Fix SC59x Ethernet PHY descriptions and ADIN PHY Driver by ozan956 · Pull Request #130 · analogdevicesinc/u-boot

ozan956 · 2026-07-19T20:46:10Z

This series updates SC594 and SC598 + EZKIT or EZLITE ethernet support for ADIN1200 and ADIN1300 PHYs.

Changes include:

Add ADIN1200 PHY registration to the U-Boot ADIN PHY driver. (tried to match the Linux Driver)
Configure ADIN1200 RMII mode and FIFO depth.
Configure ADIN1300 RGMII RX/TX internal delays to 2000 ps.
Configure ADIN1200 RMII FIFO depth to 8 bits on EZKIT boards.
Enable ADIN PHY support in the relevant board configurations.
Remove obsolete TI PHY fixups and reset handling.
Update SC594 and SC598 device trees consistently.

Hardware test is performed on SC598 with EZKIT on ADIN1300:

ADIN1300 PHY detected at addr 0
eth0 Waiting for PHY auto negotiation to complete........ done
BOOTP broadcast 1

References:

ADIN1300 datasheet:
https://www.analog.com/media/en/technical-documentation/data-sheets/ADIN1300.pdf
Analog Devices ADIN PHY documentation:
https://wiki.analog.com/resources/tools-software/linux-drivers/net-phy/adin

ozan956 · 2026-07-19T20:57:09Z

Related with: analogdevicesinc/linux#2883

ozan956 · 2026-07-20T11:10:00Z

@gastmaier the llm-review is not triggered here 😭

gastmaier · 2026-07-20T12:25:31Z

sudo systemctl --machine apps@.host --user status  webhook-action.service
sudo journalctl _UID=$(id -u apps) --user-unit=webhook-action.service -f

it did run https://github.com/analogdevicesinc/u-boot/actions/runs/29704103738/job/88238050338 but died

github-actions · 2026-07-20T12:40:27Z

LLM review

This series adds ADIN1200/1300 Ethernet PHY support for SC594/SC598 boards, fixing incorrect DP83867/DP83848 PHY descriptions in device trees, removing stale board-level GPIO-hog reset hooks, enabling CONFIG_PHY_ADIN in five defconfigs, and extending the ADIN driver with ADIN1200 support and RMII FIFO-depth configuration.

run: 29742054170

`218166f3` - configs: adi: enable ADIN PHY support

Defconfig ordering bug: CONFIG_PHY_ADIN=y is placed after CONFIG_PHY=y but savedefconfig (via qconfig.py -s) requires it immediately after CONFIG_SPI_FLASH_MTD=y. This is because drivers/net/phy/Kconfig is sourced before ETH_DESIGNWARE/DWC_ETH_QOS in drivers/net/Kconfig, so the canonical ordering puts PHY_ADIN earlier in the defconfig. All five affected defconfigs have the same issue.

Reproduced locally:

$ python3 tools/qconfig.py -s -d <list-of-5-defconfigs>
# Result: "defconfig updated" for all 5 files, moving PHY_ADIN earlier

CI warnings

The check_qconfig_sync check fails for all five modified defconfigs (sc594-som-ezkit-spl_defconfig, sc594-som-ezlite-spl_defconfig, sc598-htol-spl_defconfig, sc598-som-ezkit-spl_defconfig, sc598-som-ezlite-spl_defconfig). The build job failures (ARM/AArch64) are a CI infrastructure issue (missing cross-compiler toolchains), not a code defect.

Verification data

Datasheets for ADIN1200 and ADIN1300 were obtained from the docling mirror at https://raw.githubusercontent.com/analogdevicesinc/doctools/refs/heads/docling/media/en/technical-documentation/data-sheets/adin12{00,00}.md and used to verify:

PHY IDs: ADIN1200 = 0x0283BC20, ADIN1300 = 0x0283BC30 — both correct
GE_RMII_CFG (0xFF24) present in both PHYs with identical bit layout; FIFO depth field [6:4], encoding 001 = ±8 bits matches the driver default
GE_RGMII_CFG (0xFF23) delay encoding 000 = 8×200ps+400ps = 2000 ps — matches adi,{rx,tx}-internal-delay-ps = <2000> in DTS
RESET_N: minimum assert 10 µs (DTS: 10000 µs), register access 5 ms after deassert (DTS: 5000 µs) — both correct
EXT_REG_PTR/EXT_REG_DATA at 0x10/0x11 present in both PHYs — shared adin_ext_read/adin_ext_write mechanism is valid

