Bluetooth: Controller: Reduce assertion check code size #90599

cvinayak · 2025-05-26T12:18:52Z

Introduce development and fatal assertion classification in the Controller implementation.

Reduce Controller assertion check code size for ARM Cortex-M CPUs by raising undefined instruction exception or use faulting instruction address in the HCI vendor-specific h/w error event.

hci_ipc for nRF5340 Audio DK with nRF21540 FEM:

cmake -GNinja -DBOARD=nrf5340_audio_dk/nrf5340/cpunet -DCONF_FILE="nrf5340_cpunet_iso-bt_ll_sw_split.conf" -DEXTRA_CONF_FILE="overlay-nrf5340_cpunet_iso_nrf21540_ek-bt_ll_sw_split.conf" -DDTC_OVERLAY_FILE="boards/nrf5340_audio_dk_nrf5340_cpunet_nrf21540_ek.overlay" -DSNIPPET="bt-ll-sw-split" ../../samples/bluetooth/hci_ipc

Upstream memory use:

Memory region         Used Size  Region Size  %age Used
           FLASH:      248176 B       256 KB     94.67%
             RAM:       64016 B        64 KB     97.68%
           SRAM1:          0 GB        64 KB      0.00%
        IDT_LIST:          0 GB        32 KB      0.00%

PR memory use when Controller assertion check use ARM Cortex-M faulting instruction address (and weirdly enabling HCI vendor-specific error event reduces ~8KB):

Memory region         Used Size  Region Size  %age Used
           FLASH:      213980 B       256 KB     81.63%
             RAM:       64340 B        64 KB     98.18%
           SRAM1:          0 GB        64 KB      0.00%
        IDT_LIST:          0 GB        32 KB      0.00%

33.39 KB or ~ 13.78% reduction in code size.

Test/example Controller assertion:

> HCI Event: Vendor (0xff) plen 38
        01 24 00 00 00 01 1e 00 00 00 01 00 00 00 00 00  .$..............
        00 00 0c bc 0a 00 14 37 00 00 a1 8b 01 00 ae 8b  .......7........
        01 00 9a 00 00 81                                ......

where ae 8b 01 00 is the program counter register value in little-endian octets.
or

***** USAGE FAULT *****
  Attempt to execute undefined instruction
r0/a1:  0x00000029  r1/a2:  0x00000001  r2/a3:  0x00000000
r3/a4:  0x00000745 r12/ip:  0x00000003 r14/lr:  0x00008b71
 xpsr:  0x8100009a
Faulting instruction address (r15/pc): 0x00008b80

Commandline to use to get details:

arm-none-eabi-addr2line -e <elf file path> <faulting instruction address>

Example:

$ arm-none-eabi-addr2line -e zephyr/zephyr.elf 0x00008b80
/home/jack/workspace/zephyrproject/zephyr/subsys/bluetooth/controller/ll_sw/nordic/lll/lll_scan_aux.c:233 (discriminator 1)

When using LTO, arm-none-eabi-addr2line is not able to fetch the source file and line number. With LTO, .lst file has to be used to look up the faulting instruction address. Hence, enable CONFIG_OUTPUT_DISASSEMBLY=y to get the .lst generated for the sample.

Minimal peripheral_hr:

cmake -GNinja -DBOARD=nrf54l15dk/nrf54l15/cpuapp -DCONF_FILE=prj_minimal.conf -DEXTRA_CONF_FILE=overlay-bt_ll_sw_split-minimal.conf ../../samples/bluetooth/peripheral_hr

Memory region         Used Size  Region Size  %age Used
           FLASH:       96860 B      1428 KB      6.62%
             RAM:       17744 B       188 KB      9.22%
        IDT_LIST:          0 GB        32 KB      0.00%

This is an increase for the minimal, with this PR assertion checks are not removed in the minimal sample where as before it was removed when logging/console/serial was not included.

sonarqubecloud · 2025-06-05T13:46:03Z

Quality Gate passed

Issues
20 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
4.5% Duplication on New Code

See analysis details on SonarQube Cloud

Copilot

Pull Request Overview

Copilot reviewed 78 out of 78 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (14)

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

Thalley · 2025-10-09T12:49:38Z

doc/releases/migration-guide-4.3.rst

+    * :c:struct:`bt_hci_vs_fata_error_cpu_data_cortex_m` to
+      :c:struct:`bt_hci_vs_fatal_error_cpu_data_cortex_m` and now contains the program counter
+      value.


The "title" for this "section" is

* The following Kconfig option have been renamed:

Suggest to either start a new list, or modify the "title"

Good catch... I only read renamed and appended the new entry!

How is it with the bt_hci_vs - Are they considered stable APIs and must go through the deprecation process? Or is this good enough?

