the mode bit

The mode bit is a tiny piece of state inside the CPU — conceptually a single bit — that records which mode the processor is in right now: user (often written as 1) or kernel (often written as 0). It is the hardware's answer to a deceptively simple question that must have an instant, trustworthy answer millions of times a second: is the currently running code allowed to do privileged things? Like a wristband at an event that says staff or guest, it is checked at the door of every restricted action.

The bit lives in a special processor register and is consulted by the hardware on every attempt to execute a privileged instruction. If the bit says kernel, the instruction proceeds; if it says user, the hardware refuses and traps to the OS. Crucially, ordinary user code cannot simply set the bit to kernel — that would defeat the entire protection scheme. The bit changes only through controlled hardware events: it flips to kernel automatically when a trap, interrupt, or system call enters the OS, and the kernel flips it back to user (by executing a special return-from-trap instruction) when it hands control back to a program. Some architectures track more than two levels, so the same idea appears as a small current-privilege-level field rather than a literal single bit.

This humble bit is what makes dual-mode operation actually work — without a hardware-protected indicator of the current privilege, none of the user/kernel separation could be enforced. It is the linchpin: every privilege check, every trap, every system call ultimately consults it. And because it can only be changed by the hardware on a controlled entry to the kernel, a user program can never just promote itself to full power, which is precisely the property a secure OS depends on.