- Straight-forward call interface: `void rx_slow_hash(const char *seedhash, const void *data, size_t length, char *result_hash)`
- Consensus chain seed hash is now updated by calling `rx_set_main_seedhash` whenever a block is added/removed or a reorg happens
- `rx_slow_hash` will compute correct hash no matter if `rx_set_main_seedhash` was called or not (the only difference is performance)
- New environment variable `MONERO_RANDOMX_FULL_MEM` to force use the full dataset for PoW verification (faster block verification)
- When dataset is used for PoW verification, dataset updates don't stall other threads (verification is done in light mode then)
- When mining is running, PoW checks now also use dataset for faster verification
Unrelated, but similar code-wise to #8643. There is a check in `DNSResolver` which automatically fails to resolve hostnames which do not contain the `.` character. This PR removes that check.
Fixes#8633. The function `append_net_address` did not parse hostname + port addresses (e.g. `bar:29080`) correctly if the hostname did not contain a `'.'` character.
@vtnerd comments 1
clear up 2nd conditional statement
update_checkpoints() makes a few DNS requests and can take up to 20-30 seconds to complete (3-6 seconds on average). It is currently called from core::handle_incoming_block() which holds m_incoming_tx_lock, so it blocks all incoming transactions and blocks processing while update_checkpoints() is running. This PR moves it to until after a new block has been processed and relayed, to avoid full monerod locking.
e29b2e9 wallet2: ensure imported outputs subaddresses are created (moneromooo-monero)
1d3657a wallet2: better test on whether to allow output import (moneromooo-monero)
0cbf557 allow exporting outputs in chunks (moneromooo-monero)
b03d709 wallet2: fixes for export/import output flow (j-berman)
4b7eb57 wallet2: do not assume imported outputs must be non empty (moneromooo-monero)
5b98beb wallet2: prevent importing outputs in a hot wallet (moneromooo-monero)
0de1571 wallet2: fix missing subaddress indices in 'light' exported outputs (moneromooo-monero)
Being offline is not a good enough heuristic, so we keep track
of whether the wallet ever refreshed from a daemon, which is a
lot better, and probably the best we can do without manual user
designation (which would break existing cold wallet setups till
the user designates those wallets)
- spend secret key is no longer the sum of multisig key shares;
no need to check that is the case upon restore.
- restoring a multisig wallet from multisig info means that the
wallet must have already completed all setup rounds. Upon restore,
set the number of rounds completed accordingly.
Before the fix, it processed all transactions in the mempool which could be very slow when mempool grows to several MBs in size. I observed `get_block_template_backlog` taking up to 15 seconds of CPU time under high mempool load.
After the fix, only transactions that can potentially be mined in the next block will be processed (a bit more than the current block median weight).
ade464a ITS#9385 fix using MDB_NOSUBDIR with nonexistent file (Kris Zyp)
033a32a Remove check is_directory check on lmdb path (Howard Chu)
b096e16 Revert 'db_lmdb: test for mmap support at init time' (Howard Chu)
493577a Silence spurious fallthru warning (Howard Chu)
b46a60e Fix rawpart flag collision (Howard Chu)
4e7586c More RAWPART support (Howard Chu)
747f5d3 Preliminary raw partition support (Howard Chu)
As of OpenSSL 3.0, `SHA256_Init`, `SHA256_Update`, and `SHA256_Final`
are deprectaed in favor of the higher-level `EVP_*` class of functions.
This causes compiler warnings, and sooner or later, will cause build
errors as these functions are excluded from distro headers.
Also add some documentation.
There are vulnerabilities in multisig protocol if the parties do not
trust each other, and while there is a patch for it, it has not been
throroughly reviewed yet, so it is felt safer to disable multisig by
default for now.
If all parties in a multisig setup trust each other, then it is safe
to enable multisig.
When forced to deal with an untrusted node, a wallet will quantize
its current height to disguise the real height to the adversary, to
try and minimize the daemon's ability to distinguish returning
wallets.
Daemons will thus return more blocks than the wallet needs, starting
from earlier in the chain. These extra blocks will be disregarded
by the wallet, which had already scanned them.
However, for the purposes of reorg size detection, the wallet assumes
all blocks the daemon sends are different, which is only correct if
the wallet hasn't been coy, which is only the case for trusted
daemons (which you should use). This causes an issue when the size
of this "fake reorg" is above the sanity check threshold at which
the wallet refuses a reorg.
To fix this, the reorg size check is moved later on, when the reorg
is about to actually happen, after the wallet has checked which
blocks are actually different from the ones it expects.
While copying my data dir to another drive, I missed copying the rpc_ssl.key file b/c of the file permissions.
This change will give a much more clear, descriptive error in that scenario.
Actions:
1. Remove unused functions from misc_os_dependent.h
2. Move three remaining functions, get_gmt_time, get_ns_count, and get_tick_count into time_helper.h
3. Remove unused functions from time_helper.h
4. Refactor get_ns_count and get_internet_time_str and get_time_interval_string
5. Remove/add includes as needed
Relevant commits on the old PR:
a9fbe52b02ffab451e90c977459fea4642731cd1
9a59b131c4ed1be8afe238fff3780fe203c65a46
7fa9e2817df9b9ef3f0290f7f86357939829e588
Here lies dozens of unused files. This commit is ONLY file deletions except
for the removing of a couple of #includes and removing filenames from CmakeLists
where appropriate.
All tests were conducted on the same PC (Ryzen 5 5600X running at fixed 4.65 GHz).
Before:
test_cn_fast_hash<32> (100000 calls) - OK: 1 us/call
test_cn_fast_hash<16384> (1000 calls) - OK: 164 us/call
After:
test_cn_fast_hash<32> (100000 calls) - OK: 0 us/call
test_cn_fast_hash<16384> (1000 calls) - OK: 31 us/call
More than 5 times speedup for cn_fast_hash.
Also noticed consistent 1-2% improvement in test_construct_tx results.
Implements view tags as proposed by @UkoeHB in MRL issue
https://github.com/monero-project/research-lab/issues/73
At tx construction, the sender adds a 1-byte view tag to each
output. The view tag is derived from the sender-receiver
shared secret. When scanning for outputs, the receiver can
check the view tag for a match, in order to reduce scanning
time. When the view tag does not match, the wallet avoids the
more expensive EC operations when deriving the output public
key using the shared secret.
SChernykh