public inbox for drm-ai-reviews@public-inbox.freedesktop.org
 help / color / mirror / Atom feed
From: Claude Code Review Bot <claude-review@example.com>
To: dri-devel-reviews@example.com
Subject: Claude review: arm64: dts: rockchip: rk356x: Add NPU and its IOMMU
Date: Thu, 04 Jun 2026 16:18:12 +1000	[thread overview]
Message-ID: <review-patch3-20260529155824.3099831-4-midgy971@gmail.com> (raw)
In-Reply-To: <20260529155824.3099831-4-midgy971@gmail.com>

Patch Review

**1. Node placement violates address ordering**

The NPU nodes are inserted between `qos_rga_wr` (0xfe158300) and `qos_npu` (0xfe180000):

```diff
 	qos_rga_wr: qos@fe158300 {
 		reg = <0x0 0xfe158300 0x0 0x20>;
 	};

+	rknn_core_0: npu@fde40000 {
```

But the NPU base address is `0xfde40000`, which is *lower* than `0xfe158300`. Rockchip DTS files sort nodes by unit address. These nodes should be placed earlier in the file, near other `0xfdeXXXXX` nodes.

**2. Shared interrupt between NPU and IOMMU**

```dts
rknn_core_0: npu@fde40000 {
	interrupts = <GIC_SPI 151 IRQ_TYPE_LEVEL_HIGH>;
	...
};

rknn_mmu_0: iommu@fde4b000 {
	interrupts = <GIC_SPI 151 IRQ_TYPE_LEVEL_HIGH>;
	...
};
```

Both the NPU core and its IOMMU use GIC_SPI 151. Verify this is correct per the TRM — on RK3588 the NPU cores and IOMMUs have distinct interrupts. Sharing an interrupt line between a device and its IOMMU can work if the driver handles it, but it's unusual and worth double-checking against the hardware documentation.

**3. `assigned-clock-rates` value**

```dts
assigned-clocks = <&scmi_clk SCMI_CLK_NPU>;
assigned-clock-rates = <200000000>;
```

The initial rate is set to 200 MHz here, but the driver's `rk3568_noc_init()` immediately ramps it to 600 MHz then 1 GHz. Consider whether 200 MHz is the right boot-time default, or whether the `assigned-clock-rates` should match the target operational frequency.

**4. Otherwise clean**

The node structure (reg regions, clocks, resets, power domain, IOMMU reference) mirrors the RK3588 pattern appropriately. The `rockchip,pmu = <&pmu>` phandle is present as required by the driver.

---

---
Generated by Claude Code Patch Reviewer

  reply	other threads:[~2026-06-04  6:18 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-29 15:58 [PATCH v2 0/4] accel: rocket: Add RK3568 NPU support MidG971
2026-05-29 15:58 ` [PATCH v2 1/4] accel: rocket: Add support for Rockchip RK3568 MidG971
2026-05-29 18:19   ` Heiko Stuebner
2026-06-04  6:18     ` Claude review: " Claude Code Review Bot
2026-05-29 15:58 ` [PATCH v2 2/4] dt-bindings: npu: rockchip, rk3588-rknn-core: Add RK3568 support MidG971
2026-05-29 16:18   ` [PATCH v2 2/4] dt-bindings: npu: rockchip,rk3588-rknn-core: " Krzysztof Kozlowski
2026-06-04  6:18     ` Claude review: " Claude Code Review Bot
2026-05-29 15:58 ` [PATCH v2 3/4] arm64: dts: rockchip: rk356x: Add NPU and its IOMMU MidG971
2026-06-04  6:18   ` Claude Code Review Bot [this message]
2026-05-29 15:58 ` [PATCH v2 4/4] arm64: dts: rockchip: rk3568-rock-3b: Enable NPU MidG971
2026-06-04  6:18   ` Claude review: " Claude Code Review Bot
2026-05-29 16:17 ` [PATCH v2 0/4] accel: rocket: Add RK3568 NPU support Krzysztof Kozlowski
2026-06-04  6:18   ` Claude review: " Claude Code Review Bot
2026-05-29 18:04 ` Heiko Stuebner

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=review-patch3-20260529155824.3099831-4-midgy971@gmail.com \
    --to=claude-review@example.com \
    --cc=dri-devel-reviews@example.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox