Jeff Epler wrote: > In principle, you can do epp in software on a so-called bidirectional > port... the linux driver seems to do this. I tried it once with the > pluto driver, but it didn't work. I assume I got some detail wrong, and > at the time I didn't "need" such a feature for myself anyway. > > but if it's a problem on enough parports, maybe we should look into it > again.. > I did this on my very earliest attempts at diagnostic programs when developing my hardware. There is a SUBSTANTIAL performance penalty. When you use the EPP / IEEE-1284 transfer mode, all the strobe and handshaking is handled by the chip, and consecutive data writes or reads can be done about every 800 ns. The ISA bridge and/or the par port chip slows everything down a lot, so every register access to the par port chip is in the same range. So, each byte write requires writing the data register, asserting the strobe, deasserting the strobe, even if you assume the hardware always responds in time (no handshaking). So, it is suddenly 2.5 us or so per byte. My boards need about 35 transfers at a minimum, so 35 x 2.5 = 88 us of port traffic alone. With some of the direction flipping, it actually is a bit more than that. When you throw in the PID loops, trajectory planning, etc. it can eat up a lot of (real) time.
Jon ------------------------------------------------------------------------------ SF.Net email is Sponsored by MIX09, March 18-20, 2009 in Las Vegas, Nevada. The future of the web can't happen without you. Join us at MIX09 to help pave the way to the Next Web now. Learn more and register at http://ad.doubleclick.net/clk;208669438;13503038;i?http://2009.visitmix.com/ _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
