Hubexec instruction fetching and REP

Take the following example that I've tacked to the end of test program:
		rep     @.rend,#0
		drvnot  #33
		drvnot  #56
		waitx   ##clock_freq
.rend

I'm thinking hubexec and cogexec will take exactly the same number of sysclock cycles per loop because instruction fetching triggers a fifo refill while the WAITX is still executing.

Any long executing instruction at the end of the loop will benefit from this. Possibly even a RDLONG although there may be congestion on that one. Admittedly, most are going to be waiting type instructions but a cordic command can take advantage of it.

We have the vastness of the internet and yet billions of people decided to spend most of their time within a horribly designed, fake-news emporium of a website that sucks every possible piece of personal information out of you so it can sell it to others. And they see nothing wrong with that.

Comments

  • Bugger, doesn't pan out in testing. I'm getting an extra 24 ticks per loop instead of the expected 6. Shrug.
    We have the vastness of the internet and yet billions of people decided to spend most of their time within a horribly designed, fake-news emporium of a website that sucks every possible piece of personal information out of you so it can sell it to others. And they see nothing wrong with that.
Sign In or Register to comment.