Midterm #2 Review Topics

• Book and lecture topics (mostly Chapters 7-10)
  • semaphores in an RTOS
  • queues in an RTOS
  • other queue-like things: pipes and mailboxes
  • delaying tasks
  • the heartbeat timer
  • clock jitter, challenges of accurate timing
  • time-out functionality in RTOS function calls
  • timer callback functions
  • events, event management in an RTOS
  • tradeoffs between semaphores, queues, and events
  • using malloc/free vs fixed-size buffer management
  • problems with RTOS function calls from ISRs, alternatives
  • spotting bugs in code (problems at end of Chapter 7)
  • bug frequency in typical code; added challenges of developing, debugging real-time code
  • increasing customer confidence: process quality standards (ISO 9000, CMM)
  • Therac-25: case history of problematic software, lessons learned
  • application code design principles (8.1, 8.2)
  • supplemental design example: USB driver for HP Inkjet printers
  • problems with too many or too few tasks
  • preferred task structure (single blocking point, etc.)
  • relative overhead of task creation, task switching, other RTOS services
  • design principles reflected in detailed example in 8.3
  • encapsulating semaphores and queues
  • hard real-time scheduling issues
  • academic contributions: principally rate monotonic scheduling
  • jitter, variability in task periods
  • uC/OS scheduling approach with constant overhead
  • saving memory space, considerations for code and data
  • power saving modes commonly found in microprocessors, design implications
  • terminology: cross-compilers, cross-assemblers, linkers, loaders, locators
  • address resolution: mapping variables and functions to addresses
  • mechanisms to map code and data to ROM and RAM
  • segments, what they are and why they are important
  • assigning segments to memory address ranges
  • special cases: initialized data, constant strings, executing out of RAM
  • techniques of getting embedded software into the target system
  • terminology: ROM emulators
  • writing flash-resident code that can load new updates of itself
  • organizing code to facilitate testing and debugging
  • limitations and advantages of host-based testing
  • terminology: test scaffold code, script file parsers
  • limitations and advantages of instruction set simulators
  • "assert" macros: usefulness, implementation
  • useful tools: scopes, logic analyzers, in-circuit emulators, software-only monitors
  • challenges in extracting the data you want: signal pins very close together, ASICs, on-chip caches
• Lab and HW topics
  • synchronization and mutual exclusion: implementing acquire and release
    • special hardware support: test&set and swap instructions
    • software solutions
  • C topics
    • arrays and structs
    • pointers (to ints, arrays, structs, functions, pointers, ...)
    • linked lists
    • debugging code: common cases
  • mapping of C constructs to assembly
    • function calls, parameter passing, function return values
    • switch statements
    • global, local, and static local variables
  • YAK structure and organization
    • functional specifications
    • operational characteristics
    • organization of application code
    • required conventions in application code
    • execution details of application code
    • saving and restoring state in context switches
      • tradeoffs of alternatives to storing on stack vs ISR
      • tradeoffs of saving at different points in call sequence
      • stack frame and return address consistency
    • using semaphores, queues and events in application code
    • implementing semaphores, queues, and events; kernel internals
    • potential problems in using queues, semaphores, and events
    • general understanding of how to add new features and functions
• Assigned case studies (from supplemental reading)
  • Operational details of underlying technology
  • Facts of specific case study (what happened, why important?)
  • Implications (technological, social, political, etc.)