|SUCCESS| bpf: fix index register for classic BPF indirect loads

[dpdklab at iol.unh.edu]

Test-Label: ai-code-review
Test-Status: SUCCESS
https://dpdk.org/patch/172185

_AI Code Review_

=== Patch Review: series.patch (via Claude) ===
Review date: 2026-10-27

# DPDK Patch Review

## Summary

This patch fixes a correctness bug in the classic BPF to DPDK BPF converter where indirect loads (using variable offset X) were incorrectly using register R0 instead of R7 as the index register.

## ERRORS

None identified.

## WARNINGS

None identified.

## INFO

### 1. Code correctness
The fix is correct. The patch moves the indirect load opcodes (`BPF_LD|BPF_IND|BPF_W/H/B`) to their own case that explicitly specifies `BPF_REG_X` (register 7) as the source register. Previously these were falling through to the arithmetic instruction case which incorrectly derived the source register from the `BPF_SRC(code)` bit, yielding `BPF_REG_A` (register 0) for byte and word loads.

### 2. Alignment with switch statement structure
The new case is well-placed immediately after the absolute load case (`BPF_LD | BPF_ABS | ...`) and before the arithmetic cases, logically grouping load operations together.

### 3. Comment quality
The added comment "Indirect loads index the packet with X: A = P[X + k]" clearly explains what these opcodes do, improving code maintainability.

### 4. Minimal change principle
The patch touches only what is necessary to fix the bug. The arithmetic instruction case retains its original logic, and the new case handles only the three indirect load opcodes.

### 5. Testing consideration
While the patch description mentions that "the interpreter and the JITs already take the index from src_reg," it would be beneficial if the commit included information about how this was tested (though this is not a requirement for the code review).

Review-Result: CLEAN