Claude review: drm/imagination: Fix missing pvr_power_fw_{en, dis}able() argument

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

Patch Review

**Correctness: Good.** The fix adds the missing arguments with the correct values:

1. **`pvr_power_fw_disable(pvr_dev, false, true)`** — The third argument `rpm_suspend=true` is correct for the device suspend path. When `rpm_suspend` is true, the function calls `disable_irq(pvr_dev->irq)` (line 111 in pvr_power.c), which is the right thing to do during system suspend. This is consistent with the non-suspend caller at line 548 which passes `rpm_suspend=false` (that's the reset path, where IRQ management is handled differently).

2. **`pvr_power_fw_enable(pvr_dev, true)`** — The argument `rpm_resume=true` is correct for the device resume path. When `rpm_resume` is true, the function calls `enable_irq(pvr_dev->irq)` (line 127), re-enabling the IRQ that was disabled during suspend. Again consistent with the reset caller at line 571 which passes `false`.

**Commit message: Good.** Includes the full build error output, clearly identifies the offending commit, and has a proper `Fixes:` tag.

**Minor note:** The drm-next tree I checked already has the correct arguments at these call sites but still uses `fw_dev.booted` rather than `fw_dev.initialised`, suggesting the broken commit 42577ba79fbf hasn't landed there yet (or this fix was already folded in via a different route). The patch's `base-commit` and `branch: drm-misc-next` confirm it targets the tree where the breakage exists.

No issues found. This is a clean, obvious build fix.

---
Generated by Claude Code Patch Reviewer