nRF5340PDK Data Length Extension

RE: nRF5340PDK Data Length Extension

rick-w — Wed, 12 Aug 2020 17:46:01 GMT

Thanks Haakonsh.

I tried this and it has resolved my issue

Best Regards,

Rick

RE: nRF5340PDK Data Length Extension

haakonsh — Tue, 11 Aug 2020 08:59:23 GMT

Is this log from the net core?

Try increasing the CONFIG_SYSTEM_WORKQUEUE_STACK_SIZE or CONFIG_MAIN_STACK_SIZE in prj.conf of the hci_rpmsg sample.

RE: nRF5340PDK Data Length Extension

rick-w — Mon, 10 Aug 2020 13:16:39 GMT

My Bad - it's an assert in:

void arm_core_mpu_mark_areas_for_dynamic_regions()
__ASSERT(0, "Marking %u areas for dynamic regions failed %x %x \n",
dyn_region_areas_num, dyn_region_areas->start, dyn_region_areas->size);

it appears that if the code size trips over boundary 0x2CA00 (decimal 182748B), then the assertion is called.

Can you point ot the config file where this boundary is set.

Thansk - Rick

RE: nRF5340PDK Data Length Extension

haakonsh — Mon, 10 Aug 2020 13:10:52 GMT

Can you share the exception?

RE: nRF5340PDK Data Length Extension

rick-w — Mon, 10 Aug 2020 13:04:37 GMT

Thanks Haakonsh.

I believe that the exception I was seeing is due to the Arm MPU size, which is configured for a given project, and that by setting CONFIG_BT_DATA_LEN_UPDATE=y, I was overstepping the MPU boundary, which is currently set to 64KRam for the NET processor.

Is there any documentation on changing the MPU size of the ARM code, from within the SDK

Regards,

Rick

RE: nRF5340PDK Data Length Extension

haakonsh — Mon, 10 Aug 2020 12:14:43 GMT

I believe LLPM uses a tiny payload due to the low-latency nature. If you need more than a few bytes of payload you should not use LLPM.

[quote user=""]but this causes an exception to be issues immediately to the debug ouput UART[/quote]

Care to share?