drivers: ethernet: Add Intel® i226 Ethernet Controller Driver #88108

Vijayakannan · 2025-04-03T12:24:06Z

The Intel i226 Ethernet Controller is a PCIe Gen 2 one-lane modular endpoint device that integrates a fully featured GbE Media Access Control (MAC) and Physical Layer (PHY) port. This PR introduces a complete driver bundle to enable Intel i226 Ethernet Controller.

Components:

MAC Driver, MDIO Driver & Parent Platform Driver

Features Supported:

Full duplex support for 10/100/1000 Mbps.
Half duplex support for 10/100 Mbps.
MSI-X interrupt support for TX/RX DMA channels.
Exclusive bottom-half processing for each hardware queue.
Auto-Negotiation for 10/100/1000 Mbps.
MTU customization for flexible packet sizes.
MAC address filtering support for:
- Random MAC Address Generated.
- MAC addresses specified in the device tree.
- Pre-programmed MAC addresses in EEPROM.
MDIO Clause 22 (C22) and Clause 45 (C45) support.
Multi-instance support to scale the system. (Tested Up to 2 instances with basic coverage)

Other Notes:

The Parent Platform Driver is designed to be modular and reusable for other PCIe-based Ethernet devices.
2.5 Gbps, Time Sync and other TSN features were not enabled despite the hardware having support.

Validation Coverage:
This driver bundle has been tested on Intel Alder Lake platforms with the Intel i226 LM variant Ethernet Controller(0x125B). The ZPERF network sample application was used to validate basic functionality, performance, and stability.

Test List:

Transmit and receive UDP packet,
Transmit and receive TCP packet,
Speed: 10/100/1000Mbps, Duplex: Full, Auto-negotiation: ON, (autoneg initiated from link partner),
Speed: 10/100; Duplex: Half, Auto-negotiation: ON, (autoneg initiated from link partner),
Speed: 10/100/1000Mbps, Duplex: Full, Auto-negotiation: ON (autoneg initiated from net shell),
Speed: 10/100Mbps, Duplex: Half, Auto-negotiation: ON (autoneg initiated from net shell),
TX and RX with different MTU sizes: 256, 512, and 1024
Throughput KPI for UDP upload and download (100/1000 Mbps),
Latency KPI: 100 Mbps, 500 us packet interval with 500,1500 bytes packet size,
Latency KPI: 1000 Mbps, 125 us packet interval, 200 bytes packet size,
Ethernet Cable hot-plug with different number of hardware queues.
MDIO Clause - 22/ 45 Support,
Configure MAC address via net shell, device tree and net shell.

github-actions · 2025-04-03T12:24:48Z

Hello @Vijayakannan, and thank you very much for your first pull request to the Zephyr project!
Our Continuous Integration pipeline will execute a series of checks on your Pull Request commit messages and code, and you are expected to address any failures by updating the PR. Please take a look at our commit message guidelines to find out how to format your commit messages, and at our contribution workflow to understand how to update your Pull Request. If you haven't already, please make sure to review the project's Contributor Expectations and update (by amending and force-pushing the commits) your pull request if necessary.
If you are stuck or need help please join us on Discord and ask your question there. Additionally, you can escalate the review when applicable. 😊

maass-hamburg · 2025-04-03T16:45:25Z

On a first look:

the Link status change detection should happen in the phy. we have the phy_link_callback_set() to set a callback function in the phy from the mac. In that callback net_eth_carrier_on() and net_eth_carrier_off() should be called.
It's generall best to not have code for the phy in the ethernet driver. It's preferd if the official zephyr apis are used for communication between the phy and mac driver.

Vijayakannan · 2025-04-04T06:19:36Z

On a first look:

Hi Fin Maaß, Thanks for spending time to review and recommendations.

the Link status change detection should happen in the phy. we have the phy_link_callback_set() to set a callback function in the phy from the mac. In that callback net_eth_carrier_on() and net_eth_carrier_off() should be called.

It's generall best to not have code for the phy in the ethernet driver. It's preferd if the official zephyr apis are used for communication between the phy and mac driver.

I agree with your suggestions and work on it.

Rationale on why earlier decided to use MAC for Link detection:
Intel i226 integrates the GbE MAC and PHY into a single chip. Additionally, MAC includes automatic Link detection functionality and update its registers automatically, which simplifies usage. So, in my initial design opted to handle it directly in MAC to avoid the overhead of invoking PHY-to-MAC callback.

