CCS/TMS320F28075: Flash programming: CCS failes, Uniflash OK

Frank,

Which version of UniFlash are you using?

Maybe try updating to the latest CCS (8.0.0.00016) to see if the problem you are seeing in CCS disappears.

Thanks,
Ricky

Hi Ricky,

myUniflash version is uniflash_sl.4.2.2.1692.exe (should be the latest).

I tried CCS 7 I already have installed for another project, and I found that it succeeds to program the devices.

However, I and my colleaques prefer to continue this project with CCS 6, so it would be nice to get this run again. I checked the On-Chip Flash settings in the Debug > Control Panel window and found no difference. With the Enable Verbose Output I get the messages below. In CCS 6, I miss the line:

C28xx_CPU1: PLL configuration status = 1. PLL configured successfully.

May be this is a hint whats wrong?

Regards,

Frank

CCS6:

C28xx_CPU1: Writing Flash @ Address 0x00088000 of Length 0x00000014 (page 0)
C28xx_CPU1: Erasing Flash Bank 0, Sector A
C28xx_CPU1: Erasing Flash Bank 0, Sector B
C28xx_CPU1: Erasing Flash Bank 0, Sector C
C28xx_CPU1: Erasing Flash Bank 0, Sector D
C28xx_CPU1: Erasing Flash Bank 0, Sector E
C28xx_CPU1: Erasing Flash Bank 0, Sector F
C28xx_CPU1: Erasing Flash Bank 0, Sector G
C28xx_CPU1: Erasing Flash Bank 0, Sector H
C28xx_CPU1: Erasing Flash Bank 0, Sector I
C28xx_CPU1: Erasing Flash Bank 0, Sector J
C28xx_CPU1: Erasing Flash Bank 0, Sector K
C28xx_CPU1: Erasing Flash Bank 0, Sector L
C28xx_CPU1: Erasing Flash Bank 0, Sector M
C28xx_CPU1: Erasing Flash Bank 0, Sector N
C28xx_CPU1: Error during Flash programming (Flash algorithm returned error code). Operation cancelled.
C28xx_CPU1: File Loader: Memory write failed: Unknown error
C28xx_CPU1: GEL: File: J:\workspace\firmware\c2000\scm_firmware\debug_flash\scm_firmware.out: Load failed.

CCS7:

C28xx_CPU1: Writing Flash @ Address 0x00088000 of Length 0x00000014 (page 0)
C28xx_CPU1: PLL configuration status = 1. PLL configured successfully.
C28xx_CPU1: Erasing Flash Bank 0, Sector A
C28xx_CPU1: Erasing Flash Bank 0, Sector B
C28xx_CPU1: Erasing Flash Bank 0, Sector C
C28xx_CPU1: Erasing Flash Bank 0, Sector D
C28xx_CPU1: Erasing Flash Bank 0, Sector E
C28xx_CPU1: Erasing Flash Bank 0, Sector F
C28xx_CPU1: Erasing Flash Bank 0, Sector G
C28xx_CPU1: Erasing Flash Bank 0, Sector H
C28xx_CPU1: Erasing Flash Bank 0, Sector I
C28xx_CPU1: Erasing Flash Bank 0, Sector J
C28xx_CPU1: Erasing Flash Bank 0, Sector K
C28xx_CPU1: Erasing Flash Bank 0, Sector L
C28xx_CPU1: Erasing Flash Bank 0, Sector M
C28xx_CPU1: Erasing Flash Bank 0, Sector N
C28xx_CPU1: Verifying Flash @ Address 0x00088000 of Length 0x00000028
C28xx_CPU1: Writing Flash @ Address 0x00080000 of Length 0x00000002 (page 0)
C28xx_CPU1: Data has been buffered at the end of the current data block for 64-bit aligned writes.
C28xx_CPU1: Writing Flash @ Address 0x000929cb of Length 0x00000840 (page 0)
C28xx_CPU1: Data has been buffered at the beginning of the current data block for 64-bit aligned writes.
C28xx_CPU1: Data has been buffered at the end of the current data block for 64-bit aligned writes.
C28xx_CPU1: Verifying Flash @ Address 0x000929CC of Length 0x00001078
C28xx_CPU1: Writing Flash @ Address 0x00088014 of Length 0x00003ff8 (page 0)
C28xx_CPU1: Verifying Flash @ Address 0x00088014 of Length 0x00004000
C28xx_CPU1: Verifying Flash @ Address 0x0008A014 of Length 0x00003FF0
(etc.)

