Describe the bug
The node panics when trying to use debug_* methods, I've only identified debug_traceBlockByHash because someone reported it, there could be more.
I suspect this thing will happen on One and maybe Nova as well when the upgrade time comes, so any help to solve this issue is greatly appreciated. Thanks!
We can reproduce this with transactions affecting Stylus contracts, affecting debug_storageRangeAt breaking, the evmdis tracer has an out of bound stack as well.
Describe the bug The node panics when trying to use debug_* methods, I've only identified
debug_traceBlockByHash
because someone reported it, there could be more.To Reproduce Example request:
curl localhost:8545 -X POST -H "Content-Type: application/json" --data '{"id": 9513, "jsonrpc": "2.0", "method": "debug_traceBlockByHash", "params": ["0x4a574d301f31b516b8f5f46547b4a81fbd37ea68a6f08f794c331ef2902b5b3b", {"tracer": "prestateTracer", "tracerConfig": {"onlyTopCall": false, "diffMode": true}, "timeout": null}]}'
Output:{"jsonrpc":"2.0","id":9513,"error":{"code":-32603,"message":"method handler crashed"}}
Expected behavior The method to return a non-error response
Additional context Here is the stack-trace: https://gist.github.com/SMihai97/a664ef8786c9b69ee41cc2c3b39bda2e
Nitro version:
The node is running as a systemd unit, here is the command:
I suspect this thing will happen on One and maybe Nova as well when the upgrade time comes, so any help to solve this issue is greatly appreciated. Thanks!