Suggested patches

Apply the suggested patches with:

cd path/to/repository
export GITHUB_TOKEN=ghp_***
apply-patches --repo=analogdevicesinc/u-boot 29742054170

Install instructions

The following one-liner installs the script if not present already:

curl -fSsL "https://raw.githubusercontent.com/analogdevicesinc/doctools/refs/heads/main/ci/scripts/apply-patches.sh" \
     -o ~/.local/bin/apply-patches.sh && \
  grep -q "/apply-patches.sh" ~/.bashrc || echo "source ~/.local/bin/apply-patches.sh" >> $_ ; . $_

More information at AI Usage.

Configure the shared ADIN vendor registers for both ADIN1200 and ADIN1300 PHYs. This enables RMII mode and programs the RMII FIFO depth for ADIN1200 instead of relying only on generic PHY configuration. Keep the device-tree values in physical units. When the delay and FIFO properties are absent, use the ADIN1300 defaults of 2 ns for RGMII internal delay and 8 bits for RMII FIFO depth. Signed-off-by: Ozan Durgut <ozan.durgut@analog.com>

Enable the ADIN PHY driver in the SC594 and SC598 SPL defconfigs, which describe ADIN1200 and ADIN1300 PHYs. Signed-off-by: Ozan Durgut <ozan.durgut@analog.com>

The ADIN1300 reset GPIO is now described by the PHY node. Remove the obsolete carrier reset hooks so the PHY driver is the sole reset owner. Signed-off-by: Ozan Durgut <ozan.durgut@analog.com>

The SC598 boards use an ADIN1300 on eth0, and the EZKIT also uses an ADIN1200 on eth1. Replace the stale DP83867 description, describe the second RMII PHY, and remove the obsolete board PHY fixup. Move the ADIN1300 reset GPIO into the PHY nodes. The ADIN1300 data sheet, Table 3, requires RESET_N to be asserted for at least 10 us and specifies a 5 ms delay before MDIO registers are accessible after reset deassertion. Signed-off-by: Ozan Durgut <ozan.durgut@analog.com>

The SC594 EZKIT uses an ADIN1300 on eth0 and an ADIN1200 on eth1. Replace the stale PHY descriptions, describe the second RMII PHY, and remove the obsolete board PHY fixup. Move the ADIN1300 reset GPIO into the PHY nodes. The ADIN1300 data sheet, Table 3, requires RESET_N to be asserted for at least 10 us and specifies a 5 ms delay before MDIO registers are accessible after reset deassertion. Signed-off-by: Ozan Durgut <ozan.durgut@analog.com>

sipraga · 2026-07-22T09:11:14Z

+
+static int adin1300_config(struct phy_device *phydev)
+{
+	printf("ADIN1300 PHY detected at addr %d\n", phydev->addr);


I do wonder how useful this print really is (imagine if all devices announced their presence like this)

sipraga · 2026-07-22T09:22:18Z

+
+	dflt = adin_lookup_reg_value(tbl, dflt_cfg);
+	if (dflt < 0)
+		return dflt;


if dflt is negative, you are then implicitly casting that negative value to a u32 when returning. looks like the callers of this function expect a valid value to always be returned. but now you might be returning ((u32)(-EINVAL)) which is pretty dank

I am all for error handling, but maybe it's easier to just drop this check since it is anyway unrelated to your PR. otherwise you need to find a way for this function to unambiguously return error/success, and also update the callers to actually check for it (otherwise why bother?)

sipraga · 2026-07-22T09:26:05Z

-		printf("%s: failed to find %s, using default %d\n",
-		       __func__, prop_name, dflt);
+		printf("%s: failed to find %s, using default %u\n",
+		       __func__, prop_name, dflt_cfg);


if the property is not required by the bindings then I think such a print is a bit noisy, but maybe it's just me

sipraga · 2026-07-22T09:33:32Z

+	return adin_config(phydev);
+}
+
+U_BOOT_PHY_DRIVER(ADIN1200) = {


ah, weird, I was looking at mainline u-boot and noticed this is already there.

u-boot@52309be

Since we agreed in the Linux PR not to bother explicitly configuring RMII FIFO size, maybe you wanna just backport the above patch instead? Should achieve the same thing (except erroneously printing "ADIN1300" detected for an ADIN1200)

sipraga · 2026-07-22T09:36:23Z

+		#size-cells = <0>;
+
+		adin1200: ethernet-phy@1 {
+			compatible = "ethernet-phy-ieee802.3-c22";


not sure what u-boot policy is on whether to require this compatible or not, but worth a look

sipraga · 2026-07-22T09:37:32Z

 		#size-cells = <0>;

