Skip to content

Conversation

@ripatel-fd
Copy link
Contributor

  • Move all lock guards into their own scopes to defend against
    lock leaks
  • Reject input instead of crashing if sysvars are missing

- Move all lock guards into their own scopes to defend against
  lock leaks
- Reject input instead of crashing if sysvars are missing
@github-actions
Copy link

Performance Measurements ⏳

Suite Baseline New Change
backtest mainnet-368528500-perf per slot 0.063623 s 0.063796 s 0.272%
backtest mainnet-368528500-perf snapshot load 2.399 s 2.334 s -2.709%
backtest mainnet-368528500-perf total elapsed 63.622852 s 63.796136 s 0.272%
firedancer mem usage with mainnet.toml 1001.1 GiB 1001.1 GiB 0.000%

Comment on lines +376 to +379
if( FD_UNLIKELY( !epoch_schedule ) ) {
FD_LOG_WARNING(( "Block test vector missing epoch_schedule sysvar" ));
return NULL;
}
Copy link
Contributor

@mjain-jump mjain-jump Nov 20, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

i actually prefer crashing here because otherwise small mutator changes that alter this behavior go unnoticed

Copy link
Contributor

@mjain-jump mjain-jump left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

scoping looks good, but dont agree with the silent input rejection

@ripatel-fd
Copy link
Contributor Author

@mjain-jump there is no silent input rejection. test_sol_compat logs a warning and the shared interface reports up an error. It is up to the shared interface user to crash though if it gracefully reports an invalid input. It doesn't make sense to make the entire interface crash.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants