topic Running S32E288 code from QSPI in S32Z/E

Running S32E288 code from QSPI

HiddenSquid — Mon, 02 Mar 2026 09:54:10 GMT

Hello,

My setup:
S32DS 3.6.4
S32 Flash Tool
S32 Debug Probe
S32E288-975EVB

I have a problem running a simple LED blinking code from QuadSPI Serial Flash with the S32E288-975EVB.

The code runs and builds with no issues when I use the S32 Debug Probe "S32E288_FreeRTOS_M33_Debug_RAM_S32Debug".

The problem:

I use the IVT tool to generate a BLOB. The BLOB contains:
1. the IVT image - I set the boot target as System Manager M33
2. a DCD file - for initialising RAM
3. The Application Bootloader binary - The .elf file format is too large to use as the application bootloader, which is why I use the raw binary. As I understand, the .elf adds some additional debug information, but shouldn't modify the code itself.

I then use the S32 Flash Tool to flash my BLOB on the board at location 0x0000_0000 (The start of QuadSPI_flash).

When I power on the board and attach to it using the S32 Debug Probe, I can see, that the Program Counter has gotten stuck at 0xefff_fffe, which seems to be outside the address range in general..? Also, when I attach to the board and set the program counter to the RAM_START, then everything seems to run fine at first, but at some point, the program counter ends up again at 0xefff_fffe. The Disassembly fails to read the command at that location.

Where or how can I find what leads to this erratic program counter value or does the value itself already somehow give an indication of the error?

I am also using freeRTOS, should that change anything in the bring up sequence?

What does the DEBUG_RAM version do extra, that my application bootloader or BLOB doesn't?

When writing my own application bootloader, what steps should I write in to my initialization script and what does the compiler take care of? For Example, do I need to initialize stack pointer, do I need to initialize any other peripheral besides RAM?

Re: Running S32E288 code from QSPI

Joey_z — Tue, 03 Mar 2026 02:09:33 GMT

Hi,HiddenSquid

Sorry for the reply late.

1.Could you please check your boot mode from QSPI is normal？

2.Do you boot another demo to generate Blob? Such as the Uart/LED demo in RTD.

3.Also, could you share your configuration of IVT?

Joey

Re: Running S32E288 code from QSPI

HiddenSquid — Tue, 03 Mar 2026 07:40:23 GMT

Hey @Joey_z ,

I will share most of my project with you. I have tried to make it as simple as possible for the time being, but the issue remains.

I believe my board works fine, as I previously had the greenVIP demo running and booting successfully from flash. For BOOT_CFG pins, dip switches at index [2] and [3] are set 1 for booting from flash.

I managed to step through the code at some point, when I manually overwrite the XPSR register. Could that be part of the issue?

Re: Running S32E288 code from QSPI

HiddenSquid — Tue, 03 Mar 2026 09:27:58 GMT

I have concluded, that the issue is in fact from the XPSR register. I don't have the T flag set and the ISR_NUMBER indicates a hard_fault, but for some reason I never really get to the hard fault. Also I don't understand, why or when does the T flag get unset?

Re: Running S32E288 code from QSPI

Joey_z — Wed, 04 Mar 2026 03:05:03 GMT

Hi,HiddenSquid

Thank you for your reply.

How do you set the start/entry pointer? it also would seem that the pointer is not configured correctly.

Joey

Re: Running S32E288 code from QSPI

HiddenSquid — Wed, 04 Mar 2026 07:50:40 GMT

I set the entry and start pointer both to 0x2500_0000, which is the start of RAM, which I have initalised with the DCD.
I have also confirmed that the entry pointer gets correctly set by checking the MC_ME.PRTN0_CORE0_ADDR register, which has a value of 0x2500_0000.

But it still doesn't explain how the Program Counter ever reaches 0xefff_fffe or when or how does the T flag get cleared?

Re: Running S32E288 code from QSPI

Joey_z — Wed, 04 Mar 2026 10:51:44 GMT

Hi,HiddenSquid

Thank you for your reply and information.

I will help you to check this issue and will reply to you as soon as if I get result.

Joey

Re: Running S32E288 code from QSPI

Joey_z — Mon, 09 Mar 2026 07:14:58 GMT

Hi,HiddenSquid

Please try to set the RAM entry pointer as the following picture, it is in the .map file.

Joey