-		dp83867: ethernet-phy@0 {
+		adin1300: ethernet-phy@0 {


this doesn't look right. the PHY is on the carrier board, not the SoM

sipraga · 2026-07-22T09:38:21Z

 {
 	struct gpio_desc *eth1;
 	struct gpio_desc *eth1_reset;
-	struct gpio_desc *gige_reset;


this commit should be ordered after the update to the DT, or reworded as such, since in the commit message it suggests that the reset has already been put there

ozan956 requested a review from a team July 19, 2026 20:46

ozan956 self-assigned this Jul 19, 2026

ozan956 added this to ADSP Jul 19, 2026

ozan956 moved this to In Development in ADSP Jul 19, 2026

ozan956 mentioned this pull request Jul 19, 2026

Fix SC59x Ethernet PHY descriptions analogdevicesinc/linux#3444

Open

ozan956 added the llm review Request a review from a LLM Reviewer label Jul 19, 2026

ozan956 force-pushed the fix-phy-sc5xx branch from f7bfdea to d3da2d6 Compare July 19, 2026 22:15

ozan956 requested a review from a team July 20, 2026 06:38

gastmaier added llm review Request a review from a LLM Reviewer and removed llm review Request a review from a LLM Reviewer labels Jul 20, 2026

ozan956 added 5 commits July 21, 2026 09:34

configs: adi: enable ADIN PHY support

fa81698

Enable the ADIN PHY driver in the SC594 and SC598 SPL defconfigs, which describe ADIN1200 and ADIN1300 PHYs. Signed-off-by: Ozan Durgut <ozan.durgut@analog.com>

board: adi: somcrr: leave PHY reset to the PHY driver

3df9721

The ADIN1300 reset GPIO is now described by the PHY node. Remove the obsolete carrier reset hooks so the PHY driver is the sole reset owner. Signed-off-by: Ozan Durgut <ozan.durgut@analog.com>

ozan956 force-pushed the fix-phy-sc5xx branch 2 times, most recently from 34946ba to 036effb Compare July 21, 2026 07:40

sipraga reviewed Jul 22, 2026

View reviewed changes

Navigation Menu

Uh oh!

Fix SC59x Ethernet PHY descriptions and ADIN PHY Driver - #130

Fix SC59x Ethernet PHY descriptions and ADIN PHY Driver#130
ozan956 wants to merge 5 commits into
adi-u-boot-2025.10.yfrom
fix-phy-sc5xx

ozan956 commented Jul 19, 2026

Uh oh!

ozan956 commented Jul 19, 2026

Uh oh!

ozan956 commented Jul 20, 2026

Uh oh!

gastmaier commented Jul 20, 2026 •

edited

Loading

Uh oh!

github-actions Bot commented Jul 20, 2026

Uh oh!

sipraga Jul 22, 2026

Uh oh!

sipraga Jul 22, 2026

Uh oh!

sipraga Jul 22, 2026

Uh oh!

sipraga Jul 22, 2026

Uh oh!

sipraga Jul 22, 2026

Uh oh!

sipraga Jul 22, 2026

Uh oh!

sipraga Jul 22, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Uh oh!

Conversation

ozan956 commented Jul 19, 2026

Uh oh!

ozan956 commented Jul 19, 2026

Uh oh!

ozan956 commented Jul 20, 2026

Uh oh!

gastmaier commented Jul 20, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

github-actions Bot commented Jul 20, 2026

LLM review

218166f3 - configs: adi: enable ADIN PHY support

CI warnings

Verification data

Suggested patches

Uh oh!

sipraga Jul 22, 2026

Choose a reason for hiding this comment

Uh oh!

sipraga Jul 22, 2026

Choose a reason for hiding this comment

Uh oh!

sipraga Jul 22, 2026

Choose a reason for hiding this comment

Uh oh!

sipraga Jul 22, 2026

Choose a reason for hiding this comment

Uh oh!

sipraga Jul 22, 2026

Choose a reason for hiding this comment

Uh oh!

sipraga Jul 22, 2026

Choose a reason for hiding this comment

Uh oh!

sipraga Jul 22, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

gastmaier commented Jul 20, 2026 •

edited

Loading

`218166f3` - configs: adi: enable ADIN PHY support