OPI 4A - Allwinner T527

[JamesCL]

I just created a new image and am reporting the following:

- On the first boot, I don't have Ethernet (I don't use wireless) and
- The eMMC is not detected.

Thanks

 

Quote

root@orangepi-4a:~# cat /etc/armbian-release
# PLEASE DO NOT EDIT THIS FILE
BOARD=orangepi-4a
BOARD_NAME="Orange Pi 4A"
BOARDFAMILY=sun55iw3
BUILD_REPOSITORY_URL=https://github.com/armbian/build
BUILD_REPOSITORY_COMMIT=28862c7
LINUXFAMILY=sunxi64
ARCH=arm64
BOOT_SOC=
IMAGE_TYPE=user-built
BOARD_TYPE=csc
INITRD_ARCH=arm64
KERNEL_IMAGE_TYPE=Image
KERNEL_TARGET=edge
KERNEL_TEST_TARGET=
FORCE_BOOTSCRIPT_UPDATE=
FORCE_UBOOT_UPDATE=
OVERLAY_DIR="/boot/dtb/allwinner/overlay"
VENDOR="Armbian-unofficial"
VENDORCOLOR="247;16;0"
VENDORDOCS="https://docs.armbian.com/"
VENDORURL="https://duckduckgo.com/"
VENDORSUPPORT="https://community.armbian.com/"
VENDORBUGS="https://armbian.atlassian.net/"
BOOTSCRIPT_FORCE_UPDATE="no"
BOOTSCRIPT_DST="boot.cmd"
VERSION=26.05.0-trunk
REVISION=26.05.0-trunk
BRANCH=edge
root@orangepi-4a:~#
root@orangepi-4a:~#
root@orangepi-4a:~# dmesg | grep -i mmc
[    4.418379] sunxi-mmc 4020000.mmc: Got CD GPIO
[    4.418857] sunxi-mmc 4021000.mmc: allocated mmc-pwrseq
[    4.444666] sunxi-mmc 4022000.mmc: initialized, max. request size: 2048 KB, uses new timings mode
[    4.444692] sunxi-mmc 4021000.mmc: initialized, max. request size: 2048 KB, uses new timings mode
[    4.445215] sunxi-mmc 4020000.mmc: initialized, max. request size: 2048 KB, uses new timings mode
[    4.487006] mmc0: host does not support reading read-only switch, assuming write-enable
[    4.490389] mmc0: new high speed SDXC card at address aaaa
[    4.497740] mmcblk0: mmc0:aaaa SC64G 59.5 GiB
[    4.522286]  mmcblk0: p1 p2
[    4.585097] mmc2: error -110 whilst initialising MMC card
[    4.592377] mmc1: new high speed SDIO card at address 0001
[    4.596828] mmc2: Failed to initialize a non-removable card
[    7.193467] EXT4-fs (mmcblk0p2): mounted filesystem 757e37e7-0e5e-4683-8110-4a8276f032d5 ro with writeback data mode. Quota mode: none.
[   10.627502] EXT4-fs (mmcblk0p2): re-mounted 757e37e7-0e5e-4683-8110-4a8276f032d5 r/w.
[   12.277240] brcmfmac mmc1:0001:1: Direct firmware load for brcm/brcmfmac43456-sdio.xunlong,orangepi-4a.bin failed with error -2
root@orangepi-4a:~#
root@orangepi-4a:~#
root@orangepi-4a:~# uname -a
Linux orangepi-4a 7.0.0-rc6-edge-sunxi64 #2 SMP PREEMPT Sun Mar 29 22:40:00 UTC 2026 aarch64 GNU/Linux
root@orangepi-4a:~#

 

Edited April 6 by JamesCL

[sergiyd]

@Werner I imagine there are plenty of people here who could help solve similar problems. In theory, I could provide a pair of an old RPI (master) and a 4A (slave), with the slave serial connected to the master and power controlled via the GPIO, and they are also on the same VLAN. However, if something goes wrong it needs control to the slave's SD card, all I've found is an SD card extender, and it needs something like an emulator that connects to the master via the same USB/SPI.

