BIP 350: Implement Bech32m and use it for v1+ segwit addresses #20861

sipa · 2021-01-05T21:43:26Z

This implements BIP 350:

For segwit v1+ addresses, a new checksum algorithm called Bech32m is used.
Segwit v0 address keep using Bech32 as specified in BIP 173.

DrahtBot · 2021-01-05T22:05:47Z

The following sections might be updated with supplementary metadata relevant to reviewers and maintainers.

Conflicts

Reviewers, this pull request conflicts with the following ones:

scripted-diff: Regenerate key_io data deterministically #21279 (scripted-diff: Regenerate key_io data deterministically by MarcoFalke)

If you consider this pull request important, please also help to review the conflicting pull requests. Ideally, start with the one that should be merged first.

src/key_io.cpp

NicolasDorier · 2021-01-06T12:40:34Z

By implementing Bech32m, I found out that it is not obvious how to locate errors, because the residue was polymod^1, but now is either polimod^1 or polimod^0x2bc830a3... I am tempted to try both and take the residue with the least errors. I think this should be specified in the bip.
That said by doing this, the error patterns with failure probability in the bip are probably different. As I think you assume in your table that the decoder knows it is decoding bech32m. Unsure if this is really changing things by a lot.

This is not the right place to comment for it, but did not found elsewhere.

luke-jr · 2021-01-06T17:16:43Z

@NicolasDorier That isn't solved by sipa only allowing one polimod based on the witness version?

sipa · 2021-01-06T17:29:01Z

@luke-jr It's not, because the error may be one that affects the witness version symbol.

@NicolasDorier Good point, I will elaborate on that.

felipsoarez

Concept ACK

michaelfolkson · 2021-01-09T20:35:05Z

Concept ACK, Approach ACK. Skimmed code but will hopefully go through more thoroughly at a later date.

https://bitcoin.stackexchange.com/questions/101117/what-problems-identified-with-bech32-addresses-have-been-resolved-with-the-updat

kallewoof

Slightly tested ACK 863a5c2

I honestly don't really like the std::tuple stuff. It's pretty ugly and a simple struct containing the 3 elements passed as a copy ~~by reference~~ (edit: I meant like the tuples are passed now) in the same way would be more straightforward.

src/bech32.h

sipa · 2021-01-18T22:05:58Z

@kallewoof Agree, a struct is definitely warranted here. Updated.

btcslade

utACK 936322d

src/bech32.h

kallewoof

re-slightly-tested-ACK 2827cf8

decryp2kanon

Code Review ACK 2827cf8

benthecarman

utACK 2827cf8

gruve-p · 2021-01-21T11:41:31Z

utACK 2827cf8

sipa · 2021-01-27T00:23:01Z

Rebased now #20832 is merged.

I've also addressed some comments on the BIP draft itself, see https://lists.linuxfoundation.org/pipermail/bitcoin-dev/2021-January/018362.html.

kallewoof

re-ACK 3fee858

benthecarman · 2021-01-27T07:42:52Z

re-ACK 3fee858

…ors.py 5c0210e bugfix: fix bech32_encode calls in gen_key_io_test_vectors.py (Pieter Wuille) Pull request description: This fixes the the calls to bech32_encode in the gen_key_io_test_vectors.py script. Bug introduced in #20861. ACKs for top commit: fanquake: ACK 5c0210e Tree-SHA512: 8e8aee08741619c1700371ca1a8ca05ffdb2f48544d9fd3982f2665f6afb926b126478cf644f15a699f8c7e7da53c2777a56ce7989f05e4a3ef9fbe085f74d5a

…gwit ad…

maflcko

left some minor nits

maflcko · 2021-03-19T19:13:39Z

src/test/fuzz/bech32.cpp

-        const std::vector<uint8_t>& data = r2.second;
-        assert(hrp == "bc");
-        assert(data == input);
+    if (input.size() + 3 + 6 <= 90) {


nit: Any reason to remove the branch that checks bech32 for larger input sizes? I know this isn't relevant for bitcoin addresses, but there are other applications that use bech32 for encoding and checking it here doesn't cost us anything, no?

