Le lundi 16 novembre 2009, Gordon Henderson a écrit :
> It's doing a mighty fine job on my project.

In mine too... In the earlier setps of my project I had first started 
converting some fast interrupt code into asm, expecting more compact and 
performing code, but I soon realized that there was little optimization I 
could bring to what the compiler had done except from removing a few useless 
BANKSELs - or maybe I'm not good enough at assembler, so I had mostly stuck to 
what the compiler had cooked for me.

Noticing I was gaining so little as that was such a hassle and maintenance 
annoyance, I reverted back to "all C code" and it's working fine.

However I would be curious ton compare the code size that sdcc generates for a 
PIC18 and its performance with what I could get from another compiler, but 
unfortunately a few "PIC compilers comparisons" I could find here and there 
didn't even mention sdcc :-/

I could try compiling my code with picc (which I had considered when falling 
upon sdcc PIC16 port bugs), but I understood from the doc that it would take 
quite a lot of code adaptation, and furthermore the "free" version of this 
commercial product is supposed to be "non-optimizing", which seemingly makes 
the biggest difference between the free non-commercial version and the 
expensive optimizing version...

Anyway if anybody has seen or made by himself some PIC18 performance 
comparison between sdcc and other compilers, I would be interested in seeing 
them.

-- 
Michel Bouissou (OpenPGP ID 0xEB04D09C)

------------------------------------------------------------------------------
Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day 
trial. Simplify your report design, integration and deployment - and focus on 
what you do best, core application coding. Discover what's new with
Crystal Reports now.  http://p.sf.net/sfu/bobj-july
_______________________________________________
Sdcc-user mailing list
Sdcc-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/sdcc-user

Reply via email to