Kernel Mode explained

Key Takeaways

And you will see there is one important concept. If we want to deal with anything that has to do with kernel stuff, you cannot do it as the process itself, as the user process itself, as the user mode. You need to do it as the kernel mode. Right? So, you might say, "How?" Well, there are some thread There are some kernel threads, but most of kernel mode operations that happen to the same process are basically a bit that we flip the process. This It's the same process, but we just tell it, "Hey, you are now in kernel mode. You may read the kernel stuff." Because guess what? All the kernel stuff is technically mapped in every single process in existence. You cannot see the kernel stuff unless you switch to this kernel mode. Sorry. So, now with that mode in mind, that to switch from user mode to kernel mode, there is a cost. There is a cost of flushing the stuff from the registers to the process control block to say, "Hey, this is where we last were in. We're executing this instruction in the user space." And that instruction lived in the text block. There might be some translation going on. There might be some more fetches, memory fetches. So, that's the cost of the kernel. Essentially, the kernel switch. And we call this the kernel mode. Very critical to understand because the kernel mode executes every time the kernel needs to get involved. So, that's in a Every time you call a system call, every time you get a page fault, as we going to see, and every time the kernel takes over, essentially. And basically, these are the two cases. And I want you to understand the cost of the kernel switch. So, every time we mention a kernel switch, think of that cost. But if it happens 100,000 times without you noticing, that's a lot of time. Do the math.