Hello,

On Mon, Nov 03, 2008 at 02:31:44PM +0100, ?yvind Harboe wrote:
> I've modified the patch to treat the current formatting with a bit more
> respect :-)
Not a problem with me. I'll try to do better next time ;-)
> 
...
> > The other place is within the xsvf_add_statemove. I need to investigate this
> > further, but my target seems to have blown a gasket. This will have to be
> > fixed first.

Ok, I have my target fixed, and did a little more investigation. 

My previous patch does issue jtag_add_tlr() whenever xsvf_add_statemove(TAP_TLR)
is called. Now, due to the side effects of jtag_add_tlr() this might not be
an entirely good idea :-(. Because at the end of many xsvf operations, there
is a call to xsvf_add_statemove(xsvf_to_tap[xendir]) or [xenddr], and it
seems that xendir and xenddr are "Test-Logic-Reset" by default.

Thus, jtag_add_tlr() gets called in the middle of command sequences and
resets the whole chain and returns all devices to bypass.

Can someone give me a pointer on how to handle this? Is there a way to put
the state machine into TLR, but without the side effects? Or would it be
better to substitute TAP_RTI in these places? Or did I miss something
entirely?

Peter
-- 
"Only wimps use tape backup: _real_ men just upload their important stuff
on ftp, and let the rest of the world mirror it ;)"
(Linus Torvalds, about his failing hard drive on linux.cs.helsinki.fi)

Attachment: pgpzRW7hCnYE2.pgp
Description: PGP signature

_______________________________________________
Openocd-development mailing list
[email protected]
https://lists.berlios.de/mailman/listinfo/openocd-development

Reply via email to