Skip to content

Conversation

@Gustav-Simonsson
Copy link

NOTE: RecoverPubkey is on the consensus-critical code path. Please review carefully.

Go wrapper update is in it's own commit, files changed:

crypto/secp256k1/secp256.go
crypto/secp256k1/secp256_test.go
benchmark                                  old ns/op     new ns/op     delta
BenchmarkSignRandomInputEachRound-2        148648        149994        +0.91%
BenchmarkRecoverRandomInputEachRound-2     207831        203833        -1.92%

We'll probably see a more significant speedup when we can improve internal conversions (as in remove where possible) of signatures and pubkeys - currently we need two extra calls to the lib for each recovery - one to serialise signature and one to serialise the returned pubkey.

If we can avoid those calls + other internal conversions we could probably gain a few % more on recovery.

  • Update libsecp256k1
  • Update Go wrapper to new libsecp256k1 API
  • Refactor Go wrapper tests to have meaningful names, clearer validations and a few more edge cases.
  • Add benchmarks of Sign and RecoverPubkey

@robotally
Copy link

Vote Count Reviewers
👍 2 @fjl @obscuren
👎 0

Updated: Fri Oct 9 13:02:59 UTC 2015

@fjl
Copy link
Contributor

fjl commented Oct 1, 2015

Did you add zeroing of intermediate buffers yet?

@Gustav-Simonsson
Copy link
Author

@fjl not yet. added to Sign

@obscuren
Copy link
Contributor

obscuren commented Oct 8, 2015

LGTM 👍

@Gustav-Simonsson
Copy link
Author

@obscuren removed unneeded return matchings. SECP256K1_EC_COMPRESSED is from the name of the macro in the lib

@codecov-io
Copy link

Current coverage is 47.77%

Merging #1853 into develop will increase coverage by +0.14% as of 97d4d77

Powered by Codecov. Updated on successful CI builds.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

It's usually better to use the same input data for all iterations.
package testing runs benchmarks several times to make the
result more stable, so StartTimer/StopTimer is not required.
This also applies to BenchmarkSign.

_, seckey := GenerateKeyPair()
msg := randentropy.GetEntropyCSPRNG(32)
sig, _ := Sign(msg, seckey)
for i := 0; i < b.N; i++ {
    if _, err := RecoverPubkey(msg, sig); err != nil {
        b.Fatal(err)
    }
} 

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Actually in this benchmark it's intended to use a new key for every recovery to avoid potential caching affecting the benchmark. I'll add another benchmark where the same key is used though, that will be interesting to see if caching affects the result. In that one starting and stopping of the timer can indeed be skipped.

Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What kind of caching? libsecp256k1 shouldn't cache anything.

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't know, but it takes much less time if using the same key. See new commit - pushed a separate benchmark for using the same key.

@fjl
Copy link
Contributor

fjl commented Oct 9, 2015

👍 and please squash one more time.

obscuren added a commit that referenced this pull request Oct 15, 2015
Update libsecp256k1, Go wrapper and tests
@obscuren obscuren merged commit f466243 into ethereum:develop Oct 15, 2015
@obscuren obscuren removed the review label Oct 15, 2015
@obscuren obscuren modified the milestones: 2.0.0, 1.3.0 Oct 29, 2015
sduchesneau pushed a commit to streamingfast/go-ethereum that referenced this pull request Oct 31, 2025
sduchesneau pushed a commit to streamingfast/go-ethereum that referenced this pull request Nov 17, 2025
* Parallel block import (ethereum#1662)

* core: naive first implmenetation of parallel block import in InsertChainStateless

* core: add logic to handle dependency while parallel import

* core: handle unknown ancestor + init some UTs

* core: implement alternative way to check for unknown ancestor

* core: defer ValidateWitnessPreState

* zero commit

* core: do sequential insert if the importing chain is small

* core: fix  error handling

* core: update UTs

* core,eth,internal/cli: add config to enable/disable parallel import + refactor

* core: add metrics for seq and parallel import

* internal/cli: moved parallelstatelessimport flag to the witness config

* core: changed debug import log to info

* core: added 2 new metrics to track the number of blocks processed in sequential and parallel stateless imports

* core: init benchmarks

* core: fix benchmarks

* core: use worker pool for parallel block import

* core: try parallel import for smaller batch

* core: dedup header verification logic

* zero commit

* core,eth,internal/cli: improvements based on comments

* core: add adversarial tests for stateless insert

* core: re-add empty chain insertion case

* zero commit to trigger CI

* core,eth/catalyst: report correct execution stats for stateless insert

* CI: temp update for running stateless tests

* zero commit to trigger CI

* zero commit to trigger CI

* CI: try running on stateless branch

* Revert "CI: try running on stateless branch"

This reverts commit 98b5704fabcc6855225cc03e8f84e77f913fda08.

