USB Ethernet Gadget "Error inserting g_ether: No such device"

Hello Victor Mayoral,

 As you said you have compiled USB as g_ether.ko module but have you copied it to SDCARD of FLASH,

For OMAP processors after compiling modules following is the command to copy modules into SDCARD
  

  sudo make ARCH=arm modules_install INSTALL_MOD_PATH=/mnt/sdcard
    <assuming his SD card is mounted in this path /mnt/sdcard>

Regards,

Rajeev.

Dear Rajeev,

Thanks for your comment. Yes i've done it that.

And also tried 10s of combinations at menuconfig all so that i could use the Ethernet gadget over USB. Still no luck.

Right now i'm trying to go with the last release of the linux-davinci version from the Arago git repositories...

Tough road this one!!

Cheers!

Can you post your Linux kernel .config file?

Steve K.

Here it is Steve

1462.config.txt

Uploaded again

Hello,

I'd be interested if you could find any solution to this problem?

Thanks,

Istvan

Hi Istvan,

I'm sorry to say that i could not find the solution. I tried tenths of different settings.

Good luck with it.

Thanks for the feedback Victor..

Would be nice though, if somebody from TI could look into that...

Istvan

Can you please post the output of command 'cat /proc/driver/musb_hdrc'?

I think I have figured it out meanwhile. At least, one step further.

I check /proc/driver/musb_hdrc, but it was not even existing.

After couple of added kernel debug prints, I noticed that the call to da8xx_register_usb20 is actually missing from board-omapl138-hawk.c (I'm using an L138 LCDK board).

So I added this into omapl138_hawk_init, just before the existing omapl138_hawk_usb_init() call.

Now it looks much better.

lsusb lists the device, it's also present in /proc/driver, etc.

Still not perfect thought, I receive 

VBUS_ERROR in a_wait_vrise (80, <SessEnd), retry #3, port1 00080108

when I plug in a wifi dongle, but it's much closer now :)

Good progress!

Which kernel package do you use? I am kinda surprised the board.c does not get MUSB initialized properly, if it is TI PSP package.

I don't know much about all the different boards based on omapl138, but in general VBUS_ERROR means during USB device enumeration, the VBUS voltage dropped below the valid threshold. You could check the board schematics to make sure the VBUS line has a minimum 120uF capacitor, which is required in USB Specs.

You can also try plugging in different USB devices, like mouse, thumb drive to see if VBUS_ERROR happens all the time.

Thanks :)

I'm using DaVinci-PSP-SDK-03.22.00.02. Actually, I'm a bit surprised too about this, but anyway, that's not the first problem with this, I also had to make some changes for mmc/sd to work (http://e2e.ti.com/support/embedded/linux/f/354/t/245170.aspx). 

I actually went a step further and also merged some more usb20 init to the board support file, adding:

cfgchip2 &= ~CFGCHIP2_OTGMODE;
cfgchip2 |= CFGCHIP2_SESENDEN | CFGCHIP2_VBDTCTEN;

With that it could detect the wifi dongle, now I need to get rid of the debug prints I added.

Again, thanks for the tip!

Do you mean adding cfgchip2 changes makes VBUS_ERROR go away?

I think so. One of the other board support files in mach-davinci has some comment about this, if I remember correctly.

Now I got rid of the printk's and after a clean boot I could use my wifi dongle over the OTG port :)

nice job!

Hi

Me too facing same issue but you have solved the issue.

I'm trying to enable the USB2.0 OTG in J2 and i modified as follows

static __init void omapl138_hawk_usb_init(void)
{
int ret;
u32 cfgchip2;

...

ret = da8xx_register_usb11(&omapl138_hawk_usb11_pdata);
if (ret) {
pr_warning("%s: USB 1.1 registration failed: %d\n",
__func__, ret);
goto usb11_setup_fail;
}

//USB2.0 OTG start

/*
* We have to override VBUS/ID signals when MUSB is configured into the
* host-only mode -- ID pin will float if no cable is connected, so the
* controller won't be able to drive VBUS thinking that it's a B-device.
* Otherwise, we want to use the OTG mode and enable VBUS comparators.
*/

#ifdef CONFIG_USB_MUSB_HOST
cfgchip2 |= CFGCHIP2_FORCE_HOST;
#else
cfgchip2 |= CFGCHIP2_SESENDEN | CFGCHIP2_VBDTCTEN;
#endif

cfgchip2 &= ~CFGCHIP2_OTGMODE;
cfgchip2 |= CFGCHIP2_SESENDEN | CFGCHIP2_VBDTCTEN;

pr_info("USB 2.0 OTG initialization..!!");

ret = davinci_cfg_reg(DA830_USB0_DRVVBUS);
if (ret)
pr_warning("%s: USB 2.0 PinMux setup failed: %d\n",
__func__, ret);
else {
/*
* TPS2065 switch @ 5V supplies 1 A (sustains 1.5 A),
* with the power on to power good time of 3 ms.
*/
ret = da8xx_register_usb20(1000, 3);
if (ret)
pr_warning("%s: USB 2.0 registration failed: %d\n",
__func__, ret);
}
//USB2.0 OTG end 
return;

usb11_setup_fail:
gpio_free(DA850_USB1_OC_PIN);
usb11_setup_oc_fail:
gpio_free(DA850_USB1_VBUS_PIN);
}

and configuration file modification is

#
# USB Host Controller Drivers
#

CONFIG_USB_MUSB_HDRC=y
CONFIG_USB_MUSB_DA8XX=y
CONFIG_MUSB_PIO_ONLY=y

#
# USB Miscellaneous drivers
#

CONFIG_USB_GADGET=y
CONFIG_USB_GADGET_VBUS_DRAW=2
CONFIG_USB_GADGET_DUALSPEED=y

#
# OTG and related infrastructure
#
CONFIG_USB_OTG_UTILS=y
# CONFIG_USB_GPIO_VBUS is not set
# CONFIG_USB_ULPI is not set
# CONFIG_NOP_USB_XCEIV is not set

got built success fully and i got this error

da8xx_musb_interrupt 350: VBUS error workaround (delay coming)
da8xx_musb_interrupt 350: VBUS error workaround (delay coming)

root@omapl138-lcdk:/proc/driver# ls
rtc snd-page-alloc
root@omapl138-lcdk:/proc/driver# lsusb
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 002 Device 001: ID 1d6b:0001 Linux Foundation 1.1 root hub

there is no node for musb_hrdc. 

I dono what i missed and please let me know the issue and how to fix it.

Here i've attached log file and defconfig file.

7455.omapl138_lcdk_defconfig.txt

2772.dmesg.txt

Thanks

Sangly

Dear,

me too solved the same issue in adding the additional

cfgchip2 |= CFGCHIP2_SESENDEN | CFGCHIP2_VBDTCTEN;

even if in host only mode.

I am in kernel 3.17.0, now usb works, but note, key must be inserted at boot, an will not be detected anymore at a new re-insertion.

Without the line above, instead, detection of the usb key works properly, but  the following error doesn't allow to access the media:

VBUS_ERROR in a_wait_vrise (80, <SessEnd), retry #3, port1 00080108

Now, since the same kernel and board support config are the same of kernel 3.5.1, i am expecting a final clarification about why this drivers are not working properly anymore. From 3.5.1 to 3.17 there were some changes, one is to put CONFIG_USB_MUSB_DA8XX not selectable anymore (broken) in Kconfig (broken need to be removed to allow usb to work).

Felipe Balbi <balbi@ti.com> seems to be a contributor of last patches, that maybe can clarify.

Regards