Commit d6beb085 authored by David S. Miller's avatar David S. Miller

Merge branch 'net-phy-rust'

FUJITA Tomonori says:

====================
Rust abstractions for network PHY drivers

No functional change since v10; only comment and commit log updates.

This patchset adds Rust abstractions for phylib. It doesn't fully
cover the C APIs yet but I think that it's already useful. I implement
two PHY drivers (Asix AX88772A PHYs and Realtek Generic FE-GE). Seems
they work well with real hardware.

The first patch introduces Rust bindings for phylib.

The second patch adds a macro to declare a kernel module for PHYs
drivers.

The third adds the Rust ETHERNET PHY LIBRARY entry to MAINTAINERS
file; adds the binding file and me as a maintainer (as Andrew Lunn
suggested) with Trevor Gross as a reviewer.

The last patch introduces the Rust version of Asix PHY driver,
drivers/net/phy/ax88796b.c. The features are equivalent to the C
version. You can choose C (by default) or Rust version on kernel
configuration.

v11:
  - adds Andrew, Alice, and Trevor's Reviewed-by
  - comment update
v10: https://lore.kernel.org/netdev/20231210234924.1453917-1-fujita.tomonori@gmail.com/T/
  - adds Trevor's SoB to the third patch
  - adds Benno's Reviewed-by to the second patch
v9: https://lore.kernel.org/netdev/20231205.124531.842372711631366729.fujita.tomonori@gmail.com/T/
  - adds a workaround to access to a bit field in phy_device
  - fixes a comment typo
v8: https://lore.kernel.org/netdev/20231123050412.1012252-1-fujita.tomonori@gmail.com/
  - updates the safety comments on Device and its related code
  - uses _phy_start_aneg instead of phy_start_aneg
  - drops the patch for enum synchronization
  - moves Sync From Registration to DriverVTable
  - fixes doctest errors
  - minor cleanups
v7: https://lore.kernel.org/netdev/20231026001050.1720612-1-fujita.tomonori@gmail.com/T/
  - renames get_link() to is_link_up()
  - improves the macro format
  - improves the commit log in the third patch
  - improves comments
v6: https://lore.kernel.org/netdev/20231025.090243.1437967503809186729.fujita.tomonori@gmail.com/T/
  - improves comments
  - makes the requirement of phy_drivers_register clear
  - fixes Makefile of the third patch
v5: https://lore.kernel.org/all/20231019.094147.1808345526469629486.fujita.tomonori@gmail.com/T/
  - drops the rustified-enum option, writes match by hand; no *risk* of UB
  - adds Miguel's patch for enum checking
  - moves CONFIG_RUST_PHYLIB_ABSTRACTIONS to drivers/net/phy/Kconfig
  - adds a new entry for this abstractions in MAINTAINERS
  - changes some of Device's methods to take &mut self
  - comment improvment
v4: https://lore.kernel.org/netdev/20231012125349.2702474-1-fujita.tomonori@gmail.com/T/
  - split the core patch
  - making Device::from_raw() private
  - comment improvement with code update
  - commit message improvement
  - avoiding using bindings::phy_driver in public functions
  - using an anonymous constant in module_phy_driver macro
v3: https://lore.kernel.org/netdev/20231011.231607.1747074555988728415.fujita.tomonori@gmail.com/T/
  - changes the base tree to net-next from rust-next
  - makes this feature optional; only enabled with CONFIG_RUST_PHYLIB_BINDINGS=y
  - cosmetic code and comment improvement
  - adds copyright
v2: https://lore.kernel.org/netdev/20231006094911.3305152-2-fujita.tomonori@gmail.com/T/
  - build failure fix
  - function renaming
