p2wsh and p2sh-p2wsh address in decodescript #12321

fivepiece · 2018-02-01T01:16:59Z

Attempts to address #12244 . p2wsh addresses are returned only for scripts that are neither p2sh nor any witness program.

fivepiece · 2018-02-01T01:21:01Z

Again I'm only considering the edge cases after clicking "create pull request".. What if the script contains a checksig with an uncompressed pubkey?
Obviously not ready.

meshcollider · 2018-02-01T01:30:01Z

@fivepiece feel free to add [WIP] to the title of your PR if its work in progress :) It's usually fine to open up a PR before its completely ready because other contributors might like to give feedback on the approach or high level concept ACK/NACKs regardless. Thanks for helping out!

fivepiece · 2018-02-01T01:34:36Z

Cheers, title changed
(and now changed back :) )

fivepiece · 2018-02-01T20:24:09Z

Ideas welcome, returning a p2wsh for a p2pk or multisig with an uncompressed pubkey can result in fund loss, then again decodescript already returns p2sh addresses even for complete garbage.
What should this change to decodescript take into account when deciding whether to post a p2wsh address to the user?

instagibbs · 2018-02-01T22:14:12Z

concept ACK

I think returning garbage is fine. It can be something unprovably unspendable even, and I don't think it's the job of this tool to check.

Wonder if it's worth it to have it also return a p2sh-p2wsh version, in other words interpreting the script as a witnessscript, if applicable?

fivepiece · 2018-02-01T23:10:26Z

Agreed. Keeping the simple check that this is not a bare witness program seems to be in line with the check above it not letting p2sh be re-encoded into p2sh.
This commit adds p2sh-p2wsh as well.

promag

Missing tests for new keys.

Doesn't make sense to move this to ScriptPubKeyToUniv?

fivepiece · 2018-02-08T14:38:50Z

@promag wouldn't moving this to ScriptPubKeyToUniv cause script translation to all these types of addresses to fire on decoderawtransaction as well?
Sorry if I misunderstood your comment.

I will be looking into adding tests as well.

fivepiece · 2018-02-09T19:31:15Z

I've made some changes to this:

Segwit stuff is handled in a generic way using GetScriptForWitness(). This allows to return a more appropriate p2wpkh for p2pkh and p2pk scriptpubkeys, and (currently) p2wsh for all the rest.
The data returned for the segwit scripts is its own univalue object (@promag suggestion), so it includes more useful info like the segwit scriptpubkey and type, along with the bech32 address.

Various examples of the output can be seen here :
https://gist.github.com/fivepiece/e25b0fcdf8303fdbfd56363ae50dec15

Caveats :

For the segwit scriptpubkeys, we always have "reqSigs": 1.
It's still possible to get a segwit program address for scripts using checksigs with uncompressed pubkeys. This is because GetScriptForWitness() doesn't check vSolutions for presense of such pubkeys. It's probably be possible to change GetScriptForWitness to check for that when bare pubkeys are invloved in the script, but out of scope for this PR.

fivepiece · 2018-02-16T14:39:30Z

This is pretty much a complete rewrite taking into account @sipa's suggestion (on irc) to use the same logic from validateaddress in here.

Uncompressed pubkeys are handled correctly - meaning standard scripts containing them will not be encoded in segwit scriptpubkeys.
Added a few tests on top of the existing rpc_decodescript.py. I couldn't find a way to easily test that segwit output is not returned for standard scripts with uncompressed pubkeys. Suggestions on how to do that are welcome.
Updated https://gist.github.com/fivepiece/e25b0fcdf8303fdbfd56363ae50dec15 with examples showing segwit output is disabled for uncompressed pubkey scripts.

~~(please restart the travis job if possible, it hasn't failed, just cancelled before it began)~~ Done, thanks!

instagibbs · 2018-02-16T14:44:41Z

kicked travis.

I couldn't find a way to easily test that segwit output is not returned for standard scripts with uncompressed pubkeys.

hm? Your gist has examples that you could use.

fivepiece · 2018-02-16T14:48:20Z

Specifically, I don't know how to assert that a value (say rpc_result['segwit']) does not exist in the returned univalue.

instagibbs · 2018-02-16T14:51:06Z

in python 'segwit' not in rpc_result ?

fivepiece · 2018-02-16T14:54:49Z

Ah, I was trying to bend assert_array_result to do the check but your way works.

fivepiece · 2018-02-17T09:12:23Z

Added tests for checking 'segwit' value is not returned when inappropriate.

