Roman Khimov
5b6c5af704
*: implement EncodeBinary with pointer receivers where appropriate
...
Everywhere except ParamType (which is just a byte), reduce copying things
around for no real reason.
2019-12-09 18:25:15 +03:00
Roman Khimov
7e371588a7
core/tx: remove one layer of indirection for scripts and inouts
...
It reduces heap pressure a little for these elements as we don't have to
allocate/free them individually. And they're directly tied to transactions or
block, not being shared or anything like that, so it makes little sense for
them to be pointer-based. It only makes building transactions a little easier,
but that's obviously a minor usecase.
2019-12-09 17:14:10 +03:00
Roman Khimov
b542a5e7a0
io: add support for pointer receivers in WriteArray()
...
It's actually preferable to have pointer receivers for serializable types, so
this should be supported.
2019-12-09 16:57:25 +03:00
Roman Khimov
052ba1e94f
Merge pull request #545 from nspcc-dev/feat/optimizeio
...
Remove some reflection from the io package
2019-12-09 15:12:13 +03:00
Evgenii Stratonikov
fccb008594
io: implement ReadBytes()
2019-12-09 15:00:15 +03:00
Evgenii Stratonikov
838050f8b5
io: rename ReadBytes() to ReadVarBytes()
2019-12-09 15:00:15 +03:00
Evgenii Stratonikov
f01fc1cc29
io: optimize BinWriter.WriteArray()
...
Replace reflect.MethodByName with a simple interface cast.
2019-12-09 14:59:49 +03:00
Evgenii Stratonikov
1784a14148
io: optimize BinReader.ReadArray()
...
reflect.MethodByName is a rather expensive function especially when
called on hot path. This became obvious during profiling of db restore.
This commit replaces reflection with a cast to an interface.
2019-12-09 14:58:37 +03:00
Roman Khimov
3898aadcec
rpc: simplify uint160 handling
...
Use new Uint160DecodeStringLE() function.
2019-12-06 19:47:58 +03:00
Roman Khimov
e4d821f32d
Merge pull request #546 from nspcc-dev/write-optimizations
...
Write optimizations
2019-12-06 19:40:38 +03:00
Roman Khimov
9992a98007
core: optimize balance utxo removal
...
One of my samples had 8 (out of 30) seconds spent here, but values of 100ms
are more typical. After this change it becomes invisible.
2019-12-06 18:46:24 +03:00
Roman Khimov
844491d365
*: use more efficient WriteBytes where appropriate
...
Before this patch on block import we could easily be spending more than 6
seconds out of 30 in Uint256 encoding for UnspentBalance, now it's completely
off the radar.
2019-12-06 18:22:21 +03:00
Roman Khimov
e7687d620d
io: simplify WriteBytes()
...
Which speeds it up at least twofold for a typical 32-bytes write (and that's
for a very naïve test that allocates new BufBinWriter on every iteration):
pkg: github.com/CityOfZion/neo-go/pkg/io
BenchmarkWriteBytes-8 10000000 124 ns/op
BenchmarkWriteBytesOld-8 5000000 251 ns/op
2019-12-06 17:40:47 +03:00
Evgenii Stratonikov
aa20a95181
util: add Uint256DecodeStringBE()
2019-12-06 12:16:55 +03:00
Evgenii Stratonikov
72fe884faa
util: add Uint160DecodeStringLE()
2019-12-06 12:16:55 +03:00
Evgenii Stratonikov
07e832f046
util: add Uint160DecodeBytesLE()
2019-12-06 12:16:55 +03:00
Evgenii Stratonikov
09b295d727
util: add Uint160.Reverse()
2019-12-06 12:16:55 +03:00
Evgenii Stratonikov
9e04e61533
util: make Uint160Size public
2019-12-06 12:16:55 +03:00
Evgenii Stratonikov
7179e4ba9f
util: add LE suffix to Uint256 methods
2019-12-06 12:16:55 +03:00
Evgenii Stratonikov
57efad912c
util: add LE suffix to Uint160 methods
2019-12-06 12:16:55 +03:00
Evgenii Stratonikov
138c94eda3
consensus: sign and verify consensus messages
2019-12-06 11:35:06 +03:00
Evgenii Stratonikov
765c354793
consensus: return signed messages from recovery.Get*
2019-12-06 11:33:32 +03:00
Roman Khimov
652ede03c8
Merge pull request #543 from nspcc-dev/force-neo-vm-tests-run
...
vm: force neo-vm tests presence
2019-12-05 12:42:57 +03:00
Roman Khimov
8beb135829
Merge pull request #542 from nspcc-dev/fix-storage-interop-reads
...
core: fix wrong data being read in interops (part of #501 )
2019-12-05 09:08:38 +03:00
Roman Khimov
becd4f2333
vm: force neo-vm tests presence
...
