To do so we update CValidationInterface::BlockDisconnect to take a
CBlockIndex pointing to the block being disconnected.
This new parameter will be use in the following commit to establish
wallet height.
4671fc3d9e Expand on wallet_balance.py comment from https://github.com/bitcoin/bitcoin/pull/16766\#issuecomment-527563982 (Jeremy Rubin)
91f3073f08 Update release notes to mention changes to IsTrusted and impact on wallet (Jeremy Rubin)
8f174ef112 Systematize style of IsTrusted single line if (Jeremy Rubin)
b49dcbedf7 update variable naming conventions for IsTrusted (Jeremy Rubin)
5ffe0d1449 Update comment in test/functional/wallet_balance.py (Jeremy Rubin)
a550c58267 Update wallet_balance.py test to reflect new behavior (Jeremy Rubin)
5dd7da4ccd Reuse trustedParents in looped calls to IsTrusted (Jeremy Rubin)
595f09d6de Cache tx Trust per-call to avoid DoS (Jeremy Rubin)
dce032ce29 Make IsTrusted scan parents recursively (Jeremy Rubin)
Pull request description:
This slightly modifies the behavior of IsTrusted to recursively check the parents of a transaction. Otherwise, it's possible that a parent is not IsTrusted but a child is. If a parent is not trusted, then a child should not be either.
This recursive scan can be a little expensive, so ~it might be beneficial to have a way of caching IsTrusted state, but this is a little complex because various conditions can change between calls to IsTrusted (e.g., re-org).~ I added a cache which works per call/across calls, but does not store the results semi-permanently. Which reduces DoS risk of this change. There is no risk of untrusted parents causing a resource exploitation, as we immediately return once that is detected.
This is a change that came up as a bug-fix esque change while working on OP_SECURETHEBAG. You can see the branch where this change is important here: https://github.com/bitcoin/bitcoin/compare/master...JeremyRubin:stb-with-rpc?expand=1. Essentially, without this change, we can be tricked into accepting an OP_SECURETHEBAG output because we don't properly check the parents. As this was a change which, on its own, was not dependent on OP_SECURETHEBAG, I broke it out as I felt the change stands on its own by fixing a long standing wallet bug.
The test wallet_balance.py has been corrected to meet the new behavior. The below comment, reproduced, explains what the issue is and the edge cases that can arise before this change.
# Before `test_balance()`, we have had two nodes with a balance of 50
# each and then we:
#
# 1) Sent 40 from node A to node B with fee 0.01
# 2) Sent 60 from node B to node A with fee 0.01
#
# Then we check the balances:
#
# 1) As is
# 2) With transaction 2 from above with 2x the fee
#
# Prior to #16766, in this situation, the node would immediately report
# a balance of 30 on node B as unconfirmed and trusted.
#
# After #16766, we show that balance as unconfirmed.
#
# The balance is indeed "trusted" and "confirmed" insofar as removing
# the mempool transactions would return at least that much money. But
# the algorithm after #16766 marks it as unconfirmed because the 'taint'
# tracking of transaction trust for summing balances doesn't consider
# which inputs belong to a user. In this case, the change output in
# question could be "destroyed" by replace the 1st transaction above.
#
# The post #16766 behavior is correct; we shouldn't be treating those
# funds as confirmed. If you want to rely on that specific UTXO existing
# which has given you that balance, you cannot, as a third party
# spending the other input would destroy that unconfirmed.
#
# For example, if the test transactions were:
#
# 1) Sent 40 from node A to node B with fee 0.01
# 2) Sent 10 from node B to node A with fee 0.01
#
# Then our node would report a confirmed balance of 40 + 50 - 10 = 80
# BTC, which is more than would be available if transaction 1 were
# replaced.
The release notes have been updated to note the new behavior.
ACKs for top commit:
ariard:
Code Review ACK 4671fc3, maybe extend DoS protection in a follow-up PR.
fjahr:
Code review ACK 4671fc3d9e
ryanofsky:
Code review ACK 4671fc3d9e. Changes since last review: 2 new commits adding suggested release note and python test comment, also a clean rebase with no changes to the earlier commits. The PR description is more comprehensive now, too. Looks good!
promag:
Code review ACK 4671fc3d9e.