* zero commit to trigger CI

* core: add log for deferred exec of blocks due to invalid root err in parallel import

* core: changed log to debug

* core,eth,internal/ethapi: suppress error log for state sync when parallel import is enabled

* eth/filters: fix gomock

* eth: add missing configs

* core: retry execution in parallel import for block validator errors

* core: commit block before retrying failed execution

* core: rm redundant witness in execResult

* core: try flushing state db asap post execution

* core: add temp log when committing code

* core: enable retry for first block in batch

* Revert "core: add temp log when committing code"

This reverts commit cf63cc344ce7c0981c3410320dea04ba74dc448f.

* core: cap workers spun

* Revert "core: cap workers spun"

This reverts commit 96367eebfe94dfc76e74dced4d55b71ce8a50896.

* core: ensure statedb cleanup

* core: add temp debug logs

* core: add temp debug logs

* core: add temp log to track sidechain

* core: add temp log for contract code in commitAndFlush

* core: add retry for all processing and internal state db errors

* core: add state db error validation during deferred exec

* core: add TestParallelStateless_ContractDeployedThenCalled

* Revert "core: add temp log for contract code in commitAndFlush"

This reverts commit 174765f14ced2ccff8cd8570935b04bda478342c.

* Revert "core: add temp log to track sidechain"

This reverts commit 29b33667a61016a7f2e71817c2bd287a27852f06.

* Revert "core: add temp debug logs"

This reverts commit 2c4f8407de9c4535565d092a4310576e6aab87b0.

* internal/cli: turn off parallel import by default

* internal/cli: update flag identation

* CI: update comment

* zero commit to trigger CI

* internal/cli: rm witness prune flags

---------

Co-authored-by: Pratik Patil <pratikspatil024@gmail.com>

* emit event when write block in stateless node (ethereum#1840)

* Address Biased Trie Cache (ethereum#1837)

* Address-biased trie cache

* Stop preload early

* Print more stats

* Increase cache size

* Store by depth

* More customizable

* Add metrics

* Fix key format

* Async preload

* Simplify

* Fix linter

* Allow interrupt

* Remove free disk step in CI (ethereum#1843)

Remove this step since the runner has 600G of disk space.

* chore: bump kurtosis-pos (ethereum#1845)

* chore: bump kurtosis-pos

* chore: bump to v1.2.1

* PIP-74: state-sync txs inclusion (ethereum#1726)

* first working version of new state sync tx

* remove logs

* fix lint

* fix integration tests

* test fixes

* lint fix

* fix parallel processor

* type improvement and unit tests

* remove duplicates

* ignore data test output

* fixing method calls and fields

* build fix

* remove bor filter for after hf blocks

* sort logs just when necessary

* ssTxs: fix receiptsHash mismatch (ethereum#1829)

* ssTxs: fix receiptsHash mismatch

* chore: remove block hash and number

* revert: add block number

* core(tx): fix Hash method for StateSyncTxType (ethereum#1830)

* eth: fixes in receipt handling via p2p post HF (ethereum#1825)

* eth: include bor receipts in ReceiptHash post HF

* eth: rename to receiptListHash

* eth: extrat typecasting logic for better testing, fix type matching issue

* eth: add e2e tests for receipt delivery

* eth/protocols/eth: apply HF logic while handling receipt query over eth69

* core: skip split receipts post HF

* tests/bor: extend e2e test to check presence of state-sync in block

* core/types: return nil for chainID() call over state-sync tx

* internal/ethapi: handle bor txs and receipts post HF (ethereum#1834)

* fix: cumulativeGasUsed in insertStateSyncTransactionAndCalculateReceipt (ethereum#1835)

* fix: sort logs and remove duplicate append in Finalize (ethereum#1836)

* fix: append tx only in FinalizeAndAssemble and use state instead of wrappedState

* consensus/bor: sort logs before extracting state-sync logs

* chore: revert statedb changes

---------

Co-authored-by: Manav Darji <manavdarji.india@gmail.com>

* chore: nit (ethereum#1838)

* chore: typos

* fix: lint

* Renaming to Madhugiri HF

* chore: nits

* (fix): handle pointers to receipt list received via p2p

* (chore): rename tests with HF name

* chore: more nits

* core: add blocktime in bor receipt logs (ethereum#1848)

* core/types: derive bloom for bor receipts

---------

Co-authored-by: kamuikatsurgi <shahkrishang11@gmail.com>
Co-authored-by: Krishang Shah <109511742+kamuikatsurgi@users.noreply.github.com>
Co-authored-by: Manav Darji <manavdarji.india@gmail.com>

* core, miner, params, cmd: implement EIP-7823, EIP-7825 and EIP-7883 (ethereum#1842)