Bech32 and Bech32m do not permit encoded strings over 90 characters (the checksum properties break down past that point). BOLT11 uses a relaxed version of the spec that drops this requirement.

maflcko · 2021-03-19T19:14:47Z

test/functional/wallet_labels.py

-            '❌_VER15_PROG41': 'bcrt10qqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqzc7xyq',
-            '❌_VER16_PROB01': 'bcrt1sqqpl9r5c',
+            '❌_VER15_PROG41': 'bcrt1sqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqajlxj8',
+            '❌_VER16_PROB01': 'bcrt1sqq5r4036',


nit: obviously unrelated to your changes, but the B should be a G. (blame me)

maflcko · 2021-03-19T19:15:46Z

src/bench/bech32.cpp

@@ -19,7 +19,7 @@ static void Bech32Encode(benchmark::Bench& bench)
    tmp.reserve(1 + 32 * 8 / 5);
    ConvertBits<8, 5, true>([&](unsigned char c) { tmp.push_back(c); }, v.begin(), v.end());
    bench.batch(v.size()).unit("byte").run([&] {
-        bech32::Encode("bc", tmp);
+        bech32::Encode(bech32::Encoding::BECH32, "bc", tmp);


nit: Could add a benchmark for bech32m?

Yeah, could do. The performance should be exactly identical to Bech32 though.

Reduced version of the test from master/bitcoin#20861 by John Newbery. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

This also includes updates to the Python test framework implementation, test vectors, and release notes. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

Github-Pull: bitcoin#20861 Rebased-From: 0334602

Reduced version of the test from master/bitcoin#20861 by John Newbery. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

Github-Pull: bitcoin#20861 Rebased-From: da2bb69

Github-Pull: bitcoin#20861 Rebased-From: 25b1c6e

This also includes updates to the Python test framework implementation, test vectors, and release notes. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

Github-Pull: bitcoin#20861 Rebased-From: 2e7c80f

Github-Pull: bitcoin#20861 Rebased-From: 0334602

Reduced version of the test from master/bitcoin#20861 by John Newbery. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

…dresses (0.21 backport) f2195d7 Backport invalid address tests (Pieter Wuille) 1e96711 naming nits (Fabian Jahr) 7dfe406 Add signet support to gen_key_io_test_vectors.py (Pieter Wuille) 593e206 Use Bech32m encoding for v1+ segwit addresses (Pieter Wuille) 8944aaa Add Bech32m test vectors (Pieter Wuille) 1485533 Implement Bech32m encoding/decoding (Pieter Wuille) Pull request description: Backport of #20861. Also includes #21471. ACKs for top commit: jnewbery: utACK f2195d7 MarcoFalke: cherry-pick re-ACK f2195d7 , only change is version number in doc/bips and new test commit 🍝 fanquake: ACK f2195d7 - performed the backport, changes look sane. Have not tested extensively. Tree-SHA512: 7dc043e44d7cda07d73331a7b49666b9db98c99f2635dab0cfeb45422dbfbe75a7b44d0aff85ef6369d412d8a5041ed0826c86ffdfc13c5fbff74adfe4d91c1a

…dresses (0.20 backport) c0f85fd Backport invalid address tests (Pieter Wuille) c670986 naming nits (Fabian Jahr) 1a4e88e Use Bech32m encoding for v1+ segwit addresses (Pieter Wuille) cf18ac9 Add Bech32m test vectors (Pieter Wuille) 5f9537b Implement Bech32m encoding/decoding (Pieter Wuille) Pull request description: Backport of #20861 for 0.20. Also includes #21471. ACKs for top commit: jnewbery: utACK c0f85fd MarcoFalke: range-diff-only ACK c0f85fd 🐔 Tree-SHA512: e87e52995cb9b555109bc500dba1e844993d881821d2001443b3de9e3ef9050ddb73deefa0c1af732418fa7355a86b875789887c9611c340713b3ad26809d58e

bitcoin/bitcoin#20861 (1/5)

bitcoin/bitcoin#20861 (2/5)

This also includes updates to the Python test framework implementation, test vectors, and release notes. bitcoin/bitcoin#20861 (3/5)

bitcoin/bitcoin#20861 (1/5)

bitcoin/bitcoin#20861 (2/5)

This also includes updates to the Python test framework implementation, test vectors, and release notes. bitcoin/bitcoin#20861 (3/5)

c72c949 blech32: copy ubsan suppression for bech32 to blech32 (Andrew Poelstra) d13fb49 blech32: add test vectors for blech32 and blech32m (Andrew Poelstra) 15a826e blech32: add functional tests for blech32m (Andrew Poelstra) c01e09e blech32: add blech32m format and use it to decode witness v1+ addresses (Andrew Poelstra) 18fcec8 naming nits (Fabian Jahr) 8515f40 Add signet support to gen_key_io_test_vectors.py (Pieter Wuille) b3df66f Use Bech32m encoding for v1+ segwit addresses (Pieter Wuille) 42f43a1 Add Bech32m test vectors (Pieter Wuille) b1d1d94 Implement Bech32m encoding/decoding (Pieter Wuille) c607835 Better error messages for invalid addresses (Bezdrighin) Pull request description: Includes backports of bitcoin/bitcoin#20832 (1 commit) and bitcoin/bitcoin#20861 (5 commits) ACKs for top commit: gwillen: utACK c72c949. Tree-SHA512: af96a6ef31b1cab72b0350197dcb34761e9ffb2ec43685084408b5fafcda0adee1945045003194600372652f7cca65d721bda3ed9b6be7e9543d3199e2cbe145

DrahtBot mentioned this pull request Jan 5, 2021

rpc: Better error messages for invalid addresses #20832

Merged

sipa force-pushed the 202101_bech32m branch 3 times, most recently from 8ba1f7a to 05f8733 Compare January 5, 2021 23:01

DrahtBot added the Utils/log/libs label Jan 5, 2021

sipa force-pushed the 202101_bech32m branch 2 times, most recently from f08b0ff to 863a5c2 Compare January 5, 2021 23:55

luke-jr reviewed Jan 6, 2021

View reviewed changes

src/key_io.cpp Outdated Show resolved Hide resolved

felipsoarez reviewed Jan 9, 2021

View reviewed changes

gits7r mentioned this pull request Jan 18, 2021

Impement Bech32m addresses for v1+ segwit addresses spesmilo/electrum#6949

Closed

kallewoof reviewed Jan 18, 2021

View reviewed changes

src/bech32.h Outdated Show resolved Hide resolved

sipa force-pushed the 202101_bech32m branch from 863a5c2 to 936322d Compare January 18, 2021 22:04

btcslade reviewed Jan 19, 2021

View reviewed changes

src/bech32.h Outdated Show resolved Hide resolved

sipa force-pushed the 202101_bech32m branch from 936322d to 2827cf8 Compare January 19, 2021 03:52

kallewoof reviewed Jan 19, 2021

View reviewed changes

decryp2kanon approved these changes Jan 19, 2021

View reviewed changes

benthecarman approved these changes Jan 20, 2021

View reviewed changes

DrahtBot added the Needs rebase label Jan 26, 2021

sipa force-pushed the 202101_bech32m branch from 2827cf8 to 3fee858 Compare January 27, 2021 00:08

DrahtBot removed the Needs rebase label Jan 27, 2021

kallewoof reviewed Jan 27, 2021

View reviewed changes

sipa mentioned this pull request Mar 18, 2021

BIP 350: Implement Bech32m and use it for v1+ segwit addresses (0.19 backport) #21472

