Re:Backdoor in upstream xz/liblzma leading to SSH server com...
by ptx on Saturday March 30, 2024 @12:04PM
(#39876115)
The latest commit is interesting (f9cf4c05edd14, "Fix sabotaged Landlock sandbox check").
It looks like one of Jia Tan's commits (328c52da8a2) added a stray "." character to a piece of C code that was part of a check for sandboxing support, which I guess would cause the code to fail to compile, causing the check to fail, causing the sandboxing to be disabled.
Re:Backdoor in upstream xz/liblzma leading to SSH server com...
by junon on Saturday March 30, 2024 @12:53PM
(#39876466)
Lasse has also started his own documentation on the incident.
Re:Backdoor in upstream xz/liblzma leading to SSH server com...
by josefx on Saturday March 30, 2024 @04:28PM
(#39878307)
Shouldn't they have tests running to ensure that the check works on at least some systems?
Re:Backdoor in upstream xz/liblzma leading to SSH server com...
by windenntw on Sunday March 31, 2024 @06:30AM
(#39883011)
What do you mean "tests"?
Re:Backdoor in upstream xz/liblzma leading to SSH server com...
by josefx on Monday April 01, 2024 @09:25AM
(#39893948)
Have a system were you wxpect the sandboxing to work and have an automated check that it compiles there?
Re:Backdoor in upstream xz/liblzma leading to SSH server com...
by ptx on Tuesday April 02, 2024 @07:42AM
(#39904607)
Part of the backdoor was in the tests. The attacker in this case could easily have sabotaged the test as well if a test was required.
Re:Backdoor in upstream xz/liblzma leading to SSH server com...
by account42 on Tuesday April 02, 2024 @07:34AM
(#39904555)
If your project becomes complex enough eventually you need tests for the configure step. Even without malicious actors its easy to miss that a compiler or system change broke some check.
https://git.tukaani.org/