I am open for your suggestion in case if you have other thoughts. - Thanks, Vijay

drivers/ethernet/phy/phy_intel_igc.c

drivers/mdio/mdio_intel_igc.c

Vijayakannan · 2025-04-23T06:10:30Z

On a first look:

Hi Fin Maaß, Thanks for spending time to review and recommendations.

the Link status change detection should happen in the phy. we have the phy_link_callback_set() to set a callback function in the phy from the mac. In that callback net_eth_carrier_on() and net_eth_carrier_off() should be called.

It's generall best to not have code for the phy in the ethernet driver. It's preferd if the official zephyr apis are used for communication between the phy and mac driver.

I agree with your suggestions and work on it.

Progress Update:
The interrupt line used for this PHY link status change is shared for multiple purposes as per the hardware design. Therefore, I have started a small refactor to accommodate this review comment and all other use cases. Let me publish the changes once I finish my sanity checks.

drivers/ethernet/intel/eth_intel_igc.c

Vijayakannan · 2025-07-10T14:18:19Z

Hi @maass-hamburg,

All the review comments were addressed in this PR version 7. Please review and share feedback. Thanks, Vijay

Changes since v6:

- Use generic phy driver for i226. (Test Coverage: Got Satisfactory performance results)
	○ Replace the compatible string in device tree to use "ethernet-phy" generic driver.
	○ Remove drivers/ethernet/phy/phy_intel_igc.c
	○ Remove dts/bindings/ethernet/phy/intel,igc-phy.yaml
	○ Clean up drivers/ethernet/phy/CMakeLists.txt.
	○ Clean up drivers/ethernet/phy/Kconfig.

- Remove Miscellaneous interrupts handling codes from the mac driver.
	○ eth_intel_igc_misc_isr
	○ eth_intel_igc_connect_misc_msix_vector
	○ Clean eth_intel_igc_pcie_msix_setup
	○ Clean eth_intel_igc_map_intr_to_vector
	○ eth_intel_igc_intr_enable

- Remove INTEL_IGC_MAC_ASSIGN_ADDR unused macro.
- Move interrupt-parent to parent platform driver.
- Clean up the PR description.
- Fix 3 sonarqubecloud reported issues.

Vijayakannan · 2025-07-15T04:10:00Z

Hi @maass-hamburg, please review and approve this PR.?

drivers/ethernet/intel/eth_intel_igc.c

Vijayakannan · 2025-07-16T16:03:50Z

Hi @maass-hamburg, please review PR revision v8.

Changes since PR version v7:
• Cleanup eth_intel_igc_get_phy
• Use sys_read32 and sys_write32.
• Move the intr vector configuration to normal init function.
• Cleanup eth_intel_igc_disable_pcie_master function.
• Remove unused 20millisecond sleep from eth_intel_igc_init_mac_hw.
• Add 1 millisecond delay after MAC reset sequence to settle.
• Relocate eth_intel_igc_map_intr_to_vector and eth_intel_igc_init_speed.
• Refine the inline comments and LOG prints.

drivers/ethernet/intel/eth_intel_igc.c

Vijayakannan · 2025-07-17T15:35:19Z

Hi @maass-hamburg, please review PR revision v9.

Changes since PR version v8:

Remove 0 left and right shift usages in code.
Cleanup eth_intel_igc_get_sa_family
Remove Long packet receive support. (test the speed with this support)
Cleanup INTEL_IGC_SETUP_QUEUE_WORK_EXP macro
Cleanup INTEL_IGC_MAC_DATA
Use phy-handle to fetch ethernet-phy node in dts.

Vijayakannan · 2025-07-21T04:59:14Z

Hi @maass-hamburg, could you please share your feedback on this updated PR version?

maass-hamburg · 2025-07-21T07:43:09Z

drivers/ethernet/intel/eth_intel_igc_priv.h

+	struct eth_intel_igc_mac_tx *tx;
+	struct eth_intel_igc_mac_rx *rx;


Suggested change

struct eth_intel_igc_mac_tx *tx;

struct eth_intel_igc_mac_rx *rx;

struct eth_intel_igc_mac_tx tx;

struct eth_intel_igc_mac_rx rx;

I meant this, when I said to put it in it.

I meant this, when I said to put it in it.

Got it. This change requires updating the overall code. I can make the necessary modifications and update this PR.

Please let me know if there are any other review comments to address in the next version. I need to expedite this PR due to the long pending pull request.

I meant this, when I said to put it in it.

Good catch and suggestion. All the required changes are already resolved in the updated version.

