开云体育

Hi Quentin,

On Tue, 13 May 2025 at 21:52, Quentin Schulz <quentin.schulz@...> wrote:

Hi Anand,

On 5/13/25 6:14 PM, Anand Moon wrote:

Hi All,

On Mon, 12 May 2025 at 13:51, Anand Moon via groups.io
<linux.amoon@...> wrote:

Hi Quentin,

On Mon, 12 May 2025 at 12:36, Quentin Schulz <quentin.schulz@...> wrote:

Hi Anand,

On 5/9/25 7:45 PM, Anand Moon wrote:

Hi Quentin,

On Fri, 9 May 2025 at 17:21, Quentin Schulz <quentin.schulz@...> wrote:

Hi Neil, Anand,

On 5/9/25 9:57 AM, Neil Armstrong wrote:

On 09/05/2025 09:02, Anand Moon wrote:

Add conditional compilation for the usb5744_i2c_init() function based
on the CONFIG_DM_I2C configuration, to avoid compilation failure.

CC net/net-common.o
AR net/built-in.o
LDS u-boot.lds
LD u-boot
aarch64-linux-gnu-ld.bfd: common/usb_onboard_hub.o: in function
`usb5744_i2c_init':
/home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/
usb_onboard_hub.c:74:(.text.usb5744_i2c_init+0xfc): undefined
reference to `i2c_get_chip'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:89:(.text.usb5744_i2c_init+0x1a0):
undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:96:(.text.usb5744_i2c_init+0x1c8):
undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:104:(.text.usb5744_i2c_init+0x1f0):
undefined reference to `dm_i2c_write'
Segmentation fault (core dumped)
make: *** [Makefile:1824: u-boot] Error 139
make: *** Deleting file 'u-boot'

Signed-off-by: Anand Moon <linux.amoon@...>
---
common/usb_onboard_hub.c | 4 ++++
1 file changed, 4 insertions(+)

diff --git a/common/usb_onboard_hub.c b/common/usb_onboard_hub.c
index 1242513c631..92e3f3a92c9 100644
--- a/common/usb_onboard_hub.c
+++ b/common/usb_onboard_hub.c
@@ -30,6 +30,7 @@ struct onboard_hub_data {
int (*init)(struct udevice *dev);
};
+#if CONFIG_IS_ENABLED(DM_I2C)
static int usb5744_i2c_init(struct udevice *dev)
{
/*
@@ -109,6 +110,7 @@ static int usb5744_i2c_init(struct udevice *dev)
return 0;
}
+#endif
int usb_onboard_hub_reset(struct udevice *dev)
{
@@ -222,7 +224,9 @@ static const struct onboard_hub_data usb2514_data = {
};
static const struct onboard_hub_data usb5744_data = {
+#if CONFIG_IS_ENABLED(DM_I2C)
.init = usb5744_i2c_init,
+#endif
.power_on_delay_us = 1000,
.reset_us = 5,
};

I think you should completely disable usb5744_data in this case

I would recommend to go the same route as the Linux kernel which has a
special Kconfig option for it, c.f. USB_ONBOARD_DEV_USB5744 which adds
the necessary dependencies.

On Linux kernel, it needs to parse the device node with i2c-bus.

Seems like it does in U-Boot too? c.f. usb5744_i2c_init() with
dev_read_phandle_with_args with i2c-bus?

ok.

We probably should be careful and figure out which boards are actually
using the usb5744 and add this symbol to their defconfig so they don't
regress.

From my initial analysis configs/xilinx_zynqmp_kria_defconfig
which use CONFIG_USB_ONBOARD_HUB
I believe using DM_I2C to guard the code is the best approach.
it should be around .init() callback to guard this to work.

I disagree. The driver clearly supports the USB5744 and the driver is
compiled regardless of DM_I2C support which is required (as far as I
understood, maybe that's incorrect?) for being able to use USB5744? I
don't like hidden dependencies.

Okay, I will add a new configuration option to guard the code as done
in the Linux kernel.

Also, I believe Neil's point is we shouldn't try to probe the USB5744 if
there is no DM_I2C support, simply excluding .init() callback will
probably just render this "support" of USB5744 non-functional?

Let me debug this more, I will update the next version.
I need to check if this binds correctly to the USB hub reset.

I am trying to get the usb_onboard_hub to bind.
but this driver is not getting probed in the first place,
I am missing some configuration to get this working.

Don't know your setup, but maybe have a look at

?

Thank you for the tip. I've incorporated all the changes as
suggested in this series. However, I'm still unable to get the
usb_onboard_hub_probe to execute, which is preventing me
from further debugging.

Cheers,
Quentin

Thanks
-Anand

Hi All,

On Mon, 12 May 2025 at 13:51, Anand Moon via groups.io
<linux.amoon@...> wrote:

Hi Quentin,

On Mon, 12 May 2025 at 12:36, Quentin Schulz <quentin.schulz@...> wrote:

Hi Anand,

On 5/9/25 7:45 PM, Anand Moon wrote:

Hi Quentin,

On Fri, 9 May 2025 at 17:21, Quentin Schulz <quentin.schulz@...> wrote:

Hi Neil, Anand,

On 5/9/25 9:57 AM, Neil Armstrong wrote:

On 09/05/2025 09:02, Anand Moon wrote:

Add conditional compilation for the usb5744_i2c_init() function based
on the CONFIG_DM_I2C configuration, to avoid compilation failure.

CC net/net-common.o
AR net/built-in.o
LDS u-boot.lds
LD u-boot
aarch64-linux-gnu-ld.bfd: common/usb_onboard_hub.o: in function
`usb5744_i2c_init':
/home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/
usb_onboard_hub.c:74:(.text.usb5744_i2c_init+0xfc): undefined
reference to `i2c_get_chip'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:89:(.text.usb5744_i2c_init+0x1a0):
undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:96:(.text.usb5744_i2c_init+0x1c8):
undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:104:(.text.usb5744_i2c_init+0x1f0):
undefined reference to `dm_i2c_write'
Segmentation fault (core dumped)
make: *** [Makefile:1824: u-boot] Error 139
make: *** Deleting file 'u-boot'

Signed-off-by: Anand Moon <linux.amoon@...>
---
common/usb_onboard_hub.c | 4 ++++
1 file changed, 4 insertions(+)

diff --git a/common/usb_onboard_hub.c b/common/usb_onboard_hub.c
index 1242513c631..92e3f3a92c9 100644
--- a/common/usb_onboard_hub.c
+++ b/common/usb_onboard_hub.c
@@ -30,6 +30,7 @@ struct onboard_hub_data {
int (*init)(struct udevice *dev);
};
+#if CONFIG_IS_ENABLED(DM_I2C)
static int usb5744_i2c_init(struct udevice *dev)
{
/*
@@ -109,6 +110,7 @@ static int usb5744_i2c_init(struct udevice *dev)
return 0;
}
+#endif
int usb_onboard_hub_reset(struct udevice *dev)
{
@@ -222,7 +224,9 @@ static const struct onboard_hub_data usb2514_data = {
};
static const struct onboard_hub_data usb5744_data = {
+#if CONFIG_IS_ENABLED(DM_I2C)
.init = usb5744_i2c_init,
+#endif
.power_on_delay_us = 1000,
.reset_us = 5,
};

I think you should completely disable usb5744_data in this case

I would recommend to go the same route as the Linux kernel which has a
special Kconfig option for it, c.f. USB_ONBOARD_DEV_USB5744 which adds
the necessary dependencies.

On Linux kernel, it needs to parse the device node with i2c-bus.

Seems like it does in U-Boot too? c.f. usb5744_i2c_init() with
dev_read_phandle_with_args with i2c-bus?

ok.

We probably should be careful and figure out which boards are actually
using the usb5744 and add this symbol to their defconfig so they don't
regress.

From my initial analysis configs/xilinx_zynqmp_kria_defconfig
which use CONFIG_USB_ONBOARD_HUB
I believe using DM_I2C to guard the code is the best approach.
it should be around .init() callback to guard this to work.

I disagree. The driver clearly supports the USB5744 and the driver is
compiled regardless of DM_I2C support which is required (as far as I
understood, maybe that's incorrect?) for being able to use USB5744? I
don't like hidden dependencies.

Okay, I will add a new configuration option to guard the code as done
in the Linux kernel.

Also, I believe Neil's point is we shouldn't try to probe the USB5744 if
there is no DM_I2C support, simply excluding .init() callback will
probably just render this "support" of USB5744 non-functional?

Let me debug this more, I will update the next version.
I need to check if this binds correctly to the USB hub reset.

I am trying to get the usb_onboard_hub to bind.
but this driver is not getting probed in the first place,
I am missing some configuration to get this working.

Hit any key to stop autoboot: 0
=> usb start
starting USB...
Register 3000140 NbrPorts 3
Starting the controller
USB XHCI 1.10
Bus usb@ff500000: 4 USB Device(s) found
scanning usb for storage devices... 1 Storage Device(s) found
=> dm tree
...
mmc 1 [ + ] meson_gx_mmc | |-- mmc@ffe07000
blk 1 [ ] mmc_blk | | |-- mmc@...
bootdev 2 [ ] mmc_bootdev | | `-- mmc@...
simple_bus 11 [ + ] dwc3-meson-g12a | `-- usb@ffe09000
usb_gadget 0 [ ] dwc2-udc-otg | |-- usb@ff400000
usb 0 [ + ] xhci-dwc3 | `-- usb@ff500000
usb_hub 0 [ + ] usb_hub | `-- usb_hub
usb_hub 1 [ + ] usb_hub | |-- usb_hub
usb_hub 2 [ + ] usb_hub | `-- usb_hub
usb_mass_s 0 [ + ] usb_mass_storage |
`-- usb_mass_storage
blk 2 [ + ] usb_storage_blk |
|-- usb_mass_storage.lun0
partition 0 [ + ] blk_partition |
| `-- usb_mass_storage.lun0:1
bootdev 3 [ ] usb_bootdev |
`-- usb_mass_storage.lun0.bootdev
clk 2 [ + ] fixed_clock |-- xtal-clk

Thanks
-Anand

Hi Anand,

On 5/9/25 7:45 PM, Anand Moon wrote:

Hi Quentin,
On Fri, 9 May 2025 at 17:21, Quentin Schulz <quentin.schulz@...> wrote:

Hi Neil, Anand,

On 5/9/25 9:57 AM, Neil Armstrong wrote:

On 09/05/2025 09:02, Anand Moon wrote:

Add conditional compilation for the usb5744_i2c_init() function based
on the CONFIG_DM_I2C configuration, to avoid compilation failure.

CC net/net-common.o
AR net/built-in.o
LDS u-boot.lds
LD u-boot
aarch64-linux-gnu-ld.bfd: common/usb_onboard_hub.o: in function
`usb5744_i2c_init':
/home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/
usb_onboard_hub.c:74:(.text.usb5744_i2c_init+0xfc): undefined
reference to `i2c_get_chip'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:89:(.text.usb5744_i2c_init+0x1a0):
undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:96:(.text.usb5744_i2c_init+0x1c8):
undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:104:(.text.usb5744_i2c_init+0x1f0):
undefined reference to `dm_i2c_write'
Segmentation fault (core dumped)
make: *** [Makefile:1824: u-boot] Error 139
make: *** Deleting file 'u-boot'

Signed-off-by: Anand Moon <linux.amoon@...>
---
common/usb_onboard_hub.c | 4 ++++
1 file changed, 4 insertions(+)

diff --git a/common/usb_onboard_hub.c b/common/usb_onboard_hub.c
index 1242513c631..92e3f3a92c9 100644
--- a/common/usb_onboard_hub.c
+++ b/common/usb_onboard_hub.c
@@ -30,6 +30,7 @@ struct onboard_hub_data {
int (*init)(struct udevice *dev);
};
+#if CONFIG_IS_ENABLED(DM_I2C)
static int usb5744_i2c_init(struct udevice *dev)
{
/*
@@ -109,6 +110,7 @@ static int usb5744_i2c_init(struct udevice *dev)
return 0;
}
+#endif
int usb_onboard_hub_reset(struct udevice *dev)
{
@@ -222,7 +224,9 @@ static const struct onboard_hub_data usb2514_data = {
};
static const struct onboard_hub_data usb5744_data = {
+#if CONFIG_IS_ENABLED(DM_I2C)
.init = usb5744_i2c_init,
+#endif
.power_on_delay_us = 1000,
.reset_us = 5,
};

I think you should completely disable usb5744_data in this case

I would recommend to go the same route as the Linux kernel which has a
special Kconfig option for it, c.f. USB_ONBOARD_DEV_USB5744 which adds
the necessary dependencies.

On Linux kernel, it needs to parse the device node with i2c-bus.

Seems like it does in U-Boot too? c.f. usb5744_i2c_init() with dev_read_phandle_with_args with i2c-bus?

We probably should be careful and figure out which boards are actually
using the usb5744 and add this symbol to their defconfig so they don't
regress.

From my initial analysis configs/xilinx_zynqmp_kria_defconfig
which use CONFIG_USB_ONBOARD_HUB
I believe using DM_I2C to guard the code is the best approach.
it should be around .init() callback to guard this to work.

I disagree. The driver clearly supports the USB5744 and the driver is compiled regardless of DM_I2C support which is required (as far as I understood, maybe that's incorrect?) for being able to use USB5744? I don't like hidden dependencies.

Also, I believe Neil's point is we shouldn't try to probe the USB5744 if there is no DM_I2C support, simply excluding .init() callback will probably just render this "support" of USB5744 non-functional?