* first working version of new state sync tx

* remove logs

* fix lint

* fix integration tests

* test fixes

* lint fix

* fix parallel processor

* type improvement and unit tests

* remove duplicates

* ignore data test output

* fixing method calls and fields

* build fix

* remove bor filter for after hf blocks

* sort logs just when necessary

* ssTxs: fix receiptsHash mismatch (ethereum#1829)

* ssTxs: fix receiptsHash mismatch

* chore: remove block hash and number

* revert: add block number

* core(tx): fix Hash method for StateSyncTxType (ethereum#1830)

* eth: fixes in receipt handling via p2p post HF (ethereum#1825)

* eth: include bor receipts in ReceiptHash post HF

* eth: rename to receiptListHash

* eth: extrat typecasting logic for better testing, fix type matching issue

* eth: add e2e tests for receipt delivery

* eth/protocols/eth: apply HF logic while handling receipt query over eth69

* core: skip split receipts post HF

* tests/bor: extend e2e test to check presence of state-sync in block

* core/types: return nil for chainID() call over state-sync tx

* internal/ethapi: handle bor txs and receipts post HF (ethereum#1834)

* fix: cumulativeGasUsed in insertStateSyncTransactionAndCalculateReceipt (ethereum#1835)

* fix: sort logs and remove duplicate append in Finalize (ethereum#1836)

* fix: append tx only in FinalizeAndAssemble and use state instead of wrappedState

* consensus/bor: sort logs before extracting state-sync logs

* chore: revert statedb changes

---------

Co-authored-by: Manav Darji <manavdarji.india@gmail.com>

* chore: nit (ethereum#1838)

* chore: typos

* fix: lint

* Renaming to Madhugiri HF

* chore: nits

* (fix): handle pointers to receipt list received via p2p

* (chore): rename tests with HF name

* chore: more nits

* core: implement eip7883

* core: implement eip7825

* fix lint

* fix isOsaka condition

* fix lint and add comment for int test failing

* fix failing test / address comments

* core: add blocktime in bor receipt logs (ethereum#1848)

* core/types: derive bloom for bor receipts

* prioritize ss hf / remove todo

* fix lint

* split precompiled contracts and addresses for madhugiri HF from osaka

* revert err msg to align with geth

* eth/gasestimator: check ErrGasLimitTooHigh conditions (ethereum#32348)

This PR makes 2 changes to how
[EIP-7825](ethereum#31824) behaves.

When `eth_estimateGas` or `eth_createAccessList` is called without any
gas limit in the payload, geth will choose the block's gas limit or the
`RPCGasCap`, which can be larger than the `maxTxGas`.

When this happens for `estimateGas`, the gas estimation just errors out
and ends, when it should continue doing binary search to find the lowest
possible gas limit.

This PR will: 
- Add a check to see if `hi` is larger than `maxTxGas` and cap it to
`maxTxGas` if it's larger. And add a special case handling for gas
estimation execute when it errs with `ErrGasLimitTooHigh`

---------

Co-authored-by: Gary Rong <garyrong0905@gmail.com>

* internal/ethapi: skip tx gas limit check for calls (ethereum#32641)

This disables the tx gaslimit cap for eth_call and related RPC operations.

I don't like how this fix works. Ideally we'd be checking the tx
gaslimit somewhere else, like in the block validator, or any other place
that considers block transactions. Doing the check in StateTransition
means it affects all possible ways of executing a message.

The challenge is finding a place for this check that also triggers
correctly in tests where it is wanted. So for now, we are just combining
this with the EOA sender check for transactions. Both are disabled for
call-type messages.

* internal/ethapi: use gas from gaspool for call defaults

* eth/gasestimator: check for madhugiri HF

* consensus/bor: honour MaxTxGas for system calls

---------

Co-authored-by: Lucca Martins <lucca_martins30@yahoo.com.br>
Co-authored-by: kamuikatsurgi <shahkrishang11@gmail.com>
Co-authored-by: Krishang Shah <109511742+kamuikatsurgi@users.noreply.github.com>
Co-authored-by: Manav Darji <manavdarji.india@gmail.com>
Co-authored-by: Minhyuk Kim <kimminhyuk1004@gmail.com>
Co-authored-by: Gary Rong <garyrong0905@gmail.com>
Co-authored-by: Felix Lange <fjl@twurst.com>

* (chore): update madhugiri block number for amoy and update consensus block time (ethereum#1851)

* first working version of new state sync tx

* remove logs

* fix lint

* fix integration tests

* test fixes

* lint fix

* fix parallel processor

* type improvement and unit tests

* remove duplicates

* ignore data test output

* fixing method calls and fields

* build fix

* remove bor filter for after hf blocks

* sort logs just when necessary

* ssTxs: fix receiptsHash mismatch (ethereum#1829)

* ssTxs: fix receiptsHash mismatch

* chore: remove block hash and number

* revert: add block number

* core(tx): fix Hash method for StateSyncTxType (ethereum#1830)

