Adb isn't. Finding my kindle fire 6.2.2 NEED HELP IM TRYING TO ROOT IT
Did you set up the device drivers as outlined in this?
http://forum.xda-developers.com/showpost.php?p=22145030&postcount=6
Adb still says its offline
When you go into device manager, what does it display it as?
It says Android Composite Adb Interface
Th3-Chronikk said:
It says Android Composite Adb Interface
Click to expand...
Click to collapse
This is what I recommend: if you installed Android SDK, go ahead and uninstall it. It's only an absolute necessity if you're a developer or running it a virtual environment. Clean all traces of any part of Android drivers or leftovers files and folders. Next, uninstall the drivers from Device Manager under Computer or My Computer. Reboot your PC or laptop.
Next, download vashypoo's Kindle Fire Utility, available in the Kindle Fire's Android Development forum, and extract it. Inside of the extracted folder should be a .bat file to install drivers. Right click, select Run As Administrator, and follow the on-screen instructions. Once it's completed, reboot.
Inside one of the KFU folders (I can't remember if it's files or tools) is ADB. Right click while pressing Shift on your PC keyboard, open a command window, and type the following:
adb devices
If all goes well, your Fire should be listed. These are the steps I followed when I ran into this exact same issue. I hope it works!
Sent from my Kindle Fire using xda premium
Thats what I did I truer both the SDK method and the kindle fire utility my kindle fire isn't recognized by either one
Does anyone know what I can do
did you try to cold-reset ? hold pwr ~20sec and pwr on again
sometimes restarting the computer helps ...
Yes tryed that and still no luck
the driver (android composite adb interface) is the right one
did you check the adb_usb.ini file in .android folder in your personal folder ?
ist should include 0x1949 & 0x18d1 - if not add in a new line
restarting adb : adb kill-server
check again: adb devices
I did that already
sorry i'm really running out of ideas at this point ...
adb relies on:
adb_usb.ini in .android folder with the correct entries:
0x1949 & 0x18d1
drivers with the correct entries in android_winusb.inf (drivers of kfu do):
;Kindle Fire
%SingleAdbInterface% = USB_Install, USB\VID_1949&PID_0006
%CompositeAdbInterface% = USB_Install, USB\VID_1949&PID_0006&MI_01
%SingleAdbInterface% = USB_Install, USB\VID_18D1&PID_0100
%CompositeAdbInterface% = USB_Install, USB\VID_18D1&PID_0100&MI_01
in the 32 and 64 bit section
Did that.... still not recrecognized
Have you tried a different usb port? My Fire is finocky about which port i use.
Sent from my Kindle Fire using Tapatalk
WhoRuJudge said:
Have you tried a different usb port? My Fire is finocky about which port i use.
Sent from my Kindle Fire using Tapatalk
Click to expand...
Click to collapse
Yes I did no luck
This is before doing anything to it yet? (just to make sure we're on the same page)
Does it show up as an external drive in Windows? Windows, right?
Seems possible to me that if you haven't done anything at all yet and it doesn't show up in Windows you might have a bad USB cable or port on your device.
If you haven't done anything yet, have you tried calling Amazon for troubleshooting / possible replacement?
(Sorry, I'm just trying to cover the things that haven't been asked.)
gadgetman13 said:
This is before doing anything to it yet? (just to make sure we're on the same page)
Does it show up as an external drive in Windows? Windows, right?
Seems possible to me that if you haven't done anything at all yet and it doesn't show up in Windows you might have a bad USB cable or port on your device.
If you haven't done anything yet, have you tried calling Amazon for troubleshooting / possible replacement?
(Sorry, I'm just trying to cover the things that haven't been asked.)
Click to expand...
Click to collapse
It is recognized by windows as a storage device but what I'm saying is that it wouldnt show up on Adb even after I added the kindle fire driver
Use the app in unlockroot.com you won't need driver
Sent from my LG-P500 using xda premium
I was to B&N today, read few books in store (on nook)
Came home, tried to connect:
adb unable to connect...
run dropbear - unable to ssh
rebooted nook via terminal
adb unable to connect...
Scratched head…
rebooted nook again via terminal
adb unable to connect...
Did basic adb troubleshooting:
setprop service.adb.tcp.port 5555
stop adbd
start adbd
adb unable to connect...
stop adbd
adbd &
adb unable to connect...
netstat
showed listening (with adbd running)
127.0.0.1:5037
0.0.0.0:5555
rebooted nook via terminal
rebooted PC…
adb unable to connect...
put noogie SD, reboot, check uRamdisk – same size & timestamp
removed noogie, off/on - connected right away!
I’m puzzled…
I saw few times, after 2-3 days of reading adb is unable to connect, but reboot fixed it always.
What was different this time?
“Cold reboot”?
I spend like 30min and feel like an idiot right now…
I was unable to connect over the usb port as well. Without too long investigation I've used adbwireless app on NT and was able to connect.
You might want to refresh server
adb kill-server
adb start-server
then connect using your NT ip address:5555 and you should be done.
In my case, I think, there are too many conflicting drivers installed and NT is not being recognized when connected.
Good luck
Adapt0r said:
I was unable to connect over the usb port as well. Without too long investigation I've used adbwireless app on NT and was able to connect.
You might want to refresh server
adb kill-server
adb start-server
then connect using your NT ip address:5555 and you should be done.
In my case, I think, there are too many conflicting drivers installed and NT is not being recognized when connected.
Good luck
Click to expand...
Click to collapse
I did it as well before every adb connect, thought it not worth to mention.
Naturally, "rebooting PC" should reboot adb server too, right?
But it didn't help either...
If you see this issue again, just try toggling ADB Wireless On/Off and it should reset anything on the devices side. Also I've noticed that sometimes and completely random from what I can tell "USB Debugging" will get unchecked, which will mess with ADB access via USB.
Disabling/Enabling ADBwireless sometimes may help.
Maybe a stupid question: Do you have a firewall running on your pc?
digiflash said:
Disabling/Enabling ADBwireless sometimes may help.
