CS 452/652 Fall 2019 - Lecture 7

Memory Management

• C/C++ provide very good control over memory layout of structured data
  • use pointer casting to switch between unstructured and structured view
  • alternative: union declaration
• ARM 920T: all relevant unstructured data is 32-bit wide, so no padding or alignment problems
  • otherwise, GCC extensions could be used to increase control over memory layout
• techniques for dynamic data structures without heap
  • intrusive linkage, i.e., embed next (and/or other) pointers in data strucure
  • slab allocation: dedicated memory region for each relevant data type
    • stored freed objects in free list, use intrusive overlay (cast/union, see above) to manage
    • allocate first from free list, then from slab
  • use stack or static (file-)local memory for slab
  • when using static memory: do not share between tasks!

Mixing C and Assembler

Coding

• main.c, adder.c - basic argument passing in r0, r1
• adder6.c - beyond 4 arguments passed on stack
• main.c, asm_adder.S - manual assembler routine working with C caller
• main3.c, data_adder.c - memory-based data exchange between C and assembler
• special register access: msr/mrs for cpsr, spsr
• special register access: mcr/mrc for co-process registers cp15
• asm.c - embed assembler in C code; exchange data with local variables

Recommendations

• use embedded assembler via asm only for short code sequences that do not access sp
• use assembler routine linked as C routine for context-switch code

Kernel Start

• SVC mode, working stack, caches disabled, IRQs disabled
• set up (now, K1): UARTs, SWI handler
• set up (later): caches, timers, track, IRQs, IRQ handler
• create dedicated first user task (init task)
  • add to ready queue
• start kernel loop

Scheduling

• preemption: each kernel entry triggers re-scheduling
• but: no time-sliced preemption!
• after kernel entry and corresponding processing
  • task might be blocked → track
  • task might be ready → back on ready queue
  • but not both! single intrusive linkage element sufficient
• ready queue - keep simple: array of plain FIFO queues
  • no support for timeout/cancellation of blocking operations
  • no need to remove internal element from queue
  • only push to back of queue, pop from front