It takes 160+ seconds when using locklift for testing. Also removed contract is not accessible from explorer

Previous behavior:

Takes 2-3 seconds and accessible from explorer

Steps to reproduce:

• Build and Run Node SE from https://github.com/tonlabs/tonos-se/tree/master/ton-node-se
• Run tests from https://github.com/tonlabs/ton-client-js#run-tests

Result:

PS>$env:USE_NODE_SE="true";$env:TON_NETWORK_ADDRESS="http://localhost:8080";node run
1.11.0
✓ Test versions compatibility (1 / 0)
✓ crypto (2 / 0)
✓ encode_message (3 / 0)
✓ net (4 / 0)
✓ Block signatures (5 / 0)
✓ All Accounts (6 / 0)
✓ Ranges (7 / 0)
✓ Wait For (8 / 0)
𐄂 Subscribe for transactions with addresses (ABIv1) (8 / 1) - [
    {
        "matcherResult": {
            "pass": false
        },
        "message": "\u001b[2mexpect(\u001b[22m\u001b[31mreceived\u001b[39m\u001b[2m).\u001b[22mtoBeGreaterThan\u001b[2m(\u001b[22m\u001b[32mexpected\u001b[39m\u001b[2m)\u001b[22m\n\nExpected: > \u001b[32m0\u001b[39m\nReceived:   \u001b[31m0\u001b[39m"
    }
]
✓ Subscribe for transactions with addresses (ABIv2) (9 / 1)
✓ Subscribe for messages (ABI v1) (10 / 1)
✓ Subscribe for messages (ABI v2) (11 / 1)
✓ Transactions with addresses (12 / 1)
✓ Check shard_hashes greater then 0 (13 / 1)
---
success: 13
failure: 1

PS: I've open PR with a workaround for tests https://github.com/tonlabs/ton-client-js/pull/206

There are two of the easiest smart contracts firstContract and secondContract. Deploy 'firstContract' smart contract and calling external method contract "setValue", set value a,b=12. Data state contract: m_value = 24, version = 1. Calling 'updateContractCode' and set code from 'secondContract'. After update smart contract check values version=2 but m_value=0. Waiting that m_value=24,, because tvm.resetStorage don't calling. This is inappropriate behavior on upgraded data states to default value.

firstContract code: `pragma ton-solidity >= 0.57.0;

pragma AbiHeader time; pragma AbiHeader expire; pragma AbiHeader pubkey;

contract firstContract {

uint32 nonce_;
uint public m_version = 1;
uint public m_value;

address owner;

constructor(address owner_ ) public {
	tvm.accept();
	owner = owner_;
}

modifier checkOwnerAndAccept {
	require(msg.sender == owner, 120);
	tvm.accept();
	_;
}

function setValue(uint a, uint b) external {
	tvm.rawReserve(1 ton, 0);
	m_value = a + b; 
}

function getOwner() external view responsible returns (address) {
	return {value: 0, flag: 64, bounce: false} owner;

}

function updateContractCode(TvmCell newcode) public view checkOwnerAndAccept {
	tvm.setcode(newcode);
	tvm.setCurrentCode(newcode);
	
	TvmCell stateVars = abi.encode(m_version);
    onCodeUpgrade(stateVars);
}

function onCodeUpgrade(TvmCell stateVars) private pure {}

}`

secondContract code: `pragma ton-solidity >= 0.57.0;

pragma AbiHeader time; pragma AbiHeader expire; pragma AbiHeader pubkey;

contract secondContract {

uint32 nonce_;
uint public m_version; 
uint public m_value;

address owner;

modifier checkOwnerAndAccept {
	require(msg.sender == owner, 110);
	tvm.accept();
	_;
}

function setValue(uint a, uint b) external {
	tvm.rawReserve(1 ton, 0);
	m_value = a + b; 
}

function getOwner() external view responsible returns (address) {
	return {value: 0, flag: 64, bounce: false} owner;

}

function updateContractCode(TvmCell newcode) public checkOwnerAndAccept {
	tvm.setcode(newcode);
	tvm.setCurrentCode(newcode);
	TvmCell stateVars = abi.encode(m_version);

	onCodeUpgrade(stateVars);
}

function onCodeUpgrade(TvmCell stateVars) private {
	(uint version) = abi.decode(stateVars, (uint));
	m_version = version + 1;
}

}`

Problem

We need to destruct the contact and transfer its tokens to some destination account.

Version

tonlabs/local-node:0.35.0

Expected behaviour

• Desctructed contract has zero balance and "non exists" type.
• Account destination has positive balance

Resulting behaviour

• in Devnet
  • as expected
• in Evernode SE
  • Desctructed contract is not destructed, its balance unchanged, but transfer to destination account exists!
  • Account destination has positive balance!

