Commit graph

3866 commits

Author SHA1 Message Date
Roman Khimov
96339e9f64
Merge pull request #2290 from nspcc-dev/microopt
Trivial microoptimizations
2021-12-02 19:21:45 +03:00
Roman Khimov
eb27699dad vm: put child invocation tree nodes into call element
As neo-project/neo-modules#657 currently does.
2021-12-02 17:46:48 +03:00
Roman Khimov
41ab7a85a7 hash: avoid allocation in RipeMD160 calculation 2021-12-02 16:11:09 +03:00
Roman Khimov
53cfde62de vm: don't allocate another stack for entry scripts
vm.New() creates a new stack, then we load an entry script with
LoadScriptWithFlags and it creates another one which doesn't make much
sense. rvcount is -1 for it, so all elements are to be copied down anyway and
it's clear so a new loaded script can't dig down to anything it shouldn't be
able to.
2021-12-02 15:59:51 +03:00
Roman Khimov
ac7e922606 native: avoid big.Int allocation in burn()
addTokens and incBalance only read the amount, so it's not a problem.
2021-12-02 15:40:43 +03:00
Roman Khimov
a003879b84 native: avoid allocations in increaseBalance 2021-12-02 15:33:30 +03:00
Roman Khimov
ffeb3b8473 transaction: microoptimize Hash()
It doesn't cost much, but it's used _a lot_, so optimizing it makes sense.

name      old time/op    new time/op    delta
TxHash-8    4.89ns ± 5%    0.54ns ± 2%  -88.86%  (p=0.008 n=5+5)

name      old alloc/op   new alloc/op   delta
TxHash-8     0.00B          0.00B          ~     (all equal)

name      old allocs/op  new allocs/op  delta
TxHash-8      0.00           0.00          ~     (all equal)
2021-12-02 14:48:35 +03:00
Leonard Lyubich
3381716ebe oracle/neofs: Use function which combines all NeoFS client errors
Add function which unifies NeoFS API client's errors with failed status
reponses.

Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
2021-12-02 11:59:25 +03:00
Leonard Lyubich
0c4017cc20 oracle/neofs: Upgrade NeoFS dependencies
Replace `neofs-api-go` module with `neofs-sdk-go`. Adapt to NeoFS
response statuses in the implementation of NeoFS oracle.

Signed-off-by: Leonard Lyubich <leonard@nspcc.ru>
2021-12-02 11:59:20 +03:00
Roman Khimov
aa1da9de5a vm: preallocate for istack
Avoid growing it in runtime.
2021-12-01 21:36:35 +03:00
Roman Khimov
decff2bfb9 stackitem: simplify extracting value from map
Going through interfaces doesn't make it faster.
2021-12-01 21:36:35 +03:00
Roman Khimov
af8a4da6d9 stackitem: use type switch more effectively
Reduce the number of allocations, save some time.

name               old time/op    new time/op    delta
EncodeBinary-8       51.2µs ± 8%    44.6µs ± 7%  -12.79%  (p=0.008 n=5+5)
SerializeSimple-8     523ns ± 5%     449ns ± 4%  -14.27%  (p=0.008 n=5+5)

name               old alloc/op   new alloc/op   delta
EncodeBinary-8        329kB ± 0%     329kB ± 0%   -0.12%  (p=0.008 n=5+5)
SerializeSimple-8      432B ± 0%      408B ± 0%   -5.56%  (p=0.008 n=5+5)

name               old allocs/op  new allocs/op  delta
EncodeBinary-8         36.0 ± 0%      20.0 ± 0%  -44.44%  (p=0.008 n=5+5)
SerializeSimple-8      7.00 ± 0%      6.00 ± 0%  -14.29%  (p=0.008 n=5+5)
2021-12-01 21:36:35 +03:00
Roman Khimov
418ba1dbc3 *: use internal variables for simple big.NewInt() values
Avoid additional allocations.
2021-12-01 21:36:25 +03:00
Roman Khimov
3eed9d06f8 stackitem: add some hint to 'seen' maps
It's empirical, we usually have one container, but four is likely to fit most
of regular cases.
2021-12-01 21:36:25 +03:00
Roman Khimov
9576e10d04 storage: use map size hints to optimize subsequent persist()
We're likely to have something comparable to the current changeset in the
subsequent one. If it's bigger, no big deal, it'll be reallocated, if it's
smaller, no big deal, the next one will be preallocated smaller.
2021-12-01 21:36:25 +03:00
Roman Khimov
428ba48444 mempool: use capacity parameter to avoid map reallocations
Which is what happens in AddBlock that fills a new pool with exactly the
number of transactions we know already.
2021-12-01 21:36:25 +03:00
Roman Khimov
e236ac4067 io: optimize WriteString, avoid allocation
name                          old time/op    new time/op    delta
AppExecResult_EncodeBinary-8     477ns ± 0%     467ns ± 3%     ~     (p=0.190 n=4+5)

