I originally posted this question in XDA Assist yesterday. I was directed to this forum, and after reading more here, have added a few more questions/concerns with ***___**** that came after digging through the forum stickies. I know this has become a huge post, so apologies in advance.
I'm having some trouble mounting a Kindle Fire HD 7 Gen2 (Tate) in Fastboot mode on my system which is running Ubuntu 14.04.
I am new to any kind of Android development and have been following the excellent guides here by @kinfauns. Specifically I have the 5-scecond boot loop problem related to the power draw issue. The solution I'm pursuing is to swap the stock bootloader for FireFireFire. The guide says to do it with fastboot, since the conventional flash recovery method could be interrupted by the shutdown cycle.
******* After reading the Faq here, I am concerned that even if Fastboot were to detect the device, that simply using it to load the FireFireFire bootloader directly onto the device would cause the problem of the wall of fire because of the M-Shield issue. Seems I would have to follow these steps here or here to load FireFireFire as a 2nd bootloader (including downgrading the OS to v7.2.3), is that correct? Also one of those pages says not to use automated tools like FireFireFire... which I thought was a bootloader, so what bootloader should I use to get around the low battery cycle startup problem that FireFireFire specifically addresses with it's startup battery check *******
So now, just when I think I have a solution to my real problem, I get out the fastboot cable, connect the kindle to my PC, and I get an error in Nautilus.
******In one of the pages linked above, it makes a passing reference to some fastboot cables not putting the device in Fastboot. Since I cannot connect to the device in the first place when it is in "fastboot", I cannot use their method to confirm fastboot is working. When bootup with my fastboot cable, the "image" on the device does change from orange & flashing "Kinde Fire" to the orange "Kindle Fastboot" pic, so I'd to know if I can rule out the cable as part of my problem... *****
Here is the error from Nautilus: Unable to open MTP device '[usb:001,006]'. Attempting mtpfs at the terminal returns:
Unable to open ~/.mtpz-data for reading, MTPZ disabled.Listing raw device(s)
Device 0 (VID=1949 and PID=0007) is a Amazon Kindle Fire (ID1).
Found 1 device(s):
Amazon: Kindle Fire (ID1) (1949:0007) @ bus 1, dev 6
Attempting to connect device
LIBMTP PANIC: Unable to find interface & endpoints of device
Unable to open raw device 0
I found a tutorial on Ubuntu forums for getting MTP to work that involved using lsusb command to get the device id, making entries in /etc/udeb/rules.d and /lib/udev/rules.d, supposedly to allow MTP to access the device with proper credentials. However that didn't work. It seems that in Fastboot I actually might not MTP to handle the device...
OTOH, when the device is NOT in Fastboot mode, it mounts just fine, both in Nautilus and for purposes of checking with the adb devices command....
***** Since yesterday, I back and compared the vendor and product codes reported in lsusb and noticed that they are the same regardless of whether I connect the device using a regular cable (when it mounts correctly) and the fastboot cable (when the screen says fastboot, and I get the "raw device" failure to mount. My understanding is that the hardware ID should be different if the device is in fastboot mode vs "normal" mode. Is that correct? How do I deal with this behavior if the device is supposed to load different drivers when in different modes?*****
However, it seems to me that what I want is for it to mount in file transfer mode instead of MTP mode when in fastboot. The fix for this exact issue in Windows (as reported on other sites) is to change the driver in Windows Device Manager.
***** The other thing I tried was using the modprobe command to manually start the usb-storage driver as explained here. However the step to confirm the driver was unsuccessful because the /var/log/messages file was not found by the tail command. Even though that is really an ubuntu / linux question, perhaps someone here can help with some guidance on confirming the mass storage driver as well?*****
Can anybody help with this problem please?
Thank you
Additional detail
In case it's helpful to diagnose the issue, I am now experiencing an odd behavior where the device starts out in Fastboot (using the fastboot cable), and then after about 30+ seconds, it automatically switches to the "swipe" screen as though it was in normal mode.
This happened after I went and rem'd out all references to this device in the MTP rules files. The first time I plugged it in, Ubuntu popup asked how it should handle the device. I didn't see "mass storage" so I hit cancel. At that point, it started to flop over from fastboot into normal automatically even though it is plugged into a fastboot cable.
My understanding was that a fastboot cable forced it into fastboot mode (or else was somehow device in-compatible with certain Kindles). I don't understand how the host OS could do this without issuing an ADB / shell command?
The gnss hwbinder patch have been incorporated into the Aug 25 17.1 builds.
Please take 10 minutes to read all of post #1 and #2. DO NOT USE this thread to report any other problems, it's only for troubleshooting random reboots or reboots into TWRP for S2 exynos5433 tablet users. If this is unreasonable, then stop reading. Thank you.
If you cannot provide the logs below, please do not participate in this thread. Thank you.
I will take a look at all logs uploaded. The more the better to help find root cause or causes.
At this point, I have no solution to fix random reboots, but gathering data that might help narrow down which app or process is causing this.
Instructions
1. If you are experiencing random reboot or reboot into TWRP, you MUST supply tombstone logs. It's super easy and simple to do. Connect your device to PC using micro usb cable.
Make sure adb debugging is enabled in developer options. To enable developer options, tap settings, about tablet, build number 7 times until it says developer options unlocked. Go back to settings, system, advanced, developer options, enable android debugging and rooted debugging.
Bash:
adb root
adb shell
cd data/tombstones
ls -al
# if you see files tombstone_00, tombstone_01, etc then
exit
mkdir t
cd t
adb pull data/tombstones/
Zip all the files and upload here. Alternatively, you can send me the zipped logs via xda PM.
2. The better log would be adb logcat of the problem (random reboot, stuck at boot animation due to random restart) that just happened or happening.
Bash:
adb logcat > crash.txt
# wait 30 seconds then hit ctrl-c to stop capture
Please zip the .txt file before uploading. For some reason, I cannot see .txt files here.
FAQ
Q1. When will this random reboot be solved?
A1. Set your expectations that it will never be solved. At this point, I'm only gathering data to try and help. There may never be a fix if it's due to a proprietary Samsung blob that can never be updated. DO NOT ask when or if this can be fixed. It is not productive and I will completely ignore the question. Thank you.
Q2. Will you provide updates?
A2. I will only provide updates if I think we have found a solution. Otherwise, assume I'm not working on this problem because like you, I have other responsibilities and interests in life.
Q3. What will you do with the logs?
A3. I will look at every uploaded log and try to find a common pattern among them. If I find something that I think it's worth reporting, then I will post it here. If I need more information, I will revise the instructions and ask for more.
Q4. Why are you doing this?
A4. Let me be clear. I'm not experiencing any random reboots myself, I'm just trying to help if I can. The more logs you can provide, the more likely we can narrow down the problem. No logs from the commuity means the random reboot problems will never be fixed.
Let me be clear again, DO NOT ask when or if this can be fixed. Thank you.
Reserved.
Hi @retiredtab,
thanks for your offer to analyze the crash-logs!
I'm using a T810 Tab (gts210wifi) and are experiencing random reboots since updating to lineage-17.1 (ripee).
The reboots itself are of different types:
1) Stuck at bootanimation - only exit by three-button-reset.
2) Reboot to unlock-screen, but after unlock-code endless other unlock-screens come up - only exit by three-button-reset.
3) Reboot to unlock-screen, but after unlock-code the screen is empty and after some time (few minutes) it resets and ends in the same sequence - only exit by three-button-reset.
4) Reboot to unlock-screen, everything ok after unlocking.
For most of the reboots it was my observation, that there was an event, which did wakeup the tablet from deeper sleep (plugin to charger, switching on the screen, alarm, ...).
That caused me to write a small Tasker-app profile, which is pinging my dns-server all 2mins (to prevent a deeper sleep). With that profile active I did not experience any reboot for at least 2 months. Before that I had about one reboot per day.
To support your analysis, I disabled that Tasker-profile and the reboots are back.
During the last 24hours I had two reboots of type 4, and I was capturing the logcat and tombstones afterwards.
Looking into the captured data I got the impression that it might contain sensitive data (found my server-names, calender-events and so on) so I'm unsure if it can contain even more sensitive data like login-information or wifi-passwords.
So I'm hesitating to upload them to this public forum.
Would you agree to receive the zipped logs by PM instead?
Yogi555 said:
I'm using a T810 Tab (gts210wifi) and are experiencing random reboots since updating to lineage-17.1 (ripee).
The reboots itself are of different types:
1) Stuck at bootanimation - only exit by three-button-reset.
2) Reboot to unlock-screen, but after unlock-code endless other unlock-screens come up - only exit by three-button-reset.
3) Reboot to unlock-screen, but after unlock-code the screen is empty and after some time (few minutes) it resets and ends in the same sequence - only exit by three-button-reset.
4) Reboot to unlock-screen, everything ok after unlocking.
For most of the reboots it was my observation, that there was an event, which did wakeup the tablet from deeper sleep (plugin to charger, switching on the screen, alarm, ...).
That caused me to write a small Tasker-app profile, which is pinging my dns-server all 2mins (to prevent a deeper sleep). With that profile active I did not experience any reboot for at least 2 months. Before that I had about one reboot per day.
To support your analysis, I disabled that Tasker-profile and the reboots are back.
During the last 24hours I had two reboots of type 4, and I was capturing the logcat and tombstones afterwards.
Click to expand...
Click to collapse
Hi, @Yogi555.
Your reboot types 2 and 3 are interesting - I haven't seen them happen with my device. I only get types 1 and (occasionally) 4.
And type 5: reboot to TWRP. On my device this has happened more often than type 4, but both are much less common than type 1.
It does sound like the device waking from deep sleep could be triggering these events. Certainly, my device most often goes into type 1 when it's been left alone for a while, such as overnight (so my workaround has been to turn it off at night).
But I've also seen it drop into type 1 while I was actively using it, so waking up mustn't be the only trigger.
Hi, @Ipedia,
Maybe types 2 and 3 are somehow related to the individual homescreen app - I'm using Nova Launcher. Only guessing here.
Regarding type 5: Yes I did see it as well sometimes - but I'm not sure if it is a type of it's own. If I remember correctly type 1 - 3 are rebooting itself after some time of stalling. And - not sure here - after some unsucessful reboots in sequence we end up in TWRP. (In the embedded business where I'm working, it is the desired behaviour to jump to a flashing-bootloader after a number of resets within a certain time).
I've seen the reboots while busy using it as well some time ago, rare, but it happened - but when I activated my described tasker-profile, they were gone as well. I've no explanation for that.
At the moment I'm only facing type 4 resets. I did many changes, since I saw type 1 - 3 (update to latest 17.1 linageos, latest Magisk, exchange of the wifi-blobs) so not sure if any of this solved these reboots, or if they are yet to popup ...
Yogi555 said:
Would you agree to receive the zipped logs by PM instead?
Click to expand...
Click to collapse
@Yogi555 , you can send zipped logs using PM. Thank you.
can you help me to get logs? I receive this error message
Code:
c:\adb>adb root
c:\adb>adb shell
gts210ltexx:/ $ cd data/tombstones
gts210ltexx:/data/tombstones $ ls -al
ls: .: Permission denied
Max_C said:
ls: .: Permission denied
Click to expand...
Click to collapse
Maybe have a look how my adb root problems were solved in this thread: https://forum.xda-developers.com/t/...-version-of-andriod-10.4308193/#post-85353551
Worth a try, but maybe your issue is different
Yogi555 said:
Maybe have a look how my adb root problems were solved in this thread: https://forum.xda-developers.com/t/...-version-of-andriod-10.4308193/#post-85353551
Worth a try, but maybe your issue is different
Click to expand...
Click to collapse
i see, my device not rooted. I need install Magisk first, then enable rooted debugging. I have no rooted debugging in developer options
Magisk is hiding the option - without magisk you should see it - even with an unrooted device
Yogi555 said:
Magisk is hiding the option - without magisk you should see it - even with an unrooted device
Click to expand...
Click to collapse
oh. i see. I have no "rooted debugging" option in my developer settings in Havoc OS
another problem is i have old Android SDK on my PC. I installed latest one, and got this
Code:
c:\adb>adb root
ADB Root access is disabled by system setting - enable in Settings -> System -> Developer options
I had exactly the same issue with magisk being the one hiding the setting. Are you sure that you are not having magisk installed? Don't know Havoc ...
Max_C said:
oh. i see. I have no "rooted debugging" option in my developer settings in Havoc OS
Click to expand...
Click to collapse
Boot into TWRP then. You can get the same data/tombstones from TWRP using the same commands.
If you have the SM-T710, you will need my TWRP in order for adb to work. The official ones posted on twrp.me don't work with adb or mtp.
retiredtab - Browse /SM-T710/TWRP at SourceForge.net
sourceforge.net
Mine works, but it takes 10 seconds before mtp and adb work after the TWRP home screen is shown.
retiredtab said:
You can get the same data/tombstones from TWRP using the same commands.
Click to expand...
Click to collapse
thanks! got it, please check attached
I have downloaded them, but will examine tomorrow. It's been a long day and it's 12:40 AM. Time for sleep.
After 17 days of gathering data and troubleshooting, we think we may have solved the random reboots. Huge thanks to @lpedia, @CuckooPenguin and @Yogi555 for testing the various gnss patches until we finally figured out the hwbinder solution.
