keyboard for a minute or toss a dice 10 times and verifiably make this attack cost more than they'll ever store.

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 3, 2023

In particular I think a 2-bit code will only be able to detect one error (I guess, two, if they're consecutive), so it would mean we could detect certain character errors but not others. This is not zero value but it's fairly limited. I think we should keep thinking whether there's a better use of 2 bits before committing to this.

From BIP-173:

...ordering is chosen to minimize the number of pairs of similar characters that differ in more than 1 bit.

The most commonly substituted characters are all 1-bit errors.

For 256-bit seeds w/ 4-bit ECC as padding, the total error correction becomes 4 substitutions plus 1 substitution if it is a 1-bit error.
For 128-bit seeds w/ 2-bit ECC as padding, there's no corrections but it can narrow down a 1-bit error substitution to the upper or lower half of the code, maybe better.
512-bit seeds also gain a 1-bit substitution error correction boost.

If the same 2 or 4 bit code is added to each independent share then do all the derived shares get valid ecc padding as well?

@apoelstra
Copy link
Contributor

It's to protect against RNG backdoors in "secure element" HWWs.

OK, yeah, this sounds good to me. Though of course, verifying this is done correctly will require redundant hardware. But better to have the ability than to not.

If the same 2 or 4 bit code is added to each independent share then do all the derived shares get valid ecc padding as well?

I am 90% sure, yes

For 256-bit seeds w/ 4-bit ECC as padding, the total error correction becomes 4 substitutions plus 1 substitution if it is a 1-bit error.

Similarly, I am 90% sure this is right ... the two codes don't directly add so that you can do some combined error correction algorithm (as far as I'm aware). But I do think that you can do a form of soft error correction where you can find all the possible strings that are 5 errors away from the target string. There will be several but only one, I think, with pretty-good probability, will have the extra parity bits set correctly.

And even if there's more than one, it's easy to manually check them all. In fact, in general you should be able to do this -- "correct" more than four errors by using error correction to reduce the search space by
a factor 2^20. Then having a couple extra bits will give you a couple extra bits' worth of reduction.

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 3, 2023 via email

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 3, 2023

from electrum import bip32
for candidate_seed in range(2**20):
    bip32.BIP32Node.from_rootseed(candidate_seed.to_bytes(16, 'big'),xtype="standard").calc_fingerprint_of_this_node()

This takes 100 seconds on i5-1135G7 laptop and calculates fingerprints for 2^20 candidate seeds.

For comparison, checking the first wpkh() address of 2**16 candidate seeds:

from electrum import bip32
from electrum import ecc
from electrum.crypto import hash_160

for candidate_seed in range(2**16):
    priv_masterkey = bip32.BIP32Node.from_rootseed(candidate_seed.to_bytes(16,'big'),xtype="standard")
    hash_160(priv_masterkey.subkey_at_private_derivation([2147483732, 2147483648, 2147483648]).subkey_at_public_derivation([0,0])[1].get_public_key_bytes(compressed=True))

Took 46 seconds, so 736 seconds to check 2**20 seeds. The master_pubkey fingerprint identifier is a useful 7.4x speed up vs known wpkh addresses.

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 3, 2023

Back to the desired hardening of share payload derivation:

import hashlib
for candidate_seed in range(2**16):
    hashlib.scrypt(password=candidate_seed.to_bytes(16,'big'), salt=bytes('2test','utf'), n=2**6, r=8, p=1, maxmem=68*1024, dklen=16)

This generates the 1st share's payload for 2^16 candidate master_seed's with codex32 secret threshold = '2' and identifier = 'test' in 7 seconds.

This is 2.2x slower than checking the fingerprint (3.1 seconds) but still 7x faster than checking the first wpkh() address.

  • If the data is generated by a slow RNG/KDF, then an attacker with the [share] data has no advantage over one without.

The maxmem cannot be raised any higher as most HWWs have 128kiB of RAM and there is no reason to not support them. So following your advice, lets slow it down by doing 16 rounds. If you believe share payloads must be slower to derive than an address then we can use your idea:

(And we could emulate that somewhat by just iterating scrypt.)

import hashlib

for candidate_seed in range(2**14):
    intermediate_key = hashlib.scrypt(password=candidate_seed.to_bytes(16,'big'), salt=bytes('2test','utf'), n=2**6, r=8, p=1, maxmem=68*1024, dklen=16)
    for _ in range(15):
        intermediate_key = hashlib.scrypt(password=candidate_seed.to_bytes(16, 'big'), salt=intermediate_key, n=2 ** 6, r=8, p=1, maxmem=68 * 1024, dklen=16)
    intermediate_key

This generates the 1st share's payload for 2^14 candidate master_seed's with codex32 secret threshold = '2' and identifier = 'test' in 43 seconds, or ~172 seconds for 2^16 which is much slower than 46 seconds to derive 2^16 wpkh() address payloads.

I have no idea why iterating it 16 times caused a 25x+ slowdown, computers are magical.

Any problems here?

I'm still wondering why the shares need to be hardened since the search space is the 128-bit master_seed. Are we concerned that may some day not be long enough to be secure against brute force without the slowdown of address derivation?

Our attack scenario is: an attacker has access to the 20-bit ID, a share, multiple shares, whatever. And they are trying to access the actual master secret. The scenarios are:

If the data was generated correctly but with a fast RNG, then an attacker can grind out the secret far faster (by a large constant factor) by re-generating the data than by deriving addresses.

I thought it was not even possible to increment a 128-bit counter thru all values without 10 nuclear plants running at full capacity for a year (or something like that).