Tree-SHA512: 6b183ff425304fef49724290053514cb2770f4a2350dcb83660ef24af5c54f7c4c2c345b0f62bba60eb2d2f70625ee61a7fab76a7f491bb5a84be5c4cc86b92f
ScriptPubKeyMan is only using UnsetWalletFlagWithDB to unset the blank
wallet flag. Just make that it's own function and not expose the flag
writing directly.
This does not change behavior.
This moves CWallet members and methods dealing with keys to a new
LegacyScriptPubKeyMan class, and updates calling code to reference the new
class instead of CWallet.
Most of the changes are simple text replacements and variable substitutions
easily verified with:
git log -p -n1 -U0 --word-diff-regex=.
The only nontrivial chunk of code added is the new LegacyScriptPubKeyMan class
declaration, but this code isn't new and is just selectively copied and moved
from the previous CWallet class declaration. This can be verified with:
git log -p -n1 --color-moved=dimmed_zebra src/wallet/scriptpubkeyman.h src/wallet/wallet.h
or
git diff HEAD~1:src/wallet/wallet.h HEAD:src/wallet/scriptpubkeyman.h
This commit does not change behavior.
Start moving wallet and ismine code to scriptpubkeyman.h, scriptpubkeyman.cpp
The easiest way to review this commit is to run:
git log -p -n1 --color-moved=dimmed_zebra
And check that everything is a move (other than includes and copyrights comments).
This commit is move-only and doesn't change code or affect behavior.
9e95931865 [wallet] Remove `state` argument from CWallet::CommitTransaction (John Newbery)
d1734f9a3b [wallet] Remove return value from CommitTransaction() (John Newbery)
b6f486a02b [wallet] Add doxygen comment to CWallet::CommitTransaction() (John Newbery)
8bba91b22d [wallet] Fix whitespace in CWallet::CommitTransaction() (John Newbery)
Pull request description:
`CommitTransaction()` returns a bool to indicate success, but since commit
b3a7410 (#9302) it only returns true, even if the transaction was not
successfully broadcast. This commit changes CommitTransaction() to return
void.
All dead code in `if (!CommitTransaction())` branches has been removed.
Two additional commits fix up the idiosyncratic whitespace in `CommitTransaction` and add a doxygen comment for the function.
ACKs for top commit:
laanwj:
ACK 9e95931865
Tree-SHA512: a55a2c20369a45222fc0e02d0891495655a926e71c4f52cb72624768dd7b9c1dca716ea67d38420afb90f40c6e0fd448caa60c18fd693bb10ecb110b641820e6
facec1c643 wallet: Avoid showing GUI popups on RPC errors (MarcoFalke)
Pull request description:
RPC errors and warnings are shown as popups in the GUI instead of being returned to the RPC caller. For example,
```
$ ./src/bitcoin-cli loadwallet $(pwd)/./test/functional/data/wallets/high_minversion/
error code: -4
error message:
Wallet loading failed.
```
gives me a GUI popup and no reason why loading the wallet failed.
After this pull request:
```
$ ./src/bitcoin-cli loadwallet $(pwd)/./test/functional/data/wallets/high_minversion/
error code: -4
error message:
Wallet loading failed: Error loading /home/marco/workspace/btc_bitcoin_core/./test/functional/data/wallets/high_minversion/wallet.dat: Wallet requires newer version of Bitcoin Core
ACKs for top commit:
laanwj:
Code review ACK facec1c643
Tree-SHA512: c8274bbb02cfcf71676eeec1e773e51fb3538cf93f82e7cb8536f4716d44ed819cdc162dfc039ac7386a4db381a734cdb27fd32567043a1180c02519fbcba194
CommitTransaction returns a bool to indicate success, but since commit
b3a74100b8 it only returns true, even if the transaction was not
successfully broadcast. This commit changes CommitTransaction() to return
void.
All dead code in `if (!CommitTransaction())` branches has been removed.
The wallet should not be able to directly access global configuration
from the node. Remove access of "-limitancestorcount" and
"-limitdescendantcount".
ea4cc3a7b3 Truly decouple wallet from chainparams for -fallbackfee (Jorge Timón)
Pull request description:
Before it was 0 by default for main and 20000 for test and regtest.
Now it is 0 by default for all chains, thus there's no need to call Params().
Also now the default for main is properly documented.
