multi: recognize new standard P2A output type #2433

Roasbeef · 2025-09-27T02:38:07Z

In this PR, we add initial recognition of the new P2A output type. It actually overloads the existing Taproot witness version, but with a distinct witness program. We adopt the proposed standardness rules to make only spends without any witness data standard. Utilities to parse the script, make addresses, etc - have also been added.

Note that this doesn't include the logic to support "ephemeral dust", which enables zero fee and zero value outputs under certain conditions.

A replace dir is used as we modified btcutil in this PR.

coveralls · 2025-09-27T02:42:09Z

Pull Request Test Coverage Report for Build 20117557279

Details

101 of 111 (90.99%) changed or added relevant lines in 6 files are covered.
127 unchanged lines in 8 files lost coverage.
Overall coverage increased (+0.4%) to 55.171%

Changes Missing Coverage	Covered Lines	Changed/Added Lines	%
txscript/pkscript.go	1	2	50.0%
txscript/sign.go	0	1	0.0%
mempool/policy.go	28	30	93.33%
txscript/standard.go	24	30	80.0%

Files with Coverage Reduction	New Missed Lines	%
btcutil/bech32/bech32.go	1	99.61%
btcutil/bloom/filter.go	2	96.84%
btcutil/gcs/gcs.go	2	81.25%
txscript/taproot.go	2	97.74%
database/ffldb/blockio.go	4	88.81%
peer/peer.go	15	71.1%
rpcclient/infrastructure.go	38	47.99%
btcec/schnorr/musig2/context.go	63	80.17%

Totals
Change from base Build 20117517710:	0.4%
Covered Lines:	31296
Relevant Lines:	56725

💛 - Coveralls

sputn1ck

LGTM!

sputn1ck · 2025-09-30T11:19:46Z

txscript/pkscript.go

 		return nil, fmt.Errorf("unable to parse address: %v", err)
 	}

+	// P2A scripts don't have addresses.


there is an address for p2a outputs bc1pfeessrawgf
https://mempool.space/address/bc1pfeessrawgf
https://github.com/bitcoin/bitcoin/blob/25212dfdb4cd7291392b6a94130f658c5bfa0a48/test/functional/rpc_validateaddress.py#L171

Looks like a stray line, but ExtractPkScriptAddrs should handle that. Will add a test for this.

kcalvinalvin · 2025-12-12T07:32:39Z

txscript/standard.go

 				nilAddrErrStr)
 		}
 		return payToWitnessTaprootScript(addr.ScriptAddress())
+	case *btcutil.AddressPayToAnchor:


nit

btcutil.AddressPayToAnchor isn't introduced until commit 3f99cfe0a9b0ed9a33e0c11136127e8b6e2906ea.

So this commit doesn't build.

kcalvinalvin · 2025-12-12T09:16:06Z

txscript/p2a_test.go

+		},
+		{
+			name:   "invalid - wrong version",
+			script: []byte{OP_0, OP_DATA_2, 0x4e, 0x73},


~~Just a question: is OP_1 a version? Curious as to the test's name~~

nvm figured it out

kcalvinalvin · 2025-12-12T09:42:43Z

go.mod


 go 1.23.2
+
+replace github.com/btcsuite/btcd/btcutil => ./btcutil


Are we ok with this or is this temporary?

This is temp. It's needed as we ref the newly added addr type in txscript/standard.go.

#1825 will resolve this circualr dep module issue. After this is merged, I'll be ale to update the module dep, then remove this.

btcutil/p2a_address_test.go

kcalvinalvin · 2025-12-12T10:19:31Z

btcutil/p2a_address_test.go

+				otherNet = &chaincfg.TestNet3Params
+			}
+			if addr.IsForNet(otherNet) {
+				t.Errorf("IsForNet(other) = true, want false")


What's this testing for? Why do we want to test "othernet-ness"

Just that IsForNet properly detects if an address isn't for the target network.

I agree it's a trivial test, tbh I did it just so the test coverage reported is higher.

kcalvinalvin · 2025-12-12T10:21:55Z

btcutil/p2a_address_test.go

+				t.Errorf("IsForNet(other) = true, want false")
+			}
+
+			expectedScript := []byte{0x51, 0x02, 0x4e, 0x73}


I think it's better if I we declare something like the declaration in txscript/standard.go

var PayToAnchorScript = []byte{OP_1, OP_DATA_2, 0x4e, 0x73}

Yeah the issue with that again is the module issue I brought up which will be addressed by #1825. This is defined, but the btcutil module wasn't updated (as this was only added in this PR). So I had to add yet another replace to amek things work.

btcutil/p2a_address_test.go

kcalvinalvin · 2025-12-12T10:23:17Z

btcutil/address.go

+// script bytes: OP_1 <0x4e73>.
+func (a *AddressPayToAnchor) ScriptAddress() []byte {
+	// Return the fixed P2A script bytes.
+	return []byte{0x51, 0x02, 0x4e, 0x73}


Yeah here too. Would be better if we declared some variable and referred to it

Actually this won't work, as it causes an import cycle. We can't import txscript from btcutil. The other changed worked at it was in btcutil_test.

In this commit, we modify the execution logic, such that the P2A witness program doesn't appear as a unknown witness program version. For execution, we just make sure the script passes.

P2A is standard if the witness and the sigscript is empty. We make a smol refactor to be able to write a unit test for the input standardness.

We make sure it'll be accepted into the mempool, and can be spent without a witness.

Roasbeef · 2025-12-18T02:15:21Z

@kcalvinalvin PTAL

kcalvinalvin · 2026-02-03T04:26:41Z

btcutil/address.go

+//   - Regtest: bcrt1pfeesnyr2tx
+//   - Simnet:  sb1pfeesxv0pfa
+func (a *AddressPayToAnchor) EncodeAddress() string {
+	// For unknown networks, generate the address from the anchor data.


Weird comment. Should remove this.

kcalvinalvin · 2026-02-03T04:32:36Z

btcutil/p2a_address_test.go

+			require.Equal(t, tt.wantAddress, addr.String())
+			require.True(t, addr.IsForNet(tt.net))
+
+			// Verify it's not for a different network.


The require.True() does cover this and we don't need this extra check but it's ok.

Claude could be better but heh.

kcalvinalvin · 2026-02-03T05:03:15Z

txscript/standard.go

+	if IsPayToAnchorScript(pkScript) {
+		// P2A scripts don't have traditional addresses and require no
+		// signatures (anyone can spend them).
+		return PayToAnchorTy, nil, 0, nil


We should be returning the address here because of this. 2eadd59#r2390989822

kcalvinalvin · 2026-02-03T05:12:35Z

txscript/p2a_test.go

+			class, PayToAnchorTy)
+	}
+
+	if len(addrs) != 0 {


We'd want 1 address since they P2A has an address

kcalvinalvin · 2026-02-03T05:21:16Z

txscript/p2a_test.go

+			class.String())
+	}
+
+	// Test that regular taproot scripts are still recognized as


Weird thing to test for here but ok

kcalvinalvin · 2026-02-03T06:51:29Z

mempool/policy_test.go

+}
+
+// LookupEntry returns the entry for a given outpoint.
+func (m *mockUtxoView) LookupEntry(op wire.OutPoint) utxoEntry {


Seems like these mock utxo entry and mock utxoview is only needed for the tests but it really doesn't need this abstraction. Not a blocker but yeah I think simpler is better.

diff --git a/mempool/policy.go b/mempool/policy.go index 798d6686..46339d33 100644 --- a/mempool/policy.go +++ b/mempool/policy.go @@ -79,29 +79,6 @@ func calcMinRequiredTxRelayFee(serializedSize int64, minRelayTxFee btcutil.Amoun return minFee } -// utxoEntry is an interface that provides access to UTXO data needed for -// input standardness checks. -type utxoEntry interface { - PkScript() []byte -} - -// utxoView is an interface that provides access to UTXOs needed for -// input standardness checks. -type utxoView interface { - LookupEntry(wire.OutPoint) utxoEntry -} - -// utxoViewpointAdapter wraps a blockchain.UtxoViewpoint to implement the -// utxoView interface. -type utxoViewpointAdapter struct { - view *blockchain.UtxoViewpoint -} - -// LookupEntry returns the entry for a given outpoint. -func (u *utxoViewpointAdapter) LookupEntry(op wire.OutPoint) utxoEntry { - return u.view.LookupEntry(op) -} - // checkInputsStandard performs a series of checks on a transaction's inputs // to ensure they are "standard". A standard transaction input within the // context of this function is one whose referenced public key script is of a @@ -113,12 +90,6 @@ func (u *utxoViewpointAdapter) LookupEntry(op wire.OutPoint) utxoEntry { // accurately and concisely via the txscript.ScriptVerifyCleanStack and // txscript.ScriptVerifySigPushOnly flags. func checkInputsStandard(tx *btcutil.Tx, utxoView *blockchain.UtxoViewpoint) error { - return checkInputsStandardWithView(tx, &utxoViewpointAdapter{view: utxoView}) -} - -// checkInputsStandardWithView performs input standardness checks using the -// utxoView interface. -func checkInputsStandardWithView(tx *btcutil.Tx, utxoView utxoView) error { // NOTE: The reference implementation also does a coinbase check here, // but coinbases have already been rejected prior to calling this // function so no need to recheck. diff --git a/mempool/policy_test.go b/mempool/policy_test.go index c3259fd0..3411fa08 100644 --- a/mempool/policy_test.go +++ b/mempool/policy_test.go @@ -9,13 +9,13 @@ import ( "testing" "time" + "github.com/btcsuite/btcd/blockchain" "github.com/btcsuite/btcd/btcec/v2" "github.com/btcsuite/btcd/btcutil" "github.com/btcsuite/btcd/chaincfg" "github.com/btcsuite/btcd/chaincfg/chainhash" "github.com/btcsuite/btcd/txscript" "github.com/btcsuite/btcd/wire" - "github.com/stretchr/testify/mock" ) // TestCalcMinRequiredTxRelayFee tests the calcMinRequiredTxRelayFee API. @@ -660,44 +660,28 @@ func TestCheckTransactionStandard(t *testing.T) { } } -// mockUtxoEntry mocks the utxoEntry interface using testify/mock. -type mockUtxoEntry struct { - mock.Mock -} - -// PkScript returns the public key script. -func (m *mockUtxoEntry) PkScript() []byte { - args := m.Called() - return args.Get(0).([]byte) -} - -// mockUtxoView mocks the utxoView interface using testify/mock. -type mockUtxoView struct { - mock.Mock -} - -// LookupEntry returns the entry for a given outpoint. -func (m *mockUtxoView) LookupEntry(op wire.OutPoint) utxoEntry { - args := m.Called(op) - if args.Get(0) == nil { - return nil - } - return args.Get(0).(utxoEntry) -} - // TestP2ASpendingStandardness tests that P2A outputs require empty witness // and empty signature script to be considered standard. func TestP2ASpendingStandardness(t *testing.T) { // Create a previous transaction with a P2A output. - prevTxHash, _ := chainhash.NewHashFromStr("0101010101010101010101010101010101010101010101010101010101010101") - prevOut := wire.OutPoint{Hash: *prevTxHash, Index: 0} + prevTx := wire.NewMsgTx(2) + prevTx.AddTxIn(&wire.TxIn{ + PreviousOutPoint: wire.OutPoint{ + Hash: chainhash.Hash{}, + Index: 0, + }, + }) + prevTx.AddTxOut(&wire.TxOut{ + Value: 1000, + PkScript: txscript.PayToAnchorScript, + }) + prevBtcTx := btcutil.NewTx(prevTx) - // Create mocked UTXO entry and view. - mockEntry := new(mockUtxoEntry) - mockEntry.On("PkScript").Return(txscript.PayToAnchorScript) + // Create a UTXO view with the P2A output. + utxoView := blockchain.NewUtxoViewpoint() + utxoView.AddTxOuts(prevBtcTx, 100) - mockView := new(mockUtxoView) - mockView.On("LookupEntry", prevOut).Return(mockEntry) + prevOut := wire.OutPoint{Hash: *prevBtcTx.Hash(), Index: 0} tests := []struct { name string @@ -756,7 +740,7 @@ func TestP2ASpendingStandardness(t *testing.T) { }) btcTx := btcutil.NewTx(tx) - err := checkInputsStandardWithView(btcTx, mockView) + err := checkInputsStandard(btcTx, utxoView) if test.shouldFail { if err == nil { @@ -767,10 +751,6 @@ func TestP2ASpendingStandardness(t *testing.T) { t.Errorf("Unexpected error for valid P2A spend: %v", err) } } - - // Verify that the mock was called. - mockView.AssertExpectations(t) - mockEntry.AssertExpectations(t) }) } }

