Yes, I know ... It’s sometimes difficult for some people to convince us that it seems natural for all of us, I need your help now SO community (or I will send mail soon).
One of my colleagues is convinced that the linux kernel code is not repeated, because it reads it somewhere for the last time when it looks into it, probably 7 years ago. Probably his reading was correct at that time, remember that the multi-core architecture was not widespread some time ago, and the linux project from the very beginning or so was not very well written and is fully compatible with all the fancy features.
Today is different. Obviously, calling the same system call from different processes running in parallel in the same architecture will not lead to undefined behavior. The Linux kernel is now widespread and known for its performance, although it works on multi-core architectures. This is my argument. But what would be yours to prove that objectively ?
I thought to show him some function in the linux kernel (on the lxr website ) as a mutex_lock () system call. Eveything is set up to make it work in a parallel environment. But the code may not be as obvious to beginners (as I am).
Please help me.. -)