When passing around unsigned and signed transactions, outputs
and key images are passed along (outputs are passed along unsigned
transactions from the hot wallet to the cold wallet, key images
are passed along with signed transations from the cold wallet
to the hot wallet), to allow more user friendly syncing between
hot and cold wallets.
The vast majority of transactions will have just one tx pubkey,
but a bug with cold wallet signing caused two such keys to be
there, with the second one being the real one.
View wallets do not have the spend secret key, and are thus
unable to derive key images for incoming outputs. Moreover,
a previous patch set key images to zero as a means to mark
an output as having an unknown key image, so they could be
filled in when importing key images at a later time. That
later patch caused spurious collisions. We now use public
keys to detect duplicate outputs. Public keys obtained from
the blockchain are checked to be identical to the ones
derived locally, so can't be spoofed.
m_amount_out was sometimes getting initialized with the sum of
an transaction's outputs, and sometimes with the sum of outputs
that were not change. This caused confusion and bugs. We now
always set it to the sum of outputs. This reverts an earlier
fix for bad amounts as this used the other semantics. The wallet
data should be converted automatically in a percentage of cases
that I'm hesitant to estimate. In any case, restoring from seed
or keys or rebuilding the cache will get it right.
Compute derivation only once per tx, instead of once per output. Approx 33% faster while using 75% as much CPU on my machine. Note old functions in cryptonote_core (lookup_acc_outs and is_out_to_acc) are still used by tests.
The intended use is to export outputs from a hot wallet, which
can scan incoming transfers from the network, and import them
in the cold wallet, which can't. The cold wallet can then compute
key images for those outputs, which can then be exported with
export_key_images, etc.
Re-creating the transaction on the cold wallet was not splitting
the change, causing the transaction to be rejected by the network.
This worked on testnet since amounts do not have to be split.
Also add selected_transfers, which can now be saved since they're
size_t rather than iterators. This allows the view wallet to
properly set the sent outputs as spent and update balance.
Bump transfer file version numbers to match.
0950be9 wallet: speed up output selection, and fix bug with relatedness calculation (moneromooo-monero)
0eba133 wallet: fix mixup between mixin 2 and 4 before/after v5 (moneromooo-monero)
25% of the outputs are selected from the last 5 days (if possible),
in order to avoid the common case of sending recently received
outputs again. 25% and 5 days are subject to review later, since
it's just a wallet level change.
This was still using the old transaction creation algorithm,
coupled with a deterministic output selection scheme, which
made it ill suited to the job, since it'd loop indefinitely
in case the fee increased between the test tx and adding the
fee.
8b20cbf libwallet_api: do not use fast-refresh on recovery (Ilya Kitaev)
10fe626 libwallet_api: fast-refresh in case of opening non-synced wallet (Ilya Kitaev)
0019e31 libwallet_api: fix unhandled exception on address check (Ilya Kitaev)
1f73f80 libwallet_api: fast-refresh for new wallet (Ilya Kitaev)
4789347 libwallet_api: test for create/init wallet on mainnet (Ilya Kitaev)
bba6af9 wallet: cold wallet transaction signing (moneromooo-monero)
9872dcb wallet: fix log confusion between bytes and kilobytes (moneromooo-monero)
d9b0bf9 cryptonote_core: make extra field removal more generic (moneromooo-monero)
98f19d4 serialization: add support for serializing std::pair and std::list (moneromooo-monero)
This change adds the ability to create a new unsigned transaction
from a watch only wallet, and save it to a file. This file can
then be moved to another computer/VM where a cold wallet may load
it, sign it, and save it. That cold wallet does not need to have
a blockchain nor daemon. The signed transaction file can then be
moved back to the watch only wallet, which can load it and send
it to the daemon.
Two new simplewallet commands to use it:
sign_transfer (on the cold wallet)
submit_transfer (on the watch only wallet)
The transfer command used on a watch only wallet now writes an
unsigned transaction set in a file called 'unsigned_monero_tx'
instead of submitting the tx to the daemon as a normal wallet does.
The signed tx file is called 'signed_monero_tx'.
This is intended to catch traffic coming from a web browser,
so we avoid issues with a web page sending a transfer RPC to
the wallet. Requiring a particular user agent can act as a
simple password scheme, while we wait for 0MQ and proper
authentication to be merged.
This fixes misreporting of amount/fee in rct txes, as the rct
tx construction code was lumping all dests (whether change or
not) in the same dests vector, while the pre-rct code was
keeping it separate.
Keep the immediate direct deps at the library that depends on them,
declare deps as PUBLIC so that targets that link against that library
get the library's deps as transitive deps.
Break dep cycle between blockchain_db <-> crytonote_core.
