Panic on failed downcast from BoxError to redjubjub::Error

[teor2345]

Version

zebrad main branch, 2020-11-23.

Platform

Linux ... 5.4.75 #1-NixOS SMP Thu Nov 5 10:43:38 UTC 2020 x86_64 GNU/Linux

Description

When syncing Mainnet with checkpoint_sync = true around block 1018849:

1. Zebra panics due to a failed downcast from BoxError to redjubjub::Error
2. Using panic!(e) displays an unhelpful <non string panic payload> panic message

Note: 1018849 is the second-last checkpoint when checkpoint_sync = true.

The failing code is:

impl From<BoxError> for TransactionError {
    fn from(err: BoxError) -> Self {
        match err.downcast::<redjubjub::Error>() {
            Ok(e) => TransactionError::RedJubjub(*e),
            Err(e) => panic!(e),
        }
    }
}

The logs are:

Nov 23 15:38:42.088  INFO sync:extend_tips:download_and_verify{hash=block::Hash("00000000021615cd8f0425dfa64960804f68eeba533270c02bb44e8b9bea2420")}:checkpoint: zebra_consensus::checkpoint: verified checkpoint range block_count=400 current_range=(Excluded(Height(1018049)), Included(Height(1018449)))                                                                                  
Nov 23 15:38:46.022  INFO sync: zebrad::components::sync: waiting for pending blocks tips.len=1 in_flight=2372 lookahead_limit=2000                                                            
Nov 23 15:38:46.023  INFO sync: zebrad::components::sync: extending tips tips.len=1 in_flight=2000 lookahead_limit=2000                                                                        
Nov 23 15:38:46.479  INFO sync:extend_tips:download_and_verify{hash=block::Hash("00000000005b1fd9b021a22f493818246efd17b77a50a524e17b570c1d55bbfa")}:checkpoint: zebra_consensus::checkpoint: verified checkpoint range block_count=400 current_range=(Excluded(Height(1018449)), Included(Height(1018849)))                                                                                  
The application panicked (crashed).                                                                                                                                                            
Message:  <non string panic payload>
Location: zebra-consensus/src/error.rs:76
  
  ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ SPANTRACE ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
      
   0: zebrad::components::sync::downloads::download_and_verify with hash=block::Hash("00000000002c76519f671c6789f99af6345fac144509863fb01185cab68198f9")
      at zebrad/src/components/sync/downloads.rs:113
   1: zebrad::components::sync::extend_tips
      at zebrad/src/components/sync.rs:381
   2: zebrad::components::sync::sync
      at zebrad/src/components/sync.rs:150

Related Issues

Closed by #1363.
Cleanup underlying cause in #1186

[teor2345]

Note: 1018849 is the second-last Mainnet checkpoint when checkpoint_sync = true. The last checkpoint is 1019249.

[hdevalence]

This can be fixed as part of a fix for #1186; I think it would be good to fix now so that code we write has correct error types going forward.

I see this happening only several hundred thousand blocks after Sapling activation.

[teor2345]

This can be fixed as part of a fix for #1186; I think it would be good to fix now so that code we write has correct error types going forward.

I've moved #1186 into the first alpha milestone.