Cheers,
Quentin

Cheers,
Quentin

Thanks
-Anand

Applied patches to 2025.04 release (applied clean). Made stock
defconfig for odroid n2, built, signed, installed. The reboot issue
remains and usb ports are unavailable to linux kernel after reboot.

It is worth noting that linux does see some enumeration of devices
after a while, but even if it logs to the console about device
attachment, the devices don't actually function. This was the same
behavior with the release version.

Wayne

Yes. Will report back how it works on the n2+.

Hi Christian,

On Mon, 12 May 2025 at 15:05, Christian Hewitt
<christianshewitt@...> wrote:

On 12 May 2025, at 10:59?am, Anand Moon via groups.io <linux.amoon@...> wrote:

Hi Wayne,

On Sun, 11 May 2025 at 01:51, Wayne Schroeder <raz@...> wrote:

Applied patches to 2025.04 release (applied clean). Made stock
defconfig for odroid n2, built, signed, installed. The reboot issue
remains and usb ports are unavailable to linux kernel after reboot.

It is worth noting that linux does see some enumeration of devices
after a while, but even if it logs to the console about device
attachment, the devices don't actually function. This was the same
behavior with the release version.

Thanks for testing these patches.
We need to modify the boot order to probe USB before MMC

Why? Most Amlogic boards boot from an mmc device (SD or internal flash
storage) not USB. It smells like a workaround not a fix, and I wonder
what the performance impact on boot will be?

I'm providing a solution that may not apply to all Amlogic boards.
The goal is to ensure the USB device binds early, allowing the USB keyboard
and mass storage device to be detected promptly.
However, I'm unsure if booting from USB mass storage is supported.

Thanks
-Anand

Christian

diff --git a/include/configs/meson64.h b/include/configs/meson64.h
index f3275b37a51..3ace59880bd 100644
--- a/include/configs/meson64.h
+++ b/include/configs/meson64.h
@@ -119,8 +119,8 @@
#define BOOT_TARGET_DEVICES(func) \
func(ROMUSB, romusb, na) \
func(USB_DFU, usbdfu, na) \
- BOOT_TARGET_MMC(func) \
BOOT_TARGET_DEVICES_USB(func) \
+ BOOT_TARGET_MMC(func) \
BOOT_TARGET_NVME(func) \
BOOT_TARGET_SCSI(func) \
BOOT_TARGET_PXE(func) \

Here is the boot sequence at my end.
------------------------
U-Boot 2025.07-rc1-00223-g704a4d1fd804 (May 12 2025 - 12:15:29 +0530)
odroid-n2/n2-plus

Model: Hardkernel ODROID-N2
SoC: Amlogic Meson G12B (S922X) Revision 29:c (40:2)
DRAM: 1 GiB (total 3.8 GiB)
Core: 402 devices, 30 uclasses, devicetree: separate
MMC: mmc@ffe05000: 0, mmc@ffe07000: 1
Loading Environment from nowhere... OK
In: usbkbd,serial
Out: vidconsole,serial
Err: vidconsole,serial
Board variant: n2-plus
Net: eth0: ethernet@ff3f0000

Hit any key to stop autoboot: 0
starting USB...
Register 3000140 NbrPorts 3
Starting the controller
USB XHCI 1.10
Bus usb@ff500000: 4 USB Device(s) found
scanning usb for storage devices... 1 Storage Device(s) found

Device 0: Vendor: ASMT Rev: 0 Prod: 2115
Type: Hard Disk
Capacity: 114473.4 MB = 111.7 GB (234441648 x 512)
... is now current device
Scanning usb 0:1...
Card did not respond to voltage select! : -110
Cannot persist EFI variables without system partition
Loading Boot0000 'mmc 1' failed
Loading Boot0001 'usb 0' failed
EFI boot manager: Cannot load any image
Card did not respond to voltage select! : -110
switch to partitions #0, OK
mmc1(part 0) is current device
Scanning mmc 1:1...
Found /extlinux/extlinux.conf
Retrieving file: /extlinux/extlinux.conf
1: Manjaro ARM
Retrieving file: /Image
Retrieving file: /initramfs-linux.img
append: initrd=/initramfs-linux.img console=ttyAML0,115200n8
root=PARTUUID=0efc7c36-02 rootwait audit=0 clk_ignore_unused
plymouth.ignore-serial-consoles

Wayne

On Fri, May 9, 2025 at 10:01?AM Wayne Schroeder <raz@...> wrote:

Yes. Will report back how it works on the n2+.


On Fri, May 9, 2025 at 2:08?AM Anand Moon <linux.amoon@...> wrote:

Hi Wayne,

On Fri, 9 May 2025 at 01:25, Wayne Schroeder <raz@...> wrote:

I actually already have a working solution that I documented in my
github repo (dubschro/odroid-n2). All I do is set BOOTCOMMAND to
check/delete a flag file on my boot partition and if it is found, then
reset usb and sleep for 10 seconds. I then have a systemd shutdown
script only on reboot to touch that file.

CONFIG_BOOTCOMMAND="fatrm mmc 0:1 reboot && echo Resetting USB on
reboot... && usb reset && sleep 10; run distro_bootcmd"

It is a hack, but... the system reliably reboots remotely now, and
that's the important part for now.

Wayne

I did not modify this at my end. However, this is not required with
the above fix.

Thanks
-Anand

On Thu, May 8, 2025 at 11:22?AM Ferass El Hafidi
<funderscore@...> wrote:

On Wed May 7, 2025 at 1:03 PM UTC, Wayne Schroeder wrote:

It's hard for me to quantify if usb works in u-boot as I'm not loading
off of any usb storage devices. It certainly never lets me use the
keyboard to abort auto boot, and never has. On reboot, to reliably

You can send bytes via UART to abort boot, if you happen to have a UART
adapter around.

Can you test this series to see if this resolves the issue?



Thanks
-Anand

_______________________________________________
linux-amlogic mailing list
linux-amlogic@...

On 12 May 2025, at 10:59?am, Anand Moon via groups.io <linux.amoon@...> wrote:

Hi Wayne,

On Sun, 11 May 2025 at 01:51, Wayne Schroeder <raz@...> wrote:

Applied patches to 2025.04 release (applied clean). Made stock
defconfig for odroid n2, built, signed, installed. The reboot issue
remains and usb ports are unavailable to linux kernel after reboot.

It is worth noting that linux does see some enumeration of devices
after a while, but even if it logs to the console about device
attachment, the devices don't actually function. This was the same
behavior with the release version.

Thanks for testing these patches.
We need to modify the boot order to probe USB before MMC

Why? Most Amlogic boards boot from an mmc device (SD or internal flash
storage) not USB. It smells like a workaround not a fix, and I wonder
what the performance impact on boot will be?

