Updates for new Trezor and Ledger firmware

[achow101]

This is a general update to fix many issues related to newly released signing behavior in Trezors and Ledgers.

• The PSBT module is updated to handle PSBT inputs that include both non_witness_utxo and witness_utxo
• For Trezor, Ledger, and Bitbox, instead of checking for existence of witness_utxo, witness_script, and redeem_script to determine signing behavior, we instead inspect the scripts using existing template functions to determine their type and base our behavior on that.
• For Trezor and Ledger, the signing behavior is updated to pass in the full previous transaction for segwit inputs if they are available.
• For Trezor, all of the derivation paths are updated to fit within Trezor's "known paths".
• In the tests, a patch to bitcoind is added that allows for PSBTs with inputs that have both UTXO types. This is mostly from psbt: Include and allow both non_witness_utxo and witness_utxo for segwit inputs bitcoin/bitcoin#19215 and should be removed once that PR is merged.

Note that the Ledger emulator has not been updated to include the Bitcoin app that has the segwit requirements. Also note that the updated app with those requirements seems to have a bug when a transaction has multiple segwit inputs.

Closes #338 and #337

[molnard]

Hi there,
What do you think, when can we start to work on this in Wasabi? In my mind the procedure as follows:

1. Fix HWI
2. Create Wasabi PR that uses the fixed HWI
3. Add the fix to Wasabi
4. Tests
5. Release new HWI
6. Release new Wasabi

Let me know your thoughts.

[achow101]

Since I've started doing release candidates, I can publish an rc containing the fix that Wasabi can use and test with.

[molnard]

@achow101 thx, I will keep an eye on it.

[Sjors]

Tested spending a native segwit input with 45dbd08 along with bitcoin/bitcoin#19215

• Trezor T:
  • old firmware 2.1.8 still works with both these PRs
  • After upgrading to 2.3.1 it still works
  • signing a transaction without the prerequisite info causes rather ugly behavior: the screen gets stuck at "Signing transaction" and 1/4th progress, and returns a cryptic error: {"error": "b'\\t\\xef=Gy\\x1e\\xc4E\\x86\\xe2-\\x0f?\\x8bw\\xdd\\xf6=\\xc2\\xe3_\\xd2\\xe6\\xcb\\xcfmG\\x0f=\\xfc\\x84A'", "code": -13}.
• Ledger Nano S:
  • old firmware still works (firmware 1.5.5, bitcoin app version 1.3.9)
  • new firmware 1.6.0, bitcoin app 1.4.2
• Ledger Nano X, firmware 1.2.4-1:
  • old bitcoin app version 1.3.21 works
  • new bitcoin app version 1.4.2 works

I suggest we clean up the cryptic error when tx info is missing, because users of older Bitcoin Core versions will encounter it. If Trezor firmware version can be obtained, maybe we can even pre-empt the error.

[achow101]

I added a commit which should give a better error message

[Sjors]

Now it says "Previous transaction ... not available", which is at least a bit better.

[Sjors]

Tested and rather cursory code review ACK 1613a33. I didn't test the BitBox changes.

[jmacato]

I can confirm that fixes Wasabi's issues with Trezor (at least for Mainnet, Testnet is a bit wonky... probably a firmware fault). I compiled this via docker and used that binary in linux. Thanks for this @achow101 :D

[achow101]

@molnard 1.1.2-rc.2 is available for testing now: https://github.com/bitcoin-core/HWI/releases/tag/1.1.2-rc.2 (I messed up rc.1, so rc.2 is the first release candidate)

[molnard]

Thank you @achow101 !

[KayBeSee]

I'm not sure if this is the appropriate place to post this, but I am struggling to get this working on a Trezor One running 1.9.1. I am getting the same ""Previous transaction ... not available" that @Sjors posted earlier.

Where am I supposed to include the previous transaction in the PSBT?

EDIT:
Nevermind, I found the technical blog post about it.