Click to expand...
Click to collapse
won't it be the same as below or it does more?
Code:
setprop service.adb.tcp.port 5555
stop adbd
start adbd
digiflash said:
Maybe a stupid question: Do you have a firewall running on your pc?
Click to expand...
Click to collapse
Always
I did check it too. It was nothing worth to mention.
abd attempted to connect - got nothing back (timeout)
on NST adbd was listening, I was able to access gmail.
I can ssh neither, and dropbear was listening too.
Looks, in fact, like firewall on NST was blocking all incoming connection.
That's why I wrote "Scratched head…" - which means I checked pretty much everything.
Could somebody explain, what is the major difference for NST between:
reboot (hot reboot) & shutdown/start (cold reboot)
Or more precisely for the latter:
insert noogie SD/reboot/remove noogie SD/turn off/turn on
ApokrifX said:
insert noogie SD/reboot/remove noogie SD/turn off/turn on
Click to expand...
Click to collapse
Cannot connect today at all. Cold reboot doesn’t help anymore...
Did some more tests (including ADBwireless):
I cannot ping [nook IP] from my computer
I can ping [nook IP] from router.
From PC
>telnet [nook IP] 5555
Connecting To [nook IP]...Could not open connection to the host, on port 5555: Connect failed
From router:
telnet [nook IP] 5555
Connection didn’t timeout (i.e. it does connect!)
Again, looks like something on Nook blocking connections from all IP, except from default gateway (i.e. router)
Question: Does rooting process unblock connections from other IPs?
About Disabling/Enabling ADBwireless:
Before running ADBwireless
netstat
127.0.0.1:5037
0.0.0.0:5555
netstat -tapn doesn’t work
stop adbd
both above disappear from netstat
start adbd
both show up
run ADBwireless
get message: ADBwireless is off
pushed button, message changed to: adb connect [nook IP]:5555
(adb connct didn’t work)
Switch to terminal,
netstat
127.0.0.1:5037
I.e. no more: 0.0.0.0:5555
Subsequent "stop adbd"/"start adbd" changes nothing:
netstat
127.0.0.1:5037
I was wondering what ADBwireless supposed to do then?
Could somebody check from his nook what netstat shows, please?
Nook can connect to internet (gmail, market) + netstat shows all outgoing connections…
What else can I try?
I guess, I can switch to adb usb…
It might be helpful if you tell what OS you are running on your PC and what did you do with NT. If we can reproduce your case then we can troubleshoot it otherwise you will hear useless guesses and speculations.
I like NT as a reader and nothing else. Web and games are looking awful on it.
Adapt0r said:
It might be helpful if you tell what OS you are running on your PC and what did you do with NT. If we can reproduce your case then we can troubleshoot it otherwise you will hear useless guesses and speculations.
I like NT as a reader and nothing else. Web and games are looking awful on it.
Click to expand...
Click to collapse
One box is XP with Outpost firewall (abd set to full trust).
Another win7 (FW policy – allow outgoing connections + few exceptions).
I’m Win developer/admin and Linux(CentOs) admin.
I’m complete noob when it comes to Linux non-shell related programming.
Router runs DD-WRT.
As of now I switched to uRamdisk_1_1_rooted_usb.
When I run dropbear, I see it listening on 22, but cannot ssh still.
That’s all I guess…
Could you do one test for me please?
If your NST runs uRamdisk_1_1_rooted_wifi:
Could you ping you NST from you box
If not - could you do same from your router/AP?
1. I don't have uRamdisk_1_1_rooted_wifi on my NT and can't test it.
2. adbWireless works on my Win 7 and Mac the same way. Both comps have firewalls running and both are on the same network. I can connect and ping NT.
Just for clarification I will describe how it works on my setup:
- I start adbWireless on NT,
- it asks for root permission if it for the first time and I grant it
- then shows fat button on the screen and says adbWireless is off
- I push the button
- it says adbWireless is on, from your computer run
adb connect 192.168.15.112:5555
At this point I can ping IP address and run adb shell.
I've rooted NT for purposes of software testing and customization. I did it twice with older and newer Nooter and did not find any noticeable differences.
Let me know if you want some other tests or info.
Adapt0r said:
I can connect and ping NT.
…
At this point I can ping IP address and run adb shell.
Click to expand...
Click to collapse
To clarify (trying to get a setup as simple as possible):
We connected NST to our home wireless network and it gets IP.
We didn’t touch adbWireless yet!
Now do two tests, please:
1. From NST we ping gateway or "another device on same network", (providing it let us).
2. From "another device on same network", ping NST.
What do you get?
I thought, adbWireless just configure adbd for IP connection (as opposite for USB one) and starts it - nothing else.
If #1 works, but not #2,
and #2 starts working after you “hit adbWireless button”, than adbWireless does more then I wrote above…
Sorry, I didn't have time to do your test at home. I'll try today.
I think you are correct, adbWireless does configuration of connection by executing something like this:
Turning On:
setprop service.adb.tcp.port 5555
stop adbd
start adbd
Turning Off wifi and start listening on usb
setprop service.adb.tcp.port -1
stop adbd
start adbd
Cheers!
Edit:
Ping NST_ip_address from PC works fine.
Since I don't have terminal installed I could not test reverse pinging.
Ouch…
Just figured out:
1: I press “big red button” in adbWireless.
2: I push back button
And it pushed adbWireless into background.
If it’s the same as turning it (adbWireless) off,
it’s possible, that adbWireless reset adbd service to use USB.
Thus there is nothing listening on 5555 when I run netstat in terminal on NST.
Who’s with me?
Should be east to test:
press “big red button” in adbWireless.
I push back button
try to connect
ApokrifX said:
Ouch…
Click to expand...
Click to collapse
Updated router firmware - I can ping nook and ADB over wireless works too.
Not sure thought, if it was wireless channel problem.