We succeeded this test without running anything real which is certainly not
good and I think we should always run this tests.
2019-12-04 20:30:27 +03:00
Roman Khimov
c034aae378
core: fix wrong data being read in interops (part of #501 )
...
When 74590551
introduced this code we had no proper caching layer, so there
were these strange fallbacks in the code. fc0031e5
should'd removed them, but
failed to do so, so do it now and fix processing of transactions that touch
storage for the same key (address) in the same block.
2019-12-04 19:51:57 +03:00
Vsevolod Brekelov
ff15a0acfd
core: add trigger types as constants. Closes #509
2019-12-04 12:27:04 +03:00
Roman Khimov
e002d7b40c
Merge pull request #538 from nspcc-dev/opcode-and-related
...
To use opcode definitions you have to import whole vm package that you might
not care about at all. So this moves opcodes to their own package under vm, fixes
and deduplicate related code and moves compiler package up one level.
2019-12-03 19:31:11 +03:00
Roman Khimov
852e6a335b
compiler: move it up from vm
...
It really deserves it, I think. Especially given that it doesn't have any
direct usage of `vm` package now.
2019-12-03 18:23:46 +03:00
Roman Khimov
31add423a8
core: replace open-coded PUSHBYTES64 with the one from opcode
2019-12-03 18:23:46 +03:00
Roman Khimov
141d27795e
rpc: remove hardcoded pushbytes64 constant
...
Using our new and shiny opcode package.
2019-12-03 18:23:46 +03:00
Roman Khimov
138e125646
*: remove duplicate functions producing verification script
...
Drop wif.GetVerificationScript(), drop
smartcontract.CreateSignatureRedeemScript(), add GetVerificationScript()
directly to the PublicKey and use it everywhere.
2019-12-03 18:23:46 +03:00
Roman Khimov
8d4dd2d2e1
vm: move opcodes into their own package
...
This allows easier reuse of opcodes and in some cases allows to eliminate
dependencies on the whole vm package, like in compiler that only needs opcodes
and doesn't care about VM for any other purpose.
And yes, they're opcodes because an instruction is a whole thing with
operands, that's what context.Next() returns.
2019-12-03 18:22:14 +03:00
Vsevolod Brekelov
5a41da0e1a
io: add missing test
2019-12-03 16:20:06 +03:00
Roman Khimov
f48228ef7d
Merge pull request #467 from nspcc-dev/errcheck_297
...
This patchset closes #297 and #457 .
2019-12-03 15:06:11 +03:00
Vsevolod Brekelov
b08387efdb
vm: use BinWriter instead of Buffered version
...
In the future we could reuse emit interface without buf better
2019-12-03 13:50:23 +03:00
Vsevolod Brekelov
f5e2401984
vm: removed logging to fix #457
2019-12-03 13:49:49 +03:00
Vsevolod Brekelov
03ff2976ed
io: refactoring for using WriteVarBytes instead of WriteLE
...
goal is to be consistent with C# implementation.
For writing []byte WriteBytes used and for byte - WriteVarByte.
2019-12-03 13:49:33 +03:00
Roman Khimov
1c89c192ac
core: optimize some accesses to unsortedTxn in mempool
...
These don't need a full lock, they only read things from maps.
2019-12-02 22:39:43 +03:00
Roman Khimov
0cd3493fa5
core: fix potential locking problem in mempool
...
I think it should fix this issue mentioned in the #526 :
INFO[1456] blockchain persist completed blockHeight=63480 headerHeight=1810000 persistedBlocks=1 persistedKeys=4 took=740.7113ms
fatal error: concurrent map read and map write
goroutine 322 [running]:
runtime.throw(0xc8a6dc, 0x21)
/usr/local/go/src/runtime/panic.go:774 +0x72 fp=0xc003473958 sp=0xc003473928 pc=0x42e282
runtime.mapaccess2(0xb706a0, 0xc0001893b0, 0xc0034739c8, 0xc0028704e0, 0x3)
/usr/local/go/src/runtime/map.go:470 +0x278 fp=0xc0034739a0 sp=0xc003473958 pc=0x40dc08
github.com/CityOfZion/neo-go/pkg/core.MemPool.ContainsKey(0xc0001d0d20, 0xc0001893b0, 0xc0001893e0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0, ...)
