|
Message
From: Evan Jones <ejones@u...>
Date: Tue, 3 Jun 2003 09:02:36 -0400
Subject: Re: [usb] USB1.1 Core Address Assignment Failing. Debugging suggestions?
I'm a little embarrassed now. Just 1 day after posting my original
request for help, we got the thing working!
On Tuesday, Jun 3, 2003, at 05:05 Canada/Eastern, Rudolf Usselmann
wrote:
> Function address 0 seems to be strange. Never have seen that ...
Well, the bizarre errors that we were seeing were caused by some
strange interactions between the
We eventually realized the problem because we were reading the "state"
variable inside usb_ctrl.v, and it never changed. It turns out that
when we examined the compile messages, it was being detected as being
stuck at ground, and optimized away. The problem was that when we were
intitially compiling the core for our Altera APEX 20K200E, the Quartus
II software reported it as not meeting the timing constraints. Then we
found the "VENDOR_ALTERA" define (I can't remember which file exactly
now) and after setting that, it compiled MUCH faster (because it was
optimizing away a bunch of critical things). After all our furious
debugging, we started with the USB1.1 function controller source from
scratch and it compiled fast enough the first time, and worked the
first time!
The important question now is, is that "VENDOR_ALTERA" setting
supposed to work on the APEX 20K200E chip? Because it doesn't appear to.
Thank you for your assistance, and for the USB1.1 core. Once we get the
system working more thoroughly, we will be submitting some
documentation for the core.
Evan Jones
--
Evan Jones: http://www.eng.uwaterloo.ca/~ejones/
"Computers are useless. They can only give answers" - Pablo Picasso
|
 |
Browse