Details at
https://github.com/universal5433/android_device_samsung_gts2-common/pull/4
If you are one of those people who experienced random reboots, it's not too late to submit your data/tombstones or check them yourself for the error message that I listed above. In particular, you are looking for
Code:
#0 pc 00056baa /system/lib/[email protected] (android::hardware::gnss::V1_0::BsAGnssRilCallback::requestSetIdCb(unsigned int)+450) (BuildId: 30ee8b14270ae2b3ae675a7f9b4de136)
#1 pc 0000e8dd /system/vendor/lib/hw/[email protected] (android::hardware::gnss::V1_0::implementation::AGnssRil::requestSetId(unsigned int)+32) (BuildId: d0033c7cb8d0d1a2c99f977719d0d967)
#2 pc 0000fb05 /system/vendor/lib/hw/gps.default.so (broadcom::GpsiClient::marshal_agps_ril_request_setid(broadcom::IpcIncomingMessage&)+160) (BuildId: fe39be024fc2f98450c8d2fc07cda956)
Once the git pull requested is approved, this patch should make it into the next August security patch build for all flavours of the tab S2 including all the variant OS.
Finally, a huge thanks to @ananjaser1211 for tips, suggestions, code and acting as a sounding board for ideas.
retiredtab said:
this patch should make it into the next August security patch build for all flavours of the tab S2 including all the variant OS
Click to expand...
Click to collapse
sorry, I'm not sure i understand how it works... Should I wait when my HavocOS ROM will be updated in this thread? Or there is another way how to apply the patch
Max_C said:
sorry, I'm not sure i understand how it works... Should I wait when my HavocOS ROM will be updated in this thread? Or there is another way how to apply the patch
Click to expand...
Click to collapse
We tried modifying the manifest.xml directly, but the results were inconsistent. On my t710 the instructions below worked, but 2 people with the t810 it resulted in a system not booting properly. These 2 had to reflash the rom to get a working system back. On the t810, the below changed had to be compiled into the code.
If you want to try, backup your data first and be ready to reflash the old rom in case it fails. If it fails, you will have to wait for a new Havoc build or build it yourself.
Bash:
# Turn off gps in OS before booting into TWRP
# boot into TWRP
adb root
# touch mount, system
adb pull /system/vendor/etc/vintf/manifest.xml
# use a text editor and look for this section of code
<hal format="hidl">
<name>android.hardware.gnss</name>
<transport arch="32">passthrough</transport>
<version>1.0</version>
<interface>
<name>IGnss</name>
<instance>default</instance>
</interface>
<fqname>@1.0::IGnss/default</fqname>
</hal>
# change passthrough to hwbinder like so
<transport arch="32">passthrough</transport>
to
<transport>hwbinder</transport>
# save text file
# now push file
adb push manifest.xml /system/vendor/etc/vintf/manifest.xml
# verify that line manifest.xml has changed in the gnss section
adb shell cat /system/vendor/etc/vintf/manifest.xml | more
# Important. Wipe dalvik/art cache and cache
# reboot into OS, enable GPS and make sure GPS works
# Note you may have to go into your setting, apps and wipe cache data as well if GPS doesn't work.
retiredtab said:
We tried modifying the manifest.xml directly, but the results were inconsistent. On my t710 the instructions below worked,
Click to expand...
Click to collapse
Thanks, on my SM-T815 all working fine, I followed your instructions, pushed revised manifest.xml, GPS is working fine. I will test if it helped to solve the problem and leave message here with result in several days.
Max_C said:
Thanks, on my SM-T815 all working fine, I followed your instructions, pushed revised manifest.xml, GPS is working fine. I will test if it helped to solve the problem and leave message here with result in several days.
Click to expand...
Click to collapse
Well, I'm surprised it worked on the t815 when it failed on the t810. That's why I stated it was inconsistent.
So if you were getting daily random reboots due to gnss crashing, then the gnss hwbinder fixes that. We have now had 2 testers go > 150 hours uptime with this patch. Before, they were getting daily reboots/crashes due to gnss (as shown in adb logcat and tombstones).
At first, we thought going 72 hours without a reboot would qualify as a successful fix, but we decided to extend it > 100 hours.
Again, the data/tombstones and adb logcats will show you which process/app is crashing. That's why it's important to view and share them otherwise we would have never figured this out.
Please report back if/when your device reaches 100 hours of uptime.
Please, help, I made a mess. Let my teenager play around with my successfully rooted device (twrp + magisk canary A12) and the kid changed permission to the system folder in root explorer. That got me into a boot loop.
