fix: log full context of anyhow::Error

[hanabi1224]

Summary of changes

Changes introduced in this pull request:

• use alternative formatter (:#) for anyhow::Error in log to show full context. Check out this example

Reference issue to close (if applicable)

Closes

Other information and links

Change checklist

• I have performed a self-review of my own code,
• I have made corresponding changes to the documentation. All new code adheres to the team's documentation standards,
• I have added tests that prove my fix is effective or that my feature works (if possible),
• I have made sure the CHANGELOG is up-to-date. All user-facing changes should be reflected in this document.

Outside contributions

• I have read and agree to the CONTRIBUTING document.
• I have read and agree to the AI Policy document. I understand that failure to comply with the guidelines will lead to rejection of the pull request.

Summary by CodeRabbit

• Bug Fixes

  • Improved and standardized error and log formatting across the app for clearer, more informative diagnostics during DB, network, sync, RPC, and state operations.

• Refactor

  • Unified error-return conventions and enriched error contexts for more consistent and traceable failure reporting without changing runtime behavior or user-facing functionality.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 43968af3-1599-47f2-8c3b-f42ae4d756e0

📥 Commits

Reviewing files that changed from the base of the PR and between a5e858c and 9361c91.

📒 Files selected for processing (3)

• src/db/car/forest.rs
• src/db/car/many.rs
• src/rpc/methods/chain.rs

✅ Files skipped from review due to trivial changes (2)

• src/rpc/methods/chain.rs
• src/db/car/forest.rs

---

Walkthrough

Standardized many function error-return types to the anyhow::Result<T> alias, replaced string-based error returns with anyhow::Result, swapped ad-hoc error wrapping to .context()/anyhow::bail!/direct ? propagation, and changed numerous log/error formatters from {e} to {e:#}. No functional control-flow changes.

Changes

Cohort / File(s)	Summary
Result type standardization (core & state) src/state_manager/mod.rs, src/state_manager/*, src/interpreter/vm.rs, src/state_manager/circulating_supply.rs, src/state_manager/utils.rs	Converted many public/internal signatures from Result<..., anyhow::Error> to anyhow::Result<...>; adjusted error enrichment to use .context() and simplified propagation. Review state-manager constructors, VM APIs, and context propagation.
Result type standardization (misc modules) src/beacon/..., src/beacon/mock_beacon.rs, src/cli_shared/cli/mod.rs, src/fil_cns/*, src/genesis/mod.rs, src/message/mod.rs, src/message_pool/config.rs, src/utils/proofs_api/paramfetch.rs, src/shim/crypto.rs, src/tool/subcommands/*	Applied anyhow::Result alias across many modules (beacon, CLI, message pool, proofs API, shim, tools). Mostly signature-only changes; check public API surfaces for compile impacts.
Network / libp2p error conversions src/chain_sync/network_context.rs, src/libp2p/behaviour.rs, src/libp2p/discovery.rs, src/libp2p/chain_exchange/message.rs, src/libp2p/service.rs	Replaced Result<..., String> with anyhow::Result<_>, adapted TryFrom error types to anyhow::Error, and used anyhow::bail!/.context(); inspect chain-exchange TryFrom impls and caller sites for propagated error-type changes.
Error formatting in logs (#{...}) src/chain/store/chain_store.rs, src/chain_sync/chain_follower.rs, src/cli/subcommands/f3_cmd.rs, src/daemon/*, src/db/*, src/dev/subcommands/*, src/libp2p/service.rs, src/libp2p_bitswap/request_manager.rs, src/message_pool/msgpool/selection.rs, src/rpc/*, src/tool/subcommands/*, src/state_migration/*, src/db/car/*	Changed many "{e}" log interpolations to the alternate/pretty "{e:#}". Verify logging consumers and any tests that assert exact log strings.
Error handling pattern updates & context src/blocks/*, src/chain/store/errors.rs, src/db/parity_db.rs, src/message_pool/errors.rs, src/rpc/methods/eth.rs, src/rpc/methods/gas.rs, src/rpc/methods/state.rs, src/shim/state_tree.rs, src/utils/sqlite/mod.rs, src/db/car/many.rs	Replaced ad-hoc `map_err(
Callback / trait signature refinements src/shim/actors/builtin/market/mod.rs, src/shim/actors/builtin/miner/mod.rs, src/shim/state_tree.rs	Simplified Fn callback signatures to use anyhow::Result and removed redundant wrapping; ensure callers match updated Fn types and trait impls.
Chain-exchange / TryFrom conversions (high-impact) src/libp2p/chain_exchange/message.rs	Converted ChainExchangeResponse/TipsetBundle TryFrom/Error types from String to anyhow::Error, replaced string-based conversions with ?/.context() and anyhow::bail!. This is a dense change; review conversion failure paths and caller handling.
Misc small signature/log tweaks src/blocks/mod.rs, src/statediff/*, src/state_migration/*, src/message_pool/errors.rs, src/state_manager/errors.rs	Mix of error-message formatting changes, payload type tweaks (e.g., Cow<'static, str>), and From<anyhow::Error> for Error conversions now storing formatted debug ({e:#}).

Estimated code review effort

🎯 4 (Complex) | ⏱️ ~60 minutes

Possibly related PRs

• feat(api): Implement debug_traceTransaction API #6751 — overlaps with src/interpreter/vm.rs return-type changes to anyhow::Result.
• feat: Implement Forest.EthTraceCall #6412 — modifies src/rpc/methods/eth.rs trace/decoding error handling (context and decode paths).
• fix: improve forest-tool snapshot validate error message #5969 — touches state_manager::validate_tipsets and related error-formatting changes.

Suggested reviewers

• LesnyRumcajs
• akaladarshi

🚥 Pre-merge checks | ✅ 3

✅ Passed checks (3 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The pull request title 'fix: log full context of anyhow::Error' accurately describes the main change: updating error logging throughout the codebase to use the alternate formatter '{e:#}' for displaying the full error context chain of anyhow::Error.
Docstring Coverage	✅ Passed	Docstring coverage is 96.72% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch hm/print-anyhow-error-context

✨ Simplify code

• Create PR with simplified code
• Commit simplified code in branch hm/print-anyhow-error-context

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 1

Caution

Some comments are outside the diff and can’t be posted inline due to platform limitations.

⚠️ Outside diff range comments (1)

src/chain_sync/network_context.rs (1)

214-254: ⚠️ Potential issue | 🟡 Minor

Simplify error handling by removing unnecessary anyhow error wrapping in callers.

The calls to chain_exchange_full_tipset() and chain_exchange_full_tipsets() in src/chain_sync/chain_follower.rs (lines 443 and 463) use .map_err(|e| anyhow::anyhow!(e))? to handle anyhow::Result<T>. This pattern wraps an already-wrapped anyhow::Error, converting it to a string representation and losing the original error chain.

Since both methods already return anyhow::Result<T>, simplify these to:

• await? for direct propagation, or
• .await.context("additional context")? if additional context is needed

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.
In `@src/chain_sync/network_context.rs` around lines 214 - 254, The callers of
chain_exchange_full_tipset() and chain_exchange_full_tipsets() are wrapping an
anyhow::Error into a new anyhow::Error via .map_err(|e| anyhow::anyhow!(e))?
which loses the original error chain; locate the calls to
chain_exchange_full_tipset and chain_exchange_full_tipsets in chain_follower.rs
and replace .await.map_err(|e| anyhow::anyhow!(e))? with a simple .await? to
propagate the existing anyhow::Error, or use .await.context("additional
context")? if you need to add context.

🧹 Nitpick comments (3)

src/utils/sqlite/mod.rs (1)

150-158: Consider using {e:#} for error logging to align with PR objective.

These tracing::warn! calls still use {e} format, which only displays the top-level error message. Given this PR's objective is to show full error context/chain using the {:#} formatter, these lines should be updated as well.

♻️ Suggested change

 if let Err(e) = sqlx::query("VACUUM").execute(db).await {
- tracing::warn!("error vacuuming {name} database: {e}")+ tracing::warn!("error vacuuming {name} database: {e:#}")
}
if let Err(e) = sqlx::query("PRAGMA wal_checkpoint(TRUNCATE)")
.execute(db)
.await
{
- tracing::warn!("error checkpointing {name} database wal: {e}")+ tracing::warn!("error checkpointing {name} database wal: {e:#}")
}

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.
In `@src/utils/sqlite/mod.rs` around lines 150 - 158, Update the tracing warnings
in src/utils/sqlite/mod.rs to use the pretty/error-chain formatter so full error
context is logged: replace the "{e}" interpolations with "{e:#}" in the two
tracing::warn! calls associated with the sqlx::query("VACUUM") execute error and
the sqlx::query("PRAGMA wal_checkpoint(TRUNCATE)") execute error (the two blocks
that currently bind Err(e) and call tracing::warn!). Ensure both messages now
use {e:#} so the complete error chain is emitted.

src/db/blockstore_with_write_buffer.rs (1)

66-66: LGTM! Error formatting improvement applied correctly.

The change from {e} to {e:#} correctly applies the alternate formatter to display the full error context chain, which aligns with the PR objectives and improves debuggability.

Optional: Consider adding descriptive context to the warning message.

The warning currently logs only the error itself. Adding a brief description would improve log readability:

- tracing::warn!("{e:#}");+ tracing::warn!("Failed to flush buffer on drop: {e:#}");

This makes it immediately clear what operation failed, especially useful when scanning logs.

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.
In `@src/db/blockstore_with_write_buffer.rs` at line 66, The current warning call
uses tracing::warn!("{e:#}") which only prints the error chain; update that
tracing::warn invocation to include a short descriptive context string (e.g.,
"failed to flush write buffer" or "error writing blockstore") combined with the
error variable e so logs show both what operation failed and the error chain;
locate the tracing::warn!("{e:#}") line and modify the message to include the
operation context while keeping {e:#} for full error formatting.

src/rpc/methods/eth/trace/state_diff.rs (1)

219-219: Inconsistent error formatting within the same function.

Line 219 now uses {e:#} for richer error context, but lines 230 and 240 still use {e}. For consistency with the PR's goal of printing full error chains, consider updating those as well.

Suggested diff for consistency

 let kamt: EvmStorageKamt<&DB> = match Kamt::load_with_config(&storage_cid, store, config) {
Ok(k) => k,
Err(e) => {
- debug!("failed to load storage KAMT: {e}");+ debug!("failed to load storage KAMT: {e:#}");
return HashMap::default();
}
};
let mut entries = HashMap::default();
if let Err(e) = kamt.for_each(|key, value| {
entries.insert(key.to_big_endian(), *value);
Ok(())
}) {
- debug!("failed to iterate storage KAMT: {e}");+ debug!("failed to iterate storage KAMT: {e:#}");
return HashMap::default();
}

🤖 Prompt for AI Agents

Verify each finding against the current code and only fix it if needed.
In `@src/rpc/methods/eth/trace/state_diff.rs` at line 219, The debug log in
state_diff.rs uses rich error formatting in one place (debug!("failed to load
EVM state for storage extraction: {e:#}")) but other debug/error logs in the
same function still use plain "{e}"; update the remaining debug/error calls in
the same function (look for the other debug/error statements around the storage
extraction logic inside the function handling EVM state — e.g., the other debug
lines after the initial load attempt) to use "{e:#}" so all logs print full
error chains consistently, keeping the same message text and log level.

🤖 Prompt for all review comments with AI agents

Verify each finding against the current code and only fix it if needed.
Inline comments:
In `@src/libp2p/chain_exchange/message.rs`:
- Around line 140-143: The closure that maps TipsetBundle into T currently
discards the original conversion error by formatting a fresh string; instead
wrap the original error into an anyhow::Error and add context so the error chain
is preserved (e.g., replace the map_err(|e| anyhow::anyhow!("failed to convert
from tipset bundle: {e}")) with something that uses
anyhow::Error::new(e).context("failed to convert from tipset bundle") so the
original source (from T::try_from / FullTipset::try_from) is retained).
---
Outside diff comments:
In `@src/chain_sync/network_context.rs`:
- Around line 214-254: The callers of chain_exchange_full_tipset() and
chain_exchange_full_tipsets() are wrapping an anyhow::Error into a new
anyhow::Error via .map_err(|e| anyhow::anyhow!(e))? which loses the original
error chain; locate the calls to chain_exchange_full_tipset and
chain_exchange_full_tipsets in chain_follower.rs and replace .await.map_err(|e|
anyhow::anyhow!(e))? with a simple .await? to propagate the existing
anyhow::Error, or use .await.context("additional context")? if you need to add
context.
---
Nitpick comments:
In `@src/db/blockstore_with_write_buffer.rs`:
- Line 66: The current warning call uses tracing::warn!("{e:#}") which only
prints the error chain; update that tracing::warn invocation to include a short
descriptive context string (e.g., "failed to flush write buffer" or "error
writing blockstore") combined with the error variable e so logs show both what
operation failed and the error chain; locate the tracing::warn!("{e:#}") line
and modify the message to include the operation context while keeping {e:#} for
full error formatting.
In `@src/rpc/methods/eth/trace/state_diff.rs`:
- Line 219: The debug log in state_diff.rs uses rich error formatting in one
place (debug!("failed to load EVM state for storage extraction: {e:#}")) but
other debug/error logs in the same function still use plain "{e}"; update the
remaining debug/error calls in the same function (look for the other debug/error
statements around the storage extraction logic inside the function handling EVM
state — e.g., the other debug lines after the initial load attempt) to use
"{e:#}" so all logs print full error chains consistently, keeping the same
message text and log level.
In `@src/utils/sqlite/mod.rs`:
- Around line 150-158: Update the tracing warnings in src/utils/sqlite/mod.rs to
use the pretty/error-chain formatter so full error context is logged: replace
the "{e}" interpolations with "{e:#}" in the two tracing::warn! calls associated
with the sqlx::query("VACUUM") execute error and the sqlx::query("PRAGMA
wal_checkpoint(TRUNCATE)") execute error (the two blocks that currently bind
Err(e) and call tracing::warn!). Ensure both messages now use {e:#} so the
complete error chain is emitted.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 469e916d-5db1-4eb6-807f-b24ddb876385

📥 Commits

Reviewing files that changed from the base of the PR and between 4539320 and e54b2a2.

📒 Files selected for processing (57)

• src/beacon/drand.rs
• src/beacon/mock_beacon.rs
• src/blocks/header.rs
• src/blocks/mod.rs
• src/chain/store/chain_store.rs
• src/chain/store/errors.rs
• src/chain_sync/chain_follower.rs
• src/chain_sync/network_context.rs
• src/chain_sync/tipset_syncer.rs
• src/cli/subcommands/f3_cmd.rs
• src/cli_shared/cli/mod.rs
• src/daemon/context.rs
• src/daemon/db_util.rs
• src/daemon/mod.rs
• src/db/blockstore_with_write_buffer.rs
• src/db/car/forest.rs
• src/db/gc/snapshot.rs
• src/db/parity_db.rs
• src/dev/subcommands/mod.rs
• src/dev/subcommands/update_checkpoints_cmd.rs
• src/fil_cns/mod.rs
• src/fil_cns/validation.rs
• src/genesis/mod.rs
• src/interpreter/mod.rs
• src/interpreter/vm.rs
• src/libp2p/behaviour.rs
• src/libp2p/chain_exchange/message.rs
• src/libp2p/discovery.rs
• src/libp2p/service.rs
• src/libp2p_bitswap/request_manager.rs
• src/message/mod.rs
• src/message_pool/config.rs
• src/message_pool/errors.rs
• src/message_pool/msgpool/selection.rs
• src/rpc/error.rs
• src/rpc/methods/chain.rs
• src/rpc/methods/eth.rs
• src/rpc/methods/eth/trace/state_diff.rs
• src/rpc/methods/gas.rs
• src/rpc/methods/state.rs
• src/shim/actors/builtin/market/mod.rs
• src/shim/actors/builtin/miner/mod.rs
• src/shim/crypto.rs
• src/shim/state_tree.rs
• src/state_manager/circulating_supply.rs
• src/state_manager/errors.rs
• src/state_manager/mod.rs
• src/state_manager/utils.rs
• src/state_migration/common/state_migration.rs
• src/state_migration/nv19/power.rs
• src/state_migration/type_migrations/init/state_v9_to_v10.rs
• src/statediff/mod.rs
• src/statediff/resolve.rs
• src/tool/subcommands/api_cmd.rs
• src/tool/subcommands/snapshot_cmd.rs
• src/utils/proofs_api/paramfetch.rs
• src/utils/sqlite/mod.rs

[codecov]

Codecov Report

❌ Patch coverage is 35.00000% with 130 lines in your changes missing coverage. Please review.
✅ Project coverage is 63.81%. Comparing base (36b489e) to head (9361c91).
⚠️ Report is 2 commits behind head on main.
✅ All tests successful. No failed tests found.

Files with missing lines	Patch %	Lines
src/blocks/header.rs	3.22%	25 Missing and 5 partials ⚠️
src/chain_sync/network_context.rs	26.92%	18 Missing and 1 partial ⚠️
src/libp2p/chain_exchange/message.rs	53.33%	4 Missing and 3 partials ⚠️
src/shim/state_tree.rs	22.22%	2 Missing and 5 partials ⚠️
src/state_manager/mod.rs	46.15%	3 Missing and 4 partials ⚠️
src/chain_sync/tipset_syncer.rs	14.28%	5 Missing and 1 partial ⚠️
src/rpc/methods/eth.rs	0.00%	4 Missing and 1 partial ⚠️
src/libp2p/service.rs	0.00%	4 Missing ⚠️
src/statediff/mod.rs	20.00%	4 Missing ⚠️
src/cli/subcommands/f3_cmd.rs	0.00%	3 Missing ⚠️
... and 24 more

Additional details and impacted files

Files with missing lines	Coverage Δ
src/beacon/drand.rs	76.21% <100.00%> (-1.41%)	⬇️
src/chain/store/errors.rs	14.28% <100.00%> (ø)
src/cli_shared/cli/mod.rs	78.94% <100.00%> (ø)
src/db/car/many.rs	67.40% <100.00%> (ø)
src/fil_cns/mod.rs	64.70% <100.00%> (ø)
src/fil_cns/validation.rs	78.75% <100.00%> (ø)
src/genesis/mod.rs	75.00% <100.00%> (ø)
src/interpreter/mod.rs	77.27% <100.00%> (ø)
src/interpreter/vm.rs	81.73% <100.00%> (ø)
src/message/mod.rs	90.56% <100.00%> (ø)
... and 46 more

... and 4 files with indirect coverage changes

---

Continue to review full report in Codecov by Sentry.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 36b489e...9361c91. Read the comment docs.

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[akaladarshi]

Mostly LGTM, but there are couple of concerns