So lets confirm with a 4th person, the threat model is valid before we roast these poor Trezors. If the share payloads were fast to derive with ChaCha20 they'd become an almost unlimited source of error detection with each independent share acting as a cryptographic hash checksum the length of the secret. Is that more useful than making it faster for guessers to check a number that cannot be counted to?

These scrypt parameters with more iterations will work for hardening the master_seed against RNG backdoors with user supplied entropy on 128kiB RAM HWWs, as well as for deriving a unique identifier securely when re-sharing a master_seed when not all previous identifiers are known.

@apoelstra
Copy link
Contributor

I'm still wondering why the shares need to be hardened since the search space is the 128-bit master_seed. Are we legitimately concerned that may some day not be long enough to be secure against brute force without the slowdown of address derivation?

If we really have 128 bits of good entropy then I agree that hardening has no value. But maybe the seed isn't really that strong, or some of it is leaked, or whatever ... then my thinking is that we ought to try to "not make the situation worse" than bare BIP32, and we can do that by hardening all our randomness roughly as much as an EC mult.

Now, I guess you're suggesting that we make the 128 bits itself hard, even against weak entropy, by doing a serious amount of hardening on the user's input. This will protect us from bad input but not from key leakage.

So maybe the "should we bother hardening the shares/IDs" question comes down to "should we attempt to ameliorate the damage caused by leaked seed data"? My feeling is that the hardening doesn't add much cost so we might as well.

This is contradicting this:

Yeah, sorry, my thinking has changed a bit over the course of this thread.

@BenWestgate
Copy link
Contributor Author

So maybe the "should we bother hardening the shares/IDs" question comes down to "should we attempt to ameliorate the damage caused by leaked seed data"? My feeling is that the hardening doesn't add much cost so we might as well.

Makes sense.
This is like medicine "First, do no harm."

So then do we need to harden the master_pubkey fingerprint default identifier? Otherwise it's a 7.4x speed up with a 1 in a million chance of being a false positive when grinding out a partially leaked master_seed?

Or is that not important because the attacker could get the fingerprint from the watch-only wallet?

@apoelstra
Copy link
Contributor

If we are actually using the BIP32 fingerprint then we already have enough hardening, since the bip32 fingerprint requires an EC mult to calculate it from the seed :).

If we are using some alternate notion of "fingerprint" then yes, we should make sure that it's at least as hard as an ecmult.

@BenWestgate
Copy link
Contributor Author

If we are actually using the BIP32 fingerprint then we already have enough hardening, since the bip32 fingerprint requires an EC mult to calculate it from the seed :).

Yes, that is the fingerprint we have in mind.

True about the EC mult, but the fingerprint is hash_160(master_pubkey) while a wpkh() address is something like hash_160 of [BIP32_fingerprint/84'/0'/0']xpub.../0/0 so there are 5 additional HMAC-SHA512 child derivations to slow it down, assuming BIP84 standard derivation path used. That slowdown is 7 fold as I calculated earlier. Unless the slowdown thanks to HMAC-SHA512 can be ignored due to ASICs or something. We also don't specify implementations use a particular derivation path. It may only be 1 HMAC-SHA512 if they use m/*.

A cheeky fix would be HMAC-SHA512-ing the master_pubkey fingerprint 5 times to set the identifierto make it identical work to check as a known address. But this unfortunately breaks the ability to check the fingerprint matches the share identifier by converting from hex to bech32 in your head.

My feeling is since descriptors felt it safe to include 4 bytes of hash_160(master_pubkey) even in encrypted wallets, it should be safe to include 20-bits in our shares which are much less likely to be stolen than an online descriptor. An attacker has to consider the identifier may just be random noise as well since it's only a default.

@apoelstra
Copy link
Contributor

My feeling is since descriptors felt it safe to include 4 bytes of hash_160(master_pubkey) even in encrypted wallets, it should be safe to include 20-bits in our shares which are much less likely to be stolen than an online descriptor.

Agreed on all counts.

I would further say that, while BIP49-style keys (which I guess is what 99.9% of all keys are..) do 7 derivations, in principle, people could be using their master keys to generate addresses. Or m/* or something.

A single constant-time EC mult is something like 100x the cost of an HMAC (500ns vs 50us). I think we should try to match the 100x factor, but not bother with the extra 7x factor since we're getting into increasingly tenuous justifications for it..

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 4, 2023

The electrum bip32 library I used must be unrepresentatively slow at the HMACs, it should have been a few percent slower, but not 7x. The chat robot agrees 80%+ of the time is spent going from private to public.

One scrypt round n=2**6, r=8, p=1 would be twice as slow on my hardware, but I don't trust the numbers. Part because this library gave bad data.

But in general, there's no particular scrypt iterations we can say for sure will be more expensive than the EC mult on all hardware and software that may ever exist to attackers. Without grievous overestimation. But what if we did the exact same work, plus some?

This might sound dumb, but why not use public key hash for the share payloads?