[ioncube]

Anyone has a working image with GPU / VPU support? In the official Jammy I am using VPU is supported via gstreamer / CEDRA https://github.com/defencedog/orangepi4A/tree/main/GPU_VPU

[Edmilson Santos]

@ioncube  Were you able to use VPU with Jellyfin?

[sergiyd]

Things are still very sad with Ethernet. Is this could be related to some low-level timings in device configuration?

Edited May 2 by sergiyd

[Evgeniy evotronik]

extraargs=nvme_core.default_ps_max_latency_us=0 pcie_aspm=off pcie_port_pm=off

to /boot/armbianEnv.txt
fixes nvme detection

the temperature is
echo "$(( $(cat /sys/class/hwmon/hwmon0/temp1_input) / 1000 )) °C" 

if you copy the system to nvme and wanna boot with microsd (spi doesn't work for now):

1) it will actually boot but without boot mountpoint
2) mount /boot and /media/mmcboot manually

3) fix fstab 

 

Edited May 11 by Evgeniy evotronik

[Evgeniy evotronik]

to start ethernet

#!/bin/bash
echo -n "4510000.ethernet" > /sys/bus/platform/drivers/dwmac-sun55i/unbind 2>/dev/null
echo 271 > /sys/class/gpio/export 2>/dev/null
echo out > /sys/class/gpio/gpio271/direction 2>/dev/null
echo 0 > /sys/class/gpio/gpio271/value 2>/dev/null
sleep 0.1
echo 1 > /sys/class/gpio/gpio271/value 2>/dev/null
sleep 0.3
echo 271 > /sys/class/gpio/unexport 2>/dev/null
echo -n "4510000.ethernet" > /sys/bus/platform/drivers/dwmac-sun55i/bind 2>/dev/null

just systemd it

[Unit]
Description=Fix ethernet PHY on cold boot
After=multi-user.target

[Service]
Type=oneshot
ExecStart=/usr/local/bin/fix-eth.sh
RemainAfterExit=yes

[Install]
WantedBy=multi-user.target

Edited May 16 by Evgeniy evotronik

[JamesCL]

Hi ... 
 

Thanks @Evgeniy evotronik. Yesterday I downloaded the image "Armbian_community_26.2.0-trunk.886_Orangepi-4a_trixie_edge_7.0.5_minimal.img.xz", but the Ethernet and eMMC issues are still present. I will apply the Ethernet patch.

 

Thanks again.

 

[Evgeniy evotronik]

@JamesCl

 

Sorry, It's ok with 100m, but 1000m also doesn't receive an ip with dhcp.

Guess it is a dtb problem or maybe uboot

 

Edited May 16 by Evgeniy evotronik

[JamesCL]

The eMMC is still not detected.

Kernel log:

[    4.621568] mmc2: error -110 whilst initialising MMC card
[    4.627066] mmc2: Failed to initialize a non-removable card

Working devices:

[    4.502846] mmc0: new high speed SDXC card at address aaaa
[    4.508907] mmcblk0: mmc0:aaaa SC64G 59.5 GiB
[    4.551951] mmc1: new high speed SDIO card at address 0001

The system boots correctly from the SD card, but the onboard eMMC (mmc2) is not initialized.

There are also still Ethernet issues. After connecting through PuTTY, the system sometimes hangs when running dmesg, and the network connection becomes unstable or unresponsive.

 

System information:

Quote

 

root@orangepi-4a:~# uname -a
Linux orangepi-4a 7.0.7-edge-sunxi64 #1 SMP PREEMPT Thu May 14 13:31:20 UTC 2026 aarch64 GNU/Linux

root@orangepi-4a:~# cat /etc/os-release
PRETTY_NAME="Armbian_community 26.2.0-trunk.904 trixie"
NAME="Debian GNU/Linux"
VERSION_ID="13"
VERSION="13 (trixie)"
VERSION_CODENAME=trixie
DEBIAN_VERSION_FULL=13.4
ID=debian
HOME_URL="https://www.armbian.com"
SUPPORT_URL="https://forum.armbian.com"
BUG_REPORT_URL="https://www.armbian.com/bugs"
ARMBIAN_PRETTY_NAME="Armbian_community 26.2.0-trunk.904 trixie"
root@orangepi-4a:~#

 

Thanks

[Evgeniy evotronik]

guess it kernel/driver issue

The driver does not request the stmmaceth (emac1-25M) clock during probe. Without this clock the MDIO bus does not function on cold boot - PHY ID reads as 0x00000000 and kernel falls back to Generic PHY instead of YT8531. Ethernet does not work until warm reboot.

[Edmilson Santos]

HDMI is working out of box now?

[JamesCL]

@Evgeniy evotronik When I switch from 1 Gbps to 100 Mbps... Ethernet returns to normal.

[sergiyd]

This device is currently an outcast here, and until someone with understanding sets out to fix at least the main problems, welcome to the official image on the 5.15 kernel.

[bickns]

Greetings - just investigating an issue with the opi-4a Armbian 26.8.0 minimal (edge 7.0.12) build - not getting past starting kernel.  The board has a T527 - Hopefully this information is useful ?
 

### [Bug Report] Orange Pi 4A - Silent Hang at 'Starting kernel ...' (Trixie Minimal)

**Board:** Orange Pi 4A (Allwinner A523/T527)
**Image used:** Debian Trixie Minimal (Built via armbian-imager)
**Installation Media:** SD Card (Class 10)
**Log Type:** UART Serial Console Output (115200 baud)

---

#### Description of Issue
The recent community build of Debian Trixie Minimal for the Orange Pi 4A successfully clears the U-Boot / SPL stages, loads the Ramdisk and DTB from the SD card, but hangs completely silently immediately after hitting `Starting kernel ...`. The onboard status LED remains solid red with no flashing activity. 

#### The Technical Root Cause
Looking closely at the UART initialization text, there appears to be a mismatch between the SoC detected by the low-level firmware and the Device Tree Blob (DTB) loaded by the Armbian boot script:

1. **Firmware/U-Boot Chip Detection:** TF-A and U-Boot explicitly identify the SoC as an **Allwinner A523**:
   ```text
   NOTICE:  BL31: Detected Allwinner A523 SoC (1890)
   CPU:   Allwinner A523 (SUN55I)
   ```
2. **Script DTB Selection:** Immediately afterward, the automated script loads a **T527 DTB** profile:
   ```text
   Load fdt: /dtb/allwinner/sun55i-t527-orangepi-4a.dtb
   ```
Because the kernel tries to initialize the primary power rails and clock gates using a T527 map on an A523 layout, the CPU locks up silently at the execution handoff.

