News for hackers. Stuff that matters.
Re:KARL – kernel address randomized link by terminalcommand on Wednesday June 14, 2017 @02:36AM (#14551006)
Re:KARL – kernel address randomized link by libeclipse on Wednesday June 14, 2017 @03:38AM (#14551224)
Can't this (or similar) be achieved on a run-time level so that binaries are reproducible?

Don't know much about the topic so apologies for any wildly stupid comment.

Re:KARL – kernel address randomized link by klodolph on Wednesday June 14, 2017 @06:19AM (#14551707)
Anything possible at compile-time is possible at run-time, the difference is convenience.

The main purpose of reproducible builds is so that you can ensure that you have a binary which contains what the package maintainers claims it contains, since you can reproduce the binary bit for bit. That's not necessary if you're building the binary yourself—if you don't trust the build process on your own computer, you have already lost.

Re:KARL – kernel address randomized link by mikepurvis on Wednesday June 14, 2017 @09:00AM (#14552375)
There are other reasons— for example not wanting to unnecessarily mutate buildid values, since those are the keys for debug symbol archival.
Re:KARL – kernel address randomized link by loa_in_ on Thursday June 15, 2017 @05:41AM (#14558976)
Can using unhibernate on -current accidentally do random things to my system?
Re:KARL – kernel address randomized link by hobarrera on Friday June 16, 2017 @10:02PM (#14573891)
From what I understand, it won't work in the sense that it'll either crash or outright fail to unhibernate completely.