No code refactoring, just hide cycle from cmake so that
it doesn't complain (cycles are allowed only between
static libs, not shared libs).
This is in preparation for supproting BUILD_SHARED_LIBS cmake
built-in option for building internal libs as shared.
We keep 1, 2, 3 multipliers till the fee decrase from 0.01/kB
to 0.002/kB, where we start using 1, 20, 166 multipliers.
This ensures the higher multiplier will compensate for the
block reward penalty when pushing past 100% of the past median.
The fee-multiplier wallet setting is now rename to priority,
since it keeps its [0..3] range, but maps to different multiplier
values.
CMake issued a warming about policy CMP0026: access of LOCATION
target property at config time was disallowed. Offending code
was the code that merged static libraries to generate
libwallet_merged.a.
This patch does that same merge task in a much simpler way. And,
since it doesn't violate the policy, the warning went away.
When RingCT is enabled, outputs from coinbase transactions
are created as a single output, and stored as RingCT output,
with a fake mask. Their amount is not hidden on the blockchain
itself, but they are then able to be used as fake inputs in
a RingCT ring. Since the output amounts are hidden, their
"dustiness" is not an obstacle anymore to mixing, and this
makes the coinbase transactions a lot smaller, as well as
helping the TXO set to grow more slowly.
Also add a new "Null" type of rct signature, which decreases
the size required when no signatures are to be stored, as
in a coinbase tx.
This allows the key to be not the same for two outputs sent to
the same address (eg, if you pay yourself, and also get change
back). Also remove the key amounts lists and return parameters
since we don't actually generate random ones, so we don't need
to save them as we can recalculate them when needed if we have
the correct keys.
This plugs a privacy leak, where the wallet tells the daemon
which transactions contain outputs for the wallet by asking
for additional information for that particular transaction.
As a nice bonus, this actually makes refresh slightly faster.
With RCT, we allow 0 size outputs, to try and encourage txes
with two inputs and two outputs. Consolidation would then
have two non zero inputs, one zero output, and one larger
output.
Before the normal selection, we attempt to find either one or two
suitable outputs to use as inputs to the rct tx. The intent is that
most rct txes will have one or two inputs, and we want all to look
the same if possible.
When two outputs are needed, we try to find a pair which are not
related (ie, by being from the same or similar block height).
The "transfer" simplewallet command is renamed to "transfer_original".
"transfer_new" is renamed "transfer", "transfer_rct" is removed,
and the new "transfer" now selects rct or non rct transactions
based on the current block height.
The mixRing (output keys and commitments) and II fields (key images)
can be reconstructed from vin data.
This saves some modest amount of space in the tx.
If the blockchain gets reorganized, all outputs spent in the part
of the blockchain that's blown away need to be reset to unspent
(they may end up spent again on the blocks that replace the blocks
that are removed, however).
99dd572 libwallet_api: tests: checking for result while opening wallet (Ilya Kitaev)
bcf7b67 libwallet_api: Wallet::amountFromString fixed (Ilya Kitaev)
32bc7b4 libwallet_api: helper method to return maximumAllowedAmount (Ilya Kitaev)
cbe534d libwallet_api: tests: removed logged passwords (Ilya Kitaev)
b1a5a93 libwallet_api: do not store wallet on close if status is not ok (Ilya Kitaev)
This plugs a privacy leak from the wallet to the daemon,
as the daemon could previously see what input is included
as a transaction input, which the daemon hadn't previously
supplied. Now, the wallet requests a particular set of
outputs, including the real one.
This can result in transactions that can't be accepted if
the wallet happens to select too many outputs with non standard
unlock times. The daemon could know this and select another
output, but the wallet is blind to it. It's currently very
unlikely since I don't think anything uses non default
unlock times. The wallet requests more outputs than necessary
so it can use spares if any of the returns outputs are still
locked. If there are not enough spares to reach the desired
mixin, the transaction will fail.
This constrains the number of instances of any amount
to the unlocked ones (as defined by the default unlock time
setting: outputs with non default unlock time are not
considered, so may be counted as unlocked even if they are
not actually unlocked).
They are used to export a signed set of key images from a wallet
with a private spend key, so an auditor with the matching view key
may see which of those are spent, and which are not.
It is not clear why libunbound was added to this in the first place,
since it wasn't here before and #915 doesn't seem to introduce any
new dependency on it.
Tested build with STATIC=OFF (with and without libunbound-dev libunbound8
installed) and STATIC=ON, on Ubuntu Trusty, Debian Jessie, and Arch
Linux. For static builds, beware of #926 and #907.
If this hack was introduced to make it build on some other system
(Windows? OS X?), then it will have to be dealt with, but not this way.
