The units a scheduler hands the CPU to: separate address spaces, the process control block, and what threads share.

FCFS, SJF, SRTF and Priority on the same workload — watch the Gantt chart grow and the average waiting time move.

A ready queue that cycles every time quantum. Shrink the quantum and response improves while context switches pile up.

Page tables, the TLB and page faults — the machinery that page replacement has to manage when RAM runs out.

FIFO, LRU and Optimal deciding which frame to evict — including the reference string where more frames cause more faults.

First, best, worst and next fit carving holes in memory until a request fails with enough space free — then compaction.

Allocation, Max, Need and Available — the safety check that finds a safe sequence before any request is granted.

Collapse a resource-allocation graph into a wait-for graph and walk it for a cycle, with the four Coffman conditions on screen.

FCFS, SSTF, SCAN, C-SCAN, LOOK and C-LOOK plotted as head movement across the cylinders, with total seek distance.

Blocks, inodes, directory entries and journaling — what sits above the disk head that disk scheduling moves.

Time slicing, a first race condition, mutex locks and a two-lock deadlock — the overview the next three pages go deeper on.

Two threads run counter++ as load, add, store. Explore all twenty interleavings, then make them all correct with a mutex.

A bounded buffer guarded by empty, full and mutex semaphores — producers and consumers blocking and waking each other.

Five philosophers grab their left fork and deadlock. Number the forks, take the lower one first, and the cycle cannot close.

Protection rings and the trap into the kernel — how a process asks the OS for anything on this page.

L1, L2 and L3 cache lines and locality — the same hit-or-miss question page replacement asks, one level closer to the core.

Namespaces and cgroups: isolation built from the kernel features a scheduler and memory manager already have.