How to reproduce a bug

To reproduce this bug clone https://github.com/Artem-Zhdanov/evernode-se-selfdestruct-bug

everdev se start

Intall dependencies and run

npm i
node selfdestruct.js

Open http://localhost and finds accounts in the output

root@13485a15bdb7:/tonos-se/ton-node-se# cargo build --release Updating crates.io index Updating git repository https://github.com/tonlabs/ton-labs-block.git Updating git repository https://github.com/tonlabs/ton-labs-block-json.git Updating git repository https://github.com/tonlabs/ton-labs-executor.git Updating git repository https://github.com/tonlabs/ton-labs-types.git Updating git repository https://github.com/tonlabs/ton-labs-vm.git Updating git repository https://github.com/tonlabs/ton-labs-tl.git Updating git repository https://github.com/tonlabs/ton-labs-assembler.git Updating git repository https://github.com/paritytech/rust-secp256k1 Updating git repository https://github.com/tonlabs/TON-SDK.git Updating git repository https://github.com/tonlabs/lockfree.git error: failed to select a version for the requirement stream-cipher = "^0.3" candidate versions found which didn't match: 0.99.99, 0.7.1, 0.6.0, ... location searched: crates.io index required by package aes-ctr v0.3.0 ... which is depended on by adnl v0.0.4 (/tonos-se/ton-node-se/adnl) ... which is depended on by ton_node_startup v0.28.4 (/tonos-se/ton-node-se/ton_node_startup)

I tried to start SE with the command tondev se start but it didn't work: tondev se info shows that the default instance has state "exited". Also I tried docker run -e USER_AGREEMENT=yes -p80:80 tonlabs/local-node and the container exited with the following output:

16:08:57 system | node_workers.1 started (pid=14) 16:08:57 node_workers.1 | + set -e 16:08:57 node_workers.1 | + cd /ton-node 16:08:57 node_workers.1 | + exec /node/ton-node --config ./ton-node.conf.json --blockchain-config ./blockchain.conf.json 16:08:57 system | arango.1 started (pid=12) 16:08:57 arango.1 | + set -e 16:08:57 arango.1 | + ARANGO_INIT_PORT=8529 16:08:57 arango.1 | + ARANGO_ENDPOINT=tcp://127.0.0.1:8529 16:08:57 arango.1 | + ARANGO_INIT_PIDFILE=/run/arango-init.pid 16:08:57 arango.1 | + export GLIBCXX_FORCE_NEW=1 16:08:57 arango.1 | + GLIBCXX_FORCE_NEW=1 16:08:57 arango.1 | + NUMACTL= 16:08:57 arango.1 | + '[' -d /sys/devices/system/node/node1 -a -f /proc/self/numa_maps ']' 16:08:57 arango.1 | + arangod --config /arango/config --server.endpoint tcp://127.0.0.1:8529 --server.authentication=false --log.foreground-tty true --database.auto-upgrade true 16:08:57 system | nginx.1 started (pid=13) 16:08:57 system | q_server.1 started (pid=11) 16:08:57 q_server.1 | + set -e 16:08:57 q_server.1 | + export 'Q_DATA_MUT=http://127.0.0.1:8529' 16:08:57 q_server.1 | + export 'Q_DATA_HOT=http://127.0.0.1:8529' 16:08:57 q_server.1 | + export 'Q_SLOW_QUERIES_MUT=http://127.0.0.1:8529' 16:08:57 q_server.1 | + export 'Q_SLOW_QUERIES_HOT=http://127.0.0.1:8529' 16:08:57 q_server.1 | + export 'Q_REQUESTS_MODE=rest' 16:08:57 q_server.1 | + export 'Q_REQUESTS_SERVER=http://127.0.0.1' 16:08:57 q_server.1 | + export 'Q_HOST=127.0.0.1' 16:08:57 q_server.1 | + export 'Q_PORT=4000' 16:08:57 q_server.1 | + cd /home/node/ton-q-server 16:08:57 q_server.1 | + exec node index.js 16:08:57 arango.1 | /arango/entrypoint: line 34: 20 Illegal instruction (core dumped) $NUMACTL arangod --config /arango/config --server.endpoint $ARANGO_ENDPOINT --server.authentication=false --log.foreground-tty true --database.auto-upgrade true 16:08:57 arango.1 | + exit 1 16:08:57 system | arango.1 stopped (rc=1) 16:08:57 system | sending SIGTERM to q_server.1 (pid 11) 16:08:57 system | sending SIGTERM to nginx.1 (pid 13) 16:08:57 system | sending SIGTERM to node_workers.1 (pid 14) 16:08:57 system | node_workers.1 stopped (rc=-15) 16:08:57 system | q_server.1 stopped (rc=-15) 16:08:57 system | nginx.1 stopped (rc=0)

What's nginx for here?

It is used for routing requests for Node SE or Q-Server, according to the path. You can inspect its configuration for deeper investigation: https://github.com/tonlabs/tonos-se/blob/master/docker/nginx.conf.d

What route them for?

1. Why is honcho in particular used here?

2. What's the neccessity? Why not run all of them without honcho, as ordinary processes? Is it specific to the fact that it's all done in Docker?

Based on:

Set the image's platform explicit https://stackoverflow.com/questions/65612411/forcing-docker-to-use-linux-amd64-platform-by-default-on-macos

Fix docker memory leak during cargo build https://github.com/rust-lang/cargo/issues/10583

Use stable alpine version to prevent segmentation fault during cargo build https://github.com/mxe/mxe/issues/2621

Issue lines:

https://github.com/tonlabs/evernode-se/blob/master/contracts/giver_v2/GiverV2.sol#L57 https://github.com/tonlabs/evernode-se/blob/master/contracts/giver_v3/GiverV3.sol#L58

Description:

Giver contract passes compilation, but any sendTransaction call will be reverted with code 9 (Cell underflow) on mainnet

Proposal:

Replace all expireAt entries with uint32 instead of uint64 GiverV3.sol.zip