|WARNING| test/pcapng: add some leeway to timestamp test

[dpdklab at iol.unh.edu]

Test-Label: ai-code-review
Test-Status: WARNING
https://dpdk.org/patch/166465

_AI Code Review_

=== Patch Review: series.patch (via Claude) ===
Review date: 2026-07-02

# DPDK Patch Review

## Patch: test/pcapng: add some leeway to timestamp test

### Summary
This patch modifies the timestamp validation test for pcapng files to allow timestamps up to 1 second after capture was stopped, to accommodate timing variations on busy or emulated systems.

---

## Review Findings

### Errors

None.

### Warnings

None.

### Info

**1. Test logic consideration**

The patch adds 1 second of leeway to the `end_ns` timestamp boundary:
```c
ctx.end_ns = current_timestamp() + NS_PER_S;
```

This means the test now accepts packet timestamps that are up to 1 second in the *future* relative to when `pcap_open_offline_with_tstamp_precision()` is called, rather than relative to when capture stopped.

**Consider:** The comment says "after capture was stopped" but the code actually uses `current_timestamp()` which is called during validation, not at capture stop time. If the intent is to allow 1 second after capture stopped, the correct pattern would be:
```c
ctx.start_ns = started;
ctx.end_ns = capture_stopped_timestamp + NS_PER_S;
```

However, if the test previously worked with `ctx.end_ns = current_timestamp()` and only failed occasionally due to timing delays in file I/O or test infrastructure between stopping capture and opening the file, then adding slop to `current_timestamp()` is reasonable.

**2. Magic number documentation**

The `NS_PER_S` constant is clear, but the rationale for choosing 1 second specifically could be documented in a comment. One second is quite generous for test timing slop; most DPDK tests use millisecond-level tolerances.

Suggested addition:
```c
/* Allow up to 1 second slop for file I/O and test infrastructure
 * delays on busy/emulated systems */
ctx.end_ns = current_timestamp() + NS_PER_S;
```

---

## Conclusion

The patch is acceptable as-is. The change is localized to a test helper function and uses an appropriate pattern for timing tolerance in tests. The info-level comments above are suggestions for clarity, not required changes.

**Code Style:** Compliant  
**Correctness:** No bugs identified  
**API/Documentation:** Test-only change, no release notes needed  
**Overall:** LGTM with optional documentation enhancement