v1: https://lore.kernel.org/netdev/20231002085302.2274260-3-fujita.tomonori@gmail.com/T/
====================
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parents f3c2caac cbe0e415
......@@ -3072,6 +3072,14 @@ S: Maintained
F: Documentation/devicetree/bindings/net/asix,ax88796c.yaml
F: drivers/net/ethernet/asix/ax88796c_*
ASIX PHY DRIVER [RUST]
M: FUJITA Tomonori <fujita.tomonori@gmail.com>
R: Trevor Gross <tmgross@umich.edu>
L: netdev@vger.kernel.org
L: rust-for-linux@vger.kernel.org
S: Maintained
F: drivers/net/phy/ax88796b_rust.rs
ASPEED CRYPTO DRIVER
M: Neal Liu <neal_liu@aspeedtech.com>
L: linux-aspeed@lists.ozlabs.org (moderated for non-subscribers)
......@@ -7873,6 +7881,14 @@ F: include/uapi/linux/mdio.h
F: include/uapi/linux/mii.h
F: net/core/of_net.c
ETHERNET PHY LIBRARY [RUST]
M: FUJITA Tomonori <fujita.tomonori@gmail.com>
R: Trevor Gross <tmgross@umich.edu>
L: netdev@vger.kernel.org
L: rust-for-linux@vger.kernel.org
S: Maintained
F: rust/kernel/net/phy.rs
EXEC & BINFMT API
R: Eric Biederman <ebiederm@xmission.com>
R: Kees Cook <keescook@chromium.org>
......
......@@ -60,6 +60,14 @@ config FIXED_PHY
Currently tested with mpc866ads and mpc8349e-mitx.
config RUST_PHYLIB_ABSTRACTIONS
bool "Rust PHYLIB abstractions support"
depends on RUST
depends on PHYLIB=y
help
Adds support needed for PHY drivers written in Rust. It provides
a wrapper around the C phylib core.
config SFP
tristate "SFP cage support"
depends on I2C && PHYLINK
......@@ -104,6 +112,14 @@ config AX88796B_PHY
Currently supports the Asix Electronics PHY found in the X-Surf 100
AX88796B package.
config AX88796B_RUST_PHY
bool "Rust reference driver for Asix PHYs"
depends on RUST_PHYLIB_ABSTRACTIONS && AX88796B_PHY
help
Uses the Rust reference driver for Asix PHYs (ax88796b_rust.ko).
The features are equivalent. It supports the Asix Electronics PHY
found in the X-Surf 100 AX88796B package.
config BROADCOM_PHY
tristate "Broadcom 54XX PHYs"
select BCM_NET_PHYLIB
......
......@@ -37,7 +37,11 @@ obj-$(CONFIG_ADIN1100_PHY) += adin1100.o
obj-$(CONFIG_AMD_PHY) += amd.o
obj-$(CONFIG_AQUANTIA_PHY) += aquantia/
obj-$(CONFIG_AT803X_PHY) += at803x.o
obj-$(CONFIG_AX88796B_PHY) += ax88796b.o
ifdef CONFIG_AX88796B_RUST_PHY
obj-$(CONFIG_AX88796B_PHY) += ax88796b_rust.o
else
obj-$(CONFIG_AX88796B_PHY) += ax88796b.o
endif
obj-$(CONFIG_BCM54140_PHY) += bcm54140.o
obj-$(CONFIG_BCM63XX_PHY) += bcm63xx.o
obj-$(CONFIG_BCM7XXX_PHY) += bcm7xxx.o
......
// SPDX-License-Identifier: GPL-2.0
// Copyright (C) 2023 FUJITA Tomonori <fujita.tomonori@gmail.com>
//! Rust Asix PHYs driver
//!
//! C version of this driver: [`drivers/net/phy/ax88796b.c`](./ax88796b.c)
use kernel::{
c_str,
net::phy::{self, DeviceId, Driver},
prelude::*,
uapi,
};
kernel::module_phy_driver! {
drivers: [PhyAX88772A, PhyAX88772C, PhyAX88796B],
device_table: [
DeviceId::new_with_driver::<PhyAX88772A>(),
DeviceId::new_with_driver::<PhyAX88772C>(),
DeviceId::new_with_driver::<PhyAX88796B>()
],
name: "rust_asix_phy",
author: "FUJITA Tomonori <fujita.tomonori@gmail.com>",
description: "Rust Asix PHYs driver",
license: "GPL",
}
const MII_BMCR: u16 = uapi::MII_BMCR as u16;
const BMCR_SPEED100: u16 = uapi::BMCR_SPEED100 as u16;
const BMCR_FULLDPLX: u16 = uapi::BMCR_FULLDPLX as u16;
// Performs a software PHY reset using the standard
// BMCR_RESET bit and poll for the reset bit to be cleared.
// Toggle BMCR_RESET bit off to accommodate broken AX8796B PHY implementation
// such as used on the Individual Computers' X-Surf 100 Zorro card.
fn asix_soft_reset(dev: &mut phy::Device) -> Result {
dev.write(uapi::MII_BMCR as u16, 0)?;
dev.genphy_soft_reset()
}
struct PhyAX88772A;
#[vtable]
impl Driver for PhyAX88772A {
const FLAGS: u32 = phy::flags::IS_INTERNAL;
const NAME: &'static CStr = c_str!("Asix Electronics AX88772A");
const PHY_DEVICE_ID: DeviceId = DeviceId::new_with_exact_mask(0x003b1861);
// AX88772A is not working properly with some old switches (NETGEAR EN 108TP):
// after autoneg is done and the link status is reported as active, the MII_LPA
// register is 0. This issue is not reproducible on AX88772C.
fn read_status(dev: &mut phy::Device) -> Result<u16> {
dev.genphy_update_link()?;
if !dev.is_link_up() {
return Ok(0);
}
// If MII_LPA is 0, phy_resolve_aneg_linkmode() will fail to resolve
// linkmode so use MII_BMCR as default values.
let ret = dev.read(MII_BMCR)?;
if ret & BMCR_SPEED100 != 0 {
dev.set_speed(uapi::SPEED_100);
} else {
dev.set_speed(uapi::SPEED_10);
}
let duplex = if ret & BMCR_FULLDPLX != 0 {
phy::DuplexMode::Full
} else {
phy::DuplexMode::Half
};
dev.set_duplex(duplex);
dev.genphy_read_lpa()?;
if dev.is_autoneg_enabled() && dev.is_autoneg_completed() {
dev.resolve_aneg_linkmode();
}
Ok(0)
}
fn suspend(dev: &mut phy::Device) -> Result {
dev.genphy_suspend()
}
fn resume(dev: &mut phy::Device) -> Result {
dev.genphy_resume()
}
fn soft_reset(dev: &mut phy::Device) -> Result {
asix_soft_reset(dev)
}
fn link_change_notify(dev: &mut phy::Device) {
// Reset PHY, otherwise MII_LPA will provide outdated information.
// This issue is reproducible only with some link partner PHYs.
if dev.state() == phy::DeviceState::NoLink {
let _ = dev.init_hw();
let _ = dev.start_aneg();
}
}
}
struct PhyAX88772C;
#[vtable]
impl Driver for PhyAX88772C {
const FLAGS: u32 = phy::flags::IS_INTERNAL;
const NAME: &'static CStr = c_str!("Asix Electronics AX88772C");
const PHY_DEVICE_ID: DeviceId = DeviceId::new_with_exact_mask(0x003b1881);
fn suspend(dev: &mut phy::Device) -> Result {
dev.genphy_suspend()
}
fn resume(dev: &mut phy::Device) -> Result {
dev.genphy_resume()
}
fn soft_reset(dev: &mut phy::Device) -> Result {
asix_soft_reset(dev)
}
}
struct PhyAX88796B;
#[vtable]
impl Driver for PhyAX88796B {
const NAME: &'static CStr = c_str!("Asix Electronics AX88796B");
const PHY_DEVICE_ID: DeviceId = DeviceId::new_with_model_mask(0x003b1841);
fn soft_reset(dev: &mut phy::Device) -> Result {
asix_soft_reset(dev)
}
}
......@@ -8,6 +8,9 @@
#include <kunit/test.h>
#include <linux/errname.h>
#include <linux/ethtool.h>
#include <linux/mdio.h>
#include <linux/phy.h>
#include <linux/slab.h>
#include <linux/refcount.h>
#include <linux/wait.h>
......
......@@ -14,6 +14,7 @@
#![no_std]
#![feature(allocator_api)]
#![feature(coerce_unsized)]
#![feature(const_maybe_uninit_zeroed)]
#![feature(dispatch_from_dyn)]
#![feature(new_uninit)]
#![feature(offset_of)]
......@@ -38,6 +39,8 @@
pub mod ioctl;
#[cfg(CONFIG_KUNIT)]
pub mod kunit;
#[cfg(CONFIG_NET)]
pub mod net;
pub mod prelude;
pub mod print;
mod static_assert;
......
// SPDX-License-Identifier: GPL-2.0
//! Networking.
#[cfg(CONFIG_RUST_PHYLIB_ABSTRACTIONS)]
pub mod phy;
This diff is collapsed.
......@@ -7,3 +7,5 @@
*/
#include <uapi/asm-generic/ioctl.h>
#include <uapi/linux/mii.h>
#include <uapi/linux/ethtool.h>
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment