Re: [apache/incubator-teaclave-sgx-sdk] sgx_ucrypto::rsgx_rijndael128GCM_decrypt() panics with integer overflow (#252)

2020-06-29 Thread volcano
@NTSC-J I will fix this problem, thank you for reporting this error. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/incubator-teaclave-sgx-sdk/issues/252#issuecomment-651515702

Re: [apache/incubator-teaclave-sgx-sdk] Fix consttime_memequal integer overflow error (#253)

2020-06-29 Thread Yu Ding
Merged #253 into master. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/incubator-teaclave-sgx-sdk/pull/253#event-3495642148

[apache/incubator-teaclave-sgx-sdk] Fix consttime_memequal integer overflow error (#253)

2020-06-29 Thread volcano
`1 & ((res - 1) >> 8)` panic: attempt to subtract with overflow If res is 0, it will panic. #252 You can view, comment on, or merge this pull request online at: https://github.com/apache/incubator-teaclave-sgx-sdk/pull/253 -- Commit Summary -- * Fix consttime_memequal integer overflow bug

Re: [apache/incubator-teaclave] 怎么去证明teaclave是可行计算的环境呢? (#372)

2020-06-29 Thread Mingshen Sun
我们使用了 rustls [1],是 Rust rewrite 的 TLS 库,很大程度上避免了内存安全相关的漏洞,OpenSSL 的 heartbleed 漏洞 [2] 就是由内存安全问题引起的。 - [1] https://github.com/ctz/rustls - [2] https://heartbleed.com/ -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://g

Re: [apache/incubator-teaclave] 怎么去证明teaclave是可行计算的环境呢? (#372)

2020-06-29 Thread yc-huang
@mssun attested TLS用到的TLS的实现是谁提供的,实际是运行在什么环境的,会不会因为这块的实现有漏洞(比如以前open ssl的很多漏洞)从而导致泄密? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/incubator-teaclave/issues/372#issuecomment-651477482

Re: [apache/incubator-teaclave] Access Control 服务中的python脚本参与实际环境中的权限校验吗 (#374)

2020-06-29 Thread Mingshen Sun
通过 `include_str!` (https://doc.rust-lang.org/std/macro.include_str.html)。 https://github.com/apache/incubator-teaclave/blob/80fe5e9b78f3d5a023a07f9b44cabf4bd994a919/services/access_control/enclave/src/acs.rs#L33 -- You are receiving this because you are subscribed to this thread. Reply to this e

Re: [apache/incubator-teaclave] Access Control 服务中的python脚本参与实际环境中的权限校验吗 (#374)

2020-06-29 Thread ly137062117
请问下,model.conf 文件是如何编译到 enclave 中的呢?配置文件也能静态编译吗? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/incubator-teaclave/issues/374#issuecomment-651473037

Re: [apache/incubator-teaclave] start service error (#373)

2020-06-29 Thread Mingshen Sun
The main reason is that the kernel may be upgraded when you reboot the system. You can setup your Linux distro to manually upgrade kernel. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/incubator-t

Re: [apache/incubator-teaclave] start service error (#373)

2020-06-29 Thread gjj
> sudo systemctl restart aesmd it work. thanks but next time when i reboot the system. i have to continue to reload SGX_ Driver and restart aesmd? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/i

Re: [apache/incubator-teaclave] start service error (#373)

2020-06-29 Thread Mingshen Sun
Also, try to restart the AESM daemon which helps to manage architectural enclaves : `sudo systemctl restart aesmd`. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/incubator-teaclave/issues/373#issu

Re: [apache/incubator-teaclave] start service error (#373)

2020-06-29 Thread gjj
> There is something wrong with your SGX environment. Please check if the SGX > device (`/dev/isgx`) still exists. > > Also, I'll come up a fix to provide more information on the error. The SGX environment and teaclave of the server were running normally. Later, after the server was shut down a

Re: [apache/incubator-teaclave] Access Control 服务中的python脚本参与实际环境中的权限校验吗 (#374)

2020-06-29 Thread Mingshen Sun
`acs_engine.py` 和 `model.conf` 都是静态编译进 enclave 里的。 -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/incubator-teaclave/issues/374#issuecomment-651418057

Re: [apache/incubator-teaclave] start service error (#373)

2020-06-29 Thread Mingshen Sun
There is something wrong with your SGX environment. Please check if the SGX device (`/dev/isgx`) still exists. Also, I'll come up a fix to provide more information on the error. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHu

Re: [apache/incubator-teaclave] 怎么去证明teaclave是可行计算的环境呢? (#372)

2020-06-29 Thread xinuo
@qinkunbao 非常感谢解答! 听你这样说,我应该明白了. 我可能被example所"迷惑",是我自己误解了.因为我看到是放在2个用户的key都放在同一个py文件内. 真实的场景应该是 1,用户1,用户2(相当于A1,A2)各自提交自己的payload(通过SDK方式)来创建自己的task 2,最终由平台(B)来runTask. 3,A1,A2各自等待返回的结果. 这也是我的理解,也欢迎各位指正. -- You are receiving this because you are subscribed to this thread. Reply to this email dir

Re: [apache/incubator-teaclave] 怎么去证明teaclave是可行计算的环境呢? (#372)

2020-06-29 Thread Qinkun Bao
我最近也在看SGX,这是我的理解。如有不对欢迎指出来。 A <-> B 假设A是Client,B是encalve里面的代码,首先你信任client,因为这是在你本地上的,换言之这是你自己。B是Teaclave,但是你怎么需要保证Teaclave是你相信的那个版本(比如没有篡改。。),以及它跑在一个可信的环境里面(真的在encalve里面)呢?这就需要Remote Attestation (RA),你可以查阅Intel SGX的相关文档。 但是问题还有,比如你说的key的问题。理论上应该只有A和B知道key,A是你自己的client,你不能通过网络明文传输key是

[apache/incubator-teaclave] Access Control 服务中的python脚本参与实际环境中的权限校验吗 (#374)

2020-06-29 Thread ly137062117
最近读Access Control模块源码时,发现几个不太理解的问题: (1)acs.rs 中权限校验的逻辑最终调用了acs_enforce_request() 和acs_announce_fact() 函数,而这两个函数是在acs_engine.py 脚本中实现的。请问下,teaclave 服务启动后,acs_engine.py 是被编译为二进制参与到权限校验中吗?如果没有被编译,那么当 py 脚本被人为篡改(例如篡改了acs_enforce_request() 和 acs_announce_fact()的逻辑),那么权限校验岂不是失效了? (2)model.conf 文件被加载时

[apache/incubator-teaclave] start service error (#373)

2020-06-29 Thread gjj
docker-compose -f docker-compose-ubuntu-1804.yml up --build ![image](https://user-images.githubusercontent.com/30582212/85995064-7a396600-ba30-11ea-988d-e2938d9c31ab.png) -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: htt

Re: [apache/incubator-teaclave] 怎么去证明teaclave是可行计算的环境呢? (#372)

2020-06-29 Thread xinuo
谢谢解答. 不过对于明文的key在脚本(builtin_private_join_and_compute.py)里还存在疑惑: 是不是只能保证这个py文件不被别人窃取,不然,别人得到这些key也是可以通过cli进行解密? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/apache/incubator-teaclave/issues/372#issuecomme