* eth: fixes in receipt handling via p2p post HF (ethereum#1825)

* eth: include bor receipts in ReceiptHash post HF

* eth: rename to receiptListHash

* eth: extrat typecasting logic for better testing, fix type matching issue

* eth: add e2e tests for receipt delivery

* eth/protocols/eth: apply HF logic while handling receipt query over eth69

* core: skip split receipts post HF

* tests/bor: extend e2e test to check presence of state-sync in block

* core/types: return nil for chainID() call over state-sync tx

* internal/ethapi: handle bor txs and receipts post HF (ethereum#1834)

* fix: cumulativeGasUsed in insertStateSyncTransactionAndCalculateReceipt (ethereum#1835)

* fix: sort logs and remove duplicate append in Finalize (ethereum#1836)

* fix: append tx only in FinalizeAndAssemble and use state instead of wrappedState

* consensus/bor: sort logs before extracting state-sync logs

* chore: revert statedb changes

---------

Co-authored-by: Manav Darji <manavdarji.india@gmail.com>

* chore: nit (ethereum#1838)

* chore: typos

* fix: lint

* Renaming to Madhugiri HF

* chore: nits

* (fix): handle pointers to receipt list received via p2p

* (chore): rename tests with HF name

* chore: more nits

* core: add blocktime in bor receipt logs (ethereum#1848)

* core/types: derive bloom for bor receipts

* (chore): update madhugiri block for amoy

* Change consensus block time to 1s at Madhugiri HF (ethereum#1852)

---------

Co-authored-by: Lucca Martins <lucca_martins30@yahoo.com.br>
Co-authored-by: kamuikatsurgi <shahkrishang11@gmail.com>
Co-authored-by: Krishang Shah <109511742+kamuikatsurgi@users.noreply.github.com>
Co-authored-by: Jerry <jerrycgh@gmail.com>

* Completed reset prefetcher (ethereum#1853)

* internal/ethapi: restore original RPC gas cap (ethereum#1856)

* Restore original RPC gas cap

* Fix test

* chore: fix wrong function name in comment (ethereum#1841)

Signed-off-by: reddaisyy <reddaisy@outlook.jp>

* Revert "(chore): update madhugiri block number for amoy and update consensus block time (ethereum#1851)" (ethereum#1857)

This reverts commit 103f1f4.

* params: bump version to v2.4.0-beta

* Dont OpenTrie because HistoricDatabase doesnt implement trie (ethereum#1858)

* chore: use 2^25 as the MaxTxGas (ethereum#1859)

* params: bump version to v2.4.0-beta2

* chore: set madhugiri block for amoy and consensus block time (ethereum#1867)

* chore: set madhugiri block for amoy and consensus block time

* params: bump version

* Disable txn indexer in stateless mode

Since stateless nodes are using a different pruner from the regular pruner in geth, the transaction indexer will have problem with pruned blocks, resulting hanging goroutines in memory. This change prevents stateless node from running indexer on pruned blocks.

* core: init PrecompiledAddressesMadhugiri

* params: bump version

* params: bump version for stable release

* chore: set madhugiri block for mainnet and consensus block time

* chore: fix lint

* Include missing check of P256 on Amoy (ethereum#1877)

* checks p256 instruction

* lint fix

* v2.5.0 candidate (ethereum#1878)

* chore: MadhugiriPro hf

* params: use stable tag

* chore: set new HF heights

* Reinforce Precompile Check on any new HF (ethereum#1881)

* Include missing check of P256 on Amoy (ethereum#1877)

* checks p256 instruction

* lint fix

* reinforce precompilers check

---------

Co-authored-by: Lucca Martins <lucca_martins30@yahoo.com.br>

---------

Signed-off-by: reddaisyy <reddaisy@outlook.jp>
Co-authored-by: Raneet Debnath <35629432+Raneet10@users.noreply.github.com>
Co-authored-by: Pratik Patil <pratikspatil024@gmail.com>
Co-authored-by: Lucca Martins <lucca_martins30@yahoo.com.br>
Co-authored-by: Jerry <jerrycgh@gmail.com>
Co-authored-by: Krishang Shah <109511742+kamuikatsurgi@users.noreply.github.com>
Co-authored-by: kamuikatsurgi <shahkrishang11@gmail.com>
Co-authored-by: Manav Darji <manavdarji.india@gmail.com>
Co-authored-by: Minhyuk Kim <kimminhyuk1004@gmail.com>
Co-authored-by: Gary Rong <garyrong0905@gmail.com>
Co-authored-by: Felix Lange <fjl@twurst.com>
Co-authored-by: reddaisyy <reddaisy@outlook.jp>
Co-authored-by: Angel Valkov <avalkov@polygon.technology>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants