Possibility to deactivate HID

[ChrisGCode]

Is it possible to deactivate HID? It think it should be possible to activate/deactivate it like mtp, mass_store, ...

[pelya]

No, it's not possible - I've made it this way to make things simpler for users. To disable it - remove lines

if (hid_enabled) android_enable_function(dev, conf, "hid");

from https://github.com/pelya/android-keyboard-gadget/blob/master/kernel-3.4.patch#L119

Then you will have to execute this command to enable HID, from the root shell:

echo 0 > /sys/class/android_usb/android0/enable echo mtp,adb,hid > /sys/class/android_usb/android0/functions echo 1 > /sys/class/android_usb/android0/enable

On Mon, Sep 22, 2014 at 7:44 PM, ChrisGCode notifications@github.com wrote:

Is it possible to deactivate HID? It think it should be possible to activate/deactivate it like mtp, mass_store, ...

— Reply to this email directly or view it on GitHub https://github.com/pelya/android-keyboard-gadget/issues/9.

[lars18th]

Hi,

I applied the patch for kernel 3.4. Now I found directory "/sys/class/android_usb/android0/f_hid" in my phone, but NO "/dev/hidg0" or "/dev/hidg1".

$ cat /sys/class/android_usb/android0/functions mpt,adb

I need to modify "init.usb.rc" to enable HID-Gadget mode? How I can debug if the proper init funcion is executed?

[pelya]

If you modified the patch, you need to enable HID mode by hand. Until then, there will be no /dev/hidg0 and hidg1. Execute following commands in the root shell:

echo 0 > /sys/class/android_usb/android0/enable

echo mtp,adb,hid > /sys/class/android_usb/android0/functions

echo 1 > /sys/class/android_usb/android0/enable

You habe to do that each time you plug a cable to USB post, because Android system will override HID mode with just 'mtp,adb'. On Jan 28, 2015 12:04 PM, "lars18th" notifications@github.com wrote:

Hi,

I applied the patch for kernel 3.4. Now I found directory "/sys/class/android_usb/android0/f_hid" in my phone, but NO "/dev/hidg0" or "/dev/hidg1".

$ cat /sys/class/android_usb/android0/functions mpt,adb

I need to modify "init.usb.rc" to enable HID-Gadget mode? How I can debug if the proper init funcion is executed?

— Reply to this email directly or view it on GitHub https://github.com/pelya/android-keyboard-gadget/issues/9#issuecomment-71809073 .

[lars18th]

Hi Pelya,

Yes, the patch is using the "hid_enabled" check. See code here: https://github.com/Al3XKOoL/tinno82_S5501/commit/d6ff89b081fbd61ed3ce542f1c886d9180439433#diff-d3d3db76db0b001fddee1bdcfc972c17R1855

But when I execute this script as root:

!/system/bin/sh

echo 0 > /sys/class/android_usb/android0/enable echo mtp,adb,hid > /sys/class/android_usb/android0/functions echo 1 > /sys/class/android_usb/android0/enable

Then the ADB disconnects, no MPT, and no HID devices. When I unplug and I replug device, then Windows identifies it as "Unknown Device". No hardware ID. But, If I change from MTP to PTP, then this restores the USB functions.

Where you think is the trouble?

[alandewayne]

Stop fucking sending me those emails On Jan 28, 2015 3:04 AM, "lars18th" notifications@github.com wrote:

Hi,

I applied the patch for kernel 3.4. Now I found directory "/sys/class/android_usb/android0/f_hid" in my phone, but NO "/dev/hidg0" or "/dev/hidg1".

$ cat /sys/class/android_usb/android0/functions mpt,adb

I need to modify "init.usb.rc" to enable HID-Gadget mode? How I can debug if the proper init funcion is executed?

— Reply to this email directly or view it on GitHub https://github.com/pelya/android-keyboard-gadget/issues/9#issuecomment-71809073 .

[pelya]

@alandewayne you are watching this repository, that's why you are getting all these emails - https://github.com/pelya/android-keyboard-gadget/watchers Press 'unwatch' button at this webpage to stop receiving them. On Jan 28, 2015 5:12 PM, "alandewayne" notifications@github.com wrote:

Stop fucking sending me those emails On Jan 28, 2015 3:04 AM, "lars18th" notifications@github.com wrote:

Hi,

I applied the patch for kernel 3.4. Now I found directory "/sys/class/android_usb/android0/f_hid" in my phone, but NO "/dev/hidg0" or "/dev/hidg1".

$ cat /sys/class/android_usb/android0/functions mpt,adb

I need to modify "init.usb.rc" to enable HID-Gadget mode? How I can debug if the proper init funcion is executed?

— Reply to this email directly or view it on GitHub < https://github.com/pelya/android-keyboard-gadget/issues/9#issuecomment-71809073>

.

— Reply to this email directly or view it on GitHub https://github.com/pelya/android-keyboard-gadget/issues/9#issuecomment-71850349 .

[pelya]

@lars18th replace lines

if (hid_enabled) android_enable_function(dev, "hid");

with just

android_enable_function(dev, "hid");

Then it will always be enabled. I've removed that 'hid_enabled' check some time ago from all my patches, but the old version of the patch still flies around the net.

[lars18th]

Hi Pelya,

With current version of the kernel (without the line "if (hid_enabled)") the HID-Gadget patch is working: I can send keystrokes using the tool hid-gadget-test.

However, I have found a problem with the current patch. I comment the problem:

• By default, the USB mode in my Android device (with the patch applied) is: mpt,adb,hid. Because I need to connect the Android device to a very buggy USB device, I like to enable ONLY keyboard and mouse emulation over the USB connection. Then I execute these commands from the shell (as root):

$ echo 0 > /sys/class/android_usb/android0/enable $ echo hid > /sys/class/android_usb/android0/functions $ echo 1 > /sys/class/android_usb/android0/enable

And when I execute the last line the Android device freezes. Only with a hard-reboot (remove battery) I can restore it.

You know where is the problem? Thank you for your time!

[pelya]

No, I never tried to use it this way. Looks like you'll need to add printk() to each function, and trace where it freezes. Or you can try to debug it - http://stackoverflow.com/questions/17477258/how-to-use-gdbclient-to-debug-the-kernel-on-nexus-7

On Tue, Feb 3, 2015 at 11:09 AM, lars18th notifications@github.com wrote:

Hi Pelya,

With current version of the kernel (without the line "if (hid_enabled)") the HID-Gadget patch is working: I can send keystrokes using the tool hid-gadget-test.

However, I have found a problem with the current patch. I comment the problem:

• By default, the USB mode in my Android device (with the patch applied) is: mpt,adb,hid. Because I need to connect the Android device to a very buggy USB device, I like to enable ONLY keyboard and mouse emulation over the USB connection. Then I execute these commands from the shell (as root):

$ echo 0 > /sys/class/android_usb/android0/enable $ echo hid > /sys/class/android_usb/android0/functions $ echo 1 > /sys/class/android_usb/android0/enable

And when I execute the last line the Android device freezes. Only with a hard-reboot (remove battery) I can restore it.

You know where is the problem? Thank you for your time!

— Reply to this email directly or view it on GitHub https://github.com/pelya/android-keyboard-gadget/issues/9#issuecomment-72616651 .

[pelya]

Maybe you are initializing HID module twice? There is already one explicit initialization inside enable_functions(), and when you do

echo hid > /sys/class/android_usb/android0/functions

there is another initialization. On Feb 3, 2015 12:28 PM, "Sergii Pylypenko" x.pelya.x@gmail.com wrote:

No, I never tried to use it this way. Looks like you'll need to add printk() to each function, and trace where it freezes. Or you can try to debug it - http://stackoverflow.com/questions/17477258/how-to-use-gdbclient-to-debug-the-kernel-on-nexus-7

On Tue, Feb 3, 2015 at 11:09 AM, lars18th notifications@github.com wrote:

Hi Pelya,

With current version of the kernel (without the line "if (hid_enabled)") the HID-Gadget patch is working: I can send keystrokes using the tool hid-gadget-test.

However, I have found a problem with the current patch. I comment the problem:

• By default, the USB mode in my Android device (with the patch applied) is: mpt,adb,hid. Because I need to connect the Android device to a very buggy USB device, I like to enable ONLY keyboard and mouse emulation over the USB connection. Then I execute these commands from the shell (as root):

$ echo 0 > /sys/class/android_usb/android0/enable $ echo hid > /sys/class/android_usb/android0/functions $ echo 1 > /sys/class/android_usb/android0/enable

And when I execute the last line the Android device freezes. Only with a hard-reboot (remove battery) I can restore it.

You know where is the problem? Thank you for your time!

— Reply to this email directly or view it on GitHub https://github.com/pelya/android-keyboard-gadget/issues/9#issuecomment-72616651 .

[lars18th]

Hi Pelya,

After weeks using a kernel with your patch applied, I found that in some cases the kernel hangs when changing the USB mode. I suggest that you reimplement a safe mode for enable/disable HID mode. Perhaps this can be complex, but it minimizes the cases of a crash.

Regards!

[conrad-heimbold]

Answer has been given: to deactivate HID;

1. compile with android_enable_function() commented out
2. Execute in root shell (replace $FUNCTION with what you would like to have): echo 0 > /sys/class/android_usb/android0/enable echo $FUNCTION > /sys/class/android_usb/android0/functions echo 1 > /sys/class/android_usb/android0/enable

Possible functions are adb, mtp, midi, ptp, charging, ffs, hid; seperated with a comma (,) and no whitespace.