Christian

diff --git a/include/configs/meson64.h b/include/configs/meson64.h
index f3275b37a51..3ace59880bd 100644
--- a/include/configs/meson64.h
+++ b/include/configs/meson64.h
@@ -119,8 +119,8 @@
#define BOOT_TARGET_DEVICES(func) \
func(ROMUSB, romusb, na) \
func(USB_DFU, usbdfu, na) \
- BOOT_TARGET_MMC(func) \
BOOT_TARGET_DEVICES_USB(func) \
+ BOOT_TARGET_MMC(func) \
BOOT_TARGET_NVME(func) \
BOOT_TARGET_SCSI(func) \
BOOT_TARGET_PXE(func) \

Here is the boot sequence at my end.
------------------------
U-Boot 2025.07-rc1-00223-g704a4d1fd804 (May 12 2025 - 12:15:29 +0530)
odroid-n2/n2-plus

Model: Hardkernel ODROID-N2
SoC: Amlogic Meson G12B (S922X) Revision 29:c (40:2)
DRAM: 1 GiB (total 3.8 GiB)
Core: 402 devices, 30 uclasses, devicetree: separate
MMC: mmc@ffe05000: 0, mmc@ffe07000: 1
Loading Environment from nowhere... OK
In: usbkbd,serial
Out: vidconsole,serial
Err: vidconsole,serial
Board variant: n2-plus
Net: eth0: ethernet@ff3f0000

Hit any key to stop autoboot: 0
starting USB...
Register 3000140 NbrPorts 3
Starting the controller
USB XHCI 1.10
Bus usb@ff500000: 4 USB Device(s) found
scanning usb for storage devices... 1 Storage Device(s) found

Device 0: Vendor: ASMT Rev: 0 Prod: 2115
Type: Hard Disk
Capacity: 114473.4 MB = 111.7 GB (234441648 x 512)
... is now current device
Scanning usb 0:1...
Card did not respond to voltage select! : -110
Cannot persist EFI variables without system partition
Loading Boot0000 'mmc 1' failed
Loading Boot0001 'usb 0' failed
EFI boot manager: Cannot load any image
Card did not respond to voltage select! : -110
switch to partitions #0, OK
mmc1(part 0) is current device
Scanning mmc 1:1...
Found /extlinux/extlinux.conf
Retrieving file: /extlinux/extlinux.conf
1: Manjaro ARM
Retrieving file: /Image
Retrieving file: /initramfs-linux.img
append: initrd=/initramfs-linux.img console=ttyAML0,115200n8
root=PARTUUID=0efc7c36-02 rootwait audit=0 clk_ignore_unused
plymouth.ignore-serial-consoles

Wayne

On Fri, May 9, 2025 at 10:01?AM Wayne Schroeder <raz@...> wrote:

Yes. Will report back how it works on the n2+.


On Fri, May 9, 2025 at 2:08?AM Anand Moon <linux.amoon@...> wrote:

Hi Wayne,

On Fri, 9 May 2025 at 01:25, Wayne Schroeder <raz@...> wrote:

I actually already have a working solution that I documented in my
github repo (dubschro/odroid-n2). All I do is set BOOTCOMMAND to
check/delete a flag file on my boot partition and if it is found, then
reset usb and sleep for 10 seconds. I then have a systemd shutdown
script only on reboot to touch that file.

CONFIG_BOOTCOMMAND="fatrm mmc 0:1 reboot && echo Resetting USB on
reboot... && usb reset && sleep 10; run distro_bootcmd"

It is a hack, but... the system reliably reboots remotely now, and
that's the important part for now.

Wayne

I did not modify this at my end. However, this is not required with
the above fix.

Thanks
-Anand

On Thu, May 8, 2025 at 11:22?AM Ferass El Hafidi
<funderscore@...> wrote:

On Wed May 7, 2025 at 1:03 PM UTC, Wayne Schroeder wrote:

It's hard for me to quantify if usb works in u-boot as I'm not loading
off of any usb storage devices. It certainly never lets me use the
keyboard to abort auto boot, and never has. On reboot, to reliably

You can send bytes via UART to abort boot, if you happen to have a UART
adapter around.

Can you test this series to see if this resolves the issue?



Thanks
-Anand

_______________________________________________
linux-amlogic mailing list
linux-amlogic@...

Hi Quentin,

On Mon, 12 May 2025 at 12:36, Quentin Schulz <quentin.schulz@...> wrote:

Hi Anand,

On 5/9/25 7:45 PM, Anand Moon wrote:

Hi Quentin,

On Fri, 9 May 2025 at 17:21, Quentin Schulz <quentin.schulz@...> wrote:

Hi Neil, Anand,

On 5/9/25 9:57 AM, Neil Armstrong wrote:

On 09/05/2025 09:02, Anand Moon wrote:

Add conditional compilation for the usb5744_i2c_init() function based
on the CONFIG_DM_I2C configuration, to avoid compilation failure.

CC net/net-common.o
AR net/built-in.o
LDS u-boot.lds
LD u-boot
aarch64-linux-gnu-ld.bfd: common/usb_onboard_hub.o: in function
`usb5744_i2c_init':
/home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/
usb_onboard_hub.c:74:(.text.usb5744_i2c_init+0xfc): undefined
reference to `i2c_get_chip'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:89:(.text.usb5744_i2c_init+0x1a0):
undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:96:(.text.usb5744_i2c_init+0x1c8):
undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:104:(.text.usb5744_i2c_init+0x1f0):
undefined reference to `dm_i2c_write'
Segmentation fault (core dumped)
make: *** [Makefile:1824: u-boot] Error 139
make: *** Deleting file 'u-boot'

Signed-off-by: Anand Moon <linux.amoon@...>
---
common/usb_onboard_hub.c | 4 ++++
1 file changed, 4 insertions(+)

diff --git a/common/usb_onboard_hub.c b/common/usb_onboard_hub.c
index 1242513c631..92e3f3a92c9 100644
--- a/common/usb_onboard_hub.c
+++ b/common/usb_onboard_hub.c
@@ -30,6 +30,7 @@ struct onboard_hub_data {
int (*init)(struct udevice *dev);
};
+#if CONFIG_IS_ENABLED(DM_I2C)
static int usb5744_i2c_init(struct udevice *dev)
{
/*
@@ -109,6 +110,7 @@ static int usb5744_i2c_init(struct udevice *dev)
return 0;
}
+#endif
int usb_onboard_hub_reset(struct udevice *dev)
{
@@ -222,7 +224,9 @@ static const struct onboard_hub_data usb2514_data = {
};
static const struct onboard_hub_data usb5744_data = {
+#if CONFIG_IS_ENABLED(DM_I2C)
.init = usb5744_i2c_init,
+#endif
.power_on_delay_us = 1000,
.reset_us = 5,
};

I think you should completely disable usb5744_data in this case

I would recommend to go the same route as the Linux kernel which has a
special Kconfig option for it, c.f. USB_ONBOARD_DEV_USB5744 which adds
the necessary dependencies.

On Linux kernel, it needs to parse the device node with i2c-bus.

