Claude review: soc: cadence: Create helper functions for Cadence MHDP

[Claude Code Review Bot <claude-review@example.com>]

Patch Review

This patch extracts mailbox access code from the mhdp8546 driver into a new shared helper under `drivers/soc/cadence/`. The design is clean with per-instance `mailbox_mutex`, proper `lockdep_assert_held()` annotations, and address validation on DPCD/register reads.

**Issues:**

1. **Commit message typo**: "helper fucntions" and "MDHP8501" should be "helper functions" and "MHDP8501".

2. **HDCP receiver ID buffer overflow risk** in `cdns_mhdp_secure_mailbox_send_recv_multi()`:
```c
if (module_id == MB_MODULE_ID_HDCP_TX &&
    opcode == HDCP_TRAN_IS_REC_ID_VALID)
    ret = mhdp_mailbox_recv_data(base, resp2, 5 * resp1[0], true);
```
`resp1[0]` is firmware-controlled. If it returns a value larger than the buffer behind `resp2`, this overflows. The caller in mhdp8546-hdcp.c passes `0` for `resp2_size` with `hdcp_rx_id` buffer — but the size of that buffer is `HDCP_MAX_RECEIVERS * 5 = 160` bytes. There's no validation that `resp1[0] <= HDCP_MAX_RECEIVERS` here.

3. **Code duplication**: The non-secure and secure variants of `cdns_mhdp_mailbox_send_recv` and `cdns_mhdp_mailbox_send_recv_multi` are nearly identical except for the `secure` boolean. The internal functions already take a `bool secure` parameter — consider whether the six exported wrappers could be reduced to three with a boolean parameter, or if the current approach is intentional for API clarity.

4. **Placement under `drivers/soc/`**: This is a display/PHY helper library. While it provides mailbox access for an IP block, placing it under `drivers/soc/cadence/` may be debatable. The DRM bridge subsystem maintainers or the linux-soc list should weigh in on whether this belongs under `drivers/soc/` or perhaps `drivers/firmware/` or `drivers/mfd/`.

---
Generated by Claude Code Patch Reviewer