---

#### Complete Boot Log (Up to the Hang)

```text
U-Boot SPL 2026.01_armbian-2026.01-S127a-P60fe-Hc6a9-Vf757-Bd0d2-R448a (May 30 2026 - 05:53:33 +0000)
DRAM: 2048 MiB
Trying to boot from MMC1
NOTICE:  BL31: v2.13.0(debug):armbian
NOTICE:  BL31: Built : 05:52:43, May 30 2026
NOTICE:  BL31: Detected Allwinner A523 SoC (1890)
NOTICE:  BL31: Found U-Boot DTB at 0xa0b8ac0, model: OrangePi 4A
INFO:    Configuring SPC Controller
INFO:    BL31: Platform setup done
INFO:    GICv3 without legacy support detected.
INFO:    ARM GICv3 driver initialized in EL3
INFO:    Maximum SPI INTID supported: 287
INFO:    BL31: Initializing runtime services
INFO:    BL31: cortex_a55: CPU workaround for erratum 1530923 was applied
INFO:    PSCI: Suspend is unavailable
INFO:    BL31: Preparing for EL3 exit to normal world
INFO:    Entry point address = 0x4a000000
INFO:    SPSR = 0x3c9

U-Boot 2026.01_armbian-2026.01-S127a-P60fe-Hc6a9-Vf757-Bd0d2-R448a (May 30 2026 - 05:53:33 +0000) Allwinner Technology

CPU:   Allwinner A523 (SUN55I)
Model: OrangePi 4A
DRAM:  2 GiB
sunxi_set_gate: (CLK#35) unhandled
Core:  94 devices, 24 uclasses, devicetree: separate
WDT:   Not starting watchdog@2050000
MMC:   mmc@4020000: 0, mmc@4021000: 2, mmc@4022000: 1
Loading Environment from FAT... Unable to read "uboot.env" from mmc0:1... 
In:    serial@2500000
Out:   serial@2500000
Err:   serial@2500000
Net:   No ethernet found.
starting USB...
USB EHCI 1.00
USB OHCI 1.0
USB EHCI 1.00
USB OHCI 1.0
Bus usb@4101000: 1 USB Device(s) found
Bus usb@4101400: 1 USB Device(s) found
Bus usb@4200000: 2 USB Device(s) found
Bus usb@4200400: 1 USB Device(s) found
       scanning usb for storage devices... 0 Storage Device(s) found
Hit any key to stop autoboot: 0
PCIe: Acquiring resources...
PCIe: Failed to get switch-sel GPIO: -2
PCIe: Hardware init failed with error -2

Device 0: unknown device
switch to partitions #0, OK
mmc0 is current device
Scanning mmc 0:1...
Found U-Boot script /boot.scr
4641 bytes read in 2 ms (2.2 MiB/s)
## Executing script at 4fc00000
U-boot loaded from SD
Boot script loaded from mmc
249 bytes read in 2 ms (121.1 KiB/s)
Load fdt: /dtb/allwinner/sun55i-t527-orangepi-4a.dtb
33383 bytes read in 7 ms (4.5 MiB/s)
Working FDT set to 4fa00000
Failed to load '/dtb/allwinner/overlay/sun55i-t527-fixup.scr'
16748327 bytes read in 814 ms (19.6 MiB/s)
41171456 bytes read in 1707 ms (23 MiB/s)
Moving Image from 0x40080000 to 0x40200000, end=0x42a30000
## Loading init Ramdisk from Legacy Image at 4ff00000 ...
   Image Name:   uInitrd
   Image Type:   AArch64 Linux RAMDisk Image (gzip compressed)
   Data Size:    16748263 Bytes = 16 MiB
   Load Address: 00000000
   Entry Point:  00000000
   Verifying Checksum ... OK
## Flattened Device Tree blob at 4fa00000
   Booting using the fdt blob at 0x4fa00000
Working FDT set to 4fa00000
   Loading Ramdisk to 49007000, end 49fffee7 ... OK
   Loading Device Tree to 0000000048f96000, end 0000000049006fff ... OK
Working FDT set to 48f96000

Starting kernel ...
```
 

[bickns]

 Greetings 

### [Follow-up Update] Boot Deep-Dive via Cubie-A5E DTB Masking

Following up on my previous boot freeze post, I (with some assistance from Gemini - who formatted this post) managed to force the Armbian kernel to initialize by masking the sister chip profile (`sun55i-a523-cubie-a5e.dtb`) over the broken `sun55i-t527-orangepi-4a.dtb` target and injecting `clk_ignore_unused` / `irqpoll` parameters.

This got us past the initial handoff hang and successfully spawned the 8-core CPU array initialization and initramfs launch (`Run /init as init process`), but the system eventually hit a permanent deadlock during the device probe phase. 

 

I don't think I can get this to go any further without some many hours spent looking at how these things work and its currently 35 degrees C here in the UK so time for a beer methinks!

 

#### Full Boot Log Paste Link: - this is the log before using irqpoll parameter (which eventually caused a deadlock) - I can post the irqpoll logs although i wouldn't have thought irqpolling is the way to go ?

https://pastebin.com/TvxTxdbp

 

### Either way - having just read that this post refers to irqpoll then this is the boot logfile with irqpoll added an an extraparm

https://pastebin.com/TNS3DDTC

 
#### Key Technical Takeaways from the Log Dump:

1. **PMIC / I2C Communication Failures:**
   The kernel initializes the AXP717 and AXP323 power chips but encounters standard timeout errors right away:
   ```text
   axp20x-i2c 0-0034: AXP20x variant AXP717 found
   axp20x-i2c 0-0034: Failed to set masks in 0x40: -6
   axp20x-i2c 0-0034: failed to add irq chip: -6
   ```

2. **MMC / Storage Controller Deadlock:**
   Because the AXP717 framework drops out and the Cubie pin definitions do not line up with the Orange Pi 4A's physical board wiring, the kernel enters an endless loop waiting for internal voltage rails (`cldo3` and `bldo1`) to power up the storage controller:
   ```text
   platform 4020000.mmc: deferred probe pending: platform: wait for supplier /soc/i2c@7081400/pmic@34/regulators/cldo3
   platform 4021000.mmc: deferred probe pending: platform: wait for supplier /soc/i2c@7081400/pmic@34/regulators/bldo1
   ```
   This prevents the storage device slot from spinning up, trapping the system right as it attempts to mount the root filesystem.

Hopefully this helps isolate the unaligned power configuration loops and clock tables in the current `armbian-imager` source profile for the OPi 4A!
 

[bickns]

Title: [Working Image + Patch] Orange Pi 4A (T527) - Fast Ethernet Fixed on Mainline 7.0 Edge

 

Hi everyone,

Since the Orange Pi 4A (Allwinner T527) is currently an unmaintained community target on the bleeding-edge Mainline 7.0 Edge branch, the default images boot up completely blind with cannot attach to PHY (error: -ENODEV) errors.

After a lot of debugging, I (with assistance from Gemini - who wrote this post) managed to successfully compile a full, functional Debian Trixie image where Ethernet works perfectly (capped at 100M) and the green status LED pulses a native heartbeat.

 

The secret sauce is that the board uses a Motorcomm YT8531 transceiver, but the kernel falls back to a Generic PHY driver which fails to calibrate the high-frequency RGMII clock skews required for Gigabit speeds. Dropping the link frequency down to 100 Mbps Full Duplex bypasses the timing sensitivity completely, allowing NetworkManager to instantly pull a stable IPv4 DHCP lease!

(Note: Mainline 7.0 does not yet have internal C drivers for the T527 audio codec or the SDIO mmc1 bus clocks, so internal audio and onboard Wi-Fi remain offline. Use a USB sound/Wi-Fi dongle if needed).  I can't vouch whether HDMI is working or not as I was debugging it using a serial cable.

 

📦 Download the Image

https://drive.google.com/drive/folders/1ieu3TBn2nSP0BaJvJMIhLeLM6kt-x8bX?usp=sharing

 

🛠️ The Device Tree Patch File

I managed to diff the radxa-cubie-5e.dts file and patch it to get it to kind of work with the opi 4a.  The patchfile is also in the google drive folder and supplied as an attach here.

 

To include this in your own Armbian build environment, delete any old patches and save this file exactly as:

 

build/userpatches/kernel/archive/sunxi-7.0/zzzz-working-opi4a.patch

 

It wipes out the broken default configurations and replaces them with clean, compiling mainline mappings.  You can then build the image using this:

 

./compile.sh build BOARD=orangepi-4a BRANCH=edge BUILD_DESKTOP=no BUILD_MINIMAL=no KERNEL_CONFIGURE=no RELEASE=trixie USERPATCHES_ALLOWED=yes SYSTEM_VERBOSE=yes FORCE_REBUILD=yes

📝 Step-by-Step Instructions to Lock Down the 100M Connection

Because the Generic PHY driver cannot process 1Gbps packets correctly, the interface will sit in an unconfigured DOWN loop upon initial boot. Run these steps via your serial terminal loop to bring it online and automate it:

Force the hardware down to 100M to clear the packet queue:

 

 

sudo ethtool -s end0 speed 100 duplex full autoneg off

 

Tell NetworkManager to explicitly negotiate the port:

 

sudo nmcli device connect end0

 

 

Make the 100M setting permanent so it survives a reboot:

 

sudo nmcli connection add type ethernet con-name eth-100m ifname end0 802-3-ethernet.auto-negotiate no 802-3-ethernet.speed 100 802-3-ethernet.duplex full ipv4.method auto ipv6.method disabled

 

Automate the green status heartbeat LED:

Open /etc/rc.local

and add this line right above exit 0

 

echo heartbeat > /sys/class/leds/green:status/trigger

 

Reboot, and you should have a 100% automated, maintenance-free, fully operational network target - with some tweaking will hopefully get you to the next stage as apt will be working. Hope this helps anyone else struggling to get this community board running on modern releases!

zzzz-radxa-to-orangepi4a-fix.patch