I suppose I've used lots of things without really knowing what they do, but god damn, it's kind of like CC had one job here and they didn't even know what the function did?
C macros are hard to track, especially if you don't review for it. It's a preprocessor statement, so what you put there is processed by a little script before things get passed off to a compiler and a linker, that processes the actual code. One of the things I do when writing things that need a macro other than a simple header guard (or sometimes, complex combinations of these) is to put error macros inside these and their inverse, run a compilation, and then see if it does what I expect. That is something I learned to do, so that I don't propose bugs / bother people with poor code, after fucking up a couple of times and luckily getting called out for being a fucking retard during patch review.
Now, the question I'd ask here is: who was reviewing? Who was the backstop? Did they "LGTM" the patches or did they actually spend time. Review takes more time than coding if you have competent reviewers.
Even if that seed was generated with dice.
This is the main reason why I said read it carefully. There is a list of functions there that are important to note. Most importantly, I think, the XOR key split with random data and the device clone over SD. Any of the former type key should be rolled too, and any SD ever used for device clone should be secure wiped and/or shredded (don't throw it in the trash.)
C macros are hard to track, especially if you don't review for it. It's a preprocessor statement, so what you put there is processed by a little script before things get passed off to a compiler and a linker, that processes the actual code. One of the things I do when writing things that need a macro other than a simple header guard (or sometimes, complex combinations of these) is to put
errormacros inside these and their inverse, run a compilation, and then see if it does what I expect. That is something I learned to do, so that I don't propose bugs / bother people with poor code, after fucking up a couple of times and luckily getting called out for being a fucking retard during patch review.Now, the question I'd ask here is: who was reviewing? Who was the backstop? Did they "LGTM" the patches or did they actually spend time. Review takes more time than coding if you have competent reviewers.
This is the main reason why I said read it carefully. There is a list of functions there that are important to note. Most importantly, I think, the XOR key split with random data and the device clone over SD. Any of the former type key should be rolled too, and any SD ever used for device clone should be secure wiped and/or shredded (don't throw it in the trash.)