I've read somewhere, some NST firmwares doesn't work on high channels or certain combined modes - like works on A+B, but not A+G
Apparently, NST firmware 1.1.2 was targeted to solve some of these issues.
PROBLEM SOLVED.
I didn't have the proper folder for the adb.exe.
-----
I have a similar problem: I recently rooted my nook simple touch with touchnooter 2.1.31, and all seems well but I can't connect to the adb wireless. IP is 192.168.0.6:5555 according to adb; I can ping from my Win XP computer if I use 192.168.0.6, but not if I add the 5555, but I just can't connect to the ereader. Wireless connections seem fine otherwise. I've tried shutting off and one the wireless adb.
I'm a complete beginner in this. On the reader, the screen has this:
"from your computer run: adb connect 192.168.0.6:5555"
This suggests that I can't simply use a firefox browser to connect? I tried using the windows cmd box, but windows indicates it doesn't know what 'adb' is. Then I found a website about adbwireless: http://theunlockr.com/2011/04/12/how-to-connect-to-your-android-device-through-adb-wirelessly/
and installed Adroid SDK Tools as indicated. When I use the cmd window, go to the proper sdk tools subfolder, and type 'adb connect 192.168.0.6:5555', windows still insists it doesn't recognize 'adb'
Please help!
try adb.exe
Meter 13,
Thanks for the reply. I added a pre-script to my op. I had created a bat file to move to the proper folder but had realized that the adb.exe had been moved to the platform-tools subfolder from its old location in the tools folder. Once I realized this, all worked well - connected to my nook, and used adb to install an apk file.
Bob
Just a note:
Ping isn't the usual TCP/UDP stuff, so there are no (port) numbers on the end.
ping 192.168.1.27
ADB uses TCP and a port number but it usually defaults to 5555
adb connect 192.168.1.27
adb connect 192.168.1.27:5555
You can also telnet to check if you can connect to the nook via TCP and port 5555
telnet 192.168.1.27 5555
(Note: If this works it will tell you that you are connected, but you can't do anything because ADB is a binary protocol.)
I had the same problem, turns out the correct drivers weren't installed. Ive downloaded HTC sync from HTC.com. I installed it and removed HTC sync afterwards. The drivers should stay there. Now reconnect your phone and try "adb devices" Before ADB wouldn't recognize my device but now it does. Tried to "adb backup -all" again, and it worked instantly.
Maybe this solves the problem?
You could try going back to a stock /boot partition.
I run my nstg rooted with the addition of about 8 files (su, xbin, busybox in system, adb wireless in /data plus a launcher) when I first root.
with the nstg, at least, I don't actually need the modified kernels - adb connects from my computers to the nstg, I simply don't have a root shell when I first connect.
This is fixed by doing an
su
once I'm in
Might be worth trying a restore of just partion 1 from your backup?
Right, I don't know what kind of sorcery my nexus 7 tablet is employing, but it simply isn't getting the ADB treatment.
I've tried installing Samsung drivers, Naked drivers, all the drivers, bleh, test mode- the whole kaboodle and nothing's working.
ADB isn't recognizing my nexus 7, and I have absolutely no idea why.
I thought maybe it'd have something to do with these specs, maybe?
OS: Windows 8 Pro x64
ROM: ParanoidAndroid 3.1 (Newest, formerly 2.54 but still same problem)
Recovery: TWRP
MTP and PTP are offered at connection. Debugging is available on Nexus 7, but computer seems entirely oblivious to its existence as an ADB device. Device Manager lists Nexus as 'Android ADB Composite Device' after I installed Google's SDK Drivers
Ask questions away
Thanks
A lot of users have luck by using toolkits that install the drivers for you. I simply used the drivers from the Asus website on my Win 8 x64 laptop.
Sent from my SGH-I777 using xda premium
korockinout13 said:
A lot of users have luck by using toolkits that install the drivers for you. I simply used the drivers from the Asus website on my Win 8 x64 laptop.
Sent from my SGH-I777 using xda premium
Click to expand...
Click to collapse
Yep, that's the problem:
I've used:
1) WugFresh drivers
2) ASUS drivers
3) SDK Drivers
4) Whatever else- drivers there are on those toolkits
Three phrases that should never be used on XDA are:
a) "Nothing's working"
b) "I've tried everything"
c) "It doesn't work"
Why? Because they don't really provide any useful details to a reader that might otherwise be able to help.
Look - You said that you observed "ADB Composite Interface" in you device manager, and yet you said NOTHING about whether Windows noted that the device was not working correctly.
At a minimum, that means that you DO (or did) have a driver successfully installed, and moreover, the device on the far end of the cable is properly communicating and asserting a VID/PID hardware identifier that selects the installed Windows driver that is supposed to handle that device. Otherwise you would see nothing or an unknown device in the device manager.
What your OP does not mention is what exactly the symptoms of "doesn't work" actually are: are you saying that "running the command 'adb devices' from the command line produces an 'offline' message"?
Also, you failed to mention whether you are trying to communicate via adb to the regular OS or the custom recovery. It makes a difference because most windows wingnuts end up installing TWO drivers for adb - the custom recoveries claim a completely different VID/PID device identifier than what is used by the OS, and the unmodified Asus/Google drivers don't claim the custom recovery VID/PID hardware ID.
One other thing to note: with jb4.2.2, if /default.prop has
ro.adb.secure=1
then you need to have the screen unlocked when you fire up adb on the PC so that you can accept the connection from the tablet side, otherwise it will report an "offline" device. (This only applies to the OS, not the recoveries. As I noted, you didn't' say which & I don't know what PA does in its ramdisk anyway.)
good luck
Hi. Simple solution for this. Uninstall all the nexus drivers you have and reboot your machine, then go here:
http://support.asus.com/Download.aspx?SLanguage=en&m=Nexus+7&p=28&s=2
download the driver software and reinstall. That should sort you out
Thanks for the response
Well, to clear things up...
I mean, the Nexus 7 says 'USB Debugging connected' when connected.
adb devices displays 'List of devices attached' as nothing, and adb reboot gives 'error: device not found'.
