|WARNING| net/vhost: fix SW checksum when zero

[dpdklab at iol.unh.edu]

Test-Label: ai-code-review
Test-Status: WARNING
https://dpdk.org/patch/171583

_AI Code Review_

=== Patch Review: series.patch (via Claude) ===
Review date: 2026-10-17

# DPDK Patch Review

## ERRORS

None identified.

## WARNINGS

**1. Explicit comparison order inconsistency**

**Line 349 (and 392):**
```c
if (unlikely(csum == 0) && (mbuf->packet_type & RTE_PTYPE_L4_MASK) == RTE_PTYPE_L4_UDP)
```

While the patch correctly fixes the masking bug, the comparison ordering violates DPDK style.
The explicit `csum == 0` check is placed inside `unlikely()`, while the packet type comparison is outside.

**Why it matters:** DPDK style requires consistent placement of `likely()`/`unlikely()` wrappers relative to the comparison. The standard pattern places the wrapper around the entire condition or around a truthiness test, not asymmetrically around one comparison in a compound condition.

**Suggested fix:** Move `unlikely()` to wrap the entire compound condition:
```c
if (unlikely(csum == 0 && (mbuf->packet_type & RTE_PTYPE_L4_MASK) == RTE_PTYPE_L4_UDP))
    csum = 0xffff;
```

Or alternatively, make both comparisons explicit and place `unlikely()` consistently:
```c
if (unlikely((csum == 0) && ((mbuf->packet_type & RTE_PTYPE_L4_MASK) == RTE_PTYPE_L4_UDP)))
    csum = 0xffff;
```

The first form is cleaner and matches typical DPDK usage where the branch prediction hint applies to the overall condition rather than individual sub-expressions.

**Applies to:** Both occurrences at lines 349 and 392.

Review-Result: WARNING