🔒 Locked

Optimizing my blitter

Started by biskit Dec 1, 2000 at 9:11 AM 1 replies 250+ views

Original Post

Dec 1, 2000 at 9:11 AM

I have written a blitter for my library, and i was just hoping to get some input from some other people as to how I could possible optimize it further. Please don''t tell my to write in asm, because after I get a pretty optimized version in c, ill go ahead and rewrite the function in asm, with the inline assembler. All help will be appreciated. ----------------------------------- Implementation is everything. Period. Particle Toast

TwoFlower

Dec 1, 2000 at 9:26 AM

I think that really depends on what exactly your blitter needs to do. Obviously it would be quicker to read and write from aligned memory. Sequential access would be faster, as would seperate routines for different bit depths.

RLE encoding of the source could also be a good benefit because of less reads (or tightly packed data).

If it supports alphablending, then that is a whole set of optimisations in itself (lookups versus calculation etc.)

biskit

Dec 1, 2000 at 10:00 AM

Yes, I understand completely. My apology, the function code was supposed to be in the post. I reposted the post, with the code in it. If you could, take a look at it see if you see anything. Thanks a lot.

------------------------------------------------

Implementation is everything. Period.
Particle Toast

Topic Locked

This topic has been locked by a moderator. New replies are not allowed.