Seems like it does in U-Boot too? c.f. usb5744_i2c_init() with
dev_read_phandle_with_args with i2c-bus?

ok.

We probably should be careful and figure out which boards are actually
using the usb5744 and add this symbol to their defconfig so they don't
regress.

From my initial analysis configs/xilinx_zynqmp_kria_defconfig
which use CONFIG_USB_ONBOARD_HUB
I believe using DM_I2C to guard the code is the best approach.
it should be around .init() callback to guard this to work.

I disagree. The driver clearly supports the USB5744 and the driver is
compiled regardless of DM_I2C support which is required (as far as I
understood, maybe that's incorrect?) for being able to use USB5744? I
don't like hidden dependencies.

Okay, I will add a new configuration option to guard the code as done
in the Linux kernel.

Also, I believe Neil's point is we shouldn't try to probe the USB5744 if
there is no DM_I2C support, simply excluding .init() callback will
probably just render this "support" of USB5744 non-functional?

Let me debug this more, I will update the next version.
I need to check if this binds correctly to the USB hub reset.

Cheers,
Quentin

Thanks
-Anand

Hi Wayne,

On Sun, 11 May 2025 at 01:51, Wayne Schroeder <raz@...> wrote:

Applied patches to 2025.04 release (applied clean). Made stock
defconfig for odroid n2, built, signed, installed. The reboot issue
remains and usb ports are unavailable to linux kernel after reboot.

It is worth noting that linux does see some enumeration of devices
after a while, but even if it logs to the console about device
attachment, the devices don't actually function. This was the same
behavior with the release version.

Thanks for testing these patches.
We need to modify the boot order to probe USB before MMC

diff --git a/include/configs/meson64.h b/include/configs/meson64.h
index f3275b37a51..3ace59880bd 100644
--- a/include/configs/meson64.h
+++ b/include/configs/meson64.h
@@ -119,8 +119,8 @@
#define BOOT_TARGET_DEVICES(func) \
func(ROMUSB, romusb, na) \
func(USB_DFU, usbdfu, na) \
- BOOT_TARGET_MMC(func) \
BOOT_TARGET_DEVICES_USB(func) \
+ BOOT_TARGET_MMC(func) \
BOOT_TARGET_NVME(func) \
BOOT_TARGET_SCSI(func) \
BOOT_TARGET_PXE(func) \

Here is the boot sequence at my end.
------------------------
U-Boot 2025.07-rc1-00223-g704a4d1fd804 (May 12 2025 - 12:15:29 +0530)
odroid-n2/n2-plus

Model: Hardkernel ODROID-N2
SoC: Amlogic Meson G12B (S922X) Revision 29:c (40:2)
DRAM: 1 GiB (total 3.8 GiB)
Core: 402 devices, 30 uclasses, devicetree: separate
MMC: mmc@ffe05000: 0, mmc@ffe07000: 1
Loading Environment from nowhere... OK
In: usbkbd,serial
Out: vidconsole,serial
Err: vidconsole,serial
Board variant: n2-plus
Net: eth0: ethernet@ff3f0000

Hit any key to stop autoboot: 0
starting USB...
Register 3000140 NbrPorts 3
Starting the controller
USB XHCI 1.10
Bus usb@ff500000: 4 USB Device(s) found
scanning usb for storage devices... 1 Storage Device(s) found

Device 0: Vendor: ASMT Rev: 0 Prod: 2115
Type: Hard Disk
Capacity: 114473.4 MB = 111.7 GB (234441648 x 512)
... is now current device
Scanning usb 0:1...
Card did not respond to voltage select! : -110
Cannot persist EFI variables without system partition
Loading Boot0000 'mmc 1' failed
Loading Boot0001 'usb 0' failed
EFI boot manager: Cannot load any image
Card did not respond to voltage select! : -110
switch to partitions #0, OK
mmc1(part 0) is current device
Scanning mmc 1:1...
Found /extlinux/extlinux.conf
Retrieving file: /extlinux/extlinux.conf
1: Manjaro ARM
Retrieving file: /Image
Retrieving file: /initramfs-linux.img
append: initrd=/initramfs-linux.img console=ttyAML0,115200n8
root=PARTUUID=0efc7c36-02 rootwait audit=0 clk_ignore_unused
plymouth.ignore-serial-consoles

Wayne

On Fri, May 9, 2025 at 10:01?AM Wayne Schroeder <raz@...> wrote:

Yes. Will report back how it works on the n2+.


On Fri, May 9, 2025 at 2:08?AM Anand Moon <linux.amoon@...> wrote:

Hi Wayne,

On Fri, 9 May 2025 at 01:25, Wayne Schroeder <raz@...> wrote:

I actually already have a working solution that I documented in my
github repo (dubschro/odroid-n2). All I do is set BOOTCOMMAND to
check/delete a flag file on my boot partition and if it is found, then
reset usb and sleep for 10 seconds. I then have a systemd shutdown
script only on reboot to touch that file.

CONFIG_BOOTCOMMAND="fatrm mmc 0:1 reboot && echo Resetting USB on
reboot... && usb reset && sleep 10; run distro_bootcmd"

It is a hack, but... the system reliably reboots remotely now, and
that's the important part for now.

Wayne

I did not modify this at my end. However, this is not required with
the above fix.

Thanks
-Anand

On Thu, May 8, 2025 at 11:22?AM Ferass El Hafidi
<funderscore@...> wrote:

On Wed May 7, 2025 at 1:03 PM UTC, Wayne Schroeder wrote:

It's hard for me to quantify if usb works in u-boot as I'm not loading
off of any usb storage devices. It certainly never lets me use the
keyboard to abort auto boot, and never has. On reboot, to reliably

You can send bytes via UART to abort boot, if you happen to have a UART
adapter around.

Can you test this series to see if this resolves the issue?



Thanks
-Anand

_______________________________________________
linux-amlogic mailing list
linux-amlogic@...

Hi Quentin,

On Fri, 9 May 2025 at 17:21, Quentin Schulz <quentin.schulz@...> wrote:

Hi Neil, Anand,

On 5/9/25 9:57 AM, Neil Armstrong wrote:

On 09/05/2025 09:02, Anand Moon wrote:

Add conditional compilation for the usb5744_i2c_init() function based
on the CONFIG_DM_I2C configuration, to avoid compilation failure.