Hardware/PIDs for 'Android Composite ADB Interface', (comes up on Device Manager when Nexus 7 plugged in)
USB\VID_18D1&PID_4E42&REV_9999&MI_01
USB\VID_18D1&PID_4E42&MI_01
General Tab for ADB Interface properties says Device Type: Android Device, Status 'Is working properly'
Opening up recovery yields yet the same problem on ADB- nothing is found.
Plugging in my ZTE Skate with CM7= adb devices picks IT up perfectly and i can do whatever I want with it.
The result is exactly the same on my Windows 7 laptop. The adb I'm trying to communicate with is a Windows 8 x64 'Test Mode' System
MTP/PTP works fine. USB OTG with NTFS is a total failure and doesn't detect
I'm trying to communicate with the normal system OS. But both TWRP and OS are undetected by adb devices
Edit: I've never needed to 'accept' anything from the tablet side- that's odd.I've tried adding certain PIDs to the Google drivers .ini and unin/reinstalling the drivers in vain (I'm unfortunately not exactly sure what those PIDs should look like stuck in the .ini driver file anyway, sorry)
Edit 2: @BluDRed - just tried that again for good measure- still nothing.. thanks anyway
Edit 3: THIS GUIDE does not work > http://zacktutorials.blogspot.ca/2012/08/nexus7-android-development.html
tigrey said:
Hardware/PIDs for 'Android Composite ADB Interface', (comes up on Device Manager when Nexus 7 plugged in)
USB\VID_18D1&PID_4E42&REV_9999&MI_01
USB\VID_18D1&PID_4E42&MI_01
Click to expand...
Click to collapse
Last time I went through all the mode combinations was on 4.2.1 (see the VID/PID listing at the end of the first post in this thread); judging from that info the above sort of looks like ADB mode without either MTP or PTP turned on. Don't know if it is important or not, but I have noticed reports in here with slightly different results than mine - some folks saying "I couldn't get MTP to work unless ADB was turned off" ... my result was the exact opposite! But frankly, I don't remember anyone having the reverse problem where they couldn't get ADB to work and solved it by fooling with MTP/PTP toggling.
tigrey said:
Opening up recovery yields yet the same problem on ADB- nothing is found.
Click to expand...
Click to collapse
Both TWRP and CWM should show up at USB\VID_18D1&PID_D001. So the driver that loads for "adb" from the custom recovery will be different (or "unknown device") unless the .inf file in the driver install kit from Asus/Google claims this VID/PID pair... or you have a completely separate driver installed.
tigrey said:
MTP/PTP works fine.
Click to expand...
Click to collapse
This seems like a very important observation/clue. MTP seems to work extremely well on my Win 7 x64 box - transferring big files at something like 80-90 Mbits/sec! If your pair of devices is capable of those kinds of speeds, it certainly seems to rule out a scenario where the hardware link succeeds at low speed (and thus gets the driver properly set up), but then later screws up after the link transitions to high speed.
tigrey said:
Edit: I've never needed to 'accept' anything from the tablet side- that's odd.I've tried adding certain PIDs to the Google drivers .ini and unin/reinstalling the drivers in vain (I'm unfortunately not exactly sure what those PIDs should look like stuck in the .ini driver file anyway, sorry)
Click to expand...
Click to collapse
That security dialog is new in 4.2.2 stock. As I mentioned in my previous post, you might not notice it in a custom ROM (even if based on 4.2.2) if the ROM dev toggled off that adb security property.
Tell you what - typing detailed posts on my N7 is lame. I'll post this now, and then add some further comments using a real keyboard in a couple minutes...
OK, this isn't meant to be a prescription - more like a recounting of experiences to see if perhaps you get some further ideas.
When I first got my N7 in early January, I already had on my Win7 Pro x64 box the Google SDK, and (iirc) I had a variety of ADB/fastboot drivers installed for a different Android devices. (HTC and Samsung)
Fired it up, and of course no communication to the N7 with either adb or fastboot. So I updated the entire SDK and the Google driver, and now ADB & fastboot worked (with the N7's regular OS).
Rooted the device, installed a custom recovery, and - no communication with the custom recovery via ADB. WTF?
Well, it turns out that my prior Android experience had been with a custom recovery where the developer took pains to make sure that the VID/PID pair used by the custom recovery was the same as that used by the OS for ADB - so only a single "adb" driver was needed for both situations.
But that's not the case for either TWRP or CWM for the N7 - the PID is different than any values appearing in the .INF file for the Google SDK USB driver or the Asus USB drivers. That means that without modification the Google or Asus drivers simply will not work with ADB from the custom recovery.
I didn't know that at the time, so I punted and installed the XDA "Universal Naked" (and unsigned) driver - origin unknown. Ugh.
After seeing all the troubles that other people seemed to be having, and not being too happy about installing random unsigned drivers off of the internet, I decided I wanted to clean up this mess of a situation - even though it was working perfectly fine for me!
So I went and downloaded the Asus drivers (R06 2012-06-24).
Next, I hacked the .INF file "android_winusb.inf" in the Asus driver bundle.