Frank,

It sounds like we implemented a bug fix in newer versions of CCS for the TMS320F28075. If you already have all of the available updates for CCSv6, then unfortunately, you will need to upgrade to a newer version of CCS.

Thanks,
Ricky

Hi Ricky,

the upgrade to a newer version of CCS is a good solution - but only for you (the easy way out: no need to find out what's the reason, just replace the problem).
CCS 6 worked well with many of our devices over months (and still does), just in the latest days we have problems with some of the (new) devices. I am afraid that there might a problem in our hardware, and I am afraid that an update to CCS 7 only covers the problem, but does not solve it.
My question is: why CCS7 works, and why CCS6 not in every case? What was changed?

Hope you understand this point of view.

Regards,
Frank

Frank,

I want to start off by clarifying that my previous response was not meant to be an "easy way out" solution, but to determine if you already have the latest updates for CCSv6 or not, and if so, there might be bug fixes that we added in v7 that are not available in v6. This in turn could mean that you will need to upgrade to the latest version of CCS to continue with your project.

I apologize if it came out incorrectly; I definitely wanted to continue the discussion to figure out what the underlying issue is.

1) So, the first thing I wanted to confirm is if you have applied all available updates for CCS6.1.1. (You can find updates by going to 'Help -> Check for updates')

2) Next, you noted that in CCSv7, there is the following line:

"C28xx_CPU1: PLL configuration status = 1. PLL configured successfully."

This was added recently to provide additional information to users to confirm that PLL was configured correctly. This message was not available in CCSv6, but does not necessary mean that PLL was not configured correctly.

3) You had a question on "why CCS7 works, and why CCS6 not in every case? What was changed?"

We are looking through the list of changes we implemented between CCS6.1.1 and CCSv7 to see if we can figure out exactly what changed that might affect the F28075. I'll keep you updated on our findings.

4) You mentioned that it works for some of your F28075 devices and not others.

a) For the failing devices, is it 100% reproducible? and does it happen for all programs you load to it, or only specific programs?
b) Is it possible for you to generate the Debug Server logs for a passing case and a failing case in CCSv6? More information for Debug Server logs are available here (processors.wiki.ti.com/.../Troubleshooting_CCSv7

Thanks,
Ricky

Hi Ricky,

I have to apologize, I was a bit impatient and rash with my conclusions. To your questions:

1) I checked and installed TI Emulators 7.0.188.0, issues were reported with reference #: 94988.

2) You are right. The line "C28xx_CPU1: PLL configuration status = 1. PLL configured successfully." does never appear in CCS 6, even if it succeeds.

3) Looking forward...

4a) As far as I can see, the error is reproducible, and it happens for all the programs I tried.

4b) I followed your instructions and produced a lot of data for you:

ds6.log, ds7.log: (Try to) load 2 different programs to the same hardware, CCS6 fails, CCS7 succeeds.

ds6_ok.log, ds6_fail.log: load the same program to different hardware with CCS6, with success/failure.

ds_logs.zip

I hope you can find what you are looking for. Let me know if you need more information.

Thank you for the support,

Regards

Frank

Frank,

Thanks for generating the Debug Server logs. I will need some time to analyze them to see if it contains information on what is causing the difference in behaviour.

I'll keep you updated on my findings.

Thanks,
Ricky

Frank,

Sorry for the lack of updates on this.

I took a look at the logs you provided and I didn't see anything that jumps out at me when I compared ds6_ok.log and ds6_fail.log. Everything seems to be the same, except that the flash algorithm just seems to start returning an error for the failing case.

You mentioned that you generated ds7.log as well, but that was not part of the attachment you included. I'm not sure if it will help, but maybe you can attach it as well just for reference.