Suggestion for release notes:
-fallbackfee was 0 (disabled) by default for the main chain, but 20000 by default for the test chains. Now it is 0 by default for all chains. Testnet and regtest users will have to add fallbackfee=20000 to their configuration if they weren't setting it and they want it to keep working like before.
Should I propose them to the wiki for the release notes or only after merge?
For more context, see https://github.com/bitcoin/bitcoin/pull/16402#issuecomment-515701042
ACKs for top commit:
MarcoFalke:
ACK ea4cc3a7b3
Tree-SHA512: fdfaba5d813da4221e405e0988bef44f3856d10f897a94f9614386d14b7716f4326ab8a6646e26d41ef3f4fa61b936191e216b1b605e9ab0520b0657fc162e6c
Before it was 0 by default for main and 20000 for test and regtest.
Now it is 0 by default for all chains, thus there's no need to call Params().
Also now the default for main is properly documented
fa734603b7 wallet: Fix segmentation fault in CreateWalletFromFile (MarcoFalke)
fab3c34412 test: Print both messages on failure in assert_raises_message (MarcoFalke)
faa13539d5 wallet: Fix documentation around WalletParameterInteraction (MarcoFalke)
Pull request description:
Comes with a test to aid review. The test should fail without the fix to bitcoind
The following `CreateWalletFromFile` issues are fixed:
* `walletFile` refers to freed memory and will thus corrupt the debug.log and/or crash the node if read
* `WalletParameterInteraction` was moved to `CreateWalletFromFile` and `WalletInit::ParameterInteraction` without updating the documentation
ACKs for top commit:
promag:
ACK fa734603b7.
darosior:
ACK fa734603b7
meshcollider:
LGTM, code-read ACK fa734603b7
Tree-SHA512: 2aceb63a3f25b90a840cfa08d37f5874aad4eb3df8c2ebf94e2ed18b55809b185e6920bdb345b988bff1fcea5e68a214fe06c361f7da2c01a3cc29e0cc421cb4
442a87cc0a Add a test wallet_reorgsrestore (Antoine Riard)
40ede992d9 Modify wallet tx status if has been reorged out (Antoine Riard)
7e89994133 Remove SyncTransaction for conflicted txn in CWallet::BlockConnected (Antoine Riard)
a31be09bfd Encapsulate tx status in a Confirmation struct (Antoine Riard)
Pull request description:
While working on #15931, I've tried to rationalize tx state management to ease integration of block height tracking per-wallet tx. We currently rely on a combination of `hashBlock` and `nIndex` with magic value to determine tx confirmation, conflicted or abandoned state. It's hard to reason and error-prone. To solve that, we encapsulate these fields in a `TxConfirmation` struct and introduce a `TxState` member that we update accordingly at block connection/disconnection.
Following jnewbery [recommendation](https://github.com/bitcoin/bitcoin/pull/15931#discussion_r312576506), I've taken these changes in its own commit, and open a PR to get them first. It would ease review of aforementioned PR, but above all should ease fixing of long-term issues like :
* https://github.com/bitcoin/bitcoin/issues/7315 (but maybe we should abandon abandontransaction or relieve it to only free outpoints not track the transaction as abandoned in itself, need its own discussion)
* https://github.com/bitcoin/bitcoin/issues/8692 where we should cancel conflicted state of transactions chain smoothly
* `MarkConflicted` in `LoadToWallet` is likely useless if we track conflicts rights at block connection
Main changes of this PR to get right are tx update in `AddToWallet` and serialization/deserialization logic.
ACKs for top commit:
meshcollider:
Light re-Code Review ACK 442a87cc0a
ryanofsky:
utACK 442a87cc0a. Changes since last review are switching from `hasChain` to `LockChain` and removing chain lock in `WalletBatch::LoadWallet` that's redundant with the new lock still added in `CWallet::LoadWallet`, and fixing python test race condition.
Tree-SHA512: 029209e006de0240436817204e69e548c5665e2b0721b214510e7aba7eba130a5eab441d3a1ad95bd6426114dd27390492c77bf4560a9610009b32cd0a1f72f7
fa61365a13 wallet: Translate all initErrors in CreateWalletFromFile (MarcoFalke)
fa70d199d0 util: Make util/error bilingual_str (refactor) (MarcoFalke)
Pull request description:
The translations are going to close in three days (#15940), so I am submitting this as a standalone pull request.