Consider this is the time to beat,

import bitcoin.bip32 as bip32
import hashlib

def compute_bip32_fingerprint(master_seed):
    # Step 1: Calculate the extended private masterkey (m) from the master seed using HMAC-SHA512.
    m = bip32.HMAC_SHA512(b"Bitcoin seed", master_seed)

    # Step 2: Compute the extended public masterkey (M) from the extended private masterkey (m).
    M = bip32.privkey_to_pubkey(m)

    # Step 3: Take the first 4 bytes of the hash160 of the serialized extended public masterkey (M).
    M_serialized = M.serialize()
    fingerprint = hashlib.new('ripemd160', hashlib.sha256(M_serialized).digest()).digest()[:4]

    return fingerprint

then,

def compute_share_payload(master_seed, threshold, identifier, share_index):
    # Step 1: Calculate the extended private masterkey (m) from the master seed using HMAC-SHA512.
    m = bip32.HMAC_SHA512(bytes(threshold+identifier+share_index, 'utf'), master_seed)

    # Step 2: Compute the extended public masterkey (M) from the extended private masterkey (m).
    M = bip32.privkey_to_pubkey(m)

    # Step 3: Take the first 16 bytes of the hash160 of the serialized extended public masterkey (M).
    M_serialized = M.serialize()
    share_payload = hashlib.new('ripemd160', hashlib.sha256(M_serialized).digest()).digest()[:16]

    return share_payload

This is guaranteed to not be cheaper, and it's also not any slower than it needs to be. It's also very easy for wallet developers to implement, they nearly already have (perhaps too close and they forget to remove b"Bitcoin seed"...). But since it's a 128-bit match rather than 20-bits or 4 bytes, it seems there should be one more step inserted somewhere to slow it down. Open to suggestions, if you agree.

It also needs to support 16-64 byte outputs. So something like HMAC-SHA512, scrypt, pbkdf2 could work to stretch and slow it. The master_seed should be mixed back in to preserve the original entropy. The target cost to exceed is an address with 5 extra derivations.

@apoelstra
Copy link
Contributor

I tried your formula and looks like 4-bits should be able to correct a bit-error in 256-bit data.

The actual formula is the Hamming bound. For binary codes q = 2, and n is the total length of your data including checksum (so 260). To correct a single error you need d = 3, so t = 1, and the formula simplifies to give a total code size of 2^260 / (1 + 260) = 2^251.97. Unfortunately not enough to fit 256 bits of data.

The approximate formula I gave was that for 256 = 2^(2^8) you would need 8 bits to get distance 3, so even the approximate one suggests that this won't work.

The 3-bit code just needs to detect as many contiguous 3 bit-errors as possible.

I think the simplest approach is:

  • For one bit do a parity check
  • For two bits, do a parity check on the odds and a parity check on the evens
  • For three, do a parity check every third, then every third +1, then every third +2
  • In general, if you've got n bits, use them to interleave parity checks.

This approach is guaranteed to detect 1 error (which is the best the Hamming bound will let us do, for low numbers of bits) and to detect n consecutive errors.

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 12, 2023

ChatGPT explained like I was five that not even the parity bit stays valid after ms32_interpolate().

The problem with adding a checksum that won't interpolate is attacker can guess which shares are derived and which were generated better than chance. That's why I added double_sha256 bits only to master_seed, not every encode()'ed index.

So replacing that on "s" with your parity suggestion is an improvement if what I want (interpolate into more valid binary codes) is impossible.

And in your opinion would you apply the parity to all k generated shares, k-1 generated shares and the codex secret or Only the codex32 secret?

If shares they only help the user if they have that subset, while every recovery will need the secret so it's good to be certain it's valid code there.

How would they know they had an valid ECC share? Seems like the generated share indexes have to be fixed: a, c, d, up to k or k-1 have valid codes, then the remaining indexes are shuffled (as those are the ones that actually reveal the size of the backup, you can infer a exists given b, from the threshold 2.

@apoelstra
Copy link
Contributor

I'm not sure. I agree that if it's not preserved by ms32_interpolate, then this is much less useful and also a privacy loss. Will need to do some experiments since I'm not sure analytically how to approach this. Parity checks are definitely additively homomorphic (if you add two shares they're preserved) but I'm unsure how they behave under bech32 multiplication.

@apoelstra
Copy link
Contributor

Unfortunately it looks like no, in general, this sort of parity check is not preserved by interpolation. Damn.

@apoelstra
Copy link
Contributor

In fact -- even if you set these bits to fixed 0, after ms32_interpolate, they might not be 0!

I checked the BIP and it looks like we (correctly) say to choose these at random, but I had never considered this to be particularly important until now, and I think in my personal shares I may have set these bits to 0... (though for my personal shares I also generated them starting from S, A, C, D, in order, so I already don't have index privacy or privacy against initial/derived discriminators).

@BenWestgate
Copy link
Contributor Author

In fact -- even if you set these bits to fixed 0, after ms32_interpolate, they might not be 0!

There's one last thing to try, xor every 5th bit and see what happens. That should give the exact same result as adding every symbol and truncating so maybe IT will be ms32_interpolate()-able.

Or maybe better, xor the 4th bit of every symbol to pick the 1st bit of padding and the 5th bit of every symbol for the 2nd. 128-bit case. That way we're sampling in the same period of the higher field it might work. This way last partially filled symbol doesn't get included.

zeros do enhance the error correction of the codex32 checksum but only for the last character of generated strings. That's kind of flimsy when they could detect errors anywhere (and enhance chances of error correction more often).

Leaves 9 choices:

Type of padding:

  1. Zeros
  2. double_sha256
  3. parity bits

Where we should pad:

  1. k generated shares
  2. k-1 generated shares and the codex secret
  3. Only the codex32 secret

My original implementation generated len(master_seed)+1 bytes randomly for share payloads and then truncated. If generated shares are padded with coding they need fixed indexes so the user knows they have correction help, this has no impact on redundancy degree privacy. The derived shares still need deterministically random indexes.

If k-of-(k+1) or smaller backups are most common, k shares is best.

If n > k, then the best choice between s & k - 1 shares vs k shares can be calculated by number of sets in permutations(k,n) that have superior ecc. I don't think that's complex math so the function could choose based on k and n what to do. But choosing leaks a bit of redundancy privacy since the likelihood of finding early alphabet shares is k/n vs (k-1)/n and which of those it is suggests a lower or upper bound for n, since it was a function of k vs perfectly private odds if all shares are generated with random padding and indexes.

