These commands let one freeze outputs by key image, so they
do not appear in balance, nor are considered when creating
a transaction, etc
This is helpful when receiving an output from a suspected spy,
who might try to track your other outputs by seeing with what
other outputs it gets spent.
The frozen command may be used without parameters to list all
currently frozen outputs.
- import only key images generated by cold signing process
- wallet_api: trezor methods added
- wallet: button request code added
- const added to methods
- wallet2::get_tx_key_device() tries to decrypt stored tx private keys using the device.
- simplewallet supports get_tx_key and get_tx_proof on hw device using the get_tx_key feature
- live refresh enables refresh with trezor i.e. computing key images on the fly. More convenient and efficient for users.
- device: has_ki_live_refresh added
- a thread is watching whether live refresh is being computed, if not for 30 seconds, it terminates the live refresh process - switches Trezor state
- enables to perform rescan_spent / ki sync with untrusted daemon. Spent check status involves RPC calls which require trusted daemon status as it leaks information. The new call performs soft reset while preserving key images thus a sequence: refresh, ki sync / import, rescan_bc keep_ki will correctly perform spent checking without need for trusted daemon.
- useful to detect spent outputs with untrusted daemon on watch_only / multisig / hw-cold wallets after expensive key image sync.
- cli: rescan_bc keep_ki
When doing a first refresh on HW-token based wallet KI sync is required if money were received. Received money may indicate wallet was already used before the restore I.e., some transaction could have been already sent from the wallet. The spent UTXO would not be detected as spent which could lead to double spending errors on submitting a new transaction.
Thus if the wallet is HW-token based with the cold signing protocol and the first refresh detected received money the user is asked to perform the key image sync.
- simple device callback object added. Device can request passphrase/PIN entry via the callback or notify user some action is required
- callback is routed to wallet2, which routes the callback to i_wallet_callback so CLI or GUI wallets can support passphrase entry for HW tokens
- wallet: device open needs wallet callback first - passphrase protected device needs wallet callback so user can enter passphrase
* support in wallet2
* support in monero-wallet-cli
* support in monero-wallet-rpc
* support in wallet api
* support in monero-gen-trusted-multisig
* unit tests for multisig wallets creation
- device name is a new wallet property
- full device name is now a bit more structured so we can address particular device vendor + device path. Example: 'Ledger', 'Trezor:udp', 'Trezor:udp:127.0.0.1:21324', 'Trezor:bridge:usb01'. The part before ':' identifies HW device implementation, the optional part after ':' is device path to look for.
- new --hw-device parameter added to the wallet, can name the hardware device
- device reconnect added
The secret spend key is kept encrypted in memory, and
decrypted on the fly when needed.
Both spend and view secret keys are kept encrypted in a JSON
field in the keys file. This avoids leaving the keys in
memory due to being manipulated by the JSON I/O API.
If a pre-fork output is spent on both Monero and attack chain,
any post-fork output can be deduced to be a fake output, thereby
decreasing the effective ring size.
The segregate-per-fork-outputs option, on by default, allows
selecting only pre-fork outputs in this case, so that the same
ring can be used when spending it on the other side, which does
not decrease the effective ring size.
This is intended to be SET when intending to spend Monero on the
attack fork, and to be UNSET if not intending to spend Monero
on the attack fork (since it leaks the fact that the output being
spent is pre-fork).
If the user is not certain yet whether they will spend pre-fork
outputs on a key reusing fork, the key-reuse-mitigation2 option
should be SET instead.
If you use this option and intend to spend Monero on both forks,
then spend real Monero first.
This maps key images to rings, so that different forks can reuse
the rings by key image. This avoids revealing the real inputs like
would happen if two forks spent the same outputs with different
rings. This database is meant to be shared with all Monero forks
which don't bother making a new chain, putting users' privacy at
risk in the process. It is placed in a shared data directory by
default ($HOME/.shared-ringdb on UNIX like systems). You may
use --shared-ringdb-dir to override this location, and should
then do so for all Monero forks for them to share the database.
The basic approach it to delegate all sensitive data (master key, secret
ephemeral key, key derivation, ....) and related operations to the device.
As device has low memory, it does not keep itself the values
(except for view/spend keys) but once computed there are encrypted (with AES
are equivalent) and return back to monero-wallet-cli. When they need to be
manipulated by the device, they are decrypted on receive.
Moreover, using the client for storing the value in encrypted form limits
the modification in the client code. Those values are transfered from one
C-structure to another one as previously.
The code modification has been done with the wishes to be open to any
other hardware wallet. To achieve that a C++ class hw::Device has been
introduced. Two initial implementations are provided: the "default", which
remaps all calls to initial Monero code, and the "Ledger", which delegates
all calls to Ledger device.
