Thread 'IO Worker #2' panicked at 'attempted to leave type linked_hash_map::Node<ethereum_types::H512, (node_table::NodeEndpoint, std::time::Instant)> uninitialized, which is invalid'

[mondsen]

Before filing a new issue, please provide the following information.

• OpenEthereum version (>=3.1.0): v3.1.0-rc1-stable-ee2b16dfe-20201112
• Operating system: Linux aarch64-linux-gn
• Installation: built from source
• Fully synchronized: no
• Network: ethereum
• Restarted: no

Problem
I wanted to start a new ETH node based on latest openethereum version on my RPi4 (8GB). I've build from source. I started with these parameters: --cache-size 4096 --allow-ips=public --max-peers=128 --warp-barrier 11330000 --jsonrpc-hosts=xx.xx.xx.xx

Expected
Sync starts and no error occurs

stacktrace
`2020-11-28 10:39:11 26/128 peers 1 KiB chain 0 bytes queue RPC: 0 conn, 0 req/s, 0 µs
2020-11-28 10:39:41 26/128 peers 1 KiB chain 0 bytes queue RPC: 0 conn, 0 req/s, 0 µs
2020-11-28 10:40:11 25/25 peers 1 KiB chain 0 bytes queue RPC: 0 conn, 0 req/s, 0 µs
2020-11-28 10:40:41 25/25 peers 1 KiB chain 0 bytes queue RPC: 0 conn, 0 req/s, 0 µs
2020-11-28 10:40:51 Panic occured, see stderr for details

====================

stack backtrace:
0: panic_hook::set_with::{{closure}}
1: std::panicking::rust_panic_with_hook
at library/std/src/panicking.rs:581
2: std::panicking::begin_panic_handler::{{closure}}
at library/std/src/panicking.rs:484
3: std::sys_common::backtrace::__rust_end_short_backtrace
at library/std/src/sys_common/backtrace.rs:153
4: rust_begin_unwind
at library/std/src/panicking.rs:483
5: core::panicking::panic_fmt
at library/core/src/panicking.rs:85
6: core::panicking::panic
at library/core/src/panicking.rs:50
7: linked_hash_map::LinkedHashMap<K,V,S>::insert
8: lru_cache::LruCache<K,V,S>::insert
9: ethcore_network_devp2p::discovery::Discovery::on_packet
10: <ethcore_network_devp2p::host::Host as ethcore_io::IoHandler<ethcore_network::NetworkIoMessage>>::stream_readable
11: ethcore_io::worker::Worker::new::{{closure}}::{{closure}}
12: <futures::future::loop_fn::LoopFn<A,F> as futures::future::Future>::poll
13: std::sys_common::backtrace::__rust_begin_short_backtrace
14: core::ops::function::FnOnce::call_once{{vtable.shim}}
15: <alloc::boxed::Box as core::ops::function::FnOnce>::call_once
at /rustc/7eac88abb2e57e752f3302f02be5f3ce3d7adfb4/library/alloc/src/boxed.rs:1042
<alloc::boxed::Box as core::ops::function::FnOnce>::call_once
at /rustc/7eac88abb2e57e752f3302f02be5f3ce3d7adfb4/library/alloc/src/boxed.rs:1042
std::sys::unix::thread::Thread::new::thread_start
at library/std/src/sys/unix/thread.rs:87
16: start_thread
17:

Thread 'IO Worker #2' panicked at 'attempted to leave type linked_hash_map::Node<ethereum_types::H512, (node_table::NodeEndpoint, std::time::Instant)> uninitialized, which is invalid', /rustc/7eac88abb2e57e752f3302f02be5f3ce3d7adfb4/library/core/src/mem/mod.rs:658`

[ryanobjc]

Yes me too, OSX 10.15.7:
$ rustup show
Default host: x86_64-apple-darwin
rustup home: /Users/user/.rustup

stable-x86_64-apple-darwin (default)
rustc 1.48.0 (7eac88abb 2020-11-16)

I started with only a -d $(DATA_DIR) argument.

this is with commit ee2b16d from branch 'main'

EDIT:

This bug doesn't repro for me on the 'dev' branch.

[rakita]

this is fixed in dev branch

[MikeJerred]

Cherry-picking 233bab2 onto the v3.1.0 release was enough to fix this for me.

[rakita]

Duplicate #120