Pieter Conradie wrote:
> Hi Rob,
>
> I have found that if the jtag_khz setting is decreased to 1 kHz, then OpenOCD
> behaves better, but it is still not 100%.
>
> I have connected a signal analyser to capture the JTAG signals between
> OpenOCD rev 1399 and an AT91SAM9260 target using a Olimex JTAG-TINY-A. The
> jtag_khz setting was 5 kHz. See attached screenshots. There is an anomaly
> with RTCK on the first TCK pulse (see "ScreenShot0004.png"), but I don't know
> if this is a problem.
>
Pieter,
See FAQ #1 in the documentation
http://openocd.berlios.de/doc/FAQ.html#FAQ
I see two anomalies,
(1) a small pulse on rtck when reset is released - To be expected,
(screenshot 002)
(2) What is this pulse on NRST? about at about 25% of the screen?
But you speak about screen shot 004 - The clock TCK/RTCK sequence shown
in screenshot004 seems border line.
The left side of the screen seems ok...
You might set your timing analyzer up to look for this error condition,
((rising TCK) and (high RTCK)) *or* ((falling TCK) and (low RTCK))
And sample at some *HIGH* speed - ie: 1mhz w/5khz jtag clock.
Q: who makes your timing analyzer? It seems better then the cheap one I
purchased.
-Duane.
_______________________________________________
Openocd-development mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/openocd-development
