转贴:::关于DB2020(K800/K610/P990/M600/804S/904S etc)

starnetfish

普通会员
2005-10-05
205
0
0
INFORMATION REGARDING DB2020 PHONES (K800/K610/P990/M600/804S/904S etc.)


--------------------------------------------------------------------------------

This thread is intended for providing general information about the upcoming models based on the new DB2020 chip"set". It also serves the purpose of explaining the "how's" and why's regarding the unlocking/flashing of these.

First of all, DB2020 stands for Digital Baseband 2020. It's the next generation baseband "processor" from SE. It incorporates new security measures, so the old method of breaking into DB2010/DB2012 (it's predecessors) is useless against it.

Now, later SE security is/was actually very strong and therefore hard to break. It's RSA-based, meaning that it's based on public and private certificates and keys. These two are calculated based on huge prime numbers that cannot be "reversed", meaning that there is no way to calculate the private key based on the public key nor vice versa. Anything encrypted with the public key can be decrypted using the private key. The two "peers" share their public keys allowing them to decrypt eachothers "messages". For more information on RSA, please consult "http://en.wikipedia.org/wiki/Rsa".

As far as I've gathered, the "loaders" (programs) used by SE to communicate with phones either on the production line or at service centers are "signed" using a similar system (again, consult Wikipedia). An initial communication channel has to be set up, then a loader must be sent to the phone. It is through this loader that operations take place. E.g. if you want to flash a phone, you send a loader to it, this loader then listens for a flashfile to be transmitted and writes it to the memory.

Previous ways of breaking into SE phones were based on vulnerabilities in these loaders, somehow allowing developers to perform their own operations instead of their original intention. Unlocking for instance, requires an almost impossible to get access level and smartcard (holding the CSCA key, allowing a new security zone to be written and signed) for SE's own service solution, called EMMA (now counting version 3/III). Cloning this smartcard or getting the CSCA key out of it has yet to be done, if it is at all possible, the key never leaves the smartcard, calculations are made by the card itself (smartcards are very advanced, some hold 200Mhz processors, for example your average satellite-TV card).The card also holds the algorithm/mechanism for challenge/response authentication with the phone.

DB2020 is completely rewritten from scratch, I gather. They have been thinking "security" all the way through it's development as opposed to the apparent make-shift solutions found in earlier versions. This calls for new tactics by "SE-tool" developers.

The standalone version of SETool has been available for over a year now, it has served it's customers well, to say the least, but it has one major flaw; It's based on security holes found in SE security. The communication between SETool and a phone can easily be eavesdropped upon ("sniffing"), allowing other developers to copy and steal the exploitation code and then incorporate this into their own tools. The earliest public third-party SE service software was developed by "Daniel Henzulea"/zulea. At this time, SE (then simply "Ericsson") security was scarce. As it improved, new players came into the game, namely the_laser and Lead. Lately, these two have been supplying the solutions for breaking SE security, incorporating these into their own tools. zulea has been stealing these solutions and been incorporating them into his own, cheap, semi-working tool, ruining the third-party SE service software market, as anyone can afford and get it. His tool has it's own forum, filled with "less fortunate" people from the shallower end of the gene-pool living under the illusion that zulea is the biggest genious ever to have lived. When I feel down, I take a peek at that forum, reminding me how "fortunate" I really am .

Understandable enough, neither Lead nor the_laser finds this rather amusing. For SETool customers, this means that the old standalone solution must be put aside, at least for a while. The new solution has a server-client model, meaning that all important calculations are performed by the server. The communication between the software and phone will be phone-dependant, meaning it cannot be sniffed and applied to another phone. This way zulea (and others) are prevented from stealing the solution.

The new system will soon be in place, allowing at least flashing (unlocking is on it's way) of SE DB2020 models (Sharp models will not be supported at first). Some practical information on what this means for each user:

1: You will need an internet connection for servicing DB2020 models (the old models will remain supported in standalone mode). You don't need a fast connection, only small amounts of data will be transmitted.

2: You will need "credits"/"logs" for servicing DB2020 models. These prevent "stealers" from simply forwarding traffic. They also provide money for the SETool team (for buying prototype/new models, researching them and so on).

3: Each SETool card owner will have three free credits per day, allowing him/her to fully service one DB2020 phone per day without having to pay for it (if we look besides the cost of the card). One for "Complete phone" flashing, one for unlocking, and one for read/write GDFS. You could also use all three credits for unlocking, allowing you to unlock three phones per day for free. Prices for extra credits won't be steep, I gather, Unlocker will provide them soon enough. Credits won't span to the next day if you don't use them.

4: If a security hole is found in DB2020, support will be implemented into the standalone software.

Thats all I could think of, of any importance at least. I look forward to your feedback, but again; spare the crap . Also, if I'm wrong about anything, please correct me. I've taken a few educated guesses and shortcuts, it's late

I will update this post and wipe all others regularly to create better oversight and reability. [s:10]
 

starnetfish

普通会员
2005-10-05
205
0
0
是不是很多人都看过了,在别论坛上看到的,消息已经好多天了吧!!
 

starnetfish

普通会员
2005-10-05
205
0
0
说是DB2020的刷机软件很快就要推出来了,开始是SE的机子,然后才是sharp的机子,还要连接服务器才能工作,官方已经放出刷机log了!
在别地爬的,不知道事实,请大家斟酌!!!
 

lrg7986

普通会员
2004-04-08
307
0
0
好的。希望在SE。。。。。。。。。。。。。。。。 [s:20] [s:20] [s:51]
 

daxiongcn

普通会员
2006-08-08
246
0
0
关于DB2020电话的信息
(K800/K610/P990/M600/804S/904S等) 。注:文章中只有关于DB2020的,但不知为什么要包括这些机子。

这条螺纹意欲为提供关于即将来临的模型的总说明根据新DB2020 chip"set" 。它并且符合解释的需要"怎么是" 并且为什么看待这些的unlocking/flashing 。

首先, DB2020 代表数字式基带2020 年。这是SE下一代基带"处理器" 。它合并新安全措施, 因此老方法打破DB2010/DB2012 (这是前辈) 对它是无用的。

现在, 最新SE 安全is/was 实际上非常强和因此艰苦打破。它RSA 根据, 意味, 它根据公开和私有证明和钥匙。这两个被计算根据了无法"被扭转" 的巨大的质数, 意味, 没有方式计算私用密钥根据公开密钥亦不反之亦然。任何被编成密码以公开密钥可能被解码使用私用密钥。二"凝视" 份额他们的公开密钥给他们解码eachothers "消息" 。对于更多信息关于RSA, 请咨询"http://en.wikipedia.org/wiki/Rsa" 。

只要我聚集了, "装载者" (节目) 由SE 使用与电话通信或在生产线或在服务中心"签字" 使用一个相似的系统(再, 咨询Wikipedia) 。最初的通讯电路必须被设定, 装载者必须然后被送到电话。是通过这个装载者, 操作发生。即如果您想要闪动电话, 您送一个装载者到它, 这个装载者听为flashfile 被传送和给记忆然后写它。

早先方式打破SE 电话根据弱点在这些装载者, 以某种方法允许开发商进行他们自己的操作代替他们原始的意图。开锁为事例, 要求一几乎不可能得到访问级别并且smartcard (拿着CSCA 钥匙, 允许一个新安全区域被写和签字) 为SE 的自己的服务解答, 告诉埃玛(现在计数版本3/III) 。克隆这smartcard 或使CSCA 钥匙脱离它有做, 如果它是在所有可能, 钥匙从未留下smartcard, 演算由卡片做(smartcards 是非常先进, 一些拿着200Mhz 处理器, 例如您的平均卫星电视card).The 卡片并且举行algorithm/mechanism 为challenge/response 认证与电话。

DB2020 从头完全地被重写, 我会集。他们认为"安全" 所有方式通过它的发展依照被反对明显的暂时解答被发现在更加早期的版本。这要求新战术由"SE 工具" 开发商。

SETool 的独立版本是可利用的经过一年现在, 它为它服务是顾客很好, 认为最少, 但它有一个主要缺点; 它根据安全孔被发现在SE 安全。通信在SETool 和电话之间可能容易地被窃听("嗅"), 允许其它开发商复制和窃取开发代码和然后合并这他们自己的工具。最早期的公开第三方SE 服务软件由"丹尼尔开发了Henzulea"/zulea 。此时, SE (然后简单地"Ericsson") 安全是缺乏的。当它改善了, 新球员进入了比赛, 即the_laser 和主角。最近, 这两个供应解答为打破的SE 安全, 合并这些zulea 窃取这些解答和合并他们他自己的他们自己的工具, 便宜, 半工作的工具, 破坏第三方SE 服务软件市场, 因为任何人可能买得起和得到它。他的工具有它将拥有论坛, 用"较不幸运" 人民被填装从基因水池的更浅的末端居住在幻觉之下, zulea 曾经是最大genious 居住。当我感觉下来, 我采取偷看在那个论坛, 提醒我怎么"幸运" 我真正地上午。

可理解足够, 不主角亦不the_laser 发现这相当可笑。为SETool 顾客, 这意味, 老独立解答必须被投入在旁边, 至少有一阵子。新解答有一个服务器客户模型, 意味, 首要的演算由服务器执行。通信在软件和电话之间将是电话依赖, 意味它无法嗅和适用于其它电话。这方式zulea (和其他人) 防止窃取解答。

新系统很快将是到位, 允许至少闪动(开锁是对此是方式) SE DB2020 模型(锋利的模型不会支持起初) 。一些实用信息关于什么这手段为各名用户:

1: 您将需要互联网连接为为DB2020 模型服务(老模型将依然是支持在独立方式) 。您不需要快速的连接, 只有少量数据将被传送。

2: 您将需要"credits"/"logs" 为为DB2020 模型服务。这些简单地防止"stealers" 向前交通。他们为SETool 队并且提供金钱(为买prototype/new 模型, 研究他们等等) 。

3: 各个SETool 卡片所有者将有三自由信用每天, 允许him/her 充分地为一个DB2020 电话服务每天没有必须支付它(如果我们看除卡片以外的费用) 。一为"完全电话" 闪动, 一个为开锁, 和一个为读写GDFS 。您能并且使用所有三信用为开锁, 允许您打开三个电话每天为自由。价格为额外信用不会是陡峭的, 我会集, Unlocker 将提供他们足够很快。信用次日不会跨过对如果您不使用他们。

4: 如果安全孔被发现在DB2020, 支持将被实施入独立软件。

Thats 全部我能认为, 任何重要至少。我您的反馈, 但再盼望; 饶恕胡扯。并且, 如果我错误关于任何东西, 请改正我。我采取了几个猜测并且捷径, 是晚

我将更新这个岗位和通常将抹其他创造更好的失察和reability 。

[s:11] 不知道大家看不看得明~~