Claude review: Revert "drm/imagination: Warn or error on unsupported hardware"

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

Patch Review

**Correctness:** The revert is clean. It removes:
1. The `pvr_exp_hw_support` module parameter and its `module_param_named`/`MODULE_PARM_DESC` declarations
2. The `enum pvr_gpu_support_level` type and its three values
3. The `pvr_gpu_support_level()` function with the BVNC-to-support-level mapping
4. The `pvr_check_gpu_supported()` function with the support-level gating logic
5. Replaces `return pvr_check_gpu_supported(pvr_dev, gpu_id);` with `return 0;`

All removed code is dead after the change — no other callers or references remain.

**Observations:**

1. **No `Fixes:` tag:** The patch doesn't include a `Fixes:` tag pointing to the original commit `1c21f240fbc1e47b`. While this is a revert (not strictly a "fix"), if the intent is for the revert to be backported to stable kernels (since the original commit broke Renesas SoC support), a `Fixes:` tag would be appropriate to signal that. Without it, stable maintainers won't pick it up automatically.

2. **Well-documented rationale:** The below-`---` section is thorough, with 9 references justifying the revert. This is excellent for reviewer context but won't appear in the final git log (which is the right place for it, since it's background justification rather than core commit message content).

3. **Alternative approach not discussed in commit message body:** The commit message body (above the `---`) is very terse — just "Revert commit X, as it stopped the driver from working on various Renesas R-Car SoCs." It doesn't explain why a revert was chosen over adding the BVNC values. A brief note on that design choice would strengthen the commit message for future readers.

4. **Module parameter removal is an ABI change:** Removing the `exp_hw_support` module parameter is technically a user-visible ABI change. Any existing users who set this parameter in their module configuration (e.g., in `/etc/modprobe.d/`) would see a warning about an unknown parameter on the next boot after this patch lands. This is minor but worth noting.

5. **Diff context matches tree:** The removed code at lines 511–580 of the current tree matches the patch's diff context exactly, confirming this is a clean revert against the expected base.

**Verdict:** The patch is technically sound. The main question is whether the drm/imagination maintainers prefer a full revert or would rather have the Renesas BVNCs added to the allowlist. Both approaches are valid; the full revert is simpler and avoids future churn as new SoCs are enabled, but loses the safety net for truly unknown hardware.

---
Generated by Claude Code Patch Reviewer