Claude review: drm/radeon: fix internal display on iMac11, 1 (RV770/DCE3.1)

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

Patch Review

**Correctness: Looks correct.** The new macro is well-placed and the logic follows the existing DCE versioning scheme.

**Issue 1 — Missing `Fixes:` tag:** The commit message references "Fixes freedesktop issue 164" in prose, but does not include a kernel-style `Fixes:` tag pointing to the commit that introduced the regression (or the original DCE3.2 logic that omitted RV770). If this is a fix for existing behavior, it should have:
```
Fixes: <sha1> ("drm/radeon: ...")
Cc: stable@vger.kernel.org
```
If there's no single commit to blame (the logic was always wrong), a `Cc: stable` is still appropriate since this is a hardware enablement fix for a longstanding bug.

**Issue 2 — PLL flag changes affect all RV770 systems, not just iMac:**

```c
-		if (ASIC_IS_DCE32(rdev) && mode->clock > 200000)
+		if (ASIC_IS_DCE31(rdev) && mode->clock > 200000)
```

This change means *every* RV770-based system (not just the iMac11,1) will now get `RADEON_PLL_PREFER_HIGH_FB_DIV` for high clocks and `RADEON_PLL_USE_FRAC_FB_DIV` for clocks > 165000. While this is likely the correct behavior for DCE3.1 hardware in general, the commit message frames this purely as an iMac fix. The author should confirm this is intentional for all RV770 cards, or if it should be gated behind a DMI check. Given that line 588 already has `if (rdev->family < CHIP_RV770)` as a boundary, this is likely fine — but it should be stated explicitly in the commit message.

**Issue 3 — Audit of remaining `ASIC_IS_DCE32` call sites:** There are additional `ASIC_IS_DCE32()` users in `r600.c` (interrupt handling, HDMI, HPD) and `radeon_audio.c`. The author should confirm these do not also need the DCE3.1 treatment. If they are truly DCE3.2-specific features, the commit message should note this.

**Minor — Subject line wrapping:** The subject line appears wrapped across multiple header lines in the mbox (`Subject: [PATCH v2] drm/radeon: fix internal display on iMac11,\n 1 (RV770/DCE3.1)`). This may be an email client artifact, but the author should verify `git log --oneline` shows it correctly.

---

---
Generated by Claude Code Patch Reviewer