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 12, 2023

Basically I'm suggesting calculating the parity symbol in GF(32) excluding the padding symbol, then truncate the first bits to append it as padding and see if that will interpolate and remain a parity symbol for every share.

And then this will bring up question 1. again is it orthogonal to the codex32 checksum.

@apoelstra
Copy link
Contributor

apoelstra commented Aug 12, 2023

Intuitively this should work, since you're just adding the symbols in gf32 and "truncating" which is actually taking the value mod x^2 which is linear ... except that these bits aren't sitting isolated outside of the string. They're embedded inside of the last character, where they algebraically interact with the other bits of that character. So I don't think it works.

That is, we're not saying (x1 + x2 + x3) mod x^2 = Y, we're saying (x1 + x2 + x3 mod x^2) + x4 = Y. The former would be compatible with interpolation, the latter won't be.

@BenWestgate
Copy link
Contributor Author

When considering among these:

  1. k generated shares
  2. k-1 generated shares and the codex secret
  3. Only the codex32 secret

I realize from pure error detection ability, 1 & 2 are strictly superior, regardless which shares the user wants to recover with. Because they represent k * pad_len total checksum bits as opposed to 2-4. Here the non-linearity works to our advantage as threshold shares will always be able to interpolate back to index 16, 0, ... k - 1 and attempt corrections to make all k of these codes valid before checking an address.

Index 16 "s" is padded first so the unshared secret can also benefit from ecc padding.

Finally, by fixing the index of the ecc generated shares, that doesn't have to have any privacy impact at all (besides sometimes revealing this scheme is in use, which is fine if the standard) if the share set we output is randomly shuffled from the 31 total.

It's not necessary to grab the share with or without this padding in order to use it once you have k shares and can interpolate to the k padded strings.

These strings are all independent so still no error correction guarantees, but my instinct tells me triple the bits for k=3 gives 1/3 the chance of not detecting an error pattern.

The actual formula is the Hamming bound. For binary codes q = 2, and n is the total length of your data including checksum (so 260). To correct a single error you need d = 3, so t = 1, and the formula simplifies to give a total code size of 2^260 / (1 + 260) = 2^251.97. Unfortunately not enough to fit 256 bits of data.

By this formula k=4 and 128-bit shares and k=3 and 256-bit shares may be able to correct one bit error if the other k-1 strings are valid and correct (reasonable assumption if the codex32 checksum is valid).

@BenWestgate
Copy link
Contributor Author

Look at this absolute masterpiece:

def fingerprint_identifier(payload_list):
    from electrum.bip32 import BIP32Node
    from electrum.crypto import hash_160
    for data in payload_list:
        root_node = BIP32Node.from_rootseed(data, xtype="stanadrd")
        pubkeys += root_node.eckey.get_public_key_bytes(compressed=True)
    
    return ''.join([CHARSET[d] for d in convertbits(hash_160(pubkeys), 8, 5)])[:4]

Where payload list is each payload in bytes at index 16, then 0, ... k - 1.

For k = 0 or reshare=False, [master_seed] is passed to payload_list and it returns the same fingerprint as prepended in its own descriptors. :) For reshare=True it forms a unique combinable share set ident regardless of seed_length and k thanks to the prehashing. And it's expensive and can help further error detect the k generated shares from above.

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 14, 2023

Design decision fork:

Should re-shares require new app_entropy and user_entropy to create a fresh derived_key or be determined by master_seed and header alone?

master_seed and header alone means:

  1. a given master_seed has just 2^23 possible corresponding share sets (all thresholds)
  2. shares can be re-derived from master_seed rather than being independent unrecoverable secrets of their own
  3. requires hardening share payloads to be costlier to re-derive than addresses
  4. excellent error detection
  5. re-shares require no new entropy besides selecting a unique ident
  6. ident becomes a nonce to determine share payloads, vs generated payloads determine the ident above
  7. guarantees no collisions of incombinable sets for re-shared master_seed + unique ident vs hash_160(pubkeys) is best effort basis
  8. runs risk of accidentally re-using an ident when re-sharing, breaking security
  9. birthday problem is over 2^20 ident rather than avoided by generating fresh entropy for the new shares
  10. Easier to audit
  11. Less steps required to re-share

If we go header and master_seed alone:
probably rename ident to unique_id or unique_ident to emphasize the criticality it never be reused for the same master_seed unless deliberately trying to add shares to an existing backup.

If we're okay with a risk of share set collisions when the user doesn't know all previously used identifiers to guarantee a unique ident then fresh_master_seed() could simplify to this:

def fresh_master_seed(bitcoin_core_entropy, user_entropy = '', seed_length = 16, k = '2', ident = ''):
    """Derive a fresh master seed of seed length bytes with optional user-provided entropy."""
    import hashlib
    if 16 > seed_length > 64:
        return None
    if 9 < k < 2:
        return None
    master_seed = hashlib.scrypt(password=bytes(user_entropy + str(seed_length+k+ident), "utf"),
                                 salt=bytes(bitcoin_core_entropy, "utf"), n=2 ** 20, r=8, p=1, maxmem=1025 ** 3, dklen=seed_length)
    if not ident:
        ident = fingerprint_ident([master_seed])
    return encode('ms', k, ident, 's', master_seed)

Really, it's fresh_codex32_secret() but it could also return (master_seed, ident) to keep the current name accurate.

It's not terrible if fresh entropy creates an ident that matches an existing one (1% chance after 146 sets) because they have independent payloads and an incorrect recovered secret should fail the hash_160(pubkeys) last resort error detection but it is bad (as discussed earlier), if an ident nonce is accidentally reused to re-share a master_seed, which has the same odds as above.

