From mboxrd@z Thu Jan  1 00:00:00 1970
From: Claude Code Review Bot <claude-review@example.com>
To: dri-devel-reviews@example.com
Subject: Claude review: driver core: platform: count references to all kinds of firmware nodes
Date: Mon, 25 May 2026 20:33:08 +1000
Message-ID: <review-patch23-20260521-pdev-fwnode-ref-v1-23-88c324a1b8d2@oss.qualcomm.com>
In-Reply-To: <20260521-pdev-fwnode-ref-v1-23-88c324a1b8d2@oss.qualcomm.com>
References: <20260521-pdev-fwnode-ref-v1-0-88c324a1b8d2@oss.qualcomm.com> <20260521-pdev-fwnode-ref-v1-23-88c324a1b8d2@oss.qualcomm.com>
X-Mailer: Claude Code Patch Reviewer
Content-Type: text/plain; charset="utf-8"
Content-Transfer-Encoding: quoted-printable
MIME-Version: 1.0

Patch Review

This is the payoff patch. Three key changes:

**Release function**:
```c
-	of_node_put(pa->pdev.dev.of_node);
+	fwnode_handle_put(pa->pdev.dev.fwnode);
```
Correct. For OF nodes, `fwnode_handle_put()` calls through to `of_node_put(=
)` via the fwnode ops. For non-OF fwnodes, this now properly drops the refe=
rence that patch 23's updated helpers take.

**`platform_device_set_fwnode()` becomes the canonical implementation**:
```c
 void platform_device_set_fwnode(struct platform_device *pdev,
 				struct fwnode_handle *fwnode)
 {
-	if (is_of_node(fwnode))
-		platform_device_set_of_node(pdev, to_of_node(fwnode));
-	else
-		pdev->dev.fwnode =3D fwnode;
+	fwnode_handle_put(pdev->dev.fwnode);
+	pdev->dev.fwnode =3D fwnode_handle_get(fwnode);
+	pdev->dev.of_node =3D to_of_node(fwnode);
 }
```
Clean. Unified get/put for all fwnode types. `to_of_node()` returns NULL fo=
r non-OF fwnodes, correctly clearing `of_node`.

**`platform_device_set_of_node()` becomes a thin wrapper**:
```c
 void platform_device_set_of_node(struct platform_device *pdev,
 				 struct device_node *np)
 {
-	of_node_put(pdev->dev.of_node);
-	pdev->dev.of_node =3D of_node_get(np);
-	pdev->dev.fwnode =3D of_fwnode_handle(np);
+	platform_device_set_fwnode(pdev, of_fwnode_handle(np));
 }
```
Correct inversion of the delegation =E2=80=94 `set_of_node` now calls `set_=
fwnode` instead of the reverse.

**`platform_device_register_full()`**:
```c
-	pdev->dev.fwnode =3D pdevinfo->fwnode;
-	pdev->dev.of_node =3D of_node_get(to_of_node(pdev->dev.fwnode));
+	pdev->dev.fwnode =3D fwnode_handle_get(pdevinfo->fwnode);
+	pdev->dev.of_node =3D to_of_node(pdev->dev.fwnode);
```
Correct. Takes a generic fwnode ref instead of only an OF ref. The `dev_ass=
ign_of_node_reused` call on the next line is not in the current tree =E2=80=
=94 this needs a dependency note.

**Overall**: The refcounting is balanced. Every `fwnode_handle_get()` in th=
e set functions is matched by the `fwnode_handle_put()` in `platform_device=
_release()`. The `platform_device_set_of_node_from_dev()` path works becaus=
e for OF nodes, `of_node_get()` and `fwnode_handle_put()` operate on the sa=
me underlying refcount.

---
Generated by Claude Code Patch Reviewer