/neo-go/pkg/core/mem_pool.go:92 +0xcb fp=0xc003473a30 sp=0xc0034739a0 pc=0x9326db
github.com/CityOfZion/neo-go/pkg/core.(*Blockchain).HasTransaction(0xc0001ca8c0, 0x49f0b45d430e441b, 0x553db79b7072821c, 0x28969518de11976, 0xba5100efddbe79d4, 0xc003473cd0)
/neo-go/pkg/core/blockchain.go:803 +0xa1 fp=0xc003473b68 sp=0xc003473a30 pc=0x914b11
github.com/CityOfZion/neo-go/pkg/core.Blockchainer.HasTransaction-fm(0x49f0b45d430e441b, 0x553db79b7072821c, 0x28969518de11976, 0xba5100efddbe79d4, 0xc005a5d388)
/neo-go/pkg/core/blockchainer.go:28 +0x46 fp=0xc003473ba8 sp=0xc003473b68 pc=0x997326
github.com/CityOfZion/neo-go/pkg/network.(*Server).handleInvCmd(0xc00018f680, 0xd9d980, 0xc00025e190, 0xc005a5d380, 0x0, 0x0)
/neo-go/pkg/network/server.go:401 +0x3bb fp=0xc003473d38 sp=0xc003473ba8 pc=0x9924cb
github.com/CityOfZion/neo-go/pkg/network.(*Server).handleMessage(0xc00018f680, 0xd9d980, 0xc00025e190, 0xc007a0d050, 0x0, 0x0)
/neo-go/pkg/network/server.go:582 +0x1ae fp=0xc003473da0 sp=0xc003473d38 pc=0x993bbe
github.com/CityOfZion/neo-go/pkg/network.(*TCPTransport).handleConn(0xc000228420, 0xd9b880, 0xc0001b6f00)
/neo-go/pkg/network/tcp_transport.go:93 +0x202 fp=0xc003473fc8 sp=0xc003473da0 pc=0x996672
runtime.goexit()
/usr/local/go/src/runtime/asm_amd64.s:1357 +0x1 fp=0xc003473fd0 sp=0xc003473fc8 pc=0x45b3e1
created by github.com/CityOfZion/neo-go/pkg/network.(*TCPTransport).Dial
/neo-go/pkg/network/tcp_transport.go:36 +0xb4
The problem is that we're modifying `unsortedTxn` under a reader lock.
2019-12-02 22:36:59 +03:00
Roman Khimov
abc0ec33bd
Merge pull request #532 from nspcc-dev/fix/peer_count
...
Consensus fixes.
2019-12-02 18:08:36 +03:00
Evgenii Stratonikov
9ddebfdb8a
consensus: add tests for recovery message
2019-12-02 17:00:17 +03:00
Evgenii Stratonikov
b649ffb88e
consensus: set preparation hash in recovery message
2019-12-02 16:43:24 +03:00
Evgenii Stratonikov
7c900edd2d
network: count only handshaked peers as connected
2019-12-02 15:09:09 +03:00
Vsevolod Brekelov
d02673c112
vm: add bufBinWriter to emit functions in order to catch errors
2019-12-02 13:04:33 +03:00
Roman Khimov
1da76a8eb4
network: rework inventory handling, check for item presence
...
Don't ask peers about the items we already have.
2019-12-02 11:02:52 +03:00
Roman Khimov
bc5beb438f
*: gofmt
...
Somehow these crept into the repository.
2019-12-02 10:10:50 +03:00
Roman Khimov
dccf440dca
rpc: skip input/outputs mangling if no gas is attached to invocation
2019-11-29 18:00:08 +03:00
Roman Khimov
8a5ac12df8
rpc: make generic SignAndPushInvocationTx out of DeployContract
...
SignAndPushInvocationTx() is gonna be used for more than just deploying
contracts.
2019-11-29 17:59:07 +03:00
Roman Khimov
65332f5e7f
Merge pull request #529 from nspcc-dev/peer-communication-fixes
...
A set of fixes to make neo-go privnet more usable.
2019-11-29 16:29:28 +03:00