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.

This adds the debug_accountRange method which returns all accounts in
the state for a given block and transaction index.

* eth: fix error when dump config with nil checkpoint

* cmd/utils: ignore default datadir if it's already set.

cmd: encapsulate les relative cli options

cmd, eth, les: fix up ultra light config integration

appveyor: bump builder to Go 1.12.7

core: kill off managed state, use own tiny noncer for txpool

params: bump all CHTs, deploy all checkpoint oracles

* cmd/puppeth: integrate blockscout

* cmd/puppeth: expose debug namespace for blockscout

* cmd/puppeth: fix dbdir

* cmd/puppeth: run explorer in archive mode

* cmd/puppeth: ensure node is synced

* cmd/puppeth: fix explorer docker alignment + drop unneeded exec

* cmd/puppeth: polish up config saving and reloading

* cmd/puppeth: check both web and p2p port for explorer service

Make it select from all live nodes instead of selecting the heads of
random buckets.

* eth: chain config (genesis + fork) ENR entry

* core/forkid, eth: protocol independent fork ID, update to CRC32 spec

* core/forkid, eth: make forkid a struct, next uint64, enr struct, RLP

* core/forkid: change forkhash rlp encoding from int to [4]byte

* eth: fixup eth entry a bit and update it every block

* eth: fix lint

* eth: fix crash in ethclient tests

This entry was an experiment, but we're moving on to the
entry-per-protocol instead.

* cmd, common: refactor abigen command line interface

* cmd/abigen: address comment

This PR adds some hardening in the lower levels of the protocol stack, to bail early on invalid data. Primarily, attacks that this PR protects against are on the "annoyance"-level, which would otherwise write a couple of megabytes of data into the log output, which is a bit resource intensive.

* accounts, abigen: link dependent libs in deploy

* abigen: add java generation

* bind: Fix unit tests

* abigen: add unit test

* Fix CI

* Post-rebase fixes

* Fix rebase issue

* accounts/abi: Gary's review feedback

* accounts/abi: More Gary feedback

* accounts/abi: minor fixes

* core/blockchain: lessen mem-spike during 1.8->1.9 conversion

* core/blockchain.go: make levedb->freezer conversion gradually

* core/blockchain: write the batch

* accounts, cmd, contracts, les: integrate clef for transaction signing

* accounts, cmd/checkpoint-admin, signer/core: minor fixups

* les: remove half-finish APIs

* les: remove half-finish APIs

* accounts/abi, cmd/abigen: support tuple

accounts/abi/bind, cmd/abigen: add objc back

accounts/abi/bind: use byte[24] as function indicator

accounts/abi/bind: resolve struct slice or array

accounts/abi/bind: remove sort logic

accounts: fix issues in abi

* accounts/abi: address comment

This PR fixes an issue in chain indexer. Currently chain indexer will
validate whether the stored data is canonical by comparing section head
and canonical hash. But the header of the checkpoint may not exist in
the database. We should skip validation for sections below the
checkpoint.

* core: fix chain indexer reorg bug

* core: prevent reverting valid section when reorg happens

cmd, eth, les, param: drop --override.constantinople

* cmd/clef, signer: refresh tutorial, fix noticed issues

* cmd/clef, signer: support removing stored keys (delpw + rules)

* cmd/clef: polishes + Geth integration in the tutorial

* 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