Code:
$ diff android_winusb.inf.distro android_winusb.inf
30,41[b]d[/b]29
[color=red]< ;Google Nexus One
< %SingleAdbInterface% = USB_Install, USB\VID_18D1&PID_0D02
< %CompositeAdbInterface% = USB_Install, USB\VID_18D1&PID_0D02&MI_01
< %SingleAdbInterface% = USB_Install, USB\VID_18D1&PID_4E11
< %CompositeAdbInterface% = USB_Install, USB\VID_18D1&PID_4E12&MI_01
<
< ;Google Nexus S
< %SingleAdbInterface% = USB_Install, USB\VID_18D1&PID_4E21
< %CompositeAdbInterface% = USB_Install, USB\VID_18D1&PID_4E22&MI_01
< %SingleAdbInterface% = USB_Install, USB\VID_18D1&PID_4E23
< %CompositeAdbInterface% = USB_Install, USB\VID_18D1&PID_4E24&MI_01
<[/color]
47[b]a[/b]36
[color=green]> %CustomRecoveryAdbInterface% = USB_Install, USB\VID_18D1&PID_D001[/color]
51,62[b]d[/b]39
[color=red]< ;Google Nexus One
< %SingleAdbInterface% = USB_Install, USB\VID_18D1&PID_0D02
< %CompositeAdbInterface% = USB_Install, USB\VID_18D1&PID_0D02&MI_01
< %SingleAdbInterface% = USB_Install, USB\VID_18D1&PID_4E11
< %CompositeAdbInterface% = USB_Install, USB\VID_18D1&PID_4E12&MI_01
< ;Google Nexus S
< %SingleAdbInterface% = USB_Install, USB\VID_18D1&PID_4E21
< %CompositeAdbInterface% = USB_Install, USB\VID_18D1&PID_4E22&MI_01
< %SingleAdbInterface% = USB_Install, USB\VID_18D1&PID_4E23
< %CompositeAdbInterface% = USB_Install, USB\VID_18D1&PID_4E24&MI_01
<[/color]
68[b]a[/b]46
[color=green]> %CustomRecoveryAdbInterface% = USB_Install, USB\VID_18D1&PID_D001[/color]
130[b]a[/b]109
[color=green]> CustomRecoveryAdbInterface = "Custom Recovery ADB Interface"[/color]
The above might be a little hard to read, but basically I deleted all the lines pertaining to Nexus One (PIDs 0D02, 4E11, & 4E12) and Nexus S (PIDs 4E21, 4E22, 4E23, 4E24), and added to the Nexus 7 sections the line
Code:
%CustomRecoveryAdbInterface% = USB_Install, USB\VID_18D1&PID_D001
to both the [Google.NTx86] and [Google.NTamd64] sections of the .INF file
The deletions of the Nexus One and Nexus S were not necessary - if you want, they could be left in there. I wanted to be sure that this driver would ONLY be used with the Nexus 7 though.
Also - note the string "CustomRecoveryAdbInterface" - this needs to be defined in the [Strings] section at the end of the .INF file. It was shown above, but I'll show it again here:
Code:
CustomRecoveryAdbInterface = "Custom Recovery ADB Interface"
The reason I did it this way is because this is the string which will be displayed in the Windows device manager when this particular driver loads! If you want, you can customize this behavior - including putting your own name or other string in there - so that when the driver loads in Windows you will be 100% sure Windows is using the driver you intend!
For instance, I could have used
%MyFearsomeHackedDriver% = USB_Install, USB\VID_18D1&PID_D001
in the [Google.NTx86] and [Google.NTamd64] sections of the .INF file, and
MyFearsomeHackedDriver = "TWRP/CWM Recovery Using Asus Driver R06"
in the [Strings] section of the .INF file. Anyway, you get the idea. Change those strings to say whatever you want (but avoid using any special characters).
Important Note: I would avoid using Wordpad or any other editor which is likely to fool with line termination characters in the .INF file. - edit the .INF file using Notepad
OK, onward.
What I did next was to:
1) Disconnect my PC from the internet.
2) Disconnected N7 from the PC
3) Put "devmgr_show_nonpresent_devices=1" into the Windows Environment - system variables (search for devmgr_show_nonpresent_devices=1 if you don't know what it means. Computer->Properties->Advanced System Settings->Environment Variables)
4) Start up the device Manager, and then View->show hidden devices
#3 and #4 together will display every driver ever installed on your computer, even for devices which are not attached. (Devices which actually are attached will show up in bold, rather than washed-out colors)
Then what I did was - looking in all the usual places in the Device Manager - e.g.
"ADB Interface", or "Android Phone" or "Other Devices"
I laboriously inspected every driver's VID/PID pair one by one. For any that matched
VID_18D1&PID_4E40 - N7 bootloader
VID_18D1&PID_4E41 - ADB off + MTP on
VID_18D1&PID_4E42 - ADB on + MTP off (&REV_9999&MI_01) or MTP on (&REV_9999&MI_00)
VID_18D1&PID_4E43 - ADB off + PTP on
VID_18D1&PID_4E44 - ADB on + PTP on
VID_18D1&PID_D001 - Custom Recovery (either CWM or TWRP)
I not only removed them as the driver, but when presented with the option "do you want to also delete this driver", I clicked YES.
NOTE: You can skip over the section in the device manager labeled as "Portable Devices", as these pertain to MTP/PTP drivers. OR, depending on how you feel about them, you could also uninstall them - just don't say yes to the "do you want to also delete this driver" query. The idea is that we are trying to rid the machine of prior Asus/Google or other drivers matching the N7 - but not delete Microsoft drivers for MTP/PTP. If you uninstall them, they will get re-installed the next time the device starts up.
The reason for this is that Windows maintains a hidden database of every driver that has ever been installed on the machine, and includes with them the matching conditions that cause them to be loaded once they are installed - in particular, VID/PID USB hardware IDs. This is how Windows can re-instate an old driver as soon as you plug the hardware into the machine - even without asking you first!
After walking through all drivers that look like they might match (Note for instance you might have drivers for other Android Devices that you want to leave alone - they will show up with different VID/PID pairs however), and deleting ALL of them that match the above-listed Nexus 7 (and "Universal Naked Google VID/PID of VID_18D1&PID_D001), your machine should be expunged of old drivers.
The disconnection from the internet was to also make sure that Windows wouldn't automatically go out onto the internet and start downloading stuff. I wanted to be SURE that it only was capable of a manual driver install of my choosing.
So, I booted the N7 into the OS, made sure that ADB debugging was enabled, and plugged the N7 into my PC. I suppose it might have been even better to boot into the custom recovery and attach - if Windows installed a driver before it offered me a chance to manually install one, that would mean that I hadn't properly expunged matching drivers from the machine.
