This is necessary to avoid race conditions for low-latency consensus mechanisms
(like Raft) -- `ChainHeadEvent`s should always be posted in order, and the
existing code could violate this due to asynchronously sending-off events.
Now we sychronously send events in-order, but do so through a buffered
intermediary channel (`chainEvents`) so that publishing them is still
asynchronous.
This commit adds pluggable consensus engines to go-ethereum. In short, it
introduces a generic consensus interface, and refactors the entire codebase to
use this interface.
* core,eth,internal: Added `debug_getBadBlocks()` method
When bad blocks are discovered, these are stored within geth.
An RPC-endpoint makes them availablewithin the `debug`
namespace. This feature makes it easier to discover network forks.
```
* core, api: go format + docs
* core/blockchain: Documentation, fix minor nitpick
* core: fix failing blockchain test
* core: Made logging of reorgs more structured, also always log if reorg is > 63 blocks long
* core/blockchain: go fmt
* core/blockchain: Minor fixes to the reorg reporting
This significantly reduces the dependency closure of ethclient, which no
longer depends on core/vm as of this change.
All uses of vm.Logs are replaced by []*types.Log. NewLog is gone too,
the constructor simply returned a literal.
This field used to be assigned by the filter system and returned through
the RPC API. Now that we have a Go client that uses the underlying type,
the field needs to move. It is now assigned to true when the RemovedLogs
event is generated so the filter system doesn't need to care about the
field at all.
While here, remove the log list from ChainSideEvent. There are no users
of this field right now and any potential users could subscribe to
RemovedLogsEvent instead.
This commit implements EIP158 part 1, 2, 3 & 4
1. If an account is empty it's no longer written to the trie. An empty
account is defined as (balance=0, nonce=0, storage=0, code=0).
2. Delete an empty account if it's touched
3. An empty account is redefined as either non-existent or empty.
4. Zero value calls and zero value suicides no longer consume the 25k
reation costs.
params: moved core/config to params
Signed-off-by: Jeffrey Wilcke <jeffrey@ethereum.org>
This commit implements a dual state approach. The dual state approach
separates public and private state by making the core vm environment
context aware.
Although not currently implemented it will need to prohibit value
transfers and it must initialise all transactions from accounts on the
public state. This means that sending transactions increments the
account nonce on the public state and contract addresses are derived
from the public state when initialised by a transaction. For obvious
reasons, contract created by private contracts are still derived from
public state.
This is required in order to have consensus over the public state at all
times as non-private participants would still process the transaction on
the public state even though private payload can not be decrypted. This
means that participants of a private group must do the same in order to
have public consensus. However the creation of the contract and
interaction still occurs on the private state.
It implements support for the following calling model:
S: sender, (X): private, X: public, ->: direction, [ ]: read only mode
1. S -> A -> B
2. S -> (A) -> (B)
3. S -> (A) -> [ B -> C ]
It does not support
1. (S) -> A
2. (S) -> (A)
3. S -> (A) -> B
Implemented "read only" mode for the EVM. Read only mode is checked
during any opcode that could potentially modify the state. If such an
opcode is encountered during "read only", it throws an exception.
The EVM is flagged "read only" when a private contract calls in to
public state.
* trie: store nodes as pointers
This avoids memory copies when unwrapping node interface values.
name old time/op new time/op delta
Get 388ns ± 8% 215ns ± 2% -44.56% (p=0.000 n=15+15)
GetDB 363ns ± 3% 202ns ± 2% -44.21% (p=0.000 n=15+15)
UpdateBE 1.57µs ± 2% 1.29µs ± 3% -17.80% (p=0.000 n=13+15)
UpdateLE 1.92µs ± 2% 1.61µs ± 2% -16.25% (p=0.000 n=14+14)
HashBE 2.16µs ± 6% 2.18µs ± 6% ~ (p=0.436 n=15+15)
HashLE 7.43µs ± 3% 7.21µs ± 3% -2.96% (p=0.000 n=15+13)
* trie: close temporary databases in GetDB benchmark
* trie: don't keep []byte from DB load around
Nodes decoded from a DB load kept hashes and values as sub-slices of
the DB value. This can be a problem because loading from leveldb often
returns []byte with a cap that's larger than necessary, increasing
memory usage.
* trie: unload old cached nodes
* trie, core/state: use cache unloading for account trie
* trie: use explicit private flags (fixes Go 1.5 reflection issue).
* trie: fixup cachegen overflow at request of nick
* core/state: rename journal size constant
We used to have reporting of bad blocks, but it was disabled
before the Frontier release. We need it back because users
are usually unable to provide the full RLP data of a bad
block when it occurs.
A shortcoming of this particular implementation is that the
origin peer is not tracked for blocks received during eth/63
sync. No origin peer info is still better than no report at
all though.
This is necessary for external users of the go-ethereum code who want to, for instance, build a custom node that plays back transactions, as core.ApplyTransaction requires a ChainConfig as a parameter.
Added chain configuration options and write out during genesis database
insertion. If no "config" was found, nothing is written to the database.
Configurations are written on a per genesis base. This means
that any chain (which is identified by it's genesis hash) can have their
own chain settings.
Added a future lock which prevents the anything being added or removed
from or to the set when looping over the set of blocks. This fixes a nil
pointer in the range loop when trying to retrieve a block from the set
which was previously available but removed due to regular chain
processing.
Fixes#2305
Previously all blocks that were already in our chain were never re
announced as potential uncle block (e.g. ChainSideEvent). This is
problematic during mining where you want to gather as much possible
uncles as possible increasing the profit. This is now addressed in this
PR where during reorganisations of chains the old chain is regarded as
uncles.
Fixed#2298
Assuming the following scenario where a miner has 15% of all hashing
power and the ability to exert a moderate control over the network to
the point where if the attacker sees a message A, it can't stop A from
propagating, but what it **can** do is send a message B and ensure that
most nodes see B before A. The attacker can then selfish mine and
augment selfish mining strategy by giving his own blocks an advantage.
This change makes the time at which a block is received less relevant
and so the level of control an attacker has over the network no longer
makes a difference.
This change changes the current td algorithm `B_td > C_td` to the new
algorithm `B_td > C_td || B_td == C_td && rnd < 0.5`.
Pending logs are now filterable through the Go API. Filter API changed
such that each filter type has it's own bucket and adding filter
explicitly requires you specify the bucket to put it in.
When a chain reorganisation occurs we collect the logs that were deleted
during the chain reorganisation. The removed logs are posted to the
event mux indicating that those were deleted during the reorg.
This removes the burden on a single object to take care of all
validation and state processing. Now instead the validation is done by
the `core.BlockValidator` (`types.Validator`) that takes care of both
header and uncle validation through the `ValidateBlock` method and state
validation through the `ValidateState` method. The state processing is
done by a new object `core.StateProcessor` (`types.Processor`) and
accepts a new state as input and uses that to process the given block's
transactions (and uncles for rewords) to calculate the state root for
the next block (P_n + 1).
Log filtering is now using a MIPmap like approach where addresses of
logs are added to a mapped bloom bin. The current levels for the MIP are
in ranges of 1.000.000, 500.000, 100.000, 50.000, 1.000. Logs are
therefor filtered in batches of 1.000.
Péter Szilágyi
Brian Schroeder
Justin
Joel Burget
Felföldi Zsolt
Jeffrey Wilcke
Martin Holst Swende
Vivek Anand
Nick Johnson
Felix Lange
Bas van Kervel
Jeffrey Wilcke
Bas van Kervel
Leif Jurvetson
Gustav Simonsson