The Ethernet device model consists of multiple subsystem components, such as MDIO, PHY, MAC and PTP_CLOCK. These components are mapped into a single PCIe BAR location with same base address. This platform driver retrieves the MMIO mapping details and provides a framework to share it with all the child subsystem components. This approach avoid the duplicate remapping, ensures efficient re-use of MMIO mappings across related devices. Example device tree structure for first ethernet instance: parent0: parent0 { compatible = "intel,eth-plat"; interrupt-parent = <&intc>; vendor-id = <0x8086>; device-id = <0xXXXX>; igc0: igc0 { compatible = "intel,igc-mac"; /* * MAC specific properties. */ status = "okay"; }; mdio0: mdio0 { compatible = "intel,igc-mdio"; #address-cells = <1>; #size-cells = <0>; ethphy0: ethernet-phy@0 { compatible = "ethernet-phy"; /* * PHY specific properties. */ reg = <0x0>; }; }; }; This framework is modular and re-usable for other PCIe based Ethernet devices. It can also be extended to support additional platform specific information shared across child nodes. Signed-off-by: Vijayakannan Ayyathurai <vijayakannan.ayyathurai@intel.com>

Intel i226 MAC supports MDIO C22 and MDIO C45. Standard PHY registers are accessible through MDIO C22, whereas PMAPMD and PCS are accssible through MDIO C45. Signed-off-by: Vijayakannan Ayyathurai <vijayakannan.ayyathurai@intel.com>

The Intel i226 Ethernet Controller is a PCIe Gen 2 one-lane modular endpoint device that integrates a GbE Media Access Control (MAC) and Physical Layer (PHY) port. This driver provides support for MAC and DMA-specific initialization and runtime TX/RX operations. Key features: - MSI-X interrupts for TX/RX DMA channels. - Multiple TX/RX DMA channel support with exclusive bottom-half. - Implements a circular descriptor ring architechture with producer-consumer semantics for high performance pkt processing. - Full duplex support for 10/100/1000 Mbps. - Half duplex support for 10/100 Mbps. - Auto-negotiation for 10/100/1000 Mbps. - MTU customization for flexible packet sizes. - MAC address filtering based on: - Random MAC generation. - Local-mac-address mentioned in device tree. - EEPROM pre-programmed mac address. - Setting mac address via net shell. - Support for multiple Ethernet interface instances. Signed-off-by: Vijayakannan Ayyathurai <vijayakannan.ayyathurai@intel.com> Signed-off-by: Ling Pei Lee <pei.lee.ling@intel.com>

Foxville LM (0x125B) i226 variant and Intel Alder Lake platform was used for developing and stabilizing the i226 Ethernet device driver. However, users can reuse the provided device tree models as a reference when enabling the support for other i226 variants and platforms. This device-tree model include essential configurations for the i226 Ethernet controller, such as PCIe settings, interrupt mappings, Phy MDIO, and DMA descriptor configurations. Signed-off-by: Vijayakannan Ayyathurai <vijayakannan.ayyathurai@intel.com>

The Zperf sample application was chosen to demonstrate basic network functionality and high-performance use cases. This application serves as a reference for users who wish to enable other network sample applications. In addition to the essential configurations for the Intel i226 Ethernet controller, stack-specific configurations were added to ensure stability under heavy network loads. These configurations include adjustments to buffer sizes, interrupt handling, and DMA descriptor management. Signed-off-by: Vijayakannan Ayyathurai <vijayakannan.ayyathurai@intel.com>

sonarqubecloud · 2025-07-22T09:38:24Z

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

Vijayakannan · 2025-07-22T10:24:25Z

Hi @jukkar, @tbursztyka, @dcpleung, @pdgendt, @decsny, @nashif

Please add this PR in your review list and share your feedback. Thanks in advance.

Vijayakannan force-pushed the dev/vijay/i226_mainline branch from f78f26a to 0e25402 Compare April 3, 2025 14:15