Closed

sidhujag pushed a commit to syscoin/syscoin that referenced this pull request Mar 19, 2021

Merge bitcoin#20861: BIP 350: Implement Bech32m and use it for v1+ se…

752dba1

…gwit ad…

maflcko reviewed Mar 19, 2021

View reviewed changes

sipa added a commit to sipa/bitcoin that referenced this pull request Mar 24, 2021

Backport invalid address tests

ce14a52

Reduced version of the test from master/bitcoin#20861 by John Newbery. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

sipa added a commit to sipa/bitcoin that referenced this pull request Mar 27, 2021

Backport invalid address tests

66009fc

Reduced version of the test from master/bitcoin#20861 by John Newbery. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

sipa added a commit to sipa/bitcoin that referenced this pull request Mar 27, 2021

Backport invalid address tests

26347db

Reduced version of the test from master/bitcoin#20861 by John Newbery. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

sipa added a commit to sipa/bitcoin that referenced this pull request Mar 27, 2021

Use Bech32m encoding for v1+ segwit addresses

1a4e88e

This also includes updates to the Python test framework implementation, test vectors, and release notes. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

sipa pushed a commit to sipa/bitcoin that referenced this pull request Mar 27, 2021

naming nits

c670986

Github-Pull: bitcoin#20861 Rebased-From: 0334602

sipa added a commit to sipa/bitcoin that referenced this pull request Mar 27, 2021

Backport invalid address tests

c0f85fd

Reduced version of the test from master/bitcoin#20861 by John Newbery. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

sipa added a commit to sipa/bitcoin that referenced this pull request Mar 27, 2021

Implement Bech32m encoding/decoding

1485533

Github-Pull: bitcoin#20861 Rebased-From: da2bb69

sipa added a commit to sipa/bitcoin that referenced this pull request Mar 27, 2021

Add Bech32m test vectors

8944aaa

Github-Pull: bitcoin#20861 Rebased-From: 25b1c6e

sipa added a commit to sipa/bitcoin that referenced this pull request Mar 27, 2021

Use Bech32m encoding for v1+ segwit addresses

593e206

This also includes updates to the Python test framework implementation, test vectors, and release notes. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

sipa added a commit to sipa/bitcoin that referenced this pull request Mar 27, 2021

Add signet support to gen_key_io_test_vectors.py

7dfe406

Github-Pull: bitcoin#20861 Rebased-From: 2e7c80f

sipa pushed a commit to sipa/bitcoin that referenced this pull request Mar 27, 2021

naming nits

1e96711

Github-Pull: bitcoin#20861 Rebased-From: 0334602

sipa added a commit to sipa/bitcoin that referenced this pull request Mar 27, 2021

Backport invalid address tests

f2195d7

Reduced version of the test from master/bitcoin#20861 by John Newbery. Github-Pull: bitcoin#20861 Rebased-From: fe5e495

ryanofsky mentioned this pull request Apr 1, 2021

Handle exceptions instead of crash bitcoin-core/gui#260

Merged

hebasto mentioned this pull request Apr 17, 2021

Move GetDataDir to ArgsManager #21244

Merged

apoelstra pushed a commit to apoelstra/elements that referenced this pull request Aug 5, 2021

Implement Bech32m encoding/decoding

1142976

bitcoin/bitcoin#20861 (1/5)

apoelstra pushed a commit to apoelstra/elements that referenced this pull request Aug 5, 2021

Add Bech32m test vectors

44dff10

bitcoin/bitcoin#20861 (2/5)

apoelstra pushed a commit to apoelstra/elements that referenced this pull request Aug 5, 2021

