Process Management- Fork: Each
fork() creates a new process (2^n processes after n forks). - Fork Output:
for(i=0; i<3; i++) fork(); printf("hello\n"); prints "hello" 8 times (2^3). - Child/Parent Sharing: Only code, static data, and disk content are shared; heap, stack, and registers are copied.
- Process vs. Thread Switch: Process switching is more expensive due to address space switches, TLB flushes, and cache pollution.
- Multithreading Benefits: Even on single-core CPUs, threads are useful for blocking I/O operations.
- Trap Steps:
- Hardware: Save state → Change to kernel mode → Jump to trap handler.
- OS: Identify cause → Handle trap → Return to user mode.
| Memory & Paging- Address Translation (Virtual → Physical): Segment Number + Offset → (Segment Table) → Base Address + Offset.
- Page Table Entries: Calculate with (address space size / page size) × sparsity factor.
- Optimal Page Replacement: Replace the page that won't be used for the longest time.
- Page Table Structure: Multi-level tables reduce memory waste for sparse address spaces.
- TLB Hit: Virtual address → TLB → Physical address.
- TLB Miss: Virtual address → Page table → Update TLB → Physical address.
|
Synchronization- Safety Properties: Something bad never happens (e.g., mutual exclusion).
- Liveness Properties: Something good eventually happens (e.g., progress, bounded waiting).
- Semaphore Bug: Non-atomic
wait() allows multiple processes to pass, violating mutual exclusion. - Deadlock Avoidance: Break the circular wait condition (e.g., Dining Philosophers: limit philosophers at the table).
- Escape Room Strategy: One person counts all visitors (turns switch ON when they first visit; every other visitor turns it OFF once).
| File Systems- Hard Links: Point to the same inode; survive original file deletion.
- Soft Links: Point to a file path; break if the original file is deleted.
- Max File Size: Direct blocks + Indirect blocks + Double indirect + Triple indirect.
- Formula:
PD×S + PI×P×S + PDI×P²×S (PD=direct, PI=indirect, S=block size, P=pointers per block). - File I/O Operations:
- Linked Allocation: Begin(1), Mid(n/2), End(n)
- FAT: Begin(1), Mid(1), End(1)
- Indexed: Begin(6), Mid(6), End(4)
- Caching: Write-through (immediate write, consistency) vs. Write-behind (batched writes, performance).
- Log-Structured FS: Checkpointing reduces recovery time but impacts performance.
|
English with a size of 4.38 KB