Vijayakannan marked this pull request as ready for review April 3, 2025 15:55

github-actions bot added area: MDIO area: Ethernet area: Networking area: Samples Samples platform: X86 x86 and x86-64 labels Apr 3, 2025

github-actions bot requested review from ceolin, dcpleung, decsny, edersondisouza, jukkar, kartben, lmajewski, maass-hamburg, najumon1980, nashif, pdgendt, ssharks, tbursztyka and teburd April 3, 2025 15:56

github-actions bot assigned maass-hamburg Apr 3, 2025

maass-hamburg reviewed Apr 7, 2025

View reviewed changes

drivers/ethernet/phy/phy_intel_igc.c Outdated Show resolved Hide resolved

drivers/mdio/mdio_intel_igc.c Outdated Show resolved Hide resolved

Vijayakannan marked this pull request as draft April 23, 2025 05:58

Vijayakannan force-pushed the dev/vijay/i226_mainline branch 2 times, most recently from 288ebf7 to 03d2fc2 Compare May 13, 2025 20:26

maass-hamburg reviewed Jul 10, 2025

View reviewed changes

drivers/ethernet/intel/eth_intel_igc.c Outdated Show resolved Hide resolved

Vijayakannan force-pushed the dev/vijay/i226_mainline branch from 0f1a01e to 7210410 Compare July 10, 2025 14:14

github-actions bot added the platform: Intel Intel Corporation label Jul 10, 2025

Vijayakannan force-pushed the dev/vijay/i226_mainline branch from 7210410 to 3bd0d2e Compare July 11, 2025 06:19

Vijayakannan requested a review from maass-hamburg July 14, 2025 10:24

maass-hamburg requested changes Jul 15, 2025

View reviewed changes

Vijayakannan force-pushed the dev/vijay/i226_mainline branch from 3bd0d2e to 75fd2a6 Compare July 16, 2025 08:16

Vijayakannan requested a review from maass-hamburg July 16, 2025 08:17

Vijayakannan force-pushed the dev/vijay/i226_mainline branch from 75fd2a6 to 771b9cc Compare July 16, 2025 15:52

maass-hamburg requested changes Jul 16, 2025

View reviewed changes

Vijayakannan force-pushed the dev/vijay/i226_mainline branch from 771b9cc to 5f500c7 Compare July 17, 2025 15:32

Vijayakannan requested a review from maass-hamburg July 17, 2025 15:35

Vijayakannan force-pushed the dev/vijay/i226_mainline branch from 5f500c7 to e3e912d Compare July 18, 2025 11:38

maass-hamburg requested changes Jul 21, 2025

View reviewed changes

Vijayakannan force-pushed the dev/vijay/i226_mainline branch from e3e912d to 45305c5 Compare July 21, 2025 10:47

Vijayakannan requested a review from maass-hamburg July 21, 2025 10:48

maass-hamburg approved these changes Jul 22, 2025

View reviewed changes

Vijayakannan added 5 commits July 22, 2025 16:55

drivers: mdio: Add Intel i226 MDIO driver support

8b6fdef

Intel i226 MAC supports MDIO C22 and MDIO C45. Standard PHY registers are accessible through MDIO C22, whereas PMAPMD and PCS are accssible through MDIO C45. Signed-off-by: Vijayakannan Ayyathurai <vijayakannan.ayyathurai@intel.com>

Vijayakannan force-pushed the dev/vijay/i226_mainline branch from 45305c5 to 884a682 Compare July 22, 2025 09:13

		struct eth_intel_igc_mac_tx *tx;
		struct eth_intel_igc_mac_rx *rx;

drivers: ethernet: Add Intel® i226 Ethernet Controller Driver #88108

Are you sure you want to change the base?

drivers: ethernet: Add Intel® i226 Ethernet Controller Driver #88108

Uh oh!

Conversation

Vijayakannan commented Apr 3, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions bot commented Apr 3, 2025

Uh oh!

maass-hamburg commented Apr 3, 2025

Uh oh!

Vijayakannan commented Apr 4, 2025

Uh oh!

Uh oh!

Uh oh!

Vijayakannan commented Apr 23, 2025

Uh oh!

Uh oh!

Vijayakannan commented Jul 10, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Vijayakannan commented Jul 15, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Vijayakannan commented Jul 16, 2025

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Vijayakannan commented Jul 17, 2025

Uh oh!

Vijayakannan commented Jul 21, 2025

Uh oh!

maass-hamburg Jul 21, 2025

Choose a reason for hiding this comment

Uh oh!

Vijayakannan Jul 21, 2025

Choose a reason for hiding this comment

Uh oh!

Vijayakannan Jul 21, 2025

Choose a reason for hiding this comment

Uh oh!

sonarqubecloud bot commented Jul 22, 2025

Quality Gate passed

Uh oh!

Vijayakannan commented Jul 22, 2025

Uh oh!

Uh oh!

Vijayakannan commented Apr 3, 2025 •

edited

Loading

Vijayakannan commented Jul 10, 2025 •

edited

Loading