In the event a previous ident used is not known, the user has to generate fresh entropy anyway (to minimize collision odds). So using header as nonce is really only advantageous when all are known. Which is always true the first re-share.

Because of the small identifier that was not intended to be used as a nonce (or it'd be 32 or 64-bits) this cannot be reduced to insignificant odds. ≈ 2 * 10^-6 is the chance of 2 identifier's colliding when selected randomly. There's a reason why the bech32 checksum is 6 characters. If everyone on earth made a codex32 backup and re-shared an existing master_seed twice, all forgetting the original ident used (having a default may help extend this to a 3rd reshare since we can assume the fresh_codex32_secret() used bip32 fingerprint) 16,000 of them are going to accidentally reshare the same share set even though they did everything correctly (except not forgetting old ident). If the identifier was increased to 6 characters this could be 16 self-victims.

I'm also open to a combination approach where it uses master_seed and header if the user can affirm the new ident is absolutely unique but requires fresh entropy and makes a new derived_key when they cannot make give such assurance. This minimizes the risk of colliding share sets by having 2 ^ ((k - 1) * seed_length) possible re-share sets instead of just 2^20 while also minimizing labor and complexity when the user is absolutely sure they picked a unique id, which is true at least the first re-share and sometimes the 2nd.

ChatGPT informs me appending a unix timestamp to the header before generating new shares for an existing master_seed will prevent share payload collisions without needing to ask the user for fresh entropy. The time used should be displayed unconditionally for auditing purposes though. Malicious implementations could reuse a previous timestamp so it's better to display a human readable date time. Precision down to the minute is enough as no one will write and confirm multiple re-shares per minute. Or maybe just the date is enough and has better odds of being recalled if there's a need to recreate an existing re-share set.

Make the call here and I'll code for_an_existing_master_seed().

@apoelstra
Copy link
Contributor

By this formula k=4 and 128-bit shares and k=3 and 256-bit shares may be able to correct one bit error if the other k-1 strings are valid and correct (reasonable assumption if the codex32 checksum is valid).

This error correction code would only work for a single blessed share set, not arbitrary sets of k. And it would be visible to a privacy attacker which share set was blessed. And it's still not guaranteed that the error we can correct will be distinct from the errors that the full codex32 code can correct. TBH I think this line of reasoning is just too much complexity for too little benefit and we should just use the bits unused.

Look at this absolute masterpiece:

Heh, the overlap in the k=0 case with BIP32 is pretty nice. And it is easy to implement in terms of off-the-shelf BIP32 methods, even though the bare algorithm feels a little overcomplicated.

Should re-shares require new app_entropy and user_entropy to create a fresh derived_key or be determined by master_seed and header alone?

Great question. Thanks for the very thorough post. I am leaning toward the "stick an external counter/timestamp/weak randomness in there" approach. Requiring user_entropy is too much burden on the user (as is requiring the user to keep track of uniqueness, if we were considering that) but I think it's fine to require the app to generate a unique token. I think our specific recommendation should be: either (a) a monotonic counter, or (b) at least 12 bytes (96 bits) of fresh randomness. The UNIX timestamp isn't a great choice because it only changes every second, and it won't be accessible to most HWWs.

Regarding error detection abilities: my feeling is that resharing will be fairly uncommon. So I think there's a reasonable amount of value to having error detection in the initial sharing but greatly diminished value in subsequent sharings. Remember that the checksum itself provides overwhelming protection against random errors; the fingerprint detection gives us the additional ability to detect malicious errors, which is itself a fringe/uncommon case. And of course, we always have this ability if we're willing to derive full addresses.

So basically, if we use the bip32 fingerprint for the initial sharing, this gives us a bunch of nice benefits (cheap/local error detection mainly) at basically no cost. But for re-sharing, the costs become nontrivial ... and especially, there is a risk of ident reuse leading to shareset reuse, which is probable enough that it'll happen to a nonzero number of users.

As a further, point, even with the master_seed/header scheme, the error detection is not great. We no longer have the BIP32 fingerprint to compare against. Instead we have a weird hash of the share data itself, which is not something the wallet can produce on demand if the user has a single share they want to check on. Plus it's extra implementation complexity, reduced auditability, etc. So the benefit isn't even as big.

(I had a much longer reply suggesting the opposite, which took me half an hour to write but I decided that I disagreed with it :). This is quite a time-consuming exercise..)

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 15, 2023

This error correction code would only work for a single blessed share set, not arbitrary sets of k. And it would be visible to a privacy attacker which share set was blessed. And it's still not guaranteed that the error we can correct will be distinct from the errors that the full codex32 code can correct. TBH I think this line of reasoning is just too much complexity for too little benefit and we should just use the bits unused.

This will be my last question on the padding shares topic:

If a secret S, share A, C have ecc padding, the threshold is 3 and the user finds D, E, F but F has an uncorrectable errors, can he ms32_interpolate() using the candidate share F over-corrections to secret S, share A and share C and then check if their ECC padding is valid to save work?

Look at this absolute masterpiece:

Heh, the overlap in the k=0 case with BIP32 is pretty nice. And it is easy to implement in terms of off-the-shelf BIP32 methods, even though the bare algorithm feels a little overcomplicated.

I could've concatenated them as seeds to use def calc_fingerprint_of_this_node() which is a method in every bip32 library but is ambiguous for different seed_length. It doesn't matter as ident is to differentiate incombinable share sets and a length diff does that for free.

I have an even better idea for re-share ident however: encrypt the bip32 fingerprint by the birthdate/timestamp.