instagibbs · 2018-03-06T14:05:15Z

test/functional/rpc_decodescript.py


        # 3) multisig scriptPubKey
        # <m> <A pubkey> <B pubkey> <C pubkey> <n> OP_CHECKMULTISIG
        # just imagine that the pub keys used below are different.
        # for our purposes here it does not matter that they are the same even though it is unrealistic.
        rpc_result = self.nodes[0].decodescript('52' + push_public_key + push_public_key + push_public_key + '53ae')
        assert_equal('2 ' + public_key + ' ' + public_key + ' ' + public_key +  ' 3 OP_CHECKMULTISIG', rpc_result['asm'])
+        # multisig in P2WSH
+        assert_equal('0 0f320decca000221b611117448134c78dd220cd22384ac80025b59e43c8698e8', rpc_result['segwit']['asm'])


could you generate this dynamically instead of hardcoding the long hex?

instagibbs · 2018-03-06T14:05:24Z

test/functional/rpc_decodescript.py

@@ -103,6 +109,8 @@ def decodescript_script_pub_key(self):
        # lock until block 500,000
        rpc_result = self.nodes[0].decodescript('63' + push_public_key + 'ad670320a107b17568' + push_public_key + 'ac')
        assert_equal('OP_IF ' + public_key + ' OP_CHECKSIGVERIFY OP_ELSE 500000 OP_CHECKLOCKTIMEVERIFY OP_DROP OP_ENDIF ' + public_key + ' OP_CHECKSIG', rpc_result['asm'])
+        # CLTV script in P2WSH
+        assert_equal('0 01f79ceb012cbcefb251f7843b827a3e285f40ddc5014d35e2486716cbacf5f1', rpc_result['segwit']['asm'])


could you generate this dynamically instead of hardcoding the long hex?

instagibbs · 2018-03-06T14:06:50Z

utACK 785bd02

I think you can squash almost all these commits.

fivepiece · 2018-03-06T19:39:12Z

Cool, I'll squash it all into a single commit if this current change looks good, or two if it's better to separate work done on core from the tests. Let me know which is better.

instagibbs · 2018-03-06T19:41:29Z

whatever's easier to you, 1 or 2

fivepiece · 2018-03-06T19:55:24Z

Two commits now. No problem changing later on if needed.

instagibbs

utACK

instagibbs · 2018-03-06T20:00:55Z

src/rpc/rawtransaction.cpp

+            if ((which_type == TX_PUBKEY) || (which_type == TX_MULTISIG)) {
+                for (const auto& solution : solutions_data) {
+                    if ((solution.size() != 1) && !CPubKey(solution).IsCompressed()) {
+                        fFoundUncompressedPubkeys = true;


just return r here. then remove the conditional later?

Ah I see what you mean, looking into this.

plus tests

fivepiece · 2018-03-06T21:11:11Z

Good call. All in a single commit now.

instagibbs · 2018-03-06T21:26:19Z

re-utACK 4f933b3

laanwj · 2018-04-07T19:19:15Z

src/rpc/rawtransaction.cpp

@@ -563,6 +563,38 @@ UniValue decodescript(const JSONRPCRequest& request)
        // P2SH cannot be wrapped in a P2SH. If this script is already a P2SH,
        // don't return the address for a P2SH of the P2SH.
        r.pushKV("p2sh", EncodeDestination(CScriptID(script)));
+        // P2SH and witness programs cannot be wrapped in P2WSH, if this script
+        // is a witness program, don't return addresses for a segwit programs.
+        if (type.get_str().find("witness") == std::string::npos) {


Searching for a string in the type name seems brittle - if we have to work with strings here (there seems to be no other accessible way) let's explicitly list the options. e.g. according to current selection that would be

"nonstandard" "pubkey" "pubkeyhash" "scripthash" "multisig" "nulldata"

fivepiece · 2018-04-07T22:03:24Z

Updated to address @laanwj's comment

laanwj · 2018-04-08T17:34:45Z

Thank you
utACK 41ff967

41ff967 list the types of scripts we should consider for a witness program (fivepiece) 4f933b3 p2wpkh, p2wsh and p2sh-nested scripts in decodescript (fivepiece) Pull request description: Attempts to address #12244 . `p2wsh` addresses are returned only for scripts that are neither `p2sh` nor any witness program. Tree-SHA512: eb47f094c1a4c2ad2bcf27a8032307e43cf787d50bf739281aeb4101d97316a2f307b05118bf138298c937fa34e15f91436443a9b313f809fad2c43e94cd1831

