diff options
| author | Wladimir J. van der Laan <laanwj@gmail.com> | 2019-01-10 15:56:44 +0100 |
|---|---|---|
| committer | Wladimir J. van der Laan <laanwj@gmail.com> | 2019-01-10 15:56:54 +0100 |
| commit | cebe910718ae4f099f292736192a4e725ad02b94 (patch) | |
| tree | 6740bdadac251e354a5a7b4fe4498f426c6290c0 /doc | |
| parent | d7943bd008357eed01cc3a4fb7b24afcd5a523d6 (diff) | |
| parent | fa48baf23eb2ec5e9b71e3e07c24432fc1fed39c (diff) | |
Merge #15039: wallet: Avoid leaking nLockTime fingerprint when anti-fee-sniping
fa48baf23eb2ec5e9b71e3e07c24432fc1fed39c wallet: Avoid leaking locktime fingerprint when anti-fee-sniping (MarcoFalke)
453803adc9325b83fc0532a8328d455e8bdf4de6 [test] wallet_txn_clone: Correctly clone txin sequence (MarcoFalke)
Pull request description:
The wallet sets the locktime to the current height of our active chain. This is fine, as long as our node is connected to other nodes. However, when we fall back and get stuck at a particular height (e.g. taking the wallet offline), the same (potentially unique) locktime is used for all transactions. This makes it easier for passive observers to cluster transactions by wallet.
For reference, I visualized "locktime-reuse" with the data:
* blocks 545k-555k (both inclusive)
* locktimes<=60k
* excluding coinbase txs
Tree-SHA512: 2af259dd8f9f863312e2732d80ca8ba6a20c8d6d1c486b10a48479e1c85ccf13b0c38723740ebadde0f28d321cd9c133ad3e5d1e925472eb27681143bda2d0e7
Diffstat (limited to 'doc')
0 files changed, 0 insertions, 0 deletions