AURIX™ Forum Discussions

[User14192]

Hello,
I'm trying to load .cmm file through TRACE32. When ever i tried, it showing "Emulator Debug Port Secure" error. Can any one tell me,
what could be the reason and how to solve this issue? Please find attached the screen shot for the same. Thanks in advance.

[#2]

Are you using the Trace on a brand new controller without software or bootloader?

[User14192]

Yes. I am flashing software in new controller(first time flashing in to TC275 controller). Could you please tell what could be the reason.

Regards,
Pradeep

[#4]

I had a similar problem von TC277, you don't have a bootloader so that there is no function able the Trace. The JTAG is a seperate module which isn't able to communicate with the rest of the uC.

My solution was to use the bootstrap loader, which exists on every device hardcoded for a first connection.
You have to use pins 14.0 and 14.1 either for ASC (UART) or CAN connection to flash a bootloader.
Infineon offers an application note with example software to flash by bootstrap (AP32272).

When you flashed it for a first time with the bootloader from the AP the connection with Trace works afterwards.

[User19107]

Hello,

I have written the invalid values to the BMHD0 where there is no values in the other BMHDs. This invalid value locks the ECU.


According to the schematic, it is not possible for the other startup procedure. The following image shows the schematic.


I need the support to come solve this issue. Thanks in advance.

[VincentWan]

Hi

What value did u program the BMHD0?
Did u erase the BMHD0 without programming?

What is the status of ESR0?
Is it HIGH or LOW?

How do you control HWCFG3? By DIP switch or the onboard pull-up? HWCFG3 need to be LOW to use HWCFG4 & HWCFG5 for booting.
Can you then change DIP switch of HWCFG4 & HWCFG5 to OFF(HIGH) & reset the target board.
Now try to connect using the debugger again.

[User19107]

Hello,

The BMHD1, BMHD2, BMHD3 are erased. Instead of writing the valid values, accidently i have written the invalid values in the BMHD0.


Since, we are using "Internal Start from Flash" method to flash the ECU, the HWCFG[4] and HWCFG[5] pins are left open. It is not possible to follow bootstrap methods.

While giving CPU Reset via Trace32, the PORST pin get toggled from HIGH-LOW-HIGH and the ESR0 pin is toggled from LOW-HIGH-LOW.

[VincentWan]

Hi

there are total 4 sets of BMHDx_ORIG & BMHDx_COPY, if BMHD0_ORIG is wrong, it will look at BMHD0_COPY, and the sequence follows to BMHD1/2/3.
you mentioned u erased BMHD1/2/3, this will prevent the TC29x from booting because of ERRORED condition in UCB for confirmation code (ie, 0x43211234 = unlocked), normally the debugger advises u from erasing them.

in an UCB errored state, the SSW recognizes this as failed Flash startup and consequently doesn’t boot the device as indicated by ESR0 LOW.

[User19107]

Hello,

Thanks for the information.
Is there any way to recover from this issue?

[VincentWan]

VincentWan wrote:
Hi

there are total 4 sets of BMHDx_ORIG & BMHDx_COPY, if BMHD0_ORIG is wrong, it will look at BMHD0_COPY, and the sequence follows to BMHD1/2/3.
you mentioned u erased BMHD1/2/3, this will prevent the TC29x from booting because of ERRORED condition in UCB for confirmation code (ie, 0x43211234 = unlocked), normally the debugger advises u from erasing them.

in an UCB errored state, the SSW recognizes this as failed Flash startup and consequently doesn’t boot the device as indicated by ESR0 LOW.

Hi,

the answer above is valid for TC39x.

Correction to my reply above:

it should read:

there are total 4 sets of BMHDx, if BMHD0 is wrong, it will look at BMHD1/2/3 in sequence.
you mentioned u erased BMHD1/2/3, this will prevent the TC29x from booting because of ERRORED condition in UCB for confirmation code (ie, 0x43211234 = unlocked), normally the debugger advises u from erasing them.

in an UCB errored state, the SSW recognizes this as failed Flash startup and consequently doesn’t boot the device as indicated by ESR0 LOW.

[VincentWan]

shirly wrote:
Hello,

Thanks for the information.
Is there any way to recover from this issue?

Hi,

if the UCB is errored, there is no way to recover it.
we advises customer to never erase the UCB.
For development purpose, only need to always use unlocked code (ie, 0x43211234) for confirmation code.

[User19107]

Hello,

Is it possible to come out of this issue?
Do you have any solution for this?

[VincentWan]

Hi,

if u have erased the UCB confirmation code. there is no way to recover the chip.

[User19107]

Hello,

I have erased only the BMHDs, not UCB. The UCB blocks are configured as NOP which means cannot be erased. Please check the following configurations in the .cmm file.

I believe that UCB portions are not erased. But i am not able to see the content of the memory.

[VincentWan]

Hi,

I mixed up BMHDx for TC39x & TC29x.
No UCB and confirmation code for TC29x BMHDx.

I see that u have configured for ABM mode in the BMHD with PINDIS = 1
since the memory range to be checked is wrongly programmed.
the SSW will enter Generic bootstrap loader mode @ P14.0 / P14.1

if you have access to Memtool & port pins 14.0 & P14.1, u can connect using Memtool to reprogram the TC29x.

[User19107]

Hello,

As i have already told that the pins are not taken out in the ECU Schematic.

[User19149]

L1800Turbo wrote:
I had a similar problem von TC277, you don't have a bootloader so that there is no function able the Trace. The JTAG is a seperate module which isn't able to communicate with the rest of the uC.

My solution was to use the bootstrap loader, which exists on every device hardcoded for a first connection.
You have to use pins 14.0 and 14.1 either for ASC (UART) or CAN connection to flash a bootloader.
Infineon offers an application note with example software to flash by bootstrap (AP32272).

When you flashed it for a first time with the bootloader from the AP the connection with Trace works afterwards.

Hi: L1800Turbo!
Can u send the AP32272 to me ! I can not find it !
Thanks!

[VincentWan]

shirly wrote:
Hello,

As i have already told that the pins are not taken out in the ECU Schematic.

Hi

I tried programming the values in the BMHD0 with BMHD1/2/3 erased and still can connect using the debugger.
not sure are u using script to connect to the board or using the CPU -> System window to connect?

[User19107]

Hello,

what is the State of HWCFG4 and HWCFG5 in your schematic?

[VincentWan]

Hi

I am using the IFX starterkit.
Both HWCFG4/5 are HIGH.

On your schematic, if u pulled HWCFG3 to LOW & with HWCFG6 pulled HIGH, HWCFG4/5 will be HIGH.

[User19107]

Hello,

The ALM3[6] (SRI frequency Out of Range) is raised during runtime. This Alarm is triggered sporadically.
What are the root causes for this Alarm?
Note: There are no other alarms triggered.

[NeMa_4793301]

This could be a problem with the clock monitor setup, or (rarely) with the internal 100 MHz backup oscillator.

Could this be an older AURIX that is affected by CCU_TC.002 in the errata sheet? (i.e., don't use CCUCON3/CCUCON4.xxxSEL=2 or 3)

[User19107]

Hello,
The CCUCON3 and CCUCON4 configurations are seems to be correct. The following are the values loaded in the Clock Monitor Registers.

SCU_CCUCON3.U |= 0x00281028;
SCU_CCUCON4.U |= 0x00141414;

Any other things to be checked?