3 ms·
Not really a great "delay" implementation. If someone were to use such a crude looping system to delay a program (and I doubt any semi-modern operating system d
by dkersten 16y ago
Not really a great "delay" implementation. If someone were to use such a crude looping system to delay a program (and I doubt any semi-modern operating system does this) some NOP instructions would be much simpler.
Real delay operations context switch to another process or thread for a minimum of the desired amount of time. A better implementation for single-tasking systems would be something like the following pseudocode:
; we assume the realtime clock is already setup to trigger timer interrupts at a regular interval (short enough that our delay function has the desired precision)
delay:
; Assume time to delay (in timestamp ticks, converted from milliseconds elsewhere) is in ebx
RDTSC eax ; Read timestamp into ebx (not the real way to use RDSTC!)
ADD ebx, eax ; ebx now contains the timestamp coutner value of when the delay should end
.delay_loop:
CMP eax, ebx ; compare the current timestamp to the end time
JG .delay_end ; if the current time is greater than the end time, we're done
; Otherwise we wait...
HLT ; Put the processor in lower power "sleep" state
; When the timer interrupt fires (at regular intervals), the processor is woken up
RDTSC eax ; Read new timestamp counter value (not the real way to use RDSTC!)
jmp .delay_loop ; Compare again and keep waiting
.delay_end:
RET ; return from delay function
- deadlock 16y agoHi.. Thanks for your comments and it really gave me a good thought. but i was trying to explain "how the processor waste it time when delay commands fired from OS". i gave an example related to 8085 system where it does no job operation. because processor has to do something. As per modern OS concern, it takes up some other job to increase its throughput.. as you mentioned . I do agree with you.. but lower power does not apply to processor. i feel, the processor switch power of monitor, mouse and other external device.. but CPU hardly go for sleep on delay state.. Yes, certainly it decreases processing..