netdev - Re: [PATCH] net: dsa: realtek: rtl8365mb: fix return value in rtl8365mb_phy_ocp

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <1c0ee3d4-2b24-48ea-b34f-b5653abe11d4@lunn.ch>
Date: Sun, 16 Nov 2025 20:32:30 +0100
From: Andrew Lunn <andrew@...n.ch>
To: Mietek N <namiltd@...oo.com>
Cc: "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
	"linus.walleij@...aro.org" <linus.walleij@...aro.org>,
	"edumazet@...gle.com" <edumazet@...gle.com>,
	"alsi@...g-olufsen.dk" <alsi@...g-olufsen.dk>,
	"olteanv@...il.com" <olteanv@...il.com>,
	"kuba@...nel.org" <kuba@...nel.org>,
	"pabeni@...hat.com" <pabeni@...hat.com>,
	"davem@...emloft.net" <davem@...emloft.net>
Subject: Re: [PATCH] net: dsa: realtek: rtl8365mb: fix return value in
 rtl8365mb_phy_ocp_write

On Sun, Nov 16, 2025 at 05:18:31PM +0000, Mietek N wrote:
> Function rtl8365mb_phy_ocp_write() always returns 0, even when an erroroccurs during register access. This patch fixes the return value to
> propagate the actual error code from regmap operations.
> 
> Fixes: 964a56e ("net: dsa: realtek: add RTL8365MB switch driver")
> Signed-off-by: Mieczyslaw Nalewaj <namiltd@...oo.com>

Please take a read of

https://www.kernel.org/doc/html/latest/process/maintainer-netdev.html
https://docs.kernel.org/process/submitting-patches.html

Netdev requests that you only submit one version of a patch a day. If
you notice something wrong in your patch, reply to it, and say what is
wrong. And then post a new version the next day.

This is also v3 of the patch. Please put the version number in the
[patch vX] part.

Your hash in the Fixes: is also too short. The kernel requests 12
character hashes. There is a git configuration for this.

Also, 964a56e does not seem correct.

The history is a bit complex. It appears it was first added in

commit 4af2950c50c8634ed2865cf81e607034f78b84aa
Author: Alvin Šipraga <alsi@...g-olufsen.dk>
Date:   Mon Oct 18 11:38:01 2021 +0200

    net: dsa: realtek-smi: add rtl8365mb subdriver for RTL8365MB-VC

which added the file drivers/net/dsa/rtl8365mb.c. That version is
correct, it ends with:

+       ret = rtl8365mb_phy_poll_busy(smi);
+       if (ret)
+               return ret;
+
+       return 0;
+}

We have:

commit 319a70a5fea9590e9431dd57f56191996c4787f4
Author: Luiz Angelo Daros de Luca <luizluca@...il.com>
Date:   Fri Jan 28 03:04:58 2022 -0300

    net: dsa: realtek-smi: move to subdirectory

and if you look at that version, the return value is still O.K.

What actually seems to break it is:

commit 2796728460b822d549841e0341752b263dc265c4
Author: Alvin Šipraga <alsi@...g-olufsen.dk>
Date:   Mon Feb 21 19:46:31 2022 +0100

    net: dsa: realtek: rtl8365mb: serialize indirect PHY register access

        ret = rtl8365mb_phy_poll_busy(priv);
        if (ret)
-               return ret;
+               goto out;
+
+out:
+       mutex_unlock(&priv->map_lock);
 
        return 0;
 }

So instead of directly returning the error code, it now does an
unlock, and return 0. So this is what your Fixes tag needs to be.

    Andrew

---
pw-bot: cr