Use Bech32m encoding for v1+ segwit addresses

a4e778c

This also includes updates to the Python test framework implementation, test vectors, and release notes. bitcoin/bitcoin#20861 (3/5)

apoelstra mentioned this pull request Aug 5, 2021

Introduce blech32m format for v1+ witness programs ElementsProject/elements#1025

Merged

apoelstra pushed a commit to apoelstra/elements that referenced this pull request Aug 18, 2021

Implement Bech32m encoding/decoding

b1d1d94

bitcoin/bitcoin#20861 (1/5)

apoelstra pushed a commit to apoelstra/elements that referenced this pull request Aug 18, 2021

Add Bech32m test vectors

42f43a1

bitcoin/bitcoin#20861 (2/5)

apoelstra pushed a commit to apoelstra/elements that referenced this pull request Aug 18, 2021

Use Bech32m encoding for v1+ segwit addresses

b3df66f

This also includes updates to the Python test framework implementation, test vectors, and release notes. bitcoin/bitcoin#20861 (3/5)

tgalal mentioned this pull request May 19, 2022

Bech32 encoding and SegWit addresses. tgalal/motoko-bitcoin#8

Merged

bitcoin locked as resolved and limited conversation to collaborators Aug 18, 2022

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

BIP 350: Implement Bech32m and use it for v1+ segwit addresses #20861

BIP 350: Implement Bech32m and use it for v1+ segwit addresses #20861

sipa commented Jan 5, 2021 •

edited

Loading

DrahtBot commented Jan 5, 2021 •

edited

Loading

NicolasDorier commented Jan 6, 2021 •

edited

Loading

luke-jr commented Jan 6, 2021

sipa commented Jan 6, 2021

felipsoarez left a comment

michaelfolkson commented Jan 9, 2021

kallewoof left a comment •

edited

Loading

sipa commented Jan 18, 2021

btcslade left a comment

kallewoof left a comment

decryp2kanon left a comment

benthecarman left a comment

gruve-p commented Jan 21, 2021

sipa commented Jan 27, 2021

kallewoof left a comment

benthecarman commented Jan 27, 2021

maflcko left a comment

maflcko Mar 19, 2021

sipa Mar 19, 2021

maflcko Mar 19, 2021

maflcko Mar 19, 2021

sipa Mar 19, 2021

BIP 350: Implement Bech32m and use it for v1+ segwit addresses #20861

BIP 350: Implement Bech32m and use it for v1+ segwit addresses #20861

Conversation

sipa commented Jan 5, 2021 • edited Loading

DrahtBot commented Jan 5, 2021 • edited Loading

Conflicts

NicolasDorier commented Jan 6, 2021 • edited Loading

luke-jr commented Jan 6, 2021

sipa commented Jan 6, 2021

felipsoarez left a comment

Choose a reason for hiding this comment

michaelfolkson commented Jan 9, 2021

kallewoof left a comment • edited Loading

Choose a reason for hiding this comment

sipa commented Jan 18, 2021

btcslade left a comment

Choose a reason for hiding this comment

kallewoof left a comment

Choose a reason for hiding this comment

decryp2kanon left a comment

Choose a reason for hiding this comment

benthecarman left a comment

Choose a reason for hiding this comment

gruve-p commented Jan 21, 2021

sipa commented Jan 27, 2021

kallewoof left a comment

Choose a reason for hiding this comment

benthecarman commented Jan 27, 2021

maflcko left a comment

Choose a reason for hiding this comment

maflcko Mar 19, 2021

Choose a reason for hiding this comment

sipa Mar 19, 2021

Choose a reason for hiding this comment

maflcko Mar 19, 2021

Choose a reason for hiding this comment

maflcko Mar 19, 2021

Choose a reason for hiding this comment

sipa Mar 19, 2021

Choose a reason for hiding this comment

sipa commented Jan 5, 2021 •

edited

Loading

DrahtBot commented Jan 5, 2021 •

edited

Loading

NicolasDorier commented Jan 6, 2021 •

edited

Loading

kallewoof left a comment •

edited

Loading