CC net/net-common.o
AR net/built-in.o
LDS u-boot.lds
LD u-boot
aarch64-linux-gnu-ld.bfd: common/usb_onboard_hub.o: in function
`usb5744_i2c_init':
/home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/
usb_onboard_hub.c:74:(.text.usb5744_i2c_init+0xfc): undefined
reference to `i2c_get_chip'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:89:(.text.usb5744_i2c_init+0x1a0):
undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:96:(.text.usb5744_i2c_init+0x1c8):
undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-
maineline/common/usb_onboard_hub.c:104:(.text.usb5744_i2c_init+0x1f0):
undefined reference to `dm_i2c_write'
Segmentation fault (core dumped)
make: *** [Makefile:1824: u-boot] Error 139
make: *** Deleting file 'u-boot'

Signed-off-by: Anand Moon <linux.amoon@...>
---
common/usb_onboard_hub.c | 4 ++++
1 file changed, 4 insertions(+)

diff --git a/common/usb_onboard_hub.c b/common/usb_onboard_hub.c
index 1242513c631..92e3f3a92c9 100644
--- a/common/usb_onboard_hub.c
+++ b/common/usb_onboard_hub.c
@@ -30,6 +30,7 @@ struct onboard_hub_data {
int (*init)(struct udevice *dev);
};
+#if CONFIG_IS_ENABLED(DM_I2C)
static int usb5744_i2c_init(struct udevice *dev)
{
/*
@@ -109,6 +110,7 @@ static int usb5744_i2c_init(struct udevice *dev)
return 0;
}
+#endif
int usb_onboard_hub_reset(struct udevice *dev)
{
@@ -222,7 +224,9 @@ static const struct onboard_hub_data usb2514_data = {
};
static const struct onboard_hub_data usb5744_data = {
+#if CONFIG_IS_ENABLED(DM_I2C)
.init = usb5744_i2c_init,
+#endif
.power_on_delay_us = 1000,
.reset_us = 5,
};

I think you should completely disable usb5744_data in this case

I would recommend to go the same route as the Linux kernel which has a
special Kconfig option for it, c.f. USB_ONBOARD_DEV_USB5744 which adds
the necessary dependencies.

On Linux kernel, it needs to parse the device node with i2c-bus.

We probably should be careful and figure out which boards are actually
using the usb5744 and add this symbol to their defconfig so they don't
regress.

From my initial analysis configs/xilinx_zynqmp_kria_defconfig
which use CONFIG_USB_ONBOARD_HUB
I believe using DM_I2C to guard the code is the best approach.
it should be around .init() callback to guard this to work.

Cheers,
Quentin

Thanks
-Anand

I actually already have a working solution that I documented in my
github repo (dubschro/odroid-n2). All I do is set BOOTCOMMAND to
check/delete a flag file on my boot partition and if it is found, then
reset usb and sleep for 10 seconds. I then have a systemd shutdown
script only on reboot to touch that file.

CONFIG_BOOTCOMMAND="fatrm mmc 0:1 reboot && echo Resetting USB on
reboot... && usb reset && sleep 10; run distro_bootcmd"

It is a hack, but... the system reliably reboots remotely now, and
that's the important part for now.

Wayne

On Thu, May 8, 2025 at 11:22?AM Ferass El Hafidi
<funderscore@...> wrote:

On Wed May 7, 2025 at 1:03 PM UTC, Wayne Schroeder wrote:

It's hard for me to quantify if usb works in u-boot as I'm not loading
off of any usb storage devices. It certainly never lets me use the
keyboard to abort auto boot, and never has. On reboot, to reliably

You can send bytes via UART to abort boot, if you happen to have a UART
adapter around.

Hi Neil, Anand,

On 5/9/25 9:57 AM, Neil Armstrong wrote:

On 09/05/2025 09:02, Anand Moon wrote:

Add conditional compilation for the usb5744_i2c_init() function based
on the CONFIG_DM_I2C configuration, to avoid compilation failure.

?? CC????? net/net-common.o
?? AR????? net/built-in.o
?? LDS???? u-boot.lds
?? LD????? u-boot
aarch64-linux-gnu-ld.bfd: common/usb_onboard_hub.o: in function `usb5744_i2c_init':
/home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/ usb_onboard_hub.c:74:(.text.usb5744_i2c_init+0xfc): undefined reference to `i2c_get_chip'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot- maineline/common/usb_onboard_hub.c:89:(.text.usb5744_i2c_init+0x1a0): undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot- maineline/common/usb_onboard_hub.c:96:(.text.usb5744_i2c_init+0x1c8): undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot- maineline/common/usb_onboard_hub.c:104:(.text.usb5744_i2c_init+0x1f0): undefined reference to `dm_i2c_write'
Segmentation fault (core dumped)
make: *** [Makefile:1824: u-boot] Error 139
make: *** Deleting file 'u-boot'

Signed-off-by: Anand Moon <linux.amoon@...>
---
? common/usb_onboard_hub.c | 4 ++++
? 1 file changed, 4 insertions(+)

diff --git a/common/usb_onboard_hub.c b/common/usb_onboard_hub.c
index 1242513c631..92e3f3a92c9 100644
--- a/common/usb_onboard_hub.c
+++ b/common/usb_onboard_hub.c
@@ -30,6 +30,7 @@ struct onboard_hub_data {
????? int (*init)(struct udevice *dev);
? };
+#if CONFIG_IS_ENABLED(DM_I2C)
? static int usb5744_i2c_init(struct udevice *dev)
? {
????? /*
@@ -109,6 +110,7 @@ static int usb5744_i2c_init(struct udevice *dev)
????? return 0;
? }
+#endif
? int usb_onboard_hub_reset(struct udevice *dev)
? {
@@ -222,7 +224,9 @@ static const struct onboard_hub_data usb2514_data = {
? };
? static const struct onboard_hub_data usb5744_data = {
+#if CONFIG_IS_ENABLED(DM_I2C)
????? .init = usb5744_i2c_init,
+#endif
????? .power_on_delay_us = 1000,
????? .reset_us = 5,
? };

I think you should completely disable usb5744_data in this case

I would recommend to go the same route as the Linux kernel which has a special Kconfig option for it, c.f. USB_ONBOARD_DEV_USB5744 which adds the necessary dependencies.

We probably should be careful and figure out which boards are actually using the usb5744 and add this symbol to their defconfig so they don't regress.

Cheers,
Quentin

Hi Neil,

On Fri, 9 May 2025 at 13:27, Neil Armstrong <neil.armstrong@...> wrote:

On 09/05/2025 09:02, Anand Moon wrote:

Add conditional compilation for the usb5744_i2c_init() function based
on the CONFIG_DM_I2C configuration, to avoid compilation failure.