As for looking for changes between v6 and v7 that might explain the differences in behaviour. This is proving to be a little difficult, as we have moved systems between these releases, so it is taking a little longer to determine what has changed.

I'll continue to keep you updated.

Thanks,
Ricky

Ricky,

you are right, my ds_logs.zip does not contain the ds7.log, here it is:

ds7.zip

The issue persists, and I wonder if any hardware parameters (processor production batch, supply voltages) may be the reason that some can be programmed with CCS6 and some can not.

Thank you again for your effords to solve the problem,

Frank

Frank,

We finally found a way to look up source code in our previous source control system (sorry for taking so long). I started looking at the different between the v6 code and the current code, and they are mostly related to new device support (ie; for new device families like F28004x), but there are some changes that might be relevant. To narrow down the changes, can you let me know the exact version of v7 that was working for your devices? I want to make sure some of those changes were not added after a version that is working for you.

Thank you,
Ricky

Ricky,

you find the installation details of my CCS v7 below. Let me know if you need more information.

Thank you for your support,

Frank

  am335x PDK    1.0.4    com.ti.pdk.am335x.product_1.0.4.feature.group    Texas Instruments Inc.
  Analysis Suite    4.1.0.201705251038    com.ti.dvt2.analysis.suite.feature.group    Texas Instruments
  ARM Compiler Help    16.9.0.20161026170303    com.ti.cgt.arm.16.9.help.feature.group    Texas Instruments
  ARM Compiler Tools    16.9.7    com.ti.cgt.tms470.16.win32.feature.group    Texas Instruments
  ARM Compiler Tools    16.9.0.LTS    com.ti.cgt.tms470.15.win32.feature.group    Texas Instruments
  BeagleBone FTDI Drivers    2.8.30.20140130    com.ti.ftdi.drivers.win.feature.group    Texas Instruments
  C2000 Compiler Help    16.9.0.20161026171152    com.ti.cgt.c2000.16.9.help.feature.group    Texas Instruments
  C2000 Emulation Flash    1.0.0.5    com.ti.c2800.flash.feature.group    Texas Instruments
  C2800 Compiler Tools    16.9.7    com.ti.cgt.c2000.16.win32.feature.group    Texas Instruments
  C2800 Compiler Tools    16.9.1.LTS    com.ti.cgt.c2000.15.win32.feature.group    Texas Instruments
  C6000 7.4.0 Help Feature    7.4.0.201303051055    com.ti.cgt.c6000.7.4.help.feature.group    Texas Instruments
  C6000 Compiler Tools    8.1.6    com.ti.cgt.c6000.8.1.win32.feature.group    Texas Instruments
  CCS and SAT Common Components Feature    7.2.0.201705101800    com.ti.tools.common.feature.group    Texas Instruments
  CCS Documentation    7.0.0.20161213-3    com.ti.ccstudio.doc.win32.feature.group    Texas Instruments
  CCS Flash Content    8.0.0.1213    com.ti.dsflash.win32.feature.group    Texas Instruments
  CCS Launcher    7.0.0.20161020    com.ti.ccstudio.launcher.win32.feature.group    Texas Instruments
  CCS Utilities    1.0.0.20160210    com.ti.ccstudio.utils.cli.win32.feature.group    Texas Instruments
  CCStudio p2 Tool Feature    7.0.0.201611291427    com.ti.ccstudio.p2tool.feature.group    Texas Instruments
  CCSv7 Service Release Windows    7.2.0.00013    com.ti.ccstudio.installer.win32.feature.group    Texas Instruments
