Skip to main content

Section 3.5 Translation Lookaside Buffers

In this section we discuss how to speed up the paging process. Start from Paging: Faster Lookups (TLBs) chapter, intro and sections 19.1, 19.2.

Practice 3.5.1.

What is another name for the translation lookaside buffer?
  • address translation cache
  • memory management unit
  • hardware cache

Practice 3.5.2.

Practice 3.5.3.

Read section 19.3, about who handles the TLB misses.

Practice 3.5.4.

    True or False: A TLB miss must always be handled by some appropriate software code.
  • True.

  • False.

Practice 3.5.5.

    True or False: When a TLB miss is handled by the software via a trap handler, then the return from that trap handler works as "normal", i.e. it returns to execute the instruction after the one that caused the interrupt.
  • True.

  • False.

Read section 19.4, which discusses what is contained in a TLB entry. Make sure you also read the aside about "TLB valid bit vs page table valid bit"

Practice 3.5.6.

How does the hardware look for a possible entry in the TLB?
  • It goes through each row one at a time checking for a matching VPN.
  • The buffer is indexed by the VPNs, and the hardware jumps to the correct row right away.
  • It checks all rows at the same time, in search of a match.
Read section 19.5 about what happens during a context switch.

Practice 3.5.7.

Which of these are valid approaches to handling the problems occurring during a context switch?
  • Completely flush the TLB on each context switch.
  • Add information on each row connecting the row to the address space it comes from.
  • Use read/write/execute bits in the cache rows.
Read sections 19.6-19.8
You have attempted of activities on this page.