Should re-shares require new app_entropy and user_entropy to create a fresh derived_key or be determined by master_seed and header alone?

I am leaning toward the "stick an external counter/timestamp/weak randomness in there" approach.

I coded the unique salt for the derived_key as follows:

# date used to create a unique nonce if user forgets any past identifiers
date = datetime.date.today().strftime("%Y%m%d") if forgot else '19700101'
# must be displayed unconditionally for auditing purposes if forgot=True
fingerprint = fingerprint_ident([master_seed])
salt = codex32_secret[:9] + fingerprint + date
derived_key = hashlib.pbkdf2_hmac('sha512', password=master_seed, salt=salt,
                    iterations=2048, dklen=64)

I used the 8 digit year date with zero padding to maximize the chances someone would be able to recover the salt. While if someone forgets an identifier within 24 hours and reuses it they don't deserve security.

However, I can imagine someone performing this operation more than once in a day so it might need precision down to the hour or minute, although that makes it harder to regenerate this particular set in the future.

I'd appreciate your feedback on the precision of the timestamp (trade off between repeat generation ease and avoiding nonce reuse) and whether the timestamp should just by default (adds extra info to remember/grind to regenerate and audit, but would cover the "forgot I forgot worst case scenario") and eliminate the forgot parameter.

I also changed fingerprint_ident() to not truncate so it adds the whole hash_160 to the salt, a nice slowdown if master_seed is partially leaked and being brute forced against known shares, for free.

Requiring user_entropy is too much burden on the user (as is requiring the user to keep track of uniqueness, if we were considering that) but I think it's fine to require the app to generate a unique token. I think our specific recommendation should be: either (a) a monotonic counter, or (b) at least 12 bytes (96 bits) of fresh randomness. The UNIX timestamp isn't a great choice because it only changes every second, and it won't be accessible to most HWWs.

Who can write and confirm 2 codex32 share backups in 1 second?
If HWWs don't have a clock, why don't we just ask the user for the date (or other unique string they've never typed before, avoids trusting the hardware to not leak data by chosen nonce.) And if the hardware has a clock, display the date used unconditionally so it can be audited.

So I think there's a reasonable amount of value to having error detection in the initial sharing but greatly diminished value in subsequent sharings.

I think we can get it in both now with encryption, by simply remembering or recovering the nonce used on resharing.

I wrote the existing_master_seed() so the master_seed provides all the entropy, even for re-shares. It's just simpler this way, using app entropy + user entropy and deriving shares, index_seed and master_seed from that can never regenerate a share set without a threshold of that particular share set's shares, vs master_seed as the "seed" of the share set allows adding shares to other re-share sets without having to access a threshold of the set, just knowing their threshold&identifier is enough, and possibly now, their birthdate.

(which we can brute force out of the new re-share identifier, with a known bip32 fingerprint maybe, if you don't make it too precise!)

So basically, if we use the bip32 fingerprint for the initial sharing, this gives us a bunch of nice benefits (cheap/local error detection mainly) at basically no cost. But for re-sharing, the costs become nontrivial ... and especially, there is a risk of ident reuse leading to share set reuse, which is probable enough that it'll happen to a nonzero number of users.

As a further, point, even with the master_seed/header scheme, the error detection is not great. We no longer have the BIP32 fingerprint to compare against. Instead we have a weird hash of the share data itself, which is not something the wallet can produce on demand if the user has a single share they want to check on. Plus it's extra implementation complexity, reduced auditability, etc. So the benefit isn't even as big.

Why don't we encrypt the bip32 fingerprint by the timestamp (unique new entropy used for the reshare). Two things the wallet/user can probably can produce on demand (or brute force).

That way we get a random ident and don't lose the error detection the fingerprint provides if we know the birth date.

I think this is a MUCH better idea than that convoluted concatenation of share payloads, and because this "encrypted" fingerprint feeds into the derived_key derivation, it still shuffles the indexes and generates new randomness for payloads just fine.

@BenWestgate
Copy link
Contributor Author

    date = datetime.date.today().strftime("%Y%m%d") if forgot else '19700101'
    # must be displayed unconditionally for auditing purposes if forgot=True
    fingerprint = fingerprint_ident([master_seed])
    salt = codex32_secret[:9] + fingerprint + date
    derived_key = hashlib.pbkdf2_hmac('sha512', password=master_seed, salt=salt,
                        iterations=2048, dklen=64)
    if reshare:
        encryption_key = hashlib.pbkdf2_hmac('sha512', password=date, salt=salt,
                        iterations=2048, dklen=3)

        new_ident = old_ident ^ encryption_key

Optionally, the hash_160(master_pubkey) may be dropped from salt as the wallet might not have access to it. Then the identifier can be decrypted / error detected with the correct fingerprint, the header used to generate the reshare and the birthdate.

@BenWestgate BenWestgate linked a pull request Aug 15, 2023 that will close this issue
21 tasks
@apoelstra
Copy link
Contributor

If a secret S, share A, C have ecc padding, the threshold is 3 and the user finds D, E, F but F has an uncorrectable errors, can he ms32_interpolate() using the candidate share F over-corrections to secret S, share A and share C and then check if their ECC padding is valid to save work?

I'm not sure what this means but I don't think so.

I used the 8 digit year date with zero padding to maximize the chances someone would be able to recover the salt. While if someone forgets an identifier within 24 hours and reuses it they don't deserve security.

They don't need to forget it to reuse it. They need to fail to actively override their wallet's default ident computation, because they are aware of this "24 hour" rule that doesn't exist in any other scheme. I think this is too much of a burden on users. And yes, users will want to do this more than once per day (likely more than once per second, if their hardware can compute fast enough).