  Chromium Browser Feature    7.1.0.201705191300    com.ti.chromium.browser.feature.group    Texas Instruments
  Code Composer Studio Base Components    7.2.0.02057    com.ti.ccstudio.base.feature.group    Texas Instruments
  Code Composer Studio IDE ARM Components    7.2.0.201705101800    com.ti.ccstudio.tms470.feature.group    Texas Instruments
  Code Composer Studio IDE C2000 Components    7.2.0.201705101800    com.ti.ccstudio.c2000.feature.group    Texas Instruments
  Code Composer Studio IDE Main Feature    7.2.0.201705101800    com.ti.ccstudio.idemain.feature.group    Texas Instruments
  Code Composer Studio IDE Workflow    7.2.0.201705101800    com.ti.ccstudio.workflow.feature.group    Texas Instruments
  Compiler Tools On-line Documentation    1.1.5    com.ti.cgt.dmed.win32.feature.group    Texas Instruments
  Debug Server    7.2.0.2057    com.ti.ccstudio.debugserver.win32.feature.group    Texas Instruments
  DVT - Graph Visualization    4.1.0.201705251038    com.ti.dvt2.graph.visualization.feature.group    Texas Instruments
  DVT - Profiler Analysis Manager    4.1.0.201705251038    com.ti.dvt2.profileanalysismanager.feature.group    Texas Instruments
  DVT - Resource Explorer    4.1.0.201705251038    com.ti.dvt2.resource.explorer.feature.group    Texas Instruments
  DVT - System Analyzer    4.1.0.201705251038    com.ti.dvt2.uia.feature.group    Texas Instruments
  DVT - Trace Control    4.1.0.201705251038    com.ti.dvt2.trace.control.feature.group    Texas Instruments
  Eclipse Help System    2.2.0.v20160606-1100    org.eclipse.help.feature.group    Eclipse.org
  Eclipse Platform    4.6.0.I20160607-1100    org.eclipse.platform.ide    Eclipse.org
  EDMA3    2.12.1    com.ti.sdo.edma3.product_02.12.01.feature.group    Texas Instruments Inc.
  Equinox p2, Provisioning for IDEs.    2.2.200.v20160606-1311    org.eclipse.equinox.p2.user.ui.feature.group    Eclipse.org - Equinox
  GCC ARM Compiler Tools    4.9.3.2015q3_20150921    com.ti.gcc.arm.win32.feature.group    Texas Instruments
  GEF (MVC)    3.11.0.201606061308    org.eclipse.gef.feature.group    Eclipse GEF
  Graphical Modeling Framework (GMF) Runtime Third-Party Components    1.10.0.201606071959    org.eclipse.gmf.runtime.thirdparty.feature.group    Eclipse Modeling Project
  GUI Composer Backplane Feature    6.1.0.201705101800    com.ti.gc.backplane.feature.group    Texas Instruments
  GUI Composer Program Model Feature    7.2.0.201705101800    com.ti.gc.pm.feature.group    Texas Instruments
  IDE - Debug Server Integration Feature    6.0.1.201705101800    com.ti.debug.server.feature.group    Texas Instruments
  Image Analyzer    4.1.0.201705251038    com.ti.dvt2.via.ccs.feature.group    Texas Instruments
  IPC (Multicore and I/O) (IDE Client)    1.25.3.15    com.ti.rtsc.IPC.product.ui.feature.group    Texas Instruments
  IPC (Multicore and I/O) (Target Content)    1.25.3.15    com.ti.rtsc.IPC.product_1.25.3.15.feature.group    Texas Instruments
  NDK (IDE Client)    2.25.0.09    com.ti.rtsc.NDK.product.ui.feature.group    Texas Instruments
  NDK (Target Content)    2.25.0.09    com.ti.rtsc.NDK.product_2.25.0.09.feature.group    Texas Instruments
  Node.js    6.9.1    com.ti.ccstudio.nodejs.win32.feature.group    Texas Instruments
  p2appf    7.2.0.201705101800    com.ti.ccstudio.p2appf.feature.group    Texas Instruments