name                          old alloc/op   new alloc/op   delta
AppExecResult_EncodeBinary-8      376B ± 0%      368B ± 0%   -2.13%  (p=0.008 n=5+5)

name                          old allocs/op  new allocs/op  delta
AppExecResult_EncodeBinary-8      5.00 ± 0%      4.00 ± 0%  -20.00%  (p=0.008 n=5+5)
2021-12-01 21:36:25 +03:00
Roman Khimov
748a70569a mempool: use uint256 library for fee tracking
It's very effective in avoiding allocations for big.Int, we don't have a
microbenchmark for memppol, but this improves TPS metrics by ~1-2%, so it's
noticeable.
2021-12-01 21:36:25 +03:00
Roman Khimov
ee05f73b6f vm: simplify slot and make it private
Hiding refcounter inside a slot is actually a good idea, but it makes the
structure somewhat bigger, especially given that the refcounter is the same
and belongs more to VM or Context. New structure is a bit more efficient:

name                    old time/op    new time/op    delta
ScriptFibonacci-8          672µs ± 2%     644µs ± 0%  -4.15%  (p=0.008 n=5+5)
ScriptNestedRefCount-8    1.08ms ± 1%    1.05ms ± 2%  -2.56%  (p=0.008 n=5+5)
ScriptPushPop/4-8         1.52µs ± 1%    1.47µs ± 1%  -3.14%  (p=0.008 n=5+5)
ScriptPushPop/16-8        3.66µs ± 1%    3.54µs ± 1%  -3.24%  (p=0.008 n=5+5)
ScriptPushPop/128-8       24.7µs ± 1%    23.2µs ± 1%  -6.14%  (p=0.008 n=5+5)
ScriptPushPop/1024-8       183µs ± 1%     173µs ± 1%  -5.01%  (p=0.008 n=5+5)

name                    old alloc/op   new alloc/op   delta
ScriptFibonacci-8          114kB ± 0%     114kB ± 0%    ~     (p=0.079 n=4+5)
ScriptNestedRefCount-8     241kB ± 0%     241kB ± 0%    ~     (p=0.333 n=5+4)
ScriptPushPop/4-8           160B ± 0%      160B ± 0%    ~     (all equal)
ScriptPushPop/16-8          640B ± 0%      640B ± 0%    ~     (all equal)
ScriptPushPop/128-8       8.70kB ± 0%    8.70kB ± 0%    ~     (all equal)
ScriptPushPop/1024-8      73.2kB ± 0%    73.2kB ± 0%    ~     (all equal)

name                    old allocs/op  new allocs/op  delta
ScriptFibonacci-8          3.17k ± 0%     3.17k ± 0%  -0.03%  (p=0.008 n=5+5)
ScriptNestedRefCount-8     10.7k ± 0%     10.7k ± 0%    ~     (all equal)
ScriptPushPop/4-8           8.00 ± 0%      8.00 ± 0%    ~     (all equal)
ScriptPushPop/16-8          32.0 ± 0%      32.0 ± 0%    ~     (all equal)
ScriptPushPop/128-8          259 ± 0%       259 ± 0%    ~     (all equal)
ScriptPushPop/1024-8       2.05k ± 0%     2.05k ± 0%    ~     (all equal)

