some problems about AES CCM

[LeeRJ]

1st problem:

There are some failed case when i run  \v2.9.1\modules\bsim_hw_models\nrf_hw_models\tests\zephyr_apps\54_aar_ccm_ecb\src\test_ccm.c

2nd problem:

If I reduce the amount of encrypted data in test_ccm_12b_encr_ble_255b_w_adata_nomac, the encryption process gets stuck.

[Sigurd Hellesvik]

Hi.

1st problem:

There are some failed case when i run  \v2.9.1\modules\bsim_hw_models\nrf_hw_models\tests\zephyr_apps\54_aar_ccm_ecb\src\test_ccm.c

What command do you use to run this test?

If I reduce the amount of encrypted data in test_ccm_12b_encr_ble_255b_w_adata_nomac, the encryption process gets stuck.

What is your use-case for reducing this the data?

Regards,
Sigurd Hellesvik

[LeeRJ]

Sigurd Hellesvik said:

What command do you use to run this test?

I didn't use any commands, I just opened the \v2.9.1\modules\bsim_hw_models\nrf_hw_models\tests\zephyr_apps\54_aar_ccm_ecb and build it and flash it into nrf54l15dk, it automatically executed the test and printed out the log shown in the picture. In addition, for these failed cases, I found that the code comments indicated that the failure was expected.

Sigurd Hellesvik said:

What is your use-case for reducing this the data?

I am developing an ESB program capable of encryption, but it fails when encrypting data of varying lengths. Therefore, I found this example for reference, only to discover that it also fails when encrypting data of different lengths, as illustrated in the above figure. I commented out the first 10 bytes, and the encryption failed.

[LeeRJ]

I used github.com/.../nrf52-esb-ccm-example on nRF52832 before, and recently I need to replace it with nRF54L15, so I need to adapt it. I found this problem during the adaptation process

[Sigurd Hellesvik]

LeeRJ said:

I used github.com/.../nrf52-esb-ccm-example on nRF52832 before, and recently I need to replace it with nRF54L15, so I need to adapt it. I found this problem during the adaptation process

Lets have a look at this instead of the tests for now.

What exactly goes wrong with this sample when you adapt it?

As an alternative:

Have you considered encrypting the package in the application before sending it, instead of doing it using the CCM?

[LeeRJ]

Sigurd Hellesvik said:

What exactly goes wrong with this sample when you adapt it?

When I encrypt packets of different lengths, NRF_CCM00->EVENTS_END sometimes does not generate properly