On May 21, 2009, at 10:13 AM, Dirk Behme wrote:

Michael Bruck wrote:
On Thu, May 21, 2009 at 1:39 PM, Dirk Behme <[email protected] > wrote:
Michael Bruck wrote:
Dirk, did this happen when you used the drscan/irscan commands?
Not sure, but possible. Have a look to tap-enable event of

http://svn.berlios.de/svnroot/repos/openocd/trunk/src/target/target/omap3530.cfg
Did you issue any commands before that happened in addition to what is
in the config files ?
Can you send a log with debug_level 3 ?
Does your system produce a .stackdump file during the assert/crash
that you can send ?
Does the assert happen if you disable the irscan/drscan commands in
omap3530.cfg ?

If you like you can try it your own: Configure recent trunk (1875) with --enable-dummy and then

openocd -s lib/openocd/ -f interface/dummy.cfg -f board/ ti_beagleboard.cfg

This will result in

-- cut --
Open On-Chip Debugger 0.2.0-in-development (2009-05-21-19:06) svn:1875

BUGS? Read http://svn.berlios.de/svnroot/repos/openocd/trunk/BUGS

$URL: svn://svn.berlios.de/openocd/trunk/src/openocd.c $
Warn : JTAG command queued, while TRST is low (TAP in reset)
current endstate: RUN/IDLE
Warn : Tap/Device does not have IDCODE
Error: JTAG tap: omap3.jrc got: 0x00000000 (mfg: 0x000, part: 0x0000, ver: 0x0) Error: JTAG tap: omap3.jrc expected 1 of 1: 0x0b7ae02f (mfg: 0x017, part: 0xb7ae, ver: 0x0)
Error: trying to validate configured JTAG chain anyway...
Error: Could not validate JTAG scan chain, IR mismatch, scan returned 0xFF. tap=omap3.jrc pos=0 expected 0x1 got 3
Warn : Could not validate JTAG chain, continuing anyway...
openocd: jtag.c:889: interface_jtag_add_dr_scan: Assertion `field == out_fields + scan->num_fields' failed.
-- cut --

Given that this is only an assert
Well, at least OpenOCD stopped here, so it might be more that 'only' an
assert ;) As mentioned, with r1865 it still worked.
The assert highlights that superfluous data was passed while older
versions ignored that condition. This change was deliberate to track
down errors like the one you seem to have. The problem is that the
errors can be both in the code and in user commands and for the latter
assert() isn't informative enough and irscan/drscan should inspect
their input data more closely.

Would have been nice before making this change in public svn it would have been tested with all available configurations.

Best regards

Dirk



While I agree with the sentiment, there is a practical limit on this. No one owns all of the targets or has time to do that. Patches submitted to the list rarely get tested by more than 2 or 3 people. The only way we get full testing is by committing to trunk. With no real stable releases for users, this leads to a fair amount of frustration when trunk behaves badly (doesn't currently build on OSX, assertions firing to name a few as of right now). Once 0.2.0 is out the door, we'll be doing more frequent stable releases so that only developers of OpenOCD are living on trunk.

--
Rick Altherr
[email protected]

"He said he hadn't had a byte in three days. I had a short, so I split it with him."
 -- Unsigned



Attachment: smime.p7s
Description: S/MIME cryptographic signature

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

Reply via email to