Commit graph

880 commits

Author SHA1 Message Date
Roman Khimov
85d0bc4544
Merge pull request #488 from nspcc-dev/dbdump_466
cli: compatible with NGD reading/writing blockchain dump
2019-11-11 13:31:32 +03:00
Vsevolod Brekelov
7b6756cb7e cli: compatible with NGD reading/writing blockchain dump 2019-11-11 13:15:34 +03:00
Roman Khimov
5c8fb1fc8e
Merge pull request #486 from nspcc-dev/dump-restore-skip-fix
cli: fix skip usage in DB dump/restore
2019-11-08 17:24:17 +03:00
Roman Khimov
4736dad8f0 cli: fix skip usage in DB dump/restore
It was failing to process any blocks at all when (skip >= count) and in other
cases it processed wrong number of blocks.
2019-11-08 17:00:02 +03:00
Roman Khimov
4d770e3c37
Merge pull request #485 from nspcc-dev/close-blockchain
core: add Close() to blockchainer, implement it to properly close chain
2019-11-08 12:24:42 +03:00
Roman Khimov
b05754deac core: add Close() to blockchainer, implement it to properly close chain
Before it the deferred function in Run() was actually never able to properly
close the Store, so we weren't synching the latest state to the disk.
2019-11-08 12:19:54 +03:00
Roman Khimov
d33083e1e1 update CHANGELOG and ROADMAP, release 0.62.0 2019-11-07 17:18:43 +03:00
Roman Khimov
52a78d8c3d
Merge pull request #484 from nspcc-dev/feature/max_integer
The last VM limit we need to implement is the restriction of BigInteger size.
This PR implements it.
Closes #373 .
2019-11-07 16:02:43 +03:00
Evgenii Stratonikov
816d78b5ee vm: redefine SHL/SHR limits based on integer size 2019-11-07 12:50:11 +03:00
Evgenii Stratonikov
439cd72294 vm: restrict BigInteger item size 2019-11-07 12:34:27 +03:00
Vsevolod
f686069f37
Merge pull request #481 from nspcc-dev/core-fix-init-corner-cases
Fix init corner cases
2019-11-06 18:36:40 +03:00
Vsevolod
11b372b45d
Merge pull request #480 from nspcc-dev/handshake-wait-for-version
network: wait for both Version messages before ACKing
2019-11-06 18:33:20 +03:00
Roman Khimov
c567307748 core: fix potential issue with header list init
If we're to receive some 500 headers (less than `headerBatchCount`) and quit
before receiving more of them we end up with clean `bc.headerList` that will
be inited going backwards to the `targetHash`, but code path doesn't add add
the `targetHash` itself which it should do in this particular case, otherwise
we end with no genesis block hash in the list.
2019-11-06 18:13:45 +03:00
Roman Khimov
d7f747fa9a network: wait for both Version messages before ACKing
Otherwise the node might crash in `startProtocol` because of missing Version
field in the peer. And it also keeps the sequence correct, Version MUST be
sent first and ACKs can only follow it.
2019-11-06 18:05:50 +03:00
Roman Khimov
b5199625cd core: init genesis block header if there are no other headers in DB
Fixes crash when restarted after the DB initialization and no blocks written
into the DB.
2019-11-06 17:58:19 +03:00
Roman Khimov
c16c2bf102 core: store current header reference on clean init
Missing it the following line could fail on subsequent restarts:

        currHeaderHeight, currHeaderHash, err := storage.CurrentHeaderHeight(bc.store)

if the node was stopped before any headers had been received.
2019-11-06 17:56:06 +03:00
Roman Khimov
c062a8e288
Merge pull request #470 from nspcc-dev/config_updates_460
config: use logpath and address for server/rpc/monitoring, fix #460.
2019-11-06 17:30:01 +03:00
Vsevolod Brekelov
d799c98cfe io: move common function and add unit tests for it 2019-11-06 17:12:33 +03:00
Roman Khimov
aeef395c45
Merge pull request #476 from nspcc-dev/feature/neoUT
VM: Use JSON-based tests from neoVM

After the implementation of stack limits nothing is needed for us to pass reference JSON tests :)
The only thing that differs --- we do not compare stack in case of FAULT (which matches NEO 3 behavior).
Also two commits were reverted to match 2.x VM behavior.
2019-11-06 16:42:06 +03:00
Evgenii Stratonikov
6c002297cd vm: implement json tests from neoVM
Add neo-vm submodule @master2.x .
Closes #196.
2019-11-06 16:22:56 +03:00
Roman Khimov
9f7625d699 core: don't do useless DB reads if nothing was persisted 2019-11-06 16:10:37 +03:00
Roman Khimov
90d4299f0b core: gofmt blockchain.go 2019-11-06 16:09:45 +03:00
Vsevolod Brekelov
11ce73af28 server: add log-path and address configuration
- LogPath can be configured through config
- node,rpc and monitoring address can be configured thought command line
or config
2019-11-06 15:58:54 +03:00
Vsevolod Brekelov
1c08753915 gomod: fix 2019-11-06 15:56:24 +03:00
Roman Khimov
79d0c7446a
Merge pull request #478 from nspcc-dev/handshake-and-peers-fix
Fixes #458, MaxPeers handling and some other related things.
2019-11-06 15:42:02 +03:00
Roman Khimov
7cf9a40468 network: fix MaxPeers, introduce AttemptConnPeers
Our node didn't respect the MaxPeers setting, fix it with a drop of random
connection when this limit is reached (to give a chance for newcomers to
communicate), but also introduce AttemptConnPeers setting to tune the number
of attempted connections.