No choice but to flash firmware.
MIUI Global 13.0.4 stable A12 SKHMIXM
I successfully flashed firmware but strangely I did not get re-locked. But now I cannot access the phone in adb to re-root due to a known problem "device not migrated properly".
Device Driver LOG:
-------------------------------------------
Device USB\VID_18D1&PID_4EE7\51bad43e was not migrated due to partial or ambiguous match.
Last Device Instance Id: USB\VID_1307&PID_0165\00000000000017
Class Guid: {36fc9e60-c465-11cf-8056-444553540000}
Location Path: PCIROOT(0)#PCI(1400)#USBROOT(0)#USB(1)
Migration Rank: 0xF000FFFFFFFF0023
Present: false
Status: 0xC0000719
-------------------------------------------------
tried all sort of solutions: checked usb, used another PC, changed cables, revoked and renewed authorisation, suspended driver signature in win10 etc etc
PC will recognise adb device normally just not in fast boot!
would be so very grateful for any ideas. Any regedits perhaps??? I am just an amateur. Thanks in advance
tried regedit + deleting the device from device manager and also this very handy adb command:
maybe it helps others but it did not solve my "Device was not migrated due to partial or ambiguous match" problem. Please send some help, expert, suggestions. run out of ideas. there must be a simple solution?
sadly didnt get any replies. one thing that really helped in the end was this manual: https://www.droidwin.com/unbrick-poco-f3/ So I leave it here in case other have similar need...
I'm running an older version of OMNIROM with Microg on my Zenfone 8 (US). Obviously, the bootloader is unlocked. I've got TWRP on it also. A few days ago I had a problem where the cell service just died. I couldn't enable Mobile Data and had a little x over the cell signal bars. I finally got the cell service working by wiping Dalvik Cache, Data, and internal storage. However, since I just wiped my phone, I'd like to update it to Android 12 (either OMNIROM or LineageOS). However, my Windown 10 PC does not see the phone when I try to use ADB or Fastboot. It doesn't list any device.
Things I've tried:
-In addition to the ASUS cable, tried OnePlus 6's cable
-4 different USB ports on my PC (2.0 and 3.0)
-Numerous ADB drivers, including ASUS' driver, and many generic drivers
-Three different ADB/Fastboot installations
-Ran Windows' device troubleshooter
-Tried installing driver with Windows ignoring driver certificates
I do have USB Debugging enabled. I can see the phone when it is running the ROM and set to File Transfer.
Any suggestions on what I could try to get my phone recognized?
I've been at this most of the day. I post here, then try one more driver change, and it works. Well, it worked enough to get fastboot to recognize it, so I flashed stock.
The driver change I did was to pull up Devices and Printers in Win10. I right-clicked on the Zenfone 8, selected Properties | Hardware. It showed that there was no driver installed for the device. I clicked on Update Driver, and pointed it to the folder with the ASUS official drivers (which I had installed previously in a different manner). I let it do its thing. Rebooted, and it worked. I hope this helps somebody else.
Yippee38 said:
I'm running an older version of OMNIROM with Microg on my Zenfone 8 (US). Obviously, the bootloader is unlocked. I've got TWRP on it also. A few days ago I had a problem where the cell service just died. I couldn't enable Mobile Data and had a little x over the cell signal bars. I finally got the cell service working by wiping Dalvik Cache, Data, and internal storage. However, since I just wiped my phone, I'd like to update it to Android 12 (either OMNIROM or LineageOS). However, my Windown 10 PC does not see the phone when I try to use ADB or Fastboot. It doesn't list any device.
Things I've tried:
-In addition to the ASUS cable, tried OnePlus 6's cable
-4 different USB ports on my PC (2.0 and 3.0)
-Numerous ADB drivers, including ASUS' driver, and many generic drivers
-Three different ADB/Fastboot installations
-Ran Windows' device troubleshooter
-Tried installing driver with Windows ignoring driver certificates
I do have USB Debugging enabled. I can see the phone when it is running the ROM and set to File Transfer.
Any suggestions on what I could try to get my phone recognized?
Click to expand...
Click to collapse
Any change to try Linux on your PC ? e.g. via LIveCD or in a virtual machine with access to the USB port?
bnsmb said:
Any change to try Linux on your PC ? e.g. via LIveCD or in a virtual machine with access to the USB port?
Click to expand...
Click to collapse
I could try it on my laptop which only runs linux, but at this point, I'm past that, so I really don't need to go back and try it again.