It'd be especially nice to internalize static slot, but as we can't compare
slices it's not possible.
2021-12-01 21:36:25 +03:00
Roman Khimov
a0473aca92
Merge pull request #2283 from nspcc-dev/fix-vm-tests
vm: fix cli tests
2021-12-01 11:35:41 +03:00
Roman Khimov
01d15ff473
Merge pull request #2270 from nspcc-dev/vm-invoked-contracts
Add invoked contract tracing
2021-12-01 11:27:27 +03:00
AnnaShaleva
c67d249a8d vm: fix cli tests
* Remove extra symbols during command execution.
* Fix test contract paths.
2021-12-01 11:14:29 +03:00
Roman Khimov
c663d3f39f
Merge pull request #2279 from nspcc-dev/minor-fixes
services: fix Oracle response creation algorithm
2021-11-30 18:15:50 +03:00
Roman Khimov
dbe184a335
Merge pull request #2280 from nspcc-dev/compiler-struct-cast
compiler: allow to slice struct field
2021-11-30 18:14:56 +03:00
Roman Khimov
f12f871432 compiler: use new defer stack for inlined functions, fix #2281
Inlined functions shouldn't care about outer scope defer statements, they
can't.
2021-11-30 18:04:02 +03:00
Roman Khimov
cddd5e1f6d compiler: add a panic on internal lambda inconsistency
It's a bit better than behavior observer in #2281 (which is very cryptic).
2021-11-30 18:02:38 +03:00
Evgeniy Stratonikov
66a4245bef compiler: allow to slice struct field
It makes no sense to restrict to identifiers.
2021-11-30 15:48:46 +03:00
Anna Shaleva
1d1538c566 services: fix Oracle response transaction creation
Problem: transactions with wrong hashes are accepted to the chain if
consensus nodes are designated as Oracle nodes. The result is wrong
MerkleRoot for the accepted block. Consensus nodes got such blocks
right from the dbft and store them without errors, but if
non-consensus nodes are present in the network, they just can't accept
these "bad" blocks:
```
2021-11-29T12:56:40.533+0300	WARN	blockQueue: failed adding block into the blockchain	{"error": "invalid block: MerkleRoot mismatch (expected a866b57ad637934f7a7700e3635a549387e644970b42681d865a54c3b3a46122, calculated d465aafabaf4539a3f619d373d178eeeeab7acb9847e746e398706c8c1582bf8)", "blockHeight": 17, "nextIndex": 18}
```

This problem happens because of transaction hash caching. We can't set
transaction hash if transaction construction wasn't yet completed.
2021-11-30 10:43:58 +03:00
Roman Khimov
33e37e60e5
Merge pull request #2264 from nspcc-dev/fix-win-tests
*: Windows compatibility fixes
2021-11-29 11:25:35 +03:00
AnnaShaleva
aefb6f9fee *: fix tests failing due to path.Join usage
Solution:
Use `file/filepath` package to construct expected path. This package is OS-aware, see https://github.com/golang/go/issues/30616.
2021-11-29 11:11:09 +03:00
Anna Shaleva
6cdb701a9d vm: improve error message for (*executor).checkError
It is helpful if something goes wrong.
2021-11-29 11:09:39 +03:00
AnnaShaleva
7c48177bf7 io: fix TestMakeDirForFile* tests
Problem:
```
--- FAIL: TestMakeDirForFile_HappyPath (0.01s)
    testing.go:894: TempDir RemoveAll cleanup: remove C:\Users\Anna\AppData\Local\Temp\TestMakeDirForFile_HappyPath402638411\001\testDir\testFile.test: The process cannot access the file because it is being used by another process.
--- FAIL: TestMakeDirForFile_Negative (0.01s)
    testing.go:894: TempDir RemoveAll cleanup: remove C:\Users\Anna\AppData\Local\Temp\TestMakeDirForFile_Negative672737582\001\testFile.test: The process cannot access the file because it is being used by another process.
FAIL
```

Solution:
Release resources occupied by os.Create.
2021-11-26 18:26:27 +03:00
AnnaShaleva
06beb4d534 core: fix TestMemCachedPersist test
Problem:
```
--- FAIL: TestMemCachedPersist (0.07s)
    --- FAIL: TestMemCachedPersist/BoltDBStore (0.07s)
        testing.go:894: TempDir RemoveAll cleanup: remove C:\Users\Anna\AppData\Local\Temp\TestMemCachedPersist_BoltDBStore294966711\001\test_bolt_db: The process cannot access the file because it is being used by another process.
```