  Pin Mux    4.0.1470    com.ti.ccstudio.pinmux.win32.feature.group    Texas Instruments
  powerSUITE (IDE Client)    1.6.0.00    com.devkits.product.ui.feature.group    Texas Instruments Inc.
  powerSUITE (Target Content)    1.6.0.00    com.devkits.product_1.6.0.00.feature.group    Texas Instruments Inc.
  powerSUITE (Target Content)    1.1.0.00    com.devkits.product_1.1.0.00.feature.group    Texas Instruments Inc.
  ROV    4.1.0.201705251038    com.ti.dvt2.rov.feature.group    Texas Instruments
  RTSC/XDCtools (IDE Client)    3.32.2.25    org.eclipse.rtsc.xdctools.product.ui.feature.group    Texas Instruments
  RTSC/XDCtools (Target Runtime Support)    3.32.2.25    org.eclipse.rtsc.xdctools.product_3.32.2.25.feature.group    Texas Instruments
  RTSC/XDCtools (Target Runtime Support)    3.32.1.22    org.eclipse.rtsc.xdctools.product_3.32.1.22.feature.group    Texas Instruments
  RTSC/XDCtools (Target Runtime Support)    3.32.0.06    org.eclipse.rtsc.xdctools.product_3.32.0.06.feature.group    Texas Instruments
  RTSC/XDCtools (Target Runtime Support)    3.30.6.67    org.eclipse.rtsc.xdctools.product_3.30.6.67.feature.group    Texas Instruments
  RTSC/XDCtools (Target Runtime Support)    3.30.5.60    org.eclipse.rtsc.xdctools.product_3.30.5.60.feature.group    Texas Instruments
  RTSC/XDCtools (Target Runtime Support)    3.30.1.25    org.eclipse.rtsc.xdctools.product_3.30.1.25.feature.group    Texas Instruments
  Shared Device Support    1.0.9    com.ti.ccstudio.shared.devicesupport.win32.feature.group    Texas Instruments
  Sitara device support    1.3.7    com.ti.ccstudio.sitara.devicesupport.win32.feature.group    Texas Instruments
  Spectrum Digital Emulators    5.2.0.14    com.sd.ccstudio.emulation.win32.feature.group    Texas Instruments
  SYS/BIOS (IDE Client)    6.45.1.29    com.ti.rtsc.SYSBIOS.product.ui.feature.group    Texas Instruments
  SYS/BIOS (Target Content)    6.45.1.29    com.ti.rtsc.SYSBIOS.product_6.45.1.29.feature.group    Texas Instruments
  SYS/BIOS (Target Content)    6.41.4.54    com.ti.rtsc.SYSBIOS.product_6.41.4.54.feature.group    Texas Instruments
  SYS/BIOS (Target Content)    6.41.0.26    com.ti.rtsc.SYSBIOS.product_6.41.0.26.feature.group    Texas Instruments
  TargetDB Default Props    6.0.0.20131206    com.ti.ccstudio.targetdb.props.default.win32.feature.group    Texas Instruments
  TI C2000 Device Support    4.2.4.0    com.ti.c2000.support.win32.feature.group    Texas Instruments
  TI Cloud Agent    0.5.1351    com.ti.ccstudio.cloudagent.win32.feature.group    Texas Instruments
  TI Emulators    7.0.100.1    com.ti.emulation.pack.win32.feature.group    Texas Instruments
  TI Target Content    1.0.0.201608011655    com.ti.targetcontent.feature.group    Texas Instruments
  TI-RTOS for C2000 (IDE Client)    2.16.1.14    com.ti.rtsc.TIRTOSc2000.product.ui.feature.group    Texas Instruments
  TI-RTOS for C2000 (Target Content)    2.16.1.14    com.ti.rtsc.TIRTOSc2000.product_2.16.1.14.feature.group    Texas Instruments
  TIREX Desktop    3.4.1.201706082019    com.ti.tirex.feature.group    Texas Instruments
  Trace Analyzer    4.1.0.201705251038    com.ti.dvt2.trace.ccs.feature.group    Texas Instruments
  XDCtools Core Update Feature    3.32.2.25    com.ti.xdctools_3_32.update.feature.group    Texas Instruments
  Xulrunner    24.0.0    com.ti.ccstudio.xulrunner.win32.feature.group    Texas Instruments

