Author Topic: Attempting to write uTaskerV1.3.bin with LMFlash fails to program LM6965 board  (Read 21483 times)

Offline joey

  • Newbie
  • *
  • Posts: 4
    • View Profile
I've tried with the version downloaded from the website, and a GNU make ala make_uTaskerV1.3_GNU_LM3SXXXX.bat

In both cases it gets to either 7% or 8%, then pops up an message box "**ERROR**: Programming error 0xA."

It doesn't happen with other binary files.  I've tried it on multiple machines including an x64 Vista, and a x86 XP machine, with identical results.

It is quite mysterious to me, as I would have thought to LMFlash the contents of a binary file are irrelevant.

Has anyone else experienced this problem?  Does anyone know how fix it?
 



Offline mark

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3240
    • View Profile
    • uTasker
Hi Joey

Unfortunately I don't know of such a problem.

To be sure, I just did the following:
1) Downloaded the demonstration file from the web site (uTaskerV13.bin 51k in size)
2) Using Luminary Micro Flash programmer V3.0 I loaded to the LM3S6965 Ethernet EVAL Board.

This worked as normal and the code ran correctly.

Here are the settings (you could also try erasing the entire FLASH content before by using the Erase function on the Flash utility page, to ensure that no pre-loaded program can be interfering).
- Manual configuration
- JTAG FTDI interface
- JTag speed 1M
- Clock 6000000Hz
- On programming page - erase entire FLASH and Reset MCU after program.

I see no reason why a certain binary content should cause problems. If you can't find any deviation from the above which may explain something you may like to ask on the Luminary forum whether anything is known which could explain the phenomena.

Regards

Mark





Offline joey

  • Newbie
  • *
  • Posts: 4
    • View Profile
I am running LMFlash Build 718, the version currently on luminary's website.

I also had a problem when using openOCD, but it appears to be because it expects the file to end on a 4byte word boundary.  Adding two zero bytes to the end seem to make openOCD happy, but not LMFlash.

Offline mark

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3240
    • View Profile
    • uTasker
Hi

Thanks, I will install the latest version and re-test to see whether it is related to the latest version or not.

Regards

Mark


Offline mark

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3240
    • View Profile
    • uTasker
Hi

I am sorry to have to report that I can reproduce the same problem with the LM Flash Programmer Build 718. (usually at 15% the error 0xA would appear).

I was careful to not overwrite the older version of LM Flash Programmer which I have always used up to now and I could verify that this still works correctly with the test file.

Presumably it is necessary to have the binary in a slightly different format (perhaps Motorola style binary where it is divided in to blocks with a small header (?)). However I don't understand how this can work with GCC...

I have posted in the Luminary forum with reference to this thread to see whether someone can explain what could be happening.

Regards

Mark


Offline joey

  • Newbie
  • *
  • Posts: 4
    • View Profile
I programmed uTaskerV13.bin (with 2 bytes of padding on the end) with OpenOCD.   I can then verify flash against the .bin file in LMFlash just fine.  This implies to me that the format is still a straight binary, and that most likely there is a bug in the new LMFlash.



Offline joey

  • Newbie
  • *
  • Posts: 4
    • View Profile
I emailed luminaries support yesterday, and got an email this morning that LMFlash Build 757 is now on Luminaries website.  It fixes the problem.  There is much rejoicing.


Offline mark

  • Global Moderator
  • Hero Member
  • *****
  • Posts: 3240
    • View Profile
    • uTasker
Hi

Yes, I can confirm that I also have no problem with this latest version.

Rather strange problem... but glad that it is fixed.

Regard

Mark