Those changes are part of a bugfix #16661, which includes a test. The first change (the refactor) is required, the second commit is not. I am happy to drop it, if needed.
ACKs for top commit:
laanwj:
utACK fa61365a13
hebasto:
ACK fa61365a13, I have tested the code on Linux Mint 19.2.
Tree-SHA512: a7616cc38b9ffd301c6b915ea808a65815c3d97e9f57ec091772eb260e5cf0d75a13a6e4dfa3913e236833677c7929b9a748cb7d7a0e406d51749944b614e11b
d9d8984270 wallet: Use wallet name instead of pointer on unload/release (João Barbosa)
Pull request description:
Fixes#16668. Wallet name is unique so it can be used instead of pointer.
ACKs for top commit:
meshcollider:
utACK d9d8984270
instagibbs:
utACK d9d8984270
ryanofsky:
utACK d9d8984270. Alternately I think it might be possible to use an intptr_t set instead of a string set to get around the undefined behavior described in the issue.
Tree-SHA512: eccd4d260cd4c02b52c30deeb32dbfd190a1151a5340eb3aa4ece0dc6ae3b3ed746ce5617336461f6f27c437c435629cd07d20beb1c5450f23b75edde6728598
Add a LockChain method to CWallet to know if we can lock or query
chain state safely.
At tx loading, we rely on chain to know if hashBlock of tx is still
in main chain. If not, we set its status to unconfirmed and reset
its hashBlock/nIndex.
If wallet loaded is the wallet-tool one, all wallet txn will
show up with a height of zero. It doesn't matter as status is not
used by wallet-tool.
We take lock prematurely in CWallet::LoadWallet and CWallet::Verify
to ensure that lock order is respected between cs_main an cs_wallet.
The function ExtractPubKey() checks if a given script matches the P2PK pattern
(<PubKey> OP_CHECKSIG), extracts the PubKey and additionally checks if it is
cryptographically valid (full validation with ECC library via .IsFullyValid()).
Currently this is done manually in the following order:
1) check if first script OP is data push with valid PubKey length
(first part of pattern match), extract PubKey
2) create CPubKey object with extracted PubKey
3) fully validate public key
4) check if last script OP is OP_CHECKSIG
(second part of pattern match)
Using Solver, the pattern matching and PubKey extraction can be done via a
single step, leading to the following simplified order with shorter code:
1) check if given script matches P2PK pattern with Solver
(also contains valid PubKey length check), extracts Pubkey
2) create CPubKey object with extracted Pubkey
3) fully validate public key
Translated strings should not end up in the debug log, stderr, or
returned by an RPC.
Changing the util methods in util/error to return a bilingual_str paves
the way to achieve this goal in the long term.
We shouldn't rely on this sync call to get an accurate view of txn
state, if a tx conflicts with one in mapTx we are going to update
our wallet dependencies in AddToWalletIfInvolvingMe while conflicting
txn get connected. If it doesn't conflict with one of our dependencies
we are not going to track it anyway.
This is a cleanup, as this SyncTransaction is redundant with the
following one for confirmation which is triggering the MarkConflicted
logic. We keep the loop because set of conflicted txn isn't same as txn
included in block.
Instead of relying on combination of hashBlock and nIndex
values to manage tx in its lifecycle, we introduce 4
status : CONFIRMED, UNCONFIRMED, CONFLICTED, ABANDONED.
hashBlock and nIndex magic values should only be used at
serialization/deserialization for backward-compatibility.
At block disconnection, we know flag txn as UNCONFIRMED where
previously they kept their states until being override by a
block connection or abandontransaction call. This is a change
in behavior for which user may have to call abandon twice
if transaction is disconnected and not accepted back in the mempool.
We assert status transitioning right in AddToWallet. Doing so
flagged a misbehavior in ComputeTimeSmart unit test where same
tx is confirmed twice in different block. To avoid inconsistencies
we unconfirmed tx before new connection in different block. We
also remove a cs_main lock in test, as AddToWallet and its
callees don't rely on locked chain.
Restores the confirmed/conflicted tx check removed in
8753f5652b. There should be no external
behaviour change (these txs would not get accepted to the mempool
anyway), but not having the check in the wallet causes log spam.
Also adds a comment to ResentWalletTransactions() that
confirmed/conflicted tx check is done in SubmitMemoryPoolAndRelay().