Frank,

Thank you for your patience.  

Could you please clarify two more things:

1) Did you use the same board/voltage-supply when you switch between CCSv6 and CCSv7 for the failing device?  

2) Will you be able to try XDS100 or XDS200 (if you have) instead of XDS560 for the failing devices with CCSv6 and see if that clears the issue?  Asking based on few past debugs.  

Based on your reply, we may request you to submit a few failing and passing devices to us.

Thanks and regards,

Vamsi

Vamsi,

thank you for your support. To your questions:

Vamsi Gudivada said:

1) Did you use the same board/voltage-supply when you switch between CCSv6 and CCSv7 for the failing device?

The power supply and also the emulator connection is the same in both cases. I just connect / disconnect to the device with CCSv6 and CCSv7 (different workspaces, but the same .out that is flashed). The 3V3 VDDIO is generated from a 5V5 source using a TLV62085RLTR.

Vamsi Gudivada said:

2) Will you be able to try XDS100 or XDS200 (if you have) instead of XDS560 for the failing devices with CCSv6 and see if that clears the issue?  Asking based on few past debugs.

I tried a SD XDS100v3 and a XDS 220 ISO too (no XDS 200 available). Both fail, as the XDS560v2 Traveller does.

Thanks and regards,

Frank

Frank,

Thank you for the details.

We made a fix in CCSv7 Flash Plugin to take care of below note given in 3.1.1 Fapi_initializeAPI() section of http://www.ti.com/lit/ug/spnu629a/spnu629a.pdf.  This is what is helping to always pass the program operation in CCSv7.  Will you be able to move to CCSv7?

NOTE: The accuracy of the on-chip zero-pin oscillators (INTOSC1 or INTOSC2) will not meet the accuracy requirements for Flash erase and program operations. Hence, when using INTOSC as the PLL clock source, the value initialized by the user for the u32HclkFrequency parameter should be 3% more (should be rounded to the next highest integer) than the configured SYSCLK. For example, when PLL is configured for a SYSCLK of 194MHz with INTOSC as the PLL clock source, instead of initializing the u32HclkFrequency parameter as 194MHz, initialize it as 200MHz.

CCSv6 Plugin fails to program when the INTOSC accuracy does not meet the Flash program requirements.

Thanks and regards,

Vamsi

Vamsi,

yes, we can move to CCSv7 (although there are more exiting things than software updates).

But I want to be sure that we do not have a problem with the hardware we produce. Thats why I am concerned that after a long period without problems suddenly some devices failed with CCSv6, and in our new production charge most of the devices do so.

Is our hardware OK if it can be programmed with CCSv7?

Can you please explain more detailed what happens, which parameters are important (processor production charge, power supply, etc.)?

Is there a means, e.g. a flash programming example program, we can cross check the oscillator problem you mentioned? Does it help to measure the on-chip oscillator frequency (how)? Can we find a reason to tell between passing and failing devices?

Thanks and regards,

Frank

Hello,

It looks like I have the same problem but with TMS320F28377SPZPT.

CCS v8 "solved" the issue.

I am adding details to my problem:

Problem:
Getting the very generic error just after the flash erase sequence and on the start of flash write. It stops and says:

C28xx_CPU1: Error during Flash programming (Flash algorithm returned error code). Operation cancelled.
C28xx_CPU1: File Loader: Memory write failed: Unknown error
C28xx_CPU1: GEL: File: C:\Users\ML\Documents\MATLAB\VDCU_f28377_v2_0.out: Load failed.

I cant reproduce it now but somewhere could see that it tries to write to some 8xxxxx sector and it fails at absolute beginning there.

I even tried to load the code to RAM but it loads very fast and during execution it stops rapidly on SW breakpoint even when having all breakpoints disconnected.

The same code can be loaded to the older revision PCB.

My test procedure:

• Soldered the new PCB
• Checked the voltage levels / transcients of power source that are good and similar to the last revision of my PCB
• Program XDS100v2 with the same configuration as on my previous PCB where it runs
• Looked into drivers management where I found the device recognized with the IDs are set.
• Run CCS 6.1.3 and run using the same launch configuration and did connection check, erase and blank check which were all successful
• Checked crystals that were all running
• Checked signal traces and signals from XDS that looked OK
• Resoldered MCU pins and the capacitors around to be sure that everything is connected. I think there shouldnt be soldering problems.

Between revisions I did some power source change and added pullups for FTDI2232 EEPROM as advised in reference scheme but I think this shouldnt be the case.

It would be best if this problem was written into the debugging connectivity issues to help others.