kcalvinalvin · 2026-02-03T07:27:01Z

Most things look good. But ExtractPkScriptAddrs not returning an address for P2A should be fixed before this gets merged.

Things done:

Went through each of the commits again to sanity check the changes.
Checked that the implementation here aligns with BIP-433.
Checked that the tests make sense in each of the commits.

Roasbeef added the mempool label Sep 27, 2025

saubyk assigned Roasbeef Sep 27, 2025

Roasbeef mentioned this pull request Sep 30, 2025

mempool/txgraph: add new transaction graph data structures and traversal+eviction algoritms #2436

Draft

sputn1ck approved these changes Sep 30, 2025

View reviewed changes

saubyk added the transactions v3 label Sep 30, 2025

Roasbeef requested a review from yyforyongyu October 27, 2025 23:06

Roasbeef force-pushed the p2a branch from 69fd175 to 2ee309b Compare December 11, 2025 00:18

Roasbeef changed the base branch from master to v3-tx-standard December 11, 2025 00:18

kcalvinalvin reviewed Dec 12, 2025

View reviewed changes

btcutil/p2a_address_test.go Outdated Show resolved Hide resolved

kcalvinalvin reviewed Dec 12, 2025

View reviewed changes

btcutil/p2a_address_test.go Outdated Show resolved Hide resolved

kcalvinalvin reviewed Dec 12, 2025

View reviewed changes

Roasbeef added 8 commits December 17, 2025 18:14

btcutil: add new AddressPayToAnchor addr type

7347f4c

txscript: add initial awareness of new P2A output script

9a411d6

txscript: support trivial signing for P2A

6825eaf

txscript: recognize P2A scripts in ParsePkScript

2eadd59

txscript: add tests for new P2A awareness

6216aca

txscript: recognize P2A witness program during script execution

d324aa3

In this commit, we modify the execution logic, such that the P2A witness program doesn't appear as a unknown witness program version. For execution, we just make sure the script passes.

mempool: add p2a specific policy

d466691

P2A is standard if the witness and the sigscript is empty. We make a smol refactor to be able to write a unit test for the input standardness.

rpctest: add new e2e test for new P2A behavior

ff20205

We make sure it'll be accepted into the mempool, and can be spent without a witness.

Roasbeef force-pushed the p2a branch from 2ee309b to ff20205 Compare December 18, 2025 02:14

Roasbeef requested a review from kcalvinalvin December 18, 2025 02:15

Roasbeef changed the base branch from v3-tx-standard to master December 18, 2025 02:15

kcalvinalvin reviewed Feb 3, 2026

View reviewed changes

txscript/p2a_test.go

class, PayToAnchorTy)

}

if len(addrs) != 0 {

Copy link

Collaborator

kcalvinalvin Feb 3, 2026

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We'd want 1 address since they P2A has an address

kcalvinalvin reviewed Feb 3, 2026

View reviewed changes


		go 1.23.2

		replace github.com/btcsuite/btcd/btcutil => ./btcutil

multi: recognize new standard P2A output type #2433

Are you sure you want to change the base?

multi: recognize new standard P2A output type #2433

Uh oh!

Conversation

Roasbeef commented Sep 27, 2025

Uh oh!

coveralls commented Sep 27, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Pull Request Test Coverage Report for Build 20117557279

Details

💛 - Coveralls

Uh oh!

sputn1ck left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kcalvinalvin Dec 12, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Roasbeef commented Dec 18, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kcalvinalvin Feb 3, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kcalvinalvin commented Feb 3, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

coveralls commented Sep 27, 2025 •

edited

Loading

kcalvinalvin Dec 12, 2025 •

edited

Loading

kcalvinalvin Feb 3, 2026 •

edited

Loading