Chen Xiao Fan wrote:
>>On 1/19/06, Hector Martin <hector@marcansoft.com> wrote:
>>
>>I've ported Bradley A. Minch's ASM USB firmware to "plain" 
>>asm (no structured macros, which GPASM did not seem to like). If 
>>anyone wants a copy, feel free to ask. I'll be releasing the 
>>whole project  code once I'm done with it. The USB stuff is 
>>pretty much done though.
>>
>>The original source can be found here:
>>http://pe.ece.olin.edu/ece/projects.html
> 
> 
> I'd like to see it. I'd like to see the difference between your 
> port and Ben Dugan's gpasm port. His gpasm port of Lab 2
> is posted in his web site here:
> http://www.curdes.com/gnupic/GnuPic.html
> 

I just rewrote the whole thing in plain assembly, with my own twists and 
differences. No structured macros (ENGR2210.inc) here. It is relocatable 
code and can run both from an interrupt and via polling. It uses a small 
set of standard code that I use (stddefs.asm/.inc) that contains 
definitions for a bunch of "standard" registers in the Access bank 
(R0-R7 plus T0-T3 and others). I also used symbolic constants for 
setting bits in registers ( (1<<UOWN)|(1<<DTS) instead of 0xC0 ), and 
used the Microchip (datasheet) names when possible. I made it full-speed 
code too, sicen I needed larger input reports (easy to switch back to 
low-speed). It also has been tested for non-usb-powered devices (I 
needed to do some voodoo to get the USB module to reset properly. Damn 
think kept crashing for some unknown reason related to interrupts). I'm 
also making use of the newer 18F instructions, which makes some things 
more compact (this is also my first project on the 18F by the way. I 
think I fell in love with it :) they fixed most of the great annoyances 
of the 16F core)

I am still thinking about improving it though. For example, right now I 
absolutely hate the giant nested "select" statements for processing the 
requests (right now I just have ASM equivalents of select). Especially 
for GET_DESCRIPTOR. I might include a table with start and length of 
descriptors to avoid a big chunk of code and hardcoded data. I'll also 
make it handle large descriptors (>256bytes).

Oh, and I learned a great deal of USB stuff from this :)

I'll clean it up and post it this weekend sometime.


-- 
Hector Martin (hector@marcansoft.com)
Public Key: http://www.marcansoft.com/hector.asc

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist