Reply by Randy Yates February 15, 20112011-02-15
On 02/15/2011 02:30 AM, Mitja Nemec wrote:
> On Feb 14, 5:57 pm, Randy Yates<ya...@ieee.org> wrote: >> Hello, >> >> We're using CCS 3.3 and the Signum JET-C2000 JTAG solution. >> >> I'm having trouble connecting consistently to the F2809. I can connect >> to a working F2809 the first time, flash it using the flash tool >> inside CCS, and run the code. However, from that point on I can't stop >> the emulator, and once disconnected I can't reconnect. For example, >> this is a message I get when I attempt to reconnect: >> >> Error connecting to the target: >> Error 0x80000200/-1031 >> Fatal Error during: OCS, >> Device driver: Problem with the Emulation Controller. >> It is recommended to RESET EMULATOR. This will disconnect each >> target from the emulator. The targets should then be power cycled >> or hard reset followed by an emureset and reconnect to each target. >> >> I/O Port = 240 >> >> Board Name: F2809 JTAGjet Emulator >> Cpu Name: TMS320F2809_0 >> >> Abort: Close Code Composer Studio. >> Retry: Try to connect to the target again. >> Cancel: Remain disconnected from the target >> Diagnostic: Run diagnostic utility. >> >> This has happened on two different boards now. I am suspecting the >> problem is in TI's security mode / CSM, but as far as I know I did NOT >> lock the chip with a password. Is it possible that rogue code can >> write all zeros to the password section (which would disable all >> future debug)? >> >> I am grasping at straws since I don't know this specific processor >> very well. If anyone recognizes this problem, I sure would appreciate >> a pointer or some information on it. I've already searched TI's E2E >> forum, and while it seems others have had similar problems recently >> (December 2010), I didn't see a solution. >> >> --Randy Yaets > > Hi Randy, > > Have you checked 2809 errata. I know (and have personal experience) > that one revision has some difficulties how the CPU clock is > generated. I don't recall the details, but the symptoms were similar.
Hi Mitja, Thank you for the lead. We were also told that, e.g., if the clock were not initialized correctly this problem may occur. Alas, CLKOUT is a nice 100 MHz. --Randy -- Randy Yates Digital Signal Labs 919-577-9882 http://www.digitalsignallabs.com yates@digitalsignallabs.com
Reply by Mitja Nemec February 15, 20112011-02-15
On Feb 14, 5:57&#4294967295;pm, Randy Yates <ya...@ieee.org> wrote:
> Hello, > > We're using CCS 3.3 and the Signum JET-C2000 JTAG solution. > > I'm having trouble connecting consistently to the F2809. I can connect > to a working F2809 the first time, flash it using the flash tool > inside CCS, and run the code. However, from that point on I can't stop > the emulator, and once disconnected I can't reconnect. For example, > this is a message I get when I attempt to reconnect: > > &#4294967295; Error connecting to the target: > &#4294967295; Error 0x80000200/-1031 > &#4294967295; Fatal Error during: OCS, > &#4294967295; Device driver: Problem with the Emulation Controller. > &#4294967295; It is recommended to RESET EMULATOR. &#4294967295;This will disconnect each > &#4294967295; target from the emulator. &#4294967295;The targets should then be power cycled > &#4294967295; or hard reset followed by an emureset and reconnect to each target. > > &#4294967295; I/O Port = 240 > > &#4294967295; Board Name: F2809 JTAGjet Emulator > &#4294967295; Cpu Name: TMS320F2809_0 > > &#4294967295; Abort: &#4294967295; &#4294967295; &#4294967295; &#4294967295; &#4294967295; &#4294967295; &#4294967295; &#4294967295;Close Code Composer Studio. > &#4294967295; Retry: &#4294967295; &#4294967295; &#4294967295; &#4294967295; &#4294967295; &#4294967295; &#4294967295; &#4294967295;Try to connect to the target again. > &#4294967295; Cancel: &#4294967295; &#4294967295; &#4294967295; &#4294967295; &#4294967295; &#4294967295; &#4294967295; Remain disconnected from the target > &#4294967295; Diagnostic: &#4294967295; Run diagnostic utility. > > This has happened on two different boards now. I am suspecting the > problem is in TI's security mode / CSM, but as far as I know I did NOT > lock the chip with a password. Is it possible that rogue code can > write all zeros to the password section (which would disable all > future debug)? > > I am grasping at straws since I don't know this specific processor > very well. If anyone recognizes this problem, I sure would appreciate > a pointer or some information on it. I've already searched TI's E2E > forum, and while it seems others have had similar problems recently > (December 2010), I didn't see a solution. > > --Randy Yaets
Hi Randy, Have you checked 2809 errata. I know (and have personal experience) that one revision has some difficulties how the CPU clock is generated. I don't recall the details, but the symptoms were similar. Mitja
Reply by Randy Yates February 14, 20112011-02-14
Hello,

We're using CCS 3.3 and the Signum JET-C2000 JTAG solution.

I'm having trouble connecting consistently to the F2809. I can connect
to a working F2809 the first time, flash it using the flash tool
inside CCS, and run the code. However, from that point on I can't stop
the emulator, and once disconnected I can't reconnect. For example,
this is a message I get when I attempt to reconnect:

  Error connecting to the target:
  Error 0x80000200/-1031
  Fatal Error during: OCS,
  Device driver: Problem with the Emulation Controller.
  It is recommended to RESET EMULATOR.  This will disconnect each
  target from the emulator.  The targets should then be power cycled
  or hard reset followed by an emureset and reconnect to each target.

  I/O Port = 240

  Board Name: F2809 JTAGjet Emulator
  Cpu Name: TMS320F2809_0

  Abort:		Close Code Composer Studio.
  Retry:		Try to connect to the target again.
  Cancel:		Remain disconnected from the target
  Diagnostic:	Run diagnostic utility.

This has happened on two different boards now. I am suspecting the
problem is in TI's security mode / CSM, but as far as I know I did NOT
lock the chip with a password. Is it possible that rogue code can
write all zeros to the password section (which would disable all
future debug)?

I am grasping at straws since I don't know this specific processor
very well. If anyone recognizes this problem, I sure would appreciate
a pointer or some information on it. I've already searched TI's E2E
forum, and while it seems others have had similar problems recently
(December 2010), I didn't see a solution.

--Randy Yaets