[We Need] Working Nvflash.... :(

Search This thread

compuw22c

Senior Member
Dec 26, 2007
621
237
Chicago, IL
Like who? even that would take days if not weeks plus you'd need a wicked fast internet connection

Sent from my Transformer Prime TF201 using Tapatalk

Actually hes right. We could use Amazon EC2 (i think). See this blog here. http://stacksmashing.net/2010/11/15/cracking-in-the-cloud-amazons-new-ec2-gpu-instances/

I believe we need it formulated into some kind of application of script first but I'd definitely kick in a few bucks for the server time. I'd be fine waiting a few weeks, and you wouldn't need a fast connection at all, its basically remote desktop in, set up program to run, log in next week to see if it found the key (or set up an email alert). If someone can write the cracking app/script we will find the power to run it!
 
Last edited:

Haro912

Senior Member
Jul 27, 2011
1,643
156
32
Pittsburgh

Wordlywisewiz

Senior Member
Nov 25, 2011
1,150
224
Brattleboro
Yeah I'd throw in some bucks how fast could it do this128 bit assuming it is the worst case

Sent from my Transformer Prime TF201 using Tapatalk
 

compuw22c

Senior Member
Dec 26, 2007
621
237
Chicago, IL
The way I THINK it works is you pay per compute time, ie, if they have the capacity, very little time. It doesn't matter to them if you use 100 servers for 2 minutes or 10 servers for 20. And with amazons size I'd imagine they have plenty of capacity. The real issue we need now is a cracking program or script. The horsepower we can find.
 

Haro912

Senior Member
Jul 27, 2011
1,643
156
32
Pittsburgh
Using the CUDA-Multiforce, I was able to crack all hashes from this file with a password length from 1-6 in only 49 Minutes (1 hour costs 2.10$ by the way.)

Fom that page

Sent from my Transformer Prime TF201 using xda premium
 

Diamondback

Retired Dev Committee Lead / Retired Senior Mod
Jan 17, 2010
4,476
6,631
virtuous-ten-studio.com
Guys.... if it was possible to crack an AES128 encryption with almost no money in less than a year of computing time, it wouldn't be the standard encryption for almost all encryted data... including governments of course.
Believe me guys, this won't work.
 

di11igaf

Inactive Recognized Developer
Sep 6, 2010
1,898
739
East Coast
I say we drop a bunch of twitter bombs on asus and see if that works

Sent from my Galaxy S2
This is about the best case scenario. Brute force just ain't gonna work, people have been considering this for years, and have learned its just not feasable. The only reason the atrix eventually was able to flash kernels was a leak from moto.
IPhones being jailbroken are more along the lines of us finding a root exploit, which is much easier and likely than cracking an encrypted bootloader. All's root gives is the ability to write to /system, we need an unlocked bootloader or a way around the encryption to flash to partitions(roms/kernels/recovery)
Cracking could take years, finding an expliot around the bootloader could potentially take years unless we get lucky.
Sucks either way for us and our best bet is bringing it to asus' attention an their public channels just like was done with HTC. I told myself I wouldn't buy another device where the company completely screws over the dev community and Asus is no different. The countdown has started.
 

biggem001

Senior Member
Sep 20, 2010
499
199
Guys.... if it was possible to crack an AES128 encryption with almost no money in less than a year of computing time, it wouldn't be the standard encryption for almost all encryted data... including governments of course.
Believe me guys, this won't work.

haha, i've been letting them think they can haha

considering cracking a HASH is considerably easier than 128-bit
 

Wordlywisewiz

Senior Member
Nov 25, 2011
1,150
224
Brattleboro
Well that's not nice


Let's start a thread

operation bombard Asus on twitter and support till they cave like HTC!

Sent from my Transformer Prime TF201 using Tapatalk
 

biggem001

Senior Member
Sep 20, 2010
499
199
i've already sent them an email requesting information on rooting.
i work for a university IT and we are interested in the TP for higher-access users.
of course, we need root for our VPN, Security and EMAIL apps to work... so until i can demo a device with all working applications needed for my users, then its a no-go.
i want to use the TPs in our office, but if we're forced to go Lenovo, (with their upcoming tegra 3) we will, as they are known to be much nicer to Uni's
 