41ff967 list the types of scripts we should consider for a witness program (fivepiece) 4f933b3 p2wpkh, p2wsh and p2sh-nested scripts in decodescript (fivepiece) Pull request description: Attempts to address bitcoin#12244 . `p2wsh` addresses are returned only for scripts that are neither `p2sh` nor any witness program. Tree-SHA512: eb47f094c1a4c2ad2bcf27a8032307e43cf787d50bf739281aeb4101d97316a2f307b05118bf138298c937fa34e15f91436443a9b313f809fad2c43e94cd1831

fanquake added the RPC/REST/ZMQ label Feb 1, 2018

fivepiece changed the title ~~p2wsh address in decodescript~~ [WIP] p2wsh address in decodescript Feb 1, 2018

fivepiece changed the title ~~[WIP] p2wsh address in decodescript~~ p2wsh address in decodescript Feb 3, 2018

fivepiece changed the title ~~p2wsh address in decodescript~~ p2wsh and p2sh-p2wsh address in decodescript Feb 8, 2018

promag reviewed Feb 8, 2018

View reviewed changes

fivepiece force-pushed the decodescript-p2wsh branch from 03aefb2 to d0164ba Compare February 16, 2018 13:30

instagibbs reviewed Mar 6, 2018

View reviewed changes

fivepiece force-pushed the decodescript-p2wsh branch from 85a6a66 to 19272ef Compare March 6, 2018 19:51

instagibbs approved these changes Mar 6, 2018

View reviewed changes

p2wpkh, p2wsh and p2sh-nested scripts in decodescript

4f933b3

plus tests

fivepiece force-pushed the decodescript-p2wsh branch from 1355786 to 4f933b3 Compare March 6, 2018 21:10

laanwj reviewed Apr 7, 2018

View reviewed changes

list the types of scripts we should consider for a witness program

41ff967

laanwj merged commit 41ff967 into bitcoin:master Apr 26, 2018

Bushstar mentioned this pull request Apr 27, 2018

commits from bitcoin/master FeatherCoin/Feathercoin#311

Merged

bitcoin locked as resolved and limited conversation to collaborators Sep 8, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

p2wsh and p2sh-p2wsh address in decodescript #12321

p2wsh and p2sh-p2wsh address in decodescript #12321

fivepiece commented Feb 1, 2018

fivepiece commented Feb 1, 2018

meshcollider commented Feb 1, 2018 •

edited

fivepiece commented Feb 1, 2018 •

edited

fivepiece commented Feb 1, 2018

instagibbs commented Feb 1, 2018

fivepiece commented Feb 1, 2018

promag left a comment

fivepiece commented Feb 8, 2018

fivepiece commented Feb 9, 2018

fivepiece commented Feb 16, 2018 •

edited

instagibbs commented Feb 16, 2018

fivepiece commented Feb 16, 2018

instagibbs commented Feb 16, 2018

fivepiece commented Feb 16, 2018

fivepiece commented Feb 17, 2018

instagibbs Mar 6, 2018

instagibbs Mar 6, 2018

instagibbs commented Mar 6, 2018

fivepiece commented Mar 6, 2018

instagibbs commented Mar 6, 2018

fivepiece commented Mar 6, 2018

instagibbs left a comment

instagibbs Mar 6, 2018

fivepiece Mar 6, 2018

fivepiece commented Mar 6, 2018

instagibbs commented Mar 6, 2018

laanwj Apr 7, 2018 •

edited

fivepiece commented Apr 7, 2018

laanwj commented Apr 8, 2018

p2wsh and p2sh-p2wsh address in decodescript #12321

p2wsh and p2sh-p2wsh address in decodescript #12321

Conversation

fivepiece commented Feb 1, 2018

fivepiece commented Feb 1, 2018

meshcollider commented Feb 1, 2018 • edited

fivepiece commented Feb 1, 2018 • edited

fivepiece commented Feb 1, 2018

instagibbs commented Feb 1, 2018

fivepiece commented Feb 1, 2018

promag left a comment

Choose a reason for hiding this comment

fivepiece commented Feb 8, 2018

fivepiece commented Feb 9, 2018

fivepiece commented Feb 16, 2018 • edited

instagibbs commented Feb 16, 2018

fivepiece commented Feb 16, 2018

instagibbs commented Feb 16, 2018

fivepiece commented Feb 16, 2018

fivepiece commented Feb 17, 2018

instagibbs Mar 6, 2018

Choose a reason for hiding this comment

instagibbs Mar 6, 2018

Choose a reason for hiding this comment

instagibbs commented Mar 6, 2018

fivepiece commented Mar 6, 2018

instagibbs commented Mar 6, 2018

fivepiece commented Mar 6, 2018

instagibbs left a comment

Choose a reason for hiding this comment

instagibbs Mar 6, 2018

Choose a reason for hiding this comment

fivepiece Mar 6, 2018

Choose a reason for hiding this comment

fivepiece commented Mar 6, 2018

instagibbs commented Mar 6, 2018

laanwj Apr 7, 2018 • edited

Choose a reason for hiding this comment

fivepiece commented Apr 7, 2018

laanwj commented Apr 8, 2018

meshcollider commented Feb 1, 2018 •

edited

fivepiece commented Feb 1, 2018 •

edited

fivepiece commented Feb 16, 2018 •

edited

laanwj Apr 7, 2018 •

edited