CC net/net-common.o
AR net/built-in.o
LDS u-boot.lds
LD u-boot
aarch64-linux-gnu-ld.bfd: common/usb_onboard_hub.o: in function `usb5744_i2c_init':
/home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/usb_onboard_hub.c:74:(.text.usb5744_i2c_init+0xfc): undefined reference to `i2c_get_chip'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/usb_onboard_hub.c:89:(.text.usb5744_i2c_init+0x1a0): undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/usb_onboard_hub.c:96:(.text.usb5744_i2c_init+0x1c8): undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/usb_onboard_hub.c:104:(.text.usb5744_i2c_init+0x1f0): undefined reference to `dm_i2c_write'
Segmentation fault (core dumped)
make: *** [Makefile:1824: u-boot] Error 139
make: *** Deleting file 'u-boot'

Signed-off-by: Anand Moon <linux.amoon@...>
---
common/usb_onboard_hub.c | 4 ++++
1 file changed, 4 insertions(+)

diff --git a/common/usb_onboard_hub.c b/common/usb_onboard_hub.c
index 1242513c631..92e3f3a92c9 100644
--- a/common/usb_onboard_hub.c
+++ b/common/usb_onboard_hub.c
@@ -30,6 +30,7 @@ struct onboard_hub_data {
int (*init)(struct udevice *dev);
};

+#if CONFIG_IS_ENABLED(DM_I2C)
static int usb5744_i2c_init(struct udevice *dev)
{
/*
@@ -109,6 +110,7 @@ static int usb5744_i2c_init(struct udevice *dev)

return 0;
}
+#endif

int usb_onboard_hub_reset(struct udevice *dev)
{
@@ -222,7 +224,9 @@ static const struct onboard_hub_data usb2514_data = {
};

static const struct onboard_hub_data usb5744_data = {
+#if CONFIG_IS_ENABLED(DM_I2C)
.init = usb5744_i2c_init,
+#endif
.power_on_delay_us = 1000,
.reset_us = 5,
};

I think you should completely disable usb5744_data in this case

Let me check this. I will update it in the next version.

Neil

Thanks
-Anand

Hi Neil,

On Fri, 9 May 2025 at 13:26, Neil Armstrong <neil.armstrong@...> wrote:

s/odorid-c2/odroid-c2/ in the subject

Oops, I will fix this in the next version.

On 09/05/2025 09:02, Anand Moon wrote:

Enable the onboard USB HUB driver to power the
USB_HUB_GL852G-OHG on the Odroid C2.

Signed-off-by: Anand Moon <linux.amoon@...>
---
configs/odroid-c2_defconfig | 1 +
1 file changed, 1 insertion(+)

diff --git a/configs/odroid-c2_defconfig b/configs/odroid-c2_defconfig
index bcaab46f8bc..f8a66a923c9 100644
--- a/configs/odroid-c2_defconfig
+++ b/configs/odroid-c2_defconfig
@@ -59,6 +59,7 @@ CONFIG_SYSINFO_SMBIOS=y
CONFIG_USB=y
CONFIG_USB_DWC2=y
CONFIG_USB_KEYBOARD=y
+CONFIG_USB_ONBOARD_HUB=y
CONFIG_VIDEO=y
# CONFIG_VIDEO_BPP8 is not set
# CONFIG_VIDEO_BPP16 is not set

With that fixed:
Reviewed-by: Neil Armstrong <neil.armstrong@...>

Thanks
-Anand

On 09/05/2025 09:02, Anand Moon wrote:

Enable support for the Genesys GL853G USB2.0 and USB3.1 Hub on
the Odroid N2 and N2-plus. The GL853G driver activates the hub
reset signal, which toggles the GPIO.
Signed-off-by: Anand Moon <linux.amoon@...>
---
=> dm tree
simple_bus 11 [ + ] dwc3-meson-g12a | `-- usb@ffe09000
usb_gadget 0 [ ] dwc2-udc-otg | |-- usb@ff400000
usb 0 [ + ] xhci-dwc3 | `-- usb@ff500000
usb_hub 0 [ + ] usb_hub | `-- usb_hub
usb_hub 1 [ + ] usb_hub | |-- usb_hub
usb_mass_s 0 [ + ] usb_mass_storage | | `-- usb_mass_storage
blk 2 [ + ] usb_storage_blk | | |-- usb_mass_storage.lun0
partition 0 [ + ] blk_partition | | | `-- usb_mass_storage.lun0:1
bootdev 3 [ ] usb_bootdev | | `-- usb_mass_storage.lun0.bootdev
usb_hub 2 [ + ] usb_hub | `-- usb_hub
clk 2 [ + ] fixed_clock |-- xtal-clk
---
common/usb_onboard_hub.c | 7 +++++++
1 file changed, 7 insertions(+)
diff --git a/common/usb_onboard_hub.c b/common/usb_onboard_hub.c
index 39bbc1aefa2..3914e2f773d 100644
--- a/common/usb_onboard_hub.c
+++ b/common/usb_onboard_hub.c
@@ -231,6 +231,10 @@ static const struct onboard_hub_data genesys_gl852g_data = {
.reset_us = 50,
};
+static const struct onboard_hub_data genesys_gl853g_data = {
+ .reset_us = 50,
+};
+
static const struct udevice_id usb_onboard_hub_ids[] = {
/* Use generic usbVID,PID dt-bindings (usb-device.yaml) */
{ .compatible = "usb424,2514", /* USB2514B USB 2.0 */
@@ -244,6 +248,9 @@ static const struct udevice_id usb_onboard_hub_ids[] = {
}, {
.compatible = "usb5e3,610", /* GL852G USB 2.0 */
.data = (ulong)&genesys_gl852g_data,
+ }, {
+ .compatible = "usb5e3,620", /* GL852G USB 3.1 */
+ .data = (ulong)&genesys_gl853g_data,
}
};

Reviewed-by: Neil Armstrong <neil.armstrong@...>

On 09/05/2025 09:02, Anand Moon wrote:

Enable support for the Via labs VL817 USB2.0 and USB3.1 Hub
on the Odroid C4. The Via labs VL817 USB driver activates the
hub reset signal, which toggles the GPIO.
Signed-off-by: Anand Moon <linux.amoon@...>
---
=> dm tree
simple_bus 11 [ + ] dwc3-meson-g12a | `-- usb@ffe09000
usb_gadget 0 [ ] dwc2-udc-otg | |-- usb@ff400000
usb 0 [ + ] xhci-dwc3 | `-- usb@ff500000
usb_hub 0 [ + ] usb_hub | `-- usb_hub
usb_hub 1 [ + ] usb_hub | |-- usb_hub
usb_mass_s 0 [ + ] usb_mass_storage | | `-- usb_mass_storage
blk 2 [ + ] usb_storage_blk | | |-- usb_mass_storage.lun0
partition 0 [ + ] blk_partition | | | |-- usb_mass_storage.lun0:1
partition 1 [ + ] blk_partition | | | |-- usb_mass_storage.lun0:2
partition 2 [ + ] blk_partition | | | |-- usb_mass_storage.lun0:3
partition 3 [ + ] blk_partition | | | `-- usb_mass_storage.lun0:4
bootdev 3 [ ] usb_bootdev | | `-- usb_mass_storage.lun0.bootdev
usb_hub 2 [ + ] usb_hub | `-- usb_hub
---
common/usb_onboard_hub.c | 10 ++++++++++
1 file changed, 10 insertions(+)
diff --git a/common/usb_onboard_hub.c b/common/usb_onboard_hub.c
index 3914e2f773d..1242513c631 100644
--- a/common/usb_onboard_hub.c
+++ b/common/usb_onboard_hub.c
@@ -235,6 +235,10 @@ static const struct onboard_hub_data genesys_gl853g_data = {
.reset_us = 50,
};
+static const struct onboard_hub_data vialab_vl817_data = {
+ .reset_us = 10,
+};
+
static const struct udevice_id usb_onboard_hub_ids[] = {
/* Use generic usbVID,PID dt-bindings (usb-device.yaml) */
{ .compatible = "usb424,2514", /* USB2514B USB 2.0 */
@@ -251,6 +255,12 @@ static const struct udevice_id usb_onboard_hub_ids[] = {
}, {
.compatible = "usb5e3,620", /* GL852G USB 3.1 */
.data = (ulong)&genesys_gl853g_data,
+ }, {
+ .compatible = "usb2109,817", /* Via labs VL817 2.0 */
+ .data = (ulong)&vialab_vl817_data,
+ }, {
+ .compatible = "usb2109,2817", /* Via labs VL817 3.1 */
+ .data = (ulong)&vialab_vl817_data,
}
};

