On Fri, May 18, 2012 at 5:56 AM, Andreas Fritiofson <
[email protected]> wrote:
> On Fri, May 18, 2012 at 2:18 AM, Bill Traynor <[email protected]> wrote:
> > I'm receiving Error: interface can't tri-state 'nSRST' while testing the
> new
> > ftdi driver with openocd from git, flyswatter2 JTAG, and PandaBoard-ES.
> I
> > built this OpenOCD today from git using the following patches:
> > git fetch http://openocd.zylin.com/openocd refs/changes/35/535/3 && git
> > checkout FETCH_HEAD
> >
> > and build:
> > sudo ./configure --enable-maintainer-mode --disable-werror --enable-ftdi
> >
> > Start up command is:
> > sudo ./openocd -f ../tcl/interface/ftdi/flyswatter2.cfg -f
> > ../tcl/board/ti_pandaboard_es.cfg
> >
> > result:
> >
> > Open On-Chip Debugger 0.6.0-dev-00557-g61d38c5 (2012-05-17-17:08)
> > Licensed under GNU GPL v2
> > For bug reports, read
> > http://openocd.sourceforge.net/doc/doxygen/bugs.html
> > Info : only one transport option; autoselect 'jtag'
> > RCLK - adaptive
> > Using dbgbase = 0x80000000
> > force hard breakpoints
> > trst_only separate trst_push_pull
> > Info : RCLK (adaptive clock speed)
> > Error: interface can't tri-state 'nSRST'
> >
> > I see that the new ftdi/interface/flyswatter2.cfg contains the following
> 4
> > new lines:
> >
> > ftdi_layout_init 0x0818 0x017b
> > ftdi_layout_signal nTRST -data 0x0010
> > ftdi_layout_signal nSRST -data 0x0020
> > ftdi_layout_signal LED -data 0x0400
>
> This is because reset_config selects srst_open_drain (or leaves it as
> the default), while the flyswatter2 only has a push-pull output for
> nSRST, at least according to the configuration file. I wrote the
> flyswatter2.cfg based on what was implemented in the old ft2232
> driver. If the adapter really can tristate nSRST, an -oe or -noe
> (depending on polarity) specification should be added to
> "ftdi_layout_signal nSRST". Are there any schematics available for the
> flyswatter2?
>
> The signal definition in the adapers' config files tell the new ftdi
> driver what the adapter supports. The reset_config mechanism in
> OpenOCD is totally oblivious to what the capabilities of the
> adapter/driver(/target) are. This is one of many similar architectural
> flaws in OpenOCD that needs fixing sooner or later. As of now, the
> user must make sure that the reset_config matches what the adapter can
> provide.
> Edit: I see now that your reset_config has trst_only, so nSRST
> shouldn't get asserted at all. Maybe there's a logic error in my code
> for that case, I'll take a look. You could try adding srst (w/
> push-pull) to reset_config, or comment out the nSRST line in
> flyswatter2.cfg if you must avoid hardware reset.
>
I believe trst_only this is correct for the PandaBoard_ES. The schematic
for the board is found here: http://pandaboard.org/node/223/#versions
When I comment out the nSRST line in the flyswatter2.cfg, I no longer
receive the tri-sate error but continue to see the looping through of:
Debug: 402 86 command.c:145 script_debug(): command - runtest ocd_runtest
200
Debug: 404 86 mpsse.c:760 mpsse_flush(): write 199+1, read 81
Debug: 405 86 mpsse.c:743 write_cb(): transferred 200 of 200
Debug: 406 324 mpsse.c:729 read_cb(): raw chunk 2, transferred 0 of 81
Debug: 407 579 mpsse.c:729 read_cb(): raw chunk 2, transferred 0 of 81
Debug: 409 834 mpsse.c:729 read_cb(): raw chunk 2, transferred 0 of 81
Debug: 410 1089 mpsse.c:729 read_cb(): raw chunk 2, transferred 0 of 81
etc. etc.
> /Andreas
>
------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and
threat landscape has changed and how IT managers can respond. Discussions
will include endpoint security, mobile security and the latest in malware
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
OpenOCD-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/openocd-devel