You also still have the problem that HWWs don't know the time, even to 24-hour precision.

Who can write and confirm 2 codex32 share backups in 1 second?

You don't need to write and confirm them before you start generating the next set. You might generate all the sets you need and then write/confirm them all.

If HWWs don't have a clock, why don't we just ask the user for the date (or other unique string they've never typed before, avoids trusting the hardware to not leak data by chosen nonce.) And if the hardware has a clock, display the date used unconditionally so it can be audited.

Because this is extra work for the user and has a nonobvious application (we are asking for a current date, which seems like an unnecessary and privacy-reducing thing to answer honestly, but actually we just want a unique string). And we can accomplish it by just generating random bytes.

I really don't like any of these date-based ideas.

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 16, 2023

They need to fail to actively override their wallet's default ident computation...

A wallet implementation should store the identifiers and bip32 fingerprints, ideally persistently. So that it never reuses them. Perhaps incrementing ident LSB or throwing error messages on collisions.

You also still have the problem that HWWs don't know the time, even to 24-hour precision.

Looks like HWW need (optional?) user provided uniqueness during reshares if all past IDs aren't stored/remembered.

It's the only auditable way. But the odds they'll input a unique string is higher than directly choosing a new ident.

You don't need to write and confirm them before you start generating the next set. You might generate all the sets you need and then write/confirm them all.

Without a persistent counter, complete record of previously used identifiers or a unique input, ident uniqueness and determinism for auditability can not be simultaneously guaranteed (to 1/2^10).

but actually we just want a unique string). And we can accomplish it by just generating random bytes.

Agree completely it should be "unique string" and this is a strange hassle to protect users from forgetfulness. But on the other hand, it's less work than for a fresh master seed. 20-bits of unique user entropy fully protects the reshare vs 128-bits for fresh seeds.

It's unsafe to let a wallet alone generate a random nonce for reshares, it could grind for ones that make your shares and/or the ident leak parts of the secret.

If the random bytes are generated by KDF/HMAC(app_entropy, user_unique_string) with app entropy unconditionally displayed before the user types

This is totally fine, the string "42" or "we" may be unique enough for the purpose.

Anything never typed before is unique to 1/2^10, and perfect uniqueness is only achieved by storing and/or remembering every ident.

I really don't like any of these date-based ideas.

Can date, all we need is a unique input.

@apoelstra
Copy link
Contributor

A wallet implementation should store the identifiers and bip32 fingerprints, ideally persistently

Then these would need to be backed up along with the seed data?

If we just used 96 random bits there would be no need for anyone to store anything, persistently or otherwise.

Without a persistent counter, complete record of previously used identifiers or unique input, ident uniqueness and determinism for auditability can not be simultaneously guaranteed (1/2^10).

Right. We should just drop auditability for re-sharing. It's not a common case and all the proposed solutions seem to have serious usability problems.

If the random bytes are generated by HMAC(app_entropy, user_unique_string) with app entropy unconditionally displayed before the user types

This seems good to me. Though I do think we should query for "a unique string" rather than trying to give it any other semantic content. And maybe suggest the user roll dice ten times to get the unique string, or something. If it's too short/predictable the wallet could theoretically still grind.

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 16, 2023

Then these would need to be backed up along with the seed data?

In a digital backup, yes. On paper or metal, no.

The list helps prevent ident reuse but it's not essential data, treat it like wallet metadata (ex: prevents address reuse)

it's okay for ident to collide, they're likely to every 2^10 sets generated. It's only bad if payloads collide.

I edited that to mention the app_entropy and unique_user_string should be hardened by PBKDF2 to slow grinding.

How about encrypting (still using PBKDF2) the bip32 fingerprint by the unique string and using cipher text as new_ident?

Then we don't lose the error correction and other benefits as long as user remembers the unique string.

@apoelstra
Copy link
Contributor

How about encrypting (still using PBKDF2) the bip32 fingerprint by the unique string and using cipher text as new_ident?

This is a neat idea. Yeah, let's do it. No need to use fancy encryption; just xor with the first 20 bits is fine.

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 16, 2023

This is a neat idea. Yeah, let's do it. No need to use fancy encryption; just xor with the first 20 bits is fine.

XOR is good for encryption algorithm but "unique string" shouldn't be recoverable if someone knows bip32 fingerprint and the ciphertext so PBKDF2("unique string") is needed for key derivation.

Non-zero chance people would enter names, birthdates, street numbers, PINs as a "unique string" and give a thief extra help.

Share payloads, unlike new_ident would incorporate both app entropy and unique_string so those will never collide (if RNG Honest) even if ident collides.

@apoelstra
Copy link
Contributor

Yep, agreed -- we'd xor with the (truncated) output of the PBKDF2.

@BenWestgate
Copy link
Contributor Author

Re-Share payloads derivation unlike new_ident would incorporate both app entropy (machine's random number) and user unique_string so those will not collide (if RNG Honest) even if ident does.

This seems like the best way to be able to recover the bip32 fingerprint with no risk of payload collision (if honest RNG or unique string).

Can't recover payloads with only master seed, header, and unique string. (Could brute force missing app entropy with a candidate share)

But it's a niche case:
Made a reshare set, lost some shares AND want to regenerate an identical set (without gathering a threshold from said set) to expand/repair it.

It's not a useless feature but unique payloads for all sets is far more important.

Reshare ident and its app_entropy could be stored as metadata to provide it.

Reshare app_entropy only needs to be long enough to be unique for all times a user should be able to safely reshare a master_seed and fail to provide a unique string.

1st reshare is always unique as no "unique string" was asked by fresh_master_seed.

