SoftDevice fault on nRF52840 SD S140 v6.1.1

RE: SoftDevice fault on nRF52840 SD S140 v6.1.1

Øyvind — Tue, 09 Jul 2019 12:27:14 GMT

No, this can be due to the difference in "Time to erase one page" in flash.

nRF52832 Min 2.05, Max 89,7

Difficult to say what can cause the issue. Are you able to share your code, so I can try to replicate the issue on an nRF52840 DK?

RE: SoftDevice fault on nRF52840 SD S140 v6.1.1

PrFi — Tue, 09 Jul 2019 11:38:07 GMT

The code has been written according to documentation that have you mentioned. The weird thing is that the same code compiled on nRF52832_XXAA (SD s112 V6.1.1) doesn't cause any SD fault. Is it possible that this is a bug in the SoftDevice S140?

RE: SoftDevice fault on nRF52840 SD S140 v6.1.1

Øyvind — Tue, 09 Jul 2019 11:27:57 GMT

Have a look at Timeslot API timing and the Flash API timing in the SoftDevice Scheduling documentation. It also includes suggested intervals and windows.

I could not find anything saying that it's forbidden to schedule any flash operation within a granted timeslot.

Kind regards,
Øyvind

RE: SoftDevice fault on nRF52840 SD S140 v6.1.1

PrFi — Tue, 09 Jul 2019 09:56:07 GMT

Do I understand you correctly, that scheduling any flash operation is forbidden within granted timeslot (sd_flash_write and sd_flash_page_erase)? Where can I find documentation, which describe this limitation? According to the Info Center "SoftDevice timing-activities", I assumed that SoftDevice should take care about colliding activities.

RE: SoftDevice fault on nRF52840 SD S140 v6.1.1

Øyvind — Tue, 09 Jul 2019 09:35:51 GMT

Hello,

[quote user=""]Could you let me know what causes fault from that address?[/quote]

Your FDS Garbage Collector seems to collide with the SoftDevice radio timing. Please ensure that FDS_GC is not called within a radio timeslot.

Kind regards,
Øyvind