As per the BIP 174 spec a PSBT key cannot be duplicated,
however the current code accepts key duplication.
The PSBT key/value entries can be duplicated when the value
is `empty()` or `IsNull()` for `CScript` or `CTxOut` respectively
and if those key/value entries are serialized before the non-empty ones.
For example, the following PSBT, included in the test vectors,
contains a duplicate field:
```
// magic
70736274ff
// global tx
//// key
0100
//// value
2a02000000000140420f000000000017a9146e91b72d5593e7d4391e2ff44e91e985c31641f08700000000
//// separator
00
// no inputs
// outputs
//// key PSBT_OUT_WITNESSSCRIPT
0101
//// value (empty script)
00
//// key PSBT_OUT_WITNESSSCRIPT (same as the above)
0101
//// value (an OP_RETURN script)
016a
//// separator
00
```
5df6f089b5 More tests of signer checks (Andrew Chow)
7c8bffdc24 Test that a non-witness script as witness utxo is not signed (Andrew Chow)
8254e9950f Additional sanity checks in SignPSBTInput (Pieter Wuille)
c05712cb59 Only wipe wrong UTXO type data if overwritten by wallet (Pieter Wuille)
Pull request description:
The current PSBT signing code can end up producing a non-segwit signature, while only the UTXO being spent is provided in the PSBT (as opposed to the entire transaction being spent). This may be used to trick a user to incorrectly decide a transaction has the semantics he intends to sign.
Fix this by refusing to sign if there is any mismatch between the provided data and what is being signed.
Tree-SHA512: b55790d79d8166e05513fc4c603a982a33710e79dc3c045060cddac6b48a1be3a28ebf8db63f988b6567b15dd27fd09bbaf48846e323c8635376ac20178956f4
0-input transactions can be ambiguously deserialized as being witness
transactions. Since the unsigned transaction is never serialized as
a witness transaction as it has no witnesses, we should always
deserialize it as a non-witness transaction and set the serialization
flags as such.
Also always serialize the unsigned transaction as a non-witness transaction.
When extra entropy is not specified by the caller, CKey::Sign will
now always create a signature that has a low R value and is at most
70 bytes. The resulting signature on the stack will be 71 bytes when
the sighash byte is included.
Using low R signatures means that the resulting DER encoded signature
will never need to have additional padding to account for high R
values.
Checks that all of the one byte type keys are actually one byte and
throw an error if they are not.
Add tests for each type to check for this behavior.
Added functional tests for PSBT that test the RPCs. Also added all
of the BIP 174 test vectors (except for the updater tests) in the
functional tests.
Added a Unit test for the BIP 174 updater test vector.
Antoine Poinsot