|SUCCESS| net/netvsc: fix event buffer leak on realloc failure

[dpdklab at iol.unh.edu]

Test-Label: ai-code-review
Test-Status: SUCCESS
https://dpdk.org/patch/172060

_AI Code Review_

=== Patch Review: series.patch (via Claude) ===
Review date: 2026-10-25

# DPDK Patch Review

## ERRORS

None.

## WARNINGS

None.

## INFO

### 1. Patch correctly fixes resource leak on rte_realloc() failure

The patch addresses a real bug where `rte_realloc()` failure would overwrite `rxq->event_buf` and leak the original allocation. The fix correctly:
- Saves the old pointer before calling `rte_realloc()`
- Frees the old buffer if realloc fails
- Resets `event_sz` to 0 after freeing

This is good defensive programming.

### 2. Comment style

The multi-line comment added follows DPDK style correctly:
```c
/*
 * out of memory, free the old buffer, no more events now
 */
```

### 3. .mailmap entry

The .mailmap addition is appropriate for a new contributor.

---

The patch is well-formed, fixes a legitimate resource leak, and follows DPDK coding conventions. The logic is sound: on realloc failure, the old buffer is freed (preventing the leak) and `event_sz` is reset to 0, indicating no event buffer is available. Subsequent code paths appear to handle this condition safely (the function returns and processing stops).

Review-Result: CLEAN