SDC Assertion 13, 451

RE: SDC Assertion 13, 451

Simon D-M — Tue, 26 May 2026 08:12:47 GMT

Hi,

Thanks for the update! Glad to see you were able to solve your deadlock.

Best regards,

Simon D-M

RE: SDC Assertion 13, 451

Hayden Ball — Mon, 25 May 2026 09:18:03 GMT

To follow up: we haven't seen any failures since updating to the patch linked above.

The deadlock we were seeing was an issue we had introduced, which we've now resolved.

RE: SDC Assertion 13, 451

Hayden Ball — Fri, 15 May 2026 10:47:23 GMT

Thank you Simon D-M

RE: SDC Assertion 13, 451

Simon D-M — Fri, 15 May 2026 10:20:58 GMT

Hi,

The fix (ref:DRGN-28363) is now available on the 3.3 branch : link

Best regards,

Simon D-M

RE: SDC Assertion 13, 451

Simon D-M — Tue, 12 May 2026 08:30:45 GMT

Hi,

The fix has already been merged to our internal v3.3 branch.

It is not part of the current nrfxlib-pr, but it will be exported in the next one.

I'll let you know when it will be available in the v3.3 release branch

Best regards,

Simon D-M

RE: SDC Assertion 13, 451

Hayden Ball — Fri, 08 May 2026 14:19:32 GMT

Thank you. We're still investigating, but I think the deadlock issue is on our end.

I was asking about the first option, please - the original patch, but based off the version of the SDC/MPSL that is used in nRF Connect v3.3.0.

RE: SDC Assertion 13, 451

Simon D-M — Fri, 08 May 2026 10:47:13 GMT

Hi,

First can you elaborate a bit on:

[quote user="Hayden Ball"]Having applied the patch, we're seeing bt_hci_cmd_send_sync never complete.[/quote]

We don't really understand what you meant by that. bt_hci_cmd_send_sync is a generic Zephyr Host function used for sending a lot of HCI commands from the Host to the Controller. Then it is waiting for a command complete from the controller.
Does bt_hci_cmd_send_sync fail for all commands, or is there a specific one that doesn't work?

[quote user="Simon Donnet-Monay"]

My guess is that it has not been pushed to version 3.3.0. From what I understand it will be included in v3.4.0, but not v3.3.x.

However, I'm not 100% sure. So, I've also forwarded the question internally. I'll get back to you tomorrow to confirm it.

[/quote]

I was correct, it has not been merged to any branch/version yet (not even main)

[quote user="Hayden Ball"]Are you able to provide a new SDC patch for nRF Connect v3.3.0, please?[/quote]

Are you asking about the patch same we sent you before or a different patch that would fix the bt_hci_cmd_send_sync issue?

If you are asking about the second option, we tested the already provided patch in a sample where we were able to reproduce the issue. And it passed all our regular CI tests. So it would be weird if that caused the bt_hci_cmd_send_sync issue.

We can't fix a bug we can't reproduce. Please give more information on how to reproduce this bug.

Best regards,

Simon D-M

RE: SDC Assertion 13, 451

Hayden Ball — Thu, 07 May 2026 15:38:08 GMT

Thanks Simon D-M .

Are you able to provide a new SDC patch for nRF Connect v3.3.0, please?

RE: SDC Assertion 13, 451

Simon D-M — Thu, 07 May 2026 14:40:33 GMT

Hi,

My guess is that it has not been pushed to version 3.3.0. From what I understand it will be included in v3.4.0, but not v3.3.x.

However, I'm not 100% sure. So, I've also forwarded the question internally. I'll get back to you tomorrow to confirm it.

Best regards,

Simon D-M

RE: SDC Assertion 13, 451

Hayden Ball — Thu, 07 May 2026 13:43:07 GMT

Simon D-M : Can you confirm if this patch is included in the new nRF Connect v3.3.0 release?

RE: SDC Assertion 13, 451

Hayden Ball — Thu, 07 May 2026 10:27:14 GMT

To keep this thread updated:

Having applied the patch, we're seeing bt_hci_cmd_send_sync never complete.

I'm still trying to gather more information.

RE: SDC Assertion 13, 451

Hayden Ball — Fri, 01 May 2026 07:38:59 GMT

Thanks Simon D-M

We've given that a go and are seeing some issues - we're just confirming where the fault is coming from to rule out it being something on our end.

RE: SDC Assertion 13, 451

Simon D-M — Thu, 30 Apr 2026 11:32:04 GMT

(Sorry Simon is not able to answer right now, I'll take over this case for now)

Hi,

I believe one of my colleagues sent already relayed you our proposed patch.

But just as a reference, here is the patch: (link).

If you need any help or have any question on it, feel free to ask.

Best regards,

Simon D-M

RE: SDC Assertion 13, 451

Hayden Ball — Mon, 27 Apr 2026 10:05:16 GMT

This is great news, thanks Simon.

If you can provide a patch for the current NCS version, that'd be helpful.

RE: SDC Assertion 13, 451

Simonr — Mon, 27 Apr 2026 09:24:39 GMT

Hi again Hayden

Sorry for any confusion.

Over the weekend we have gotten a good understanding of this bug, and luckily the fix isn't too complicated either. We aim to merge the fix to the main branch of the nRF Connect SDK this week, and include it as a part of the NCS 3.4.0 release.

If you require a fix in the current NCS version, let me know and we'll see what we can do on that note.

Best regards,

Simon

RE: SDC Assertion 13, 451

Hayden Ball — Fri, 24 Apr 2026 12:30:36 GMT

Thank you Simonr .

I'm not fully following - what do you mean by "terminate the sync from the sync side"?

For our application, we wouldn't be able to terminate from the advertiser side, if that's what you mean?

If we were to "squash" the assertion by overriding `bt_ctlr_assert_handle` to not reset, would this be safeish to do, or is this likely to lead to undefined behaviour?

RE: SDC Assertion 13, 451

Simonr — Fri, 24 Apr 2026 09:21:49 GMT

Another update today. We have now reproduced the issue on our end, and are looking into workarounds. You can terminate the sync from the sync side or accepting the assert and resetting, but we are working to find a proper solution/fix to this issue as well.

Best regards,

Simon

RE: SDC Assertion 13, 451

Simonr — Thu, 23 Apr 2026 12:41:57 GMT

Hi again

We are currently looking into this, but here's a progress update. We think the host on your side calls "sync terminate" right after the controller lost synchronization, but before the host was able to pull the sync lost event. We're currently trying to reproduce that to see why that is, as whenever the sync lost event is triggered the sync context should be cleaned up.

I'll let you know when I have more information.

Best regards,

Simon

RE: SDC Assertion 13, 451

Hayden Ball — Wed, 22 Apr 2026 10:47:57 GMT

Your colleague has closed out Maintaining connections while synced with PAwR , suggesting to merge it in here.

I'm not entirely sure the two are related, but I appreciate your help.

RE: SDC Assertion 13, 451

Hayden Ball — Tue, 21 Apr 2026 14:32:59 GMT

A call to `bt_le_per_adv_sync_delete` is in progress, yes.

RE: SDC Assertion 13, 451

Simonr — Tue, 21 Apr 2026 13:57:35 GMT

This assert happens when the host gets a "synchronization lost" event. Are you doing PAST or making/stopping a synchronization when this assert occurs?

Best regards,

Simon