When given the option to install a driver, I pointed it at the folder with my (modded) Asus driver, and asked it to install.
Now the Asus driver is signed - and modifying the .INF file breaks the signing. So, in my case Windows 7 complained about the broken signature - but at least it let me over-ride it's complaints and install it anyway. Windows 8 is more fussy about this sort of thing I suppose - at a minimum you would need to be in Win 8 test mode when you do this; perhaps even worse.
Anyway - long story short is that it worked. No Google USB driver, no "Universal Naked Driver" - only a single hacked Asus driver for everything: fastboot and ADB (from both the OS and the custom recoveries - I've used it with both TWRP and CWM).
If you go through this procedure correctly, you should not need to re-connect to the Internet in order to install any drivers - everything you need should already be on the machine. If you want, before you re-attache to the internet, you can toggle on and off MTP/PTP, boot into the the OS and enable/disable ADB, boot into the custom recovery - and verify that all connections are satisfied with your "hacked" Asus driver plus any drivers previously on the machine (hopefully only for the MTP and PTP modes).
I don't know if this will work for you - it sounds like youve tried similar stuff (although perhaps were fuzzy about the .INF mangling). I hope this post provides you with some things to attempt.
As I said before - if you really can transmit lots of data at high speed (via MTP), that suggests that problem is not hardware, and this entire nightmare revolves around Windows drivers issues. TBH though, I would feel a whole lot better about saying this if you were using rooted stock instead of a dev ROM. In either event, you can certainly use TWRP or CWM - that would remove any question about "is it the ROM?".
Another way to validate that it is not hardware is to attempt the whole thing using Linux: boot a Linux "live" CD/DVD Distro on the PC, and cobble together a fastboot/adb executable (plus any required shared libs not present in the Linux distro) and speak to the N7 device using exactly the same PC hardware and cabling, but eliminating Windows temporarily. If it works, the problem has to be Windows.
Jeebus. To think that neither the Mac nor any Linux box need any driver at all for this crap - it really is amazing that "easy to use Windows" is the hardest platform to get this stupid sh!t working. The ADB and fastboot protocols used "on the wire" are incredibly, incredibly bare and simplistic.
anyway, I hope this helps somebody, even if the OP's troubles can't be resolved.
cheers
bftb0 said:
-snip-
Click to expand...
Click to collapse
Good news and bad news!
First, thanks for taking time to answer my cry for help The effort was amazing.
Bad news being... it didn't work :/
I managed to mangle the INF into this:
Code:
;
; Android WinUsb driver installation.
;
[Version]
Signature = "$Windows NT$"
Class = AndroidUsbDeviceClass
ClassGuid = {3F966BD9-FA04-4ec5-991C-D326973B5128}
Provider = %ProviderName%
DriverVer = 05/24/2012,6.0.0000.00000
CatalogFile.NTx86 = androidwinusb86.cat
CatalogFile.NTamd64 = androidwinusba64.cat
;
; This section seems to be required for WinUsb driver installation.
; If this section is removed the installer will report an error
; "Required section not found in INF file".
;
[ClassInstall32]
Addreg = AndroidWinUsbClassReg
[AndroidWinUsbClassReg]
HKR,,,0,%ClassName%
HKR,,Icon,,-1
[Manufacturer]
%ProviderName% = Google, NTx86, NTamd64
[Google.NTx86]
;Google Nexus 7
%TWRPNexus7Recovery% = USB_Install, USB\VID_18D1&PID_D001
%bootloader% = USB_Install, USB\VID_18D1&PID_4E40
%ADBoff_MTPNexus7OS% = USB_Install, USB\VID_18D1&PID_4E41
%ADB_MTPNexus7OS% = USB_Install, USB\VID_18D1&PID_4E42&MI_00
%ADB_MTPoffNexus7OS% = USB_Install, USB\VID_18D1&PID_4E42&MI_01
%ADBoff_PTPoffNexus7OS% = USB_Install, USB\VID_18D1&PID_4E43
%ADB_PTPNexus7OS% = USB_Install, USB\VID_18D1&PID_4E44
[Google.NTamd64]
;Google Nexus 7
%TWRPNexus7Recovery% = USB_Install, USB\VID_18D1&PID_D001
%bootloader% = USB_Install, USB\VID_18D1&PID_4E40
%ADBoff_MTPNexus7OS% = USB_Install, USB\VID_18D1&PID_4E41
%ADB_MTPNexus7OS% = USB_Install, USB\VID_18D1&PID_4E42&MI_00
%ADB_MTPoffNexus7OS% = USB_Install, USB\VID_18D1&PID_4E42&MI_01
%ADBoff_PTPoffNexus7OS% = USB_Install, USB\VID_18D1&PID_4E43
%ADB_PTPNexus7OS% = USB_Install, USB\VID_18D1&PID_4E44
[USB_Install]
Include = winusb.inf
Needs = WINUSB.NT
[USB_Install.Services]
Include = winusb.inf
AddService = WinUSB,0x00000002,WinUSB_ServiceInstall
[WinUSB_ServiceInstall]
DisplayName = %WinUSB_SvcDesc%
ServiceType = 1
StartType = 3
ErrorControl = 1
ServiceBinary = %12%\WinUSB.sys
[USB_Install.Wdf]
KmdfService = WINUSB, WinUSB_Install
[WinUSB_Install]
KmdfLibraryVersion = 1.9
[USB_Install.HW]
AddReg = Dev_AddReg
[Dev_AddReg]
HKR,,DeviceInterfaceGUIDs,0x10000,"{F72FE0D4-CBCB-407d-8814-9ED673D0DD6B}"
[USB_Install.CoInstallers]
AddReg = CoInstallers_AddReg
CopyFiles = CoInstallers_CopyFiles
[CoInstallers_AddReg]
HKR,,CoInstallers32,0x00010000,"WdfCoInstaller01009.dll,WdfCoInstaller","WinUSBCoInstaller2.dll"
[CoInstallers_CopyFiles]
WinUSBCoInstaller2.dll
WdfCoInstaller01009.dll
[DestinationDirs]
CoInstallers_CopyFiles=11
[SourceDisksNames]
1 = %DISK_NAME%,,,\i386
2 = %DISK_NAME%,,,\amd64
[SourceDisksFiles.x86]
WinUSBCoInstaller2.dll = 1
WdfCoInstaller01009.dll = 1
[SourceDisksFiles.amd64]
WinUSBCoInstaller2.dll = 2
WdfCoInstaller01009.dll = 2
[Strings]
ProviderName = "Google, Inc."