  • Like
Reactions: Magnesus

Wordlywisewiz

Senior Member
Nov 25, 2011
1,150
224
Brattleboro
@Asus @Nvidia @Google unlock our boot loader! give us root! #TransformerPrime
Tweet it x
Sent from my Transformer Prime TF201 using Tapatalk
 

di11igaf

Inactive Recognized Developer
Sep 6, 2010
1,898
739
East Coast
rooting is a non issue, and they shouldn't be expected to leave us a root exploit. (Or Google, which is why they're patched asap). This is why Google devices ship with unlockable bootloaders, you don't need an 'exploit', which could(and would --its happened) lead to malware and malicious market apps getting root access to an unknown users device. How it should work(and does with pure Google devices), is you unlock the bootloader with fastboot-oem-unlock, agree you could potentially f#$k **** up, and bootloader is unlocked. Now all you have to do is literally change a 1 to a 0 in the boot.img (which you can now flash since your bootloader is unlocked)
Change 'ro-secure=1' to 'ro-secure=0' and now you can open a root terminal and /system is writeable(now you can safely push su binary and SU.apk.
Basically we need to be able to flash the damn kernel we can compile with the source Asus gave to us.
 

pandaball

Inactive Recognized Developer
Oct 18, 2011
2,388
1,735
Doesn't Asus release the source code for their kernel on their ftp?
 

biggem001

Senior Member
Sep 20, 2010
499
199
i actually cannot wait until we can implement CWM.

i really want to get started on some custom ROMS. since i use this device for work/school, thats what my ROM would be geared to
 

Top Liked Posts

  • There are no posts matching your filters.
  • 6
    here is an nvflash that supports cardhu(tegra 3), this is for linux, its all use so all i have, i just issued the --sync command and have not gotten an unknown device error, just getting nvflash started, then it just sits there.
    i havent used nvflash really at all, im just used to fastboot and adb, but hopefully this helps and i dont have much time today or tomorrow at all to mess with it.
    Linux users- in your rules.d(android-rules-whtever you want to make it) first line will get ADB working with the prime, second should allow it to be seen in APX mode:
    SUBSYSTEM=="usb", SYSFS{idVendor}=="0b05", MODE="0666"
    SUBSYSTEM=="usb", SYSFS{idVendor}=="0955", MODE="0666"

    Edit after messing with some nvflash commands I get a "USB write failed" error, which after Googling tells me Asus has used an SBK(secure boot key), rendering nvflash useless(unless asus gives it to us-128 bit encryption). Basically locked, encrypted bootloader and were f#$ked.
    I really hope I'm wrong.
    3
    I just sent Asus info and link to this thread and the other NVflash one. I told them unlocked users are very concerned & anxious for a new update on this issue & possible release date. hopefully Gary will let you all know something ASAP.
    2
    Lol @ all the people talking about brute forcing the keys. People have no idea at the true strength of AES encrypton. lets give some actual detail to this:

    1) Assume your CPU/GPU can try one septillion keys every second (10^24), current hardware is not even capable of a small fraction of this but let's say it can anyway. That's 1,000,000,000,000,000,000,000,000 keys per second.

    2) Now, assume you have 999,999 friends who are willing to help you, who also have the same impossibly fast CPU/GPU. So that's a million times faster, if you distribute the process over the internet.

    3) 16 bytes is 128 bits, so there's 2^128 possible keys to try. That's 3.4 * 10^48 keys.

    4) Do the math (you can use Google for this): 2^128 / 10^24 / 1000000 = 340282367 seconds.
    340282367 / 86400 = 3938.45 days
    3938.45 / 365 = 10.8 years for you and 999,999 friends to try every possible key.

    Now take into account that this assumes 0 delay from trying a key to verification that its found the correct key. This verification is a HUGE factor as well, for example:

    A GeForce GTX 460 working to break an AES-128 encrypted zip file is capable of trying 185,000 possibilities a second, however a AES-128 encrypted rar files its only capable of 3,500 possibilities a second.

    Jump up to 2 x GeForce GTX 570 working SLI and you get 495,000 possibilities a second for the AES-128 zip file and 14,000 for the AES-128 rar file.

    If you factor in the time it takes for any script to test and verify the keys its tested then its just not going to happen, not even if every single member of XDA developers devoted all their cpu and gpu time to this for years.
    1
    Yes I know you just have to restart udev service... But it seemed to me more accurate :D

    not even that much, udev as a rule automatically reloads the rules, you should only need to unplug and replug the device, and if that doesn't work then reload the rules (although most of the time completely unneeded) with
    Code:
    sudo udevadm control --reload-rules
    1
    might be interesting to try the sbkcheck from rayman
    http://xdaforums.com/showthread.php?t=1290503

    (note: IIRC there was a modified version at androidroot.mobi that could check for a "v3" SBK as well).