Firmware update without DFU

[Jithin A]

Hi, I am new to the Nordic environment. So far I have done all the integration by going through your examples code which is very good.

Now, one thing I have no knowledge about is updating the firmware image(non-secure + TF-M) without using DFU.

I do not want FW update over DFU for some personal reasons.

Please guide me to any reference or documentation for how to build a bootloader and configure the application image to run with the bootloader and do OTA update.

I do not know from where to start for implementing the OTA update.

Thank you in advance.

My setup:

NRF SDK: v2.2.0

Board: nRF5340-DK

[Marte Myrvold]

Hi,

Can you clarify what you mean by not wanting to do fw update over DFU? DFU is device firmware update, so if you want to update your device, then this will be DFU, Do you mean that you do not want to update the device over serial (e.g. UART), but only over the air?

Best regards,

Marte

[Jithin A]

Hello, just to let you know that I need to implement this onto my project with high priority.

Please let me know if I am missing anything to explain my need.

Please help me with the integration ASAP.

[Marte Myrvold]

Hi,

Thank you for clarifying. As long as you have MCUboot enabled and save the image in the correct area in flash and with the correct metadata (header + trailer), then the transport method is irrelevant. To enable MCUboot you must set the following configuration in prj.conf:

CONFIG_BOOTLOADER_MCUBOOT=y

The metadata can contain information about booting information, image hash, etc. You can find more about this, both header and trailer, in MCUboot Bootloader Design. There you will also find information about image slots, swapping, and more.

As for area in flash, when you enable MCUboot with the configuration above, the partition manager will automatically configure partitions. You can see the partitions after you have built the application by running the following command:

west build -t partition_manager_report

If you are using VS Code with our extension you can also see it by selecting the Memory report action, and going to Partitions in the generated report. As an example, adding MCUboot support to the Blinky sample, the partitions will look lilke this:

  flash_primary (0x100000 - 1024kB):
+-------------------------------------------------+
| 0x0: mcuboot (0xc000 - 48kB)                    |
+---0xc000: mcuboot_primary (0x7a000 - 488kB)-----+
| 0xc000: mcuboot_pad (0x200 - 512B)              |
+---0xc200: mcuboot_primary_app (0x79e00 - 487kB)-+
| 0xc200: app (0x79e00 - 487kB)                   |
+-------------------------------------------------+
| 0x86000: mcuboot_secondary (0x7a000 - 488kB)    |
+-------------------------------------------------+

  sram_primary (0x40000 - 256kB):
+--------------------------------------------+
| 0x20000000: sram_primary (0x40000 - 256kB) |
+--------------------------------------------+

As you can see, here the primary MCUboot image slot starts at 0xc000, and the secondary image slot at 0x86000.

Best regards,

Marte

[Jithin A]

Thank you so much for the reference

I see it’s a lot of things to take in.
Anyway after going through the bootloader reference you shared earlier, I have below 2 questions

1. How to insert header and trailer metadata to my application image?

2. How to write any data into the specific address in the flash?
—> Currently I am using PSA protected storage APIs for accessing flash inside TF-M and I see I can not specify the destination address in any of PSA PS APIs.

NOTE: I have TF-M partition in my application and anything I need to do with the flash are should be done inside TF-M partition.
I can not use non-secure partition due to some security purposes.

Please guide me through this.

I appreciate your support.

[Marte Myrvold]

Hi,

Jithin A said:

1. How to insert header and trailer metadata to my application image?

You can use imgtool to sign images and add metadata, see the imgtool documentation. Additionally, you can generate private and public keys with imgtool. Make sure to add the --pad flag for the image trailer.

For an example on how to use imgtool for manually signing you can check out one of my colleagues examples: mcuboot_manual_sign. Please not that in that example, mcumgr is used, so the necessary commands will differ somewhat from what you should use. See the documentation above for more information about the different flags and which you should use.

Jithin A said:

2. How to write any data into the specific address in the flash?

If you use the function dfu_target_write(), the firmware image will be written into MCUboots secondary slot, see MCUboot-style upgrades for more information on this.

Best regards,

Marte

[Jithin A]

Thank you so much.

I am on the right place.

let me try it on my end and get back if any.

[Jithin A]

Hi Marte,

Thank you so much for your support I ran a blinky application with mcuboot enabled and is working fine.

Please see the below logs after the boot

*** Booting Zephyr OS build v3.2.99-ncs1 ***
Attempting to boot slot 0.
Attempting to boot from address 0x8200.
Verifying signature against key 0.
Hash: 0x8d...d1
Firmware signature verified.
Firmware version 1
Setting monotonic counter (version: 1, slot: 0)
*** Booting Zephyr OS build v3.2.99-ncs1 ***
I: Starting bootloader
I: Primary image: magic=unset, swap_type=0x1, copy_done=0x3, image_ok=0x3
I: Secondary image: magic=unset, swap_type=0x1, copy_done=0x3, image_ok=0x3
I: Boot source: none
I: Swap type: none
I: Primary image: magic=unset, swap_type=0x1, copy_done=0x3, image_ok=0x3
I: Secondary image: magic=unset, swap_type=0x1, copy_done=0x3, image_ok=0x3
I: Boot source: none
I: Swap type: none
I: Bootloader chainload address offset: 0x28000
*** Booting Zephyr OS build v3.2.99-ncs1 ***

I see it has also generated hex file for FOTA "app_moved_test_update.hex"

One last question, how can I program this hex file using your nRF Connect Desktop tool or if any other method.

This is because our App is still on development and in the meanwhile I need to test OTA so I can confirm the image from the secondary slot is being booted on OTA.

I appreciate your support.

Thank you once again for guiding me all through

[Marte Myrvold]

Hi,

Great to hear that it works.

Jithin A said:

I see it has also generated hex file for FOTA "app_moved_test_update.hex"

If you are interested, you can find more information about the generated build files in Output build files. If you scroll down you will also find information about MCUboot output build files.

Jithin A said:

One last question, how can I program this hex file using your nRF Connect Desktop tool or if any other method.

You can program it using the nRF Connect Programmer in nRF Connect for Desktop, or you can use nrfjprog.

Best regards,

Marte

[Jithin A]

Hi,

I try programming the app_moved_test_update.hex image with nRF Connect Programmer and I see it does not boot the device and also I observe no logs.

Attached the screenshot of the programmer window.

In the above image you can see the Write option is disabled, I only have Erase & write option and which I think it is erasing mcuboot too.

I did not find any mcuboot image to program along side the application image.

Am I missing something here?

[Marte Myrvold]

Hi,

Please select Enable MCUboot in the Programmer.

You can also try using nrfjprog. For a list of commands see nrfjprog commands.

Best regards,

Marte