SingleAdbInterface = "Android ADB Interface"
CompositeAdbInterface = "Android Composite ADB Interface"
SingleBootLoaderInterface = "Android Bootloader Interface"
WinUSB_SvcDesc = "Android USB Driver"
DISK_NAME = "Android WinUsb installation disk"
ClassName = "Android Phone"
TWRPNexus7Recovery = "TWRP Nexus 7 Recovery"
bootloader = "Nexus 7 Bootloader"
ADBoff_MTPNexus7OS = "ADB off, MTP on Nexus 7 OS"
ADB_MTPNexus7OS = "ADB on, MTP on Nexus 7 OS"
ADB_MTPoffNexus7OS = "ADB on, MTP off Nexus 7 OS"
ADBoff_PTPoffNexus7OS = "ADB off, PTP off Nexus 7 OS"
ADB_PTPNexus7OS = "ADB on, PTP on Nexus 7 OS"
I then went around into recovery and fiddled about with the MTP PTP settings after enabling devmgr to show hidden drivers. I uninstalled them all, then reinstalled them from the Asus R06 drivers with my ethernet link disabled.
When booting into the Nexus 7 OS with ADB on and MTP off, then installing the drivers, the description ADB on, MTP off Nexus 7 OS was seen in device manager, and the hardware ID was
Code:
%ADB_MTPoffNexus7OS% = USB_Install, USB\VID_18D1&PID_4E42&MI_01
Therefore, it installed properly.
But running cmd>adb kill-server>adb start-server>adb devices
yielded... nothing.
So, I rebooted into recovery, and "TWRP Nexus 7 Recovery" showed up in devmgr. But adb devices remained oblivious.
I think I'm lost to this cause :/
P.S- you mentioned MTP could do up to 80mbit/s? Mine is very slow indeed. I use a USB OTG cable to do any transfers, and it doesn't work with NTFS, even with Paragon NTFS/HFS+ root app :/
P.P.S- word of advice, Windows 8 is the most useless, horrendous, unusable and awful OS you could ever use if you don't have a touchscreen. 1) Hacky unsigned driver install leaves you open to attacks 2) Boots a useless 'Metro' start screen with crap app selection that eats memory. 3) Clearly doesn't work with Android out of the box. Or any other Linux OS for that matter. for the love of god DON'T. GET. IT.
@tigrey
Sorry it didn't work out. Really strange though that it seems to function correctly for some things but not others.
Way OT question... is it possible to boot that Win 8 machine with anything else, or does it have mandatory signing requirements that prevents that from happening? (e.g. will it boot a Linux "live" CD?). Just curious - I haven't been keeping up with what M$ has been forcing on their hardware partners.
bftb0 said:
@tigrey
Sorry it didn't work out. Really strange though that it seems to function correctly for some things but not others.
Way OT question... is it possible to boot that Win 8 machine with anything else, or does it have mandatory signing requirements that prevents that from happening? (e.g. will it boot a Linux "live" CD?). Just curious - I haven't been keeping up with what M$ has been forcing on their hardware partners.
Click to expand...
Click to collapse
Well, it doesn't have a CD drive, but I'm sure i'd be able to make a live USB boot before windows. I haven't tried it.
I don't know anything about Linux, sorry, so I wouldn't know how to 'cobble together a fastboot/adb executable' :/
I think you can: http://mjg59.dreamwidth.org/20522.html for select distros
tigrey said:
Well, it doesn't have a CD drive, but I'm sure i'd be able to make a live USB boot before windows. I haven't tried it.
I don't know anything about Linux, sorry, so I wouldn't know how to 'cobble together a fastboot/adb executable' :/
I think you can: http://mjg59.dreamwidth.org/20522.html for select distros
Click to expand...
Click to collapse
Well the fastboot executable for Linux is dynamically linked; exactly analogous to how dynamically-linked executables for Windows require supporting .DLLs, Linux executables require ".so" files (Shared Object libraries). Usually, most of them are already part of the distro, but perhaps one of two are shipped with the executable. The last time I used a Linux Live CD with Google's fastboot for Linux, I only had to grab two files out of the Google Linux SDK - the fastboot executable itself and a single .so file - and put them on a USB key drive in order to use fastboot (from the live cd boot) without installing the Google SDK.
There is only one additional trick to getting this done - adding an environment variable LD_LIBRARY_PATH - to the command shell's environment so that when you start up "fastboot" it knows where to find the missing .so file(s).
PM me if you want some assistance - but it sort of sounds like you don't want to go down that path, especially if it were to turn out that the N7 really is the source of the trouble.
Thanks for the link - it sure looks like that UEFI/ secure boot technology is in play and some sort of signed boot shim from M$ is needed. Ugh.
Are you sure Win 8 is accepting Un-signed drivers?
I had issues before. Un-installed any android related drivers.
Accepted unsigned drivers.
Reboot.
Install Android drivers.
BAM!
bftb0 said:
Well the fastboot executable for Linux is dynamically linked; exactly analogous to how dynamically-linked executables for Windows require supporting .DLLs, Linux executables require ".so" files (Shared Object libraries). Usually, most of them are already part of the distro, but perhaps one of two are shipped with the executable. The last time I used a Linux Live CD with Google's fastboot for Linux, I only had to grab two files out of the Google Linux SDK - the fastboot executable itself and a single .so file - and put them on a USB key drive in order to use fastboot (from the live cd boot) without installing the Google SDK.
