bsc/trie/secure_trie_test.go

152 lines
4.5 KiB
Go
Raw Permalink Normal View History

2015-07-06 01:19:48 +02:00
// Copyright 2015 The go-ethereum Authors
// This file is part of the go-ethereum library.
//
// The go-ethereum library is free software: you can redistribute it and/or modify
// it under the terms of the GNU Lesser General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.
//
// The go-ethereum library is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Lesser General Public License for more details.
//
// You should have received a copy of the GNU Lesser General Public License
// along with the go-ethereum library. If not, see <http://www.gnu.org/licenses/>.
package trie
import (
"bytes"
"fmt"
"runtime"
"sync"
2015-07-06 01:19:48 +02:00
"testing"
"github.com/ethereum/go-ethereum/common"
[R4R] performance improvement in many aspects (#257) * focus on performance improvement in many aspects. 1. Do BlockBody verification concurrently; 2. Do calculation of intermediate root concurrently; 3. Preload accounts before processing blocks; 4. Make the snapshot layers configurable. 5. Reuse some object to reduce GC. add * rlp: improve decoder stream implementation (#22858) This commit makes various cleanup changes to rlp.Stream. * rlp: shrink Stream struct This removes a lot of unused padding space in Stream by reordering the fields. The size of Stream changes from 120 bytes to 88 bytes. Stream instances are internally cached and reused using sync.Pool, so this does not improve performance. * rlp: simplify list stack The list stack kept track of the size of the current list context as well as the current offset into it. The size had to be stored in the stack in order to subtract it from the remaining bytes of any enclosing list in ListEnd. It seems that this can be implemented in a simpler way: just subtract the size from the enclosing list context in List instead. * rlp: use atomic.Value for type cache (#22902) All encoding/decoding operations read the type cache to find the writer/decoder function responsible for a type. When analyzing CPU profiles of geth during sync, I found that the use of sync.RWMutex in cache lookups appears in the profiles. It seems we are running into CPU cache contention problems when package rlp is heavily used on all CPU cores during sync. This change makes it use atomic.Value + a writer lock instead of sync.RWMutex. In the common case where the typeinfo entry is present in the cache, we simply fetch the map and lookup the type. * rlp: optimize byte array handling (#22924) This change improves the performance of encoding/decoding [N]byte. name old time/op new time/op delta DecodeByteArrayStruct-8 336ns ± 0% 246ns ± 0% -26.98% (p=0.000 n=9+10) EncodeByteArrayStruct-8 225ns ± 1% 148ns ± 1% -34.12% (p=0.000 n=10+10) name old alloc/op new alloc/op delta DecodeByteArrayStruct-8 120B ± 0% 48B ± 0% -60.00% (p=0.000 n=10+10) EncodeByteArrayStruct-8 0.00B 0.00B ~ (all equal) * rlp: optimize big.Int decoding for size <= 32 bytes (#22927) This change grows the static integer buffer in Stream to 32 bytes, making it possible to decode 256bit integers without allocating a temporary buffer. In the recent commit 088da24, Stream struct size decreased from 120 bytes down to 88 bytes. This commit grows the struct to 112 bytes again, but the size change will not degrade performance because Stream instances are internally cached in sync.Pool. name old time/op new time/op delta DecodeBigInts-8 12.2µs ± 0% 8.6µs ± 4% -29.58% (p=0.000 n=9+10) name old speed new speed delta DecodeBigInts-8 230MB/s ± 0% 326MB/s ± 4% +42.04% (p=0.000 n=9+10) * eth/protocols/eth, les: avoid Raw() when decoding HashOrNumber (#22841) Getting the raw value is not necessary to decode this type, and decoding it directly from the stream is faster. * fix testcase * debug no lazy * fix can not repair * address comments Co-authored-by: Felix Lange <fjl@twurst.com>
2021-07-29 17:16:53 +08:00
"github.com/ethereum/go-ethereum/common/gopool"
"github.com/ethereum/go-ethereum/core/rawdb"
"github.com/ethereum/go-ethereum/core/types"
2015-07-06 01:19:48 +02:00
"github.com/ethereum/go-ethereum/crypto"
"github.com/ethereum/go-ethereum/trie/trienode"
2015-07-06 01:19:48 +02:00
)
func newEmptySecure() *StateTrie {
trie, _ := NewStateTrie(TrieID(types.EmptyRootHash), newTestDatabase(rawdb.NewMemoryDatabase(), rawdb.HashScheme))
2015-07-06 01:19:48 +02:00
return trie
}
// makeTestStateTrie creates a large enough secure trie for testing.
func makeTestStateTrie() (*testDb, *StateTrie, map[string][]byte) {
// Create an empty trie
triedb := newTestDatabase(rawdb.NewMemoryDatabase(), rawdb.HashScheme)
trie, _ := NewStateTrie(TrieID(types.EmptyRootHash), triedb)
// Fill it with some arbitrary data
content := make(map[string][]byte)
for i := byte(0); i < 255; i++ {
// Map the same data under multiple keys
key, val := common.LeftPadBytes([]byte{1, i}, 32), []byte{i}
content[string(key)] = val
trie.MustUpdate(key, val)
key, val = common.LeftPadBytes([]byte{2, i}, 32), []byte{i}
content[string(key)] = val
trie.MustUpdate(key, val)
// Add some other data to inflate the trie
for j := byte(3); j < 13; j++ {
key, val = common.LeftPadBytes([]byte{j, i}, 32), []byte{j, i}
content[string(key)] = val
trie.MustUpdate(key, val)
}
}
root, nodes, _ := trie.Commit(false)
if err := triedb.Update(root, types.EmptyRootHash, trienode.NewWithNodeSet(nodes)); err != nil {
panic(fmt.Errorf("failed to commit db %v", err))
}
// Re-create the trie based on the new state
trie, _ = NewStateTrie(TrieID(root), triedb)
return triedb, trie, content
}
2015-07-06 01:19:48 +02:00
func TestSecureDelete(t *testing.T) {
trie := newEmptySecure()
vals := []struct{ k, v string }{
{"do", "verb"},
{"ether", "wookiedoo"},
{"horse", "stallion"},
{"shaman", "horse"},
{"doge", "coin"},
{"ether", ""},
{"dog", "puppy"},
{"shaman", ""},
}
for _, val := range vals {
if val.v != "" {
trie.MustUpdate([]byte(val.k), []byte(val.v))
2015-07-06 01:19:48 +02:00
} else {
trie.MustDelete([]byte(val.k))
2015-07-06 01:19:48 +02:00
}
}
hash := trie.Hash()
exp := common.HexToHash("29b235a58c3c25ab83010c327d5932bcf05324b7d6b1185e650798034783ca9d")
if hash != exp {
t.Errorf("expected %x got %x", exp, hash)
}
}
func TestSecureGetKey(t *testing.T) {
trie := newEmptySecure()
trie.MustUpdate([]byte("foo"), []byte("bar"))
2015-07-06 01:19:48 +02:00
key := []byte("foo")
value := []byte("bar")
seckey := crypto.Keccak256(key)
2015-07-06 01:19:48 +02:00
if !bytes.Equal(trie.MustGet(key), value) {
2015-07-06 01:19:48 +02:00
t.Errorf("Get did not return bar")
}
if k := trie.GetKey(seckey); !bytes.Equal(k, key) {
t.Errorf("GetKey returned %q, want %q", k, key)
}
}
func TestStateTrieConcurrency(t *testing.T) {
// Create an initial trie and copy if for concurrent access
_, trie, _ := makeTestStateTrie()
threads := runtime.NumCPU()
tries := make([]*StateTrie, threads)
for i := 0; i < threads; i++ {
tries[i] = trie.Copy()
}
// Start a batch of goroutines interacting with the trie
pend := new(sync.WaitGroup)
pend.Add(threads)
for i := 0; i < threads; i++ {
[R4R] performance improvement in many aspects (#257) * focus on performance improvement in many aspects. 1. Do BlockBody verification concurrently; 2. Do calculation of intermediate root concurrently; 3. Preload accounts before processing blocks; 4. Make the snapshot layers configurable. 5. Reuse some object to reduce GC. add * rlp: improve decoder stream implementation (#22858) This commit makes various cleanup changes to rlp.Stream. * rlp: shrink Stream struct This removes a lot of unused padding space in Stream by reordering the fields. The size of Stream changes from 120 bytes to 88 bytes. Stream instances are internally cached and reused using sync.Pool, so this does not improve performance. * rlp: simplify list stack The list stack kept track of the size of the current list context as well as the current offset into it. The size had to be stored in the stack in order to subtract it from the remaining bytes of any enclosing list in ListEnd. It seems that this can be implemented in a simpler way: just subtract the size from the enclosing list context in List instead. * rlp: use atomic.Value for type cache (#22902) All encoding/decoding operations read the type cache to find the writer/decoder function responsible for a type. When analyzing CPU profiles of geth during sync, I found that the use of sync.RWMutex in cache lookups appears in the profiles. It seems we are running into CPU cache contention problems when package rlp is heavily used on all CPU cores during sync. This change makes it use atomic.Value + a writer lock instead of sync.RWMutex. In the common case where the typeinfo entry is present in the cache, we simply fetch the map and lookup the type. * rlp: optimize byte array handling (#22924) This change improves the performance of encoding/decoding [N]byte. name old time/op new time/op delta DecodeByteArrayStruct-8 336ns ± 0% 246ns ± 0% -26.98% (p=0.000 n=9+10) EncodeByteArrayStruct-8 225ns ± 1% 148ns ± 1% -34.12% (p=0.000 n=10+10) name old alloc/op new alloc/op delta DecodeByteArrayStruct-8 120B ± 0% 48B ± 0% -60.00% (p=0.000 n=10+10) EncodeByteArrayStruct-8 0.00B 0.00B ~ (all equal) * rlp: optimize big.Int decoding for size <= 32 bytes (#22927) This change grows the static integer buffer in Stream to 32 bytes, making it possible to decode 256bit integers without allocating a temporary buffer. In the recent commit 088da24, Stream struct size decreased from 120 bytes down to 88 bytes. This commit grows the struct to 112 bytes again, but the size change will not degrade performance because Stream instances are internally cached in sync.Pool. name old time/op new time/op delta DecodeBigInts-8 12.2µs ± 0% 8.6µs ± 4% -29.58% (p=0.000 n=9+10) name old speed new speed delta DecodeBigInts-8 230MB/s ± 0% 326MB/s ± 4% +42.04% (p=0.000 n=9+10) * eth/protocols/eth, les: avoid Raw() when decoding HashOrNumber (#22841) Getting the raw value is not necessary to decode this type, and decoding it directly from the stream is faster. * fix testcase * debug no lazy * fix can not repair * address comments Co-authored-by: Felix Lange <fjl@twurst.com>
2021-07-29 17:16:53 +08:00
index := i
gopool.Submit(func() {
defer pend.Done()
for j := byte(0); j < 255; j++ {
// Map the same data under multiple keys
key, val := common.LeftPadBytes([]byte{byte(index), 1, j}, 32), []byte{j}
tries[index].MustUpdate(key, val)
key, val = common.LeftPadBytes([]byte{byte(index), 2, j}, 32), []byte{j}
tries[index].MustUpdate(key, val)
// Add some other data to inflate the trie
for k := byte(3); k < 13; k++ {
key, val = common.LeftPadBytes([]byte{byte(index), k, j}, 32), []byte{k, j}
tries[index].MustUpdate(key, val)
}
}
tries[index].Commit(false)
[R4R] performance improvement in many aspects (#257) * focus on performance improvement in many aspects. 1. Do BlockBody verification concurrently; 2. Do calculation of intermediate root concurrently; 3. Preload accounts before processing blocks; 4. Make the snapshot layers configurable. 5. Reuse some object to reduce GC. add * rlp: improve decoder stream implementation (#22858) This commit makes various cleanup changes to rlp.Stream. * rlp: shrink Stream struct This removes a lot of unused padding space in Stream by reordering the fields. The size of Stream changes from 120 bytes to 88 bytes. Stream instances are internally cached and reused using sync.Pool, so this does not improve performance. * rlp: simplify list stack The list stack kept track of the size of the current list context as well as the current offset into it. The size had to be stored in the stack in order to subtract it from the remaining bytes of any enclosing list in ListEnd. It seems that this can be implemented in a simpler way: just subtract the size from the enclosing list context in List instead. * rlp: use atomic.Value for type cache (#22902) All encoding/decoding operations read the type cache to find the writer/decoder function responsible for a type. When analyzing CPU profiles of geth during sync, I found that the use of sync.RWMutex in cache lookups appears in the profiles. It seems we are running into CPU cache contention problems when package rlp is heavily used on all CPU cores during sync. This change makes it use atomic.Value + a writer lock instead of sync.RWMutex. In the common case where the typeinfo entry is present in the cache, we simply fetch the map and lookup the type. * rlp: optimize byte array handling (#22924) This change improves the performance of encoding/decoding [N]byte. name old time/op new time/op delta DecodeByteArrayStruct-8 336ns ± 0% 246ns ± 0% -26.98% (p=0.000 n=9+10) EncodeByteArrayStruct-8 225ns ± 1% 148ns ± 1% -34.12% (p=0.000 n=10+10) name old alloc/op new alloc/op delta DecodeByteArrayStruct-8 120B ± 0% 48B ± 0% -60.00% (p=0.000 n=10+10) EncodeByteArrayStruct-8 0.00B 0.00B ~ (all equal) * rlp: optimize big.Int decoding for size <= 32 bytes (#22927) This change grows the static integer buffer in Stream to 32 bytes, making it possible to decode 256bit integers without allocating a temporary buffer. In the recent commit 088da24, Stream struct size decreased from 120 bytes down to 88 bytes. This commit grows the struct to 112 bytes again, but the size change will not degrade performance because Stream instances are internally cached in sync.Pool. name old time/op new time/op delta DecodeBigInts-8 12.2µs ± 0% 8.6µs ± 4% -29.58% (p=0.000 n=9+10) name old speed new speed delta DecodeBigInts-8 230MB/s ± 0% 326MB/s ± 4% +42.04% (p=0.000 n=9+10) * eth/protocols/eth, les: avoid Raw() when decoding HashOrNumber (#22841) Getting the raw value is not necessary to decode this type, and decoding it directly from the stream is faster. * fix testcase * debug no lazy * fix can not repair * address comments Co-authored-by: Felix Lange <fjl@twurst.com>
2021-07-29 17:16:53 +08:00
})
}
// Wait for all threads to finish
pend.Wait()
}