Reviewed-by: Neil Armstrong <neil.armstrong@...>

On 09/05/2025 09:02, Anand Moon wrote:

Add conditional compilation for the usb5744_i2c_init() function based
on the CONFIG_DM_I2C configuration, to avoid compilation failure.
CC net/net-common.o
AR net/built-in.o
LDS u-boot.lds
LD u-boot
aarch64-linux-gnu-ld.bfd: common/usb_onboard_hub.o: in function `usb5744_i2c_init':
/home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/usb_onboard_hub.c:74:(.text.usb5744_i2c_init+0xfc): undefined reference to `i2c_get_chip'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/usb_onboard_hub.c:89:(.text.usb5744_i2c_init+0x1a0): undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/usb_onboard_hub.c:96:(.text.usb5744_i2c_init+0x1c8): undefined reference to `dm_i2c_write'
aarch64-linux-gnu-ld.bfd: /home/amoon/mainline/u-boot/amlogic/u-boot-maineline/common/usb_onboard_hub.c:104:(.text.usb5744_i2c_init+0x1f0): undefined reference to `dm_i2c_write'
Segmentation fault (core dumped)
make: *** [Makefile:1824: u-boot] Error 139
make: *** Deleting file 'u-boot'
Signed-off-by: Anand Moon <linux.amoon@...>
---
common/usb_onboard_hub.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/common/usb_onboard_hub.c b/common/usb_onboard_hub.c
index 1242513c631..92e3f3a92c9 100644
--- a/common/usb_onboard_hub.c
+++ b/common/usb_onboard_hub.c
@@ -30,6 +30,7 @@ struct onboard_hub_data {
int (*init)(struct udevice *dev);
};
+#if CONFIG_IS_ENABLED(DM_I2C)
static int usb5744_i2c_init(struct udevice *dev)
{
/*
@@ -109,6 +110,7 @@ static int usb5744_i2c_init(struct udevice *dev)
return 0;
}
+#endif
int usb_onboard_hub_reset(struct udevice *dev)
{
@@ -222,7 +224,9 @@ static const struct onboard_hub_data usb2514_data = {
};
static const struct onboard_hub_data usb5744_data = {
+#if CONFIG_IS_ENABLED(DM_I2C)
.init = usb5744_i2c_init,
+#endif
.power_on_delay_us = 1000,
.reset_us = 5,
};

I think you should completely disable usb5744_data in this case

Neil

On 09/05/2025 09:02, Anand Moon wrote:

Enable the onboard USB HUB driver to power the
Via labs VL817 USB 3.1 hub on the Odroid C4.
Signed-off-by: Anand Moon <linux.amoon@...>
---
configs/odroid-c4_defconfig | 1 +
1 file changed, 1 insertion(+)
diff --git a/configs/odroid-c4_defconfig b/configs/odroid-c4_defconfig
index cc3d7fc7484..3c5e53828a1 100644
--- a/configs/odroid-c4_defconfig
+++ b/configs/odroid-c4_defconfig
@@ -64,6 +64,7 @@ CONFIG_USB_DWC3=y
# CONFIG_USB_DWC3_GADGET is not set
CONFIG_USB_DWC3_MESON_G12A=y
CONFIG_USB_KEYBOARD=y
+CONFIG_USB_ONBOARD_HUB=y
CONFIG_USB_GADGET=y
CONFIG_USB_GADGET_VENDOR_NUM=0x1b8e
CONFIG_USB_GADGET_PRODUCT_NUM=0xfada

Same as patch 2, with that fixed:
Reviewed-by: Neil Armstrong <neil.armstrong@...>

On 09/05/2025 09:02, Anand Moon wrote:

Enable the onboard USB HUB driver to power the
USB_HUB_GL853G-OHG on the Odroid N2 and Odroid N2-pluss.
Signed-off-by: Anand Moon <linux.amoon@...>
---
configs/odroid-n2_defconfig | 1 +
1 file changed, 1 insertion(+)
diff --git a/configs/odroid-n2_defconfig b/configs/odroid-n2_defconfig
index a8cbaee3c96..1bf687c1e6d 100644
--- a/configs/odroid-n2_defconfig
+++ b/configs/odroid-n2_defconfig
@@ -64,6 +64,7 @@ CONFIG_USB_DWC3=y
# CONFIG_USB_DWC3_GADGET is not set
CONFIG_USB_DWC3_MESON_G12A=y
CONFIG_USB_KEYBOARD=y
+CONFIG_USB_ONBOARD_HUB=y
CONFIG_USB_GADGET=y
CONFIG_USB_GADGET_VENDOR_NUM=0x1b8e
CONFIG_USB_GADGET_PRODUCT_NUM=0xfada

Same as patch 2, with that fixed:
Reviewed-by: Neil Armstrong <neil.armstrong@...>

s/odorid-c2/odroid-c2/ in the subject

On 09/05/2025 09:02, Anand Moon wrote:

Enable the onboard USB HUB driver to power the
USB_HUB_GL852G-OHG on the Odroid C2.
Signed-off-by: Anand Moon <linux.amoon@...>
---
configs/odroid-c2_defconfig | 1 +
1 file changed, 1 insertion(+)
diff --git a/configs/odroid-c2_defconfig b/configs/odroid-c2_defconfig
index bcaab46f8bc..f8a66a923c9 100644
--- a/configs/odroid-c2_defconfig
+++ b/configs/odroid-c2_defconfig
@@ -59,6 +59,7 @@ CONFIG_SYSINFO_SMBIOS=y
CONFIG_USB=y
CONFIG_USB_DWC2=y
CONFIG_USB_KEYBOARD=y
+CONFIG_USB_ONBOARD_HUB=y
CONFIG_VIDEO=y
# CONFIG_VIDEO_BPP8 is not set
# CONFIG_VIDEO_BPP16 is not set

With that fixed:
Reviewed-by: Neil Armstrong <neil.armstrong@...>