There is only one additional trick to getting this done - adding an environment variable LD_LIBRARY_PATH - to the command shell's environment so that when you start up "fastboot" it knows where to find the missing .so file(s).
PM me if you want some assistance - but it sort of sounds like you don't want to go down that path, especially if it were to turn out that the N7 really is the source of the trouble.
Thanks for the link - it sure looks like that UEFI/ secure boot technology is in play and some sort of signed boot shim from M$ is needed. Ugh.
Click to expand...
Click to collapse
Holy mother of jesus
After a month of tinkering, I just solved the problem! I have no idea how any of this is linked together, but I updated the bootloader from 4.13 to 4.18 and bam, the device is now recognised in ADB. Since fastboot and adb are mutually exclusive, I find it odd that it worked. Thanks for the help, thought I'd end the thread with some closure
tigrey said:
Holy mother of jesus
After a month of tinkering, I just solved the problem! I have no idea how any of this is linked together, but I updated the bootloader from 4.13 to 4.18 and bam, the device is now recognised in ADB. Since fastboot and adb are mutually exclusive, I find it odd that it worked. Thanks for the help, thought I'd end the thread with some closure
Click to expand...
Click to collapse
Ouch!
Well, at least it is working now - congrats on your persistence (although I'm sure you would have preferred to have spent the time on other stuff). I don't know why that would make a difference for adb, either.
I had a similar issue as the OP with Win7 x64 presenting my Nexus 7 as an "Android Composite ADB Interface" but no devices showing up in 'adb devices'. This is using the auto-downloaded Windows Update drivers; nothing special.
On a whim, I pulled down the notification where it said, "Connected as a Media Device. Select to change" and changed it from a Media device (MTP) to a Camera (PTP). Windows beeped and booped, installed two more drivers, and I immediately got the prompt to authorize connection via ADB on the tablet. Now 'adb devices' shows the device and I'm ready to rock.
- Dave
ViperGeek said:
I had a similar issue as the OP with Win7 x64 presenting my Nexus 7 as an "Android Composite ADB Interface" but no devices showing up in 'adb devices'. This is using the auto-downloaded Windows Update drivers; nothing special.
On a whim, I pulled down the notification where it said, "Connected as a Media Device. Select to change" and changed it from a Media device (MTP) to a Camera (PTP). Windows beeped and booped, installed two more drivers, and I immediately got the prompt to authorize connection via ADB on the tablet. Now 'adb devices' shows the device and I'm ready to rock.
- Dave
Click to expand...
Click to collapse
You Sir, are a legend.
BS...
I know this is an old thread but it shows up on Google, so I thought I would add something useful to it.
If you don't want to hack your drivers and mess with unsigned drivers on Windows 8, I suggest you install the UniversalADBDriver
https://github.com/koush/UniversalAdbDriver
Your device might show up under another name in the device manager, but who cared. It works.
Continuing the work of @ofutur and @ViperGeek, there's one other suggestion that it doesn't appear to have fully been made that has been known to work...
Go into your device manager for your computer, and where the device shows up when you plug it in (Android/Portable Devices/wherever...), go to the device properties and select the Driver tab, and down at the bottom, select Uninstall. There will be a small prompt that has a check-box in it that says "Delete the driver software for this device", and check that box. When it's done removing everything, unplug the device and go into it's developer options (selecting debugging on) and going into your storage options (selecting MTP/Media Device), and then plug the device back into the computer. Drivers will start installing themselves, and when they complete, refresh your C:/ folder to see if anything's appeared there (namely, your device). If it has, in MTP mode, you're good to go. If nothing's shown there while you're in MTP mode on your device, then what's happened is that your computer used other drivers that (still) aren't compatible upon the new installation. The reason why this happened is because you have so many different driver options on your computer that it essentially doesn't know which of the multiple options you have is the right one. NakedDrivers/Cyogenmod Drivers/CompanyDrivers/WhateverDrivers, there's a lot of options out there that many people have used/installed. Effectively, you need to weed out which ones are completely compatible, and which ones aren't. So keep plugging in your device, uninstalling the drivers (and the driver software), un-plugging and re-plugging in your device (and there-by installing different drivers), until your computer installs the drivers that are the correct and most compatible drivers. Upon doing this, chances are your computer will show a different driver type in the installation, including MTP in it. From there, you should be good to go. In short, you don't need to download and install anything, you just need to uninstall whatever software and drivers you have on your machine until when you plug your device in, they communicate in the way they're supposed to.
i have spent hours stuffing around with my nexus 7 2012 (wifi) and wanted to share a tip. hope this helps someone!
i wanted to downgrade from 5.x back to 4.4.4 so i could play the old tetris (mission failed anyway, no longer available, where do you get old versions of apps & games?) anyway i found that adb 1.0.32 only shows devices once my nexus is fully powered on to the normal state
when i reboot into bootloader aka fastboot mode, i was being driven mad by adb "not showing my device"
then i realised that both the bootloader image and the 4.4.4 zip file only needed the "fastboot" command not the "adb" command. also noticed that "adb devices" has a similar command "fastboot devices" and in bootloader mode, fastboot devices DID return my device...
so i was trying to solve a problem that didn't exist! adb does not need to see devices in bootloader aka fastboot mode!! and so i was able to use fastboot to do the last steps of boot image and system zip file update
Code:
fastboot flash bootloader [bootloader file].img
fastboot -w update [image file].zip
PS. i had stumbled onto these "manual" commands at first, but they failed, so i assumed it was because of adb no devices. it turns out the nexus 7 has a 2013 model (wifi) called "flo-flo" i had downloaded from android central. well, my device was listed further down under nexus 7 2012 (wifi) called "grouper". this was the root cause of the fastboot error (not adb lack of devices in bootloader aka fastboot mode) once i used "grouper" the boot.img and image.zip worked fine. my nexus 7 actually had a little echo up top left saying "signature match" which i presume was the "flo-flo" vs. "grouper" issue.