linux-kernel - Re: [PATCH] net: phy: update fields of mii_ioctl

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: <ZiN+rBvYQ6vktNCW@shell.armlinux.org.uk>
Date: Sat, 20 Apr 2024 09:37:00 +0100
From: "Russell King (Oracle)" <linux@...linux.org.uk>
To: renjun wang <renjunw0@...mail.com>
Cc: andrew@...n.ch, hkallweit1@...il.com, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] net: phy: update fields of mii_ioctl_data for
 transferring C45 data.

On Sat, Apr 20, 2024 at 08:41:10AM +0800, renjun wang wrote:
> The phy_id is used as u32 type in function mdio_phy_id_is_c45()
> with the 30th bit for distinguishing C22 and C45. The reg_num is
> also used as u32 type in function mdiobus_c45_read() or someplace
> else. For more C45 information needed and data structure alignment
> consideration, change these two fields to __u32 type which can make
> user space program transferring clause 45 type information to kernel
> directly.

You haven't understood the user API, and are making a change that will
break existing userspace users of this API. Hence this is not
acceptable.

mdio_phy_id_is_c45() is indeed for determining if the "phy_id" field
in mii_ioctl_data is indicating a clause 45 transaction. This checks
the following bits, defined using these macros:

#define MDIO_PHY_ID_C45                 0x8000
#define MDIO_PHY_ID_PRTAD               0x03e0
#define MDIO_PHY_ID_DEVAD               0x001f

and all of these fit within the __u16.

The format used by userspace pre-dates clause 45, and bit 15 was
added to indicate clause 45 along with the "prtad" in bits 9 to
5.

However, the kernel translates this data structure into something
that the MDIO accessors can deal with. See for example phy_mii_ioctl().

Also note that the old method of shoe-horning clause 45 into just
one set of mii bus methods has gone, meaning there is now no need
to convert from mii_ioctl_data into a 32-bit reg_num.

-- 
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 80Mbps down 10Mbps up. Decent connectivity at last!