This also raises the default MaxPeers for testnet/mainnet to 100, because
neo-go nodes love making friends.
2019-11-06 15:29:58 +03:00
Roman Khimov
31954bb20c network: disallow double connections to the same peer
Makes no sense and C# node does it too.
2019-11-06 15:29:58 +03:00
Roman Khimov
d5a7ad2c47 network: fix data race in server peers map access 2019-11-06 15:29:58 +03:00
Roman Khimov
19918f471d config: fix mainnet config ports
Reference: https://docs.neo.org/docs/en-us/node/introduction.html
2019-11-06 15:29:58 +03:00
Roman Khimov
ec76ed23a5 network: rework peer handshaking, fix #458
This allows to start handshaking from both client and server (mainnet/testnet
nodes were seen to not care about string ordering for it), but still maintains
some sane checks in the process. It also makes functions thread-safe because
we have two goroutines servicing read and write side of the Peer connection,
so they can clash on access to the struct fields.

Add a test for it also.
2019-11-06 15:29:58 +03:00
Roman Khimov
e859e03240 network: split Peer's NetAddr into RemoteAddr and PeerAddr
As they are different things used for different purposes.
2019-11-06 15:26:24 +03:00
Roman Khimov
0296184da9
Merge pull request #475 from nspcc-dev/unitTests_wallet
Improve wallet unit testing.
2019-11-06 15:24:10 +03:00
Vsevolod Brekelov
75104a4a34 wallet: add unit tests 2019-11-06 13:15:47 +03:00
Evgenii Stratonikov
7d40d2f71e vm: make StepOut/StepOver match original VM behavior 2019-11-06 12:25:04 +03:00
Evgenii Stratonikov
9ebb793009 vm: revert SUBSTR offset behavior to NEO 2.x 2019-11-06 12:15:55 +03:00
Evgenii Stratonikov
4c688355bc vm: revert bool -> []byte conversion to NEO 2.x 2019-11-06 12:15:48 +03:00
Roman Khimov
5544ff1768
Merge pull request #462 from nspcc-dev/feature/stack_size
Restrict total stack item count in the VM. Refs. #373.
2019-11-06 11:09:13 +03:00
067d9655bf vm: restrict total stack item count 2019-11-06 11:03:43 +03:00
Roman Khimov
fb2fe0a408
Merge pull request #472 from nspcc-dev/feature/serialize
VM: Implement serialization and deserealization interops

This PR implements Serialize/Deserialize interops.

Closes #419.
2019-11-05 19:18:08 +03:00
ed758458d5 vm: implement serialization interops in core package 2019-11-05 17:10:52 +03:00
412582dc78 vm: add more serialization tests 2019-11-05 17:04:14 +03:00
59f9c2bddc vm: implement Array and Struct item serialization 2019-11-05 16:58:09 +03:00
e1d019e087 vm: implement Map item serialization 2019-11-05 16:57:48 +03:00
cd690803cf vm: implement BigInteger item serialization 2019-11-05 15:15:43 +03:00
25f77257ce vm: implement Boolean and ByteArray item serialization 2019-11-05 15:15:39 +03:00
Vsevolod
f48b19a305
Merge pull request #471 from nspcc-dev/rpc-minor-fixes
Minor RPC fixes
2019-11-02 22:13:17 +03:00
Roman Khimov
de2eeb4671 rpc: add one to the block height for the getblockcount response
There is a difference in interpretation of what a block count is. neo-go nodes
currently respond to this request with the latest block number which is the
same number that neoscan.io shows. However, C# nodes deliberately do add one
to this number when answering to the getblockcount request to account for the
genesis block number 0.

This patch makes us consistent with C# nodes wrt to getblockcount behaviour.
2019-11-01 20:13:00 +03:00
Roman Khimov
70e4529848 rpc: add 0x prefix to the getbestblockhash answer
For consistency with C# nodes that do add it to their answers.
2019-11-01 20:12:44 +03:00
Roman Khimov
54bb9b94c7 CHANGELOG: fix Prometheus bug number 2019-11-01 15:33:46 +03:00
Roman Khimov
8730f56785 update CHANGELOG and ROADMAP, release 0.61.0 2019-11-01 14:36:53 +03:00