A single random word from a word list is enough to allow a user resharing a master seed to fail 90 times to provide a unique string before they get into trouble.

Larger reshare app entropy is more bandwidth to attempt to leak secrets by grinding if unique_string is low entropy and predictable. Fortunately, the most redundant channel, ident, is under full user "unique string" control.

So a balance seems important.

@apoelstra
Copy link
Contributor

This seems like the best way to be able to recover the bip32 fingerprint with no risk of payload collision (if honest RNG or unique string).

Yep, agreed with everything in this comment. I think it's more important to be resilient and avoid collisions than to support increasingly-niche regeneration usecases. Ultimately, if the user has the master seed, if need be they can just redo everything. (And if they don't, none of our regeneration schemes would help them anyway.)

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 18, 2023

Encrypting the bip32 fingerprint, all I have to salt "unique string" with is ms1<threshold> and codex32 string length. It's going to be easy to brute force low entropy "unique strings" if attacker knows the bip32 fingerprint. Is that okay? Might need to emphasize if the string is if low entropy it is public (writing it along w/ the share makes the share as good as fingerprint labeled). Nothing secret, private, personally identifiable or useful to an attacker should be used for it. Or sufficient dice rolls must be appended to protect it.

Once it contains over 20-bits entropy, exhaustive attack begins to return 2^ (entropy - 20) possible passwords. Easiest to not use information they wouldn't mind telling a thief or adversary.

The inclusion of length and threshold in salt gives the re-shares some error detection for these when the "unique string" and fingerprint are known. Corrections can be quickly brute forced as there's under 9-bits total entropy for every length and threshold 4.6-bits if lengths are limited to 16, 32, 64.

Is this relationship between "unique string", fingerprint, length and threshold useful enough original shares should have it? (by using '' as the unique string.) or is the ID being the direct bech32 encoding of the fingerprint directly most useful?

@apoelstra
Copy link
Contributor

You need to know more than the bip32 fingerprint - you need to know app_entropy as well, no?

@BenWestgate
Copy link
Contributor Author

You need to know more than the bip32 fingerprint - you need to know app_entropy as well, no?

It's possible to use password = unique_string + app_entropy, to encrypt fingerprint.

But it's not necessary since the ident doesn't need to be random, it gives an evil wallet an opportunity to leak secrets into the ident by grinding and it would require storing/remembering the app entropy to decrypt the fingerprint to know what seed shares belonged to.

So I only used app_entropy for payload generation and index shuffling. The identifier is set from fingerprint and user selected info only (unique string, threshold, share length).

@apoelstra
Copy link
Contributor

But it's not necessary since the ident doesn't need to be random, it gives an evil wallet an opportunity to leak secrets into the ident by grinding and it would require storing/remembering the app entropy to decrypt the fingerprint to know what seed shares belonged to.

True, but it also puts the onus on the user to produce an actually-unique string, without the wallet "bailing him out" by providing its own unique input. I guess this is okay. It feels a little bit fragile. But the tradeoff is that it avoids grinding attacks. So I could go either way.

Nothing secret, private, personally identifiable or useful to an attacker should be used for it. Or sufficient dice rolls must be appended to protect it.

Probably best to just say "An attacker can learn this data! Do not use secrets." rather than trying to communicate something more nuanced. We could word it in a friendlier way ("This data is not treated as secret. If you input secret data it will harm your security" or something).

@BenWestgate
Copy link
Contributor Author

BenWestgate commented Aug 18, 2023

But it's not necessary since the ident doesn't need to be random, it gives an evil wallet an opportunity to leak secrets into the ident by grinding and it would require storing/remembering the app entropy to decrypt the fingerprint to know what seed shares belonged to.

True, but it also puts the onus on the user to produce an actually-unique string, without the wallet "bailing him out" by providing its own unique input. I guess this is okay. It feels a little bit fragile. But the tradeoff is that it avoids grinding attacks. So I could go either way.

Not just grinding secrets into the ident, it makes the password to decrypt the fingerprint easier to remember & store. The worst case is they reuse unique_string for the same master_seed and get the same ident, but if shares from different sets of the master_seed are combined, they'll produce a fingerprint that doesn't match the ident and unique_string.

Like I said earlier, the device could choose & display a random word or 20-bits of entropy before user types unique_string to protect the user from reused ident, or password cracking, it's not that bad either way.

I'll code it both ways since they differ by a couple lines and a third opinion or alpha tester feedback can decide.

It might be confusing "what's this extra word all about?" Now suddenly language and wordlists become a (possible) part of resharing, if app_entropy is 4 random bech32 characters, I'm positive it'll be hard to remember, begs the question too: why not just memorize or write the 4 char encryption key of the fingerprint directly.

So the only useful option seems to be deriving the fingerprint encryption key from:

  1. Less than 20-bits length (saves writing) or
  2. Memorable strings (increase odds of recovering fingerprint) or
  3. prepended app_entropy so small it doesn't interfere with the above, like 1-2 bech32 characters, or a single hex byte.
    • significantly protects against unique_string & master_seed reuse at least once.

The best protection against reuse is keeping a list of already used ident.

We do not define how to choose the identifier, beyond noting that it SHOULD be distinct for every master seed the user may need to disambiguate.

Either solution complies, and checking ident reuse against a list is better but harder to audit and backup. This is how Bitcoin wallets solved the similar "address reuse" problem.

ident list needs to be unconditionally displayed for auditability. (ie so it would not blacklist every ident except one leaking 20-bits of secret.), but now how do you know the list is correct, it ends up being something the auditor has to store and update independent of the wallet.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

3 participants
@apoelstra @BenWestgate and others