Price your future right. 60% off with code BLACKSCHOLES at USD checkout. Offer ends in 8d 09h 54m 24s8 days 9 hours 54 minutes 24 seconds.
Price your future right. 60% off with code BLACKSCHOLES at USD checkout. Offer ends in 8d 09h 54m 24s8 days 9 hours 54 minutes 24 seconds.

Sign in to watch this lesson
Memory Virtualization · 4:24
The VAX/VMS ideas (from OSTEP) that every modern kernel still uses: an invalid page 0 that turns null-pointer bugs into traps, the kernel mapped into every address space, demand-zero pages that get a frame only on first touch, and copy-on-write, which makes fork nearly free. Measured on an Apple M3: one minor fault per 16 KB page on first touch, fork of a 256 MB process in under 100 us versus milliseconds to copy it, and a few microseconds per copy-on-write fault later - which is why low-latency code pre-faults its memory and never forks a hot process.