Fix: CallErrorTracer Logging and Update Revert/Halt Handling

[dutterbutter]

Problem

The CallErrorTracer produces repetitive, unstructured logs (tracing::error!("!! Got error flags:")) without context, making them hard to interpret. Additionally, handling for ExecutionResult::Revert and ExecutionResult::Halt lacks actionable insights or proper structuring.

Solution

1. Structured Logging
   Replace unstructured error logs with clear, context-rich messages:

   • Include VM error flags.
   • Add transaction revert reasons where applicable.
   • Ensure logs provide actionable insights.

2. Update Handling for Revert/Halt Results
   Enhance ExecutionResult::Revert and ExecutionResult::Halt handling to:

   • Log meaningful context (e.g., transaction details, reason for revert/halt).
   • Structure messages to align with improved logging standards.

Action Items

1. Refactor CallErrorTracer to:
   • Replace tracing::error!("!! Got error flags:") with structured logs.
   • Include VM error flags and relevant details.
2. Update handling for ExecutionResult::Revert and ExecutionResult::Halt:
   • Extract and log transaction context and revert/halt reasons.
   • Ensure concise, actionable output.

Example of current output:

01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS
01:49:17 ERROR !! Got error flags: 
01:49:17 ERROR NOT ENOUGH ERGS

[dutterbutter]

WIP and related work:

• https://github.com/matter-labs/era-test-node/tree/db/better-error-messages
• Ideally we make use of [NEVER MERGE] Example of migration to unified failure handling system  #428

Also relevant are the error messages that we need to distinguish, they primarily made up of:

• https://github.com/matter-labs/zksync-era/blob/main/core/lib/multivm/src/glue/types/vm/tx_revert_reason.rs
• https://github.com/matter-labs/zksync-era/blob/main/core/lib/multivm/src/glue/types/vm/vm_revert_reason.rs
• API error responses

Ideally, we can even be more granular and distinguish TxRevertReason::ValidationFailed from incorrect signature to insufficient funds.

Initial view of errors: