* 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>
* all: add thousandths separators for big numbers on log messages
* p2p/sentry: drop accidental file
* common, log: add fast number formatter
* common, eth/protocols/snap: simplifty fancy num types
* log: handle nil big ints
This PR implements the first one of the "lespay" UDP queries which
is already useful in itself: the capacity query. The server pool is making
use of this query by doing a cheap UDP query to determine whether it is
worth starting the more expensive TCP connection process.
Both Hash and Address have a String method, which returns the value as
hex with 0x prefix. They also had a Format method which tried to print
the value using printf of []byte. The way Format worked was at odds with
String though, leading to a situation where fmt.Sprintf("%v", hash)
returned the decimal notation and hash.String() returned a hex string.
This commit makes it consistent again. Both types now support the %v,
%s, %q format verbs for 0x-prefixed hex output. %x, %X creates
unprefixed hex output. %d is also supported and returns the decimal
notation "[1 2 3...]".
For Address, the case of hex characters in %v, %s, %q output is
determined using the EIP-55 checksum. Using %x, %X with Address
disables checksumming.
Co-authored-by: Felix Lange <fjl@twurst.com>
ToHex was deprecated a couple years ago. The last remaining use
was in ToHexArray, which itself only had a single call site.
This just moves ToHexArray near its only remaining call site and
implements it using hexutil.Encode. This changes the default behaviour
of ToHexArray and with it the behaviour of eth_getProof. Previously we
encoded an empty slice as 0, now the empty slice is encoded as 0x.
* accounts, signer: implement gnosis safe support
* common/math: add type for marshalling big to dec
* accounts, signer: properly sign gnosis requests
* signer, clef: implement account_signGnosisTx
* signer: fix auditlog print, change rpc-name (signGnosisTx to signGnosisSafeTx)
* signer: pass validation-messages/warnings to the UI for gnonsis-safe txs
* signer/core: minor change to validationmessages of typed data
Reverting because this change started handling account balances as
uint64 in the transaction pool, which is incorrect.
This reverts commit af5c97aebe1d37486635521ef553cb8bd4bada13.
* core: use uint64 for total tx costs instead of big.Int
* core: added local tx pool test case
* core, crypto: various allocation savings regarding tx handling
* Update core/tx_list.go
* core: added tx.GasPriceIntCmp for comparison without allocation
adds a method to remove unneeded allocation in comparison to tx.gasPrice
* core: handle pools full of locals better
* core/tests: benchmark for tx_list
* core/txlist, txpool: save a reheap operation, avoid some bigint allocs
Co-authored-by: Martin Holst Swende <martin@swende.se>
The leaks were mostly in unit tests, and could all be resolved by
adding suitably-sized channel buffers or by restructuring the test
to not send on a channel after an error has occurred.
There is an unavoidable goroutine leak in Console.Interactive: when
we receive a signal, the line reader cannot be unblocked and will get
stuck. This leak is now documented and I've tried to make it slightly
less bad by adding a one-element buffer to the output channels of
the line-reading loop. Should the reader eventually awake from its
blocked state (i.e. when stdin is closed), at least it won't get stuck
trying to send to the interpreter loop which has quit long ago.
Co-authored-by: Felix Lange <fjl@twurst.com>
This change adds tests for the virtual clock and aligns the interface
with the time package by renaming Cancel to Stop. It also removes the
binary search from Stop because it complicates the code unnecessarily.
Gollvm has very aggressive dead code elimination that completely
removes one of these two benchmarks. To prevent this, use the
result of the benchmark (a boolean), and to be "fair", make the
transformation to both benchmarks.
To be reliably assured of not removing the code, "use" means
assigning to an exported global. Non-exported globals and
//go:noinline functions are possibly subject to this optimization.
* accounts/abi/bind: Accept function ptr parameter
They are translated as [24]byte
* Add Java template version
* accounts/abi/bind: fix merge issue
* Fix CI
* core: reinit chain from freezer in batches
* core/rawdb: concurrent database reinit from freezer dump
* core/rawdb: reinit from freezer in sequential order
* core, eth: some fixes for freezer
* vendor, core/rawdb, cmd/geth: add db inspector
* core, cmd/utils: check ancient store path forceily
* cmd/geth, common, core/rawdb: a few fixes
* cmd/geth: support windows file rename and fix rename error
* core: support ancient plugin
* core, cmd: streaming file copy
* cmd, consensus, core, tests: keep genesis in leveldb
* core: write txlookup during ancient init
* core: bump database version
* core/vm: remove function call for stack validation from evm runloop
* core/vm: separate gas calc into static + dynamic
* core/vm: optimize push1
* core/vm: reuse pooled bigints for ADDRESS, ORIGIN and CALLER
* core/vm: use generic error message for jump/jumpi, to avoid string interpolation
* testdata: fix tests for new error message
* core/vm: use 64-bit memory calculations
* core/vm: fix error in memory calculation
* core/vm: address review concerns
* core/vm: avoid unnecessary use of big.Int:BitLen()
* common/fdlimit: cap on MacOS file limits, fixes#18994
* common/fdlimit: fix Maximum-check to respect OPEN_MAX
* common/fdlimit: return error if OPEN_MAX is exceeded in Raise()
* common/fdlimit: goimports
* common/fdlimit: check value after setting fdlimit
* common/fdlimit: make comment a bit more descriptive
* cmd/utils: make fdlimit happy path a bit cleaner