Solution:
Release the resources occupied by the DB.
2021-11-26 18:26:27 +03:00
Anna Shaleva
e2ef904901 config: use http for oracle nodes
Otherwise the following error occurs:
```
first path segment in URL cannot contain colon
```
2021-11-26 17:17:10 +03:00
Roman Khimov
b455e9338c
Merge pull request #2275 from nspcc-dev/add-more-oracle-logs
oracle: unify filtering and add log for error case
2021-11-25 09:51:48 +03:00
Evgeniy Stratonikov
fac595bbdf core: remove old storage items asynchronously
Signed-off-by: Evgeniy Stratonikov <evgeniy@nspcc.ru>
2021-11-24 16:25:11 +03:00
Evgeniy Stratonikov
582d489c90 dao: add P2PStateExchangeExtensions setting to dao.Version
Signed-off-by: Evgeniy Stratonikov <evgeniy@nspcc.ru>
2021-11-24 16:22:27 +03:00
Evgeniy Stratonikov
f1767f361d dao: add KeepOnlyLatestState setting to dao.Version
Signed-off-by: Evgeniy Stratonikov <evgeniy@nspcc.ru>
2021-11-24 16:22:27 +03:00
Evgeniy Stratonikov
f7e2d3d717 dao: add stateroot-related settings to Version
Signed-off-by: Evgeniy Stratonikov <evgeniy@nspcc.ru>
2021-11-24 16:22:27 +03:00
Evgeniy Stratonikov
856e9cf67b statesync: copy state by swapping prefix
Signed-off-by: Evgeniy Stratonikov <evgeniy@nspcc.ru>
2021-11-24 16:22:27 +03:00
Evgeniy Stratonikov
6c5a7d9b29 dao: include settings in Version
Signed-off-by: Evgeniy Stratonikov <evgeniy@nspcc.ru>
2021-11-24 16:22:26 +03:00
Evgeniy Stratonikov
856385b106 dao: use custom storage prefix
We use 2 prefixes for storing items because of state synchronization.
This commit allows to parametrize dao with the default prefix.

Signed-off-by: Evgeniy Stratonikov <evgeniy@nspcc.ru>
2021-11-24 16:22:07 +03:00
Roman Khimov
6de6415a53 oracle: unify filtering and add log for error case
Just to simplify debugging.
2021-11-23 18:43:15 +03:00
Roman Khimov
019af7de46
Merge pull request #2271 from nspcc-dev/fix-sync-oom
core: add and use synchronous persist to avoid OOM
2021-11-23 12:37:48 +03:00
Roman Khimov
f58d424c6d rpc/client: fix custom* scope processing in CreateTxFromScript
Copying just the scope doesn't work for CustomContracts, CustomGroups and
Rules because they all contain additional metadata. Thanks @mialbu for
reporting this.
2021-11-22 12:44:42 +03:00
Roman Khimov
5fbe838fd4 core: add and use synchronous persist to avoid OOM
b9be892bf9 has made Persist asynchronous which
is very effective in allowing the system to continue processing
blocks/transactions while flushing things to disk. It at the same time is very
dangerous in that if the disk is slow and it takes much time to flush KV set
(more than persisting interval), there might be even bigger new KV set in
MemCachedStore by the time it finishes. Even if the system immediately starts
to flush this new data set it (being bigger) can take more time than the
previous one. And while doing so a new data set will appear in memory,
potentially again bigger than this.

So we can easily end up with the system going out of control, consuming more
and more memory and taking more and more time to persist a single set of
data. To avoid this we need to detect such condition and just wait for Persist
to really finish its job and release the resources.
2021-11-22 10:41:40 +03:00
Roman Khimov
c01427ca65 rpc: add verbose parameter to invokefunction and invokescript 2021-11-20 21:57:41 +03:00
Roman Khimov
5b470f14cc rpc: pass single Param to CreateFunctionInvocationScript
N3 contracts require method and array of arguments to function correctly, but
CreateFunctionInvocationScript can produce scripts that won't work correctly
if one is to pass anything but a single array parameter to it. This doesn't
make much sense, so we can always expect there to be an array of parameters in
the third positional invokefunction argument (the same way C# node does).
2021-11-20 21:57:41 +03:00
Roman Khimov
d01f9da8f3 vm: add invocation tree tracing 2021-11-20 21:57:41 +03:00
Roman Khimov
eb96f5a5a2 vm: drop CheckReturnState definition
It's something obsolete from RC phase.
2021-11-20 21:57:41 +03:00