The whole Zephyr Vendor-Specific is so just between Zephyr Host and Zephyr Controller :-) .... not sure if this ever considered stable/unstable! Atleast, definitely the error event if generated makes the device/application unstable ;-)

I will not go with deprecation process for literal renames, these changes are caught straight at compile time, mandating a immediate migration task.

We should probably consider documenting how the hci headers should in general be considered in terms of public/internal APIs :D

Introduce development and fatal assertion classification in the Controller implementation. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]>

Copilot

Pull Request Overview

Copilot reviewed 78 out of 78 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (4)

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

Reduce Controller assertion check code size for ARM Cortex-M CPUs by using the undefined instruction exception. `arm-none-eabi-addr2line` commandline can be used to get the source file and line number. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]>

Make overlay-all-bt_ll_sw_split configs consistent with the one used in hci_ipc for nRF5340 SoC. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]>

Reuse hci_vs_err_assert implementation to generate the HCI vendor-specific hardware error event. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]>

Enable HCI vendor-specific h/w error event generation in samples and tests. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]>

Enable OUTPUT_DISASSEMBLY_WITH_SOURCE when using LTO to help find the assertion check in the source code corresponding to the faulting instruction address. arm-none-eabi-addr2line tool is not able to resolve the source file and line number when using LTO. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]>

Copilot

Pull Request Overview

Copilot reviewed 79 out of 79 changed files in this pull request and generated no new comments.

Comments suppressed due to low confidence (4)

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

sonarqubecloud · 2025-10-10T06:37:12Z

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
4.0% Duplication on New Code

See analysis details on SonarQube Cloud

cvinayak force-pushed the github_ll_assert_dbg branch 4 times, most recently from d9e0e27 to 4ff5213 Compare May 26, 2025 18:57

cvinayak mentioned this pull request May 27, 2025

Bluetooth: Controller: Fix missing connection handle invalidate #89481

Merged

cvinayak changed the title ~~Bluetooth: Controller: Introduce LL_ASSERT_DBG/ERR~~ RFC: Bluetooth: Controller: Introduce LL_ASSERT_DBG/ERR Jun 5, 2025

cvinayak marked this pull request as ready for review June 5, 2025 13:17

github-actions bot added area: Bluetooth Controller area: Bluetooth labels Jun 5, 2025

github-actions bot requested review from carlescufi, erbr-ot, thoh-ot, Tronil and wopu-ot June 5, 2025 13:18

github-actions bot assigned cvinayak Jun 5, 2025

cvinayak force-pushed the github_ll_assert_dbg branch from 4ff5213 to 61f7bbb Compare June 5, 2025 13:21

github-actions bot added the Stale label Aug 5, 2025

github-actions bot closed this Aug 20, 2025

cvinayak reopened this Sep 6, 2025

zephyrbot requested a review from ppryga September 6, 2025 04:29

cvinayak force-pushed the github_ll_assert_dbg branch 2 times, most recently from a11a45e to 1c6db0b Compare September 6, 2025 05:26

cvinayak changed the title ~~RFC: Bluetooth: Controller: Introduce LL_ASSERT_DBG/ERR~~ Bluetooth: Controller: Introduce LL_ASSERT_DBG/ERR Sep 6, 2025

zephyrproject-rtos deleted a comment from github-actions bot Sep 6, 2025

cvinayak changed the title ~~Bluetooth: Controller: Introduce LL_ASSERT_DBG/ERR~~ Bluetooth: Controller: Reduce assertion check code size Sep 6, 2025

cvinayak requested a review from Copilot September 6, 2025 05:50

This comment was marked as outdated.

Sign in to view

github-actions bot removed the Stale label Sep 7, 2025

cvinayak requested a review from Thalley September 7, 2025 08:46

cvinayak requested a review from carlescufi October 8, 2025 14:16

cvinayak dismissed Thalley’s stale review via 901bf11 October 8, 2025 20:30

cvinayak force-pushed the github_ll_assert_dbg branch from efe0faa to 901bf11 Compare October 8, 2025 20:30

cvinayak requested a review from Copilot October 8, 2025 20:30

Copilot AI reviewed Oct 8, 2025

View reviewed changes

zephyrbot added the area: Tests Issues related to a particular existing or missing test label Oct 8, 2025

cvinayak force-pushed the github_ll_assert_dbg branch 3 times, most recently from 5f4ddf4 to 6fc003d Compare October 9, 2025 08:28

zephyrbot added the Release Notes To be mentioned in the release notes label Oct 9, 2025

cvinayak requested a review from Thalley October 9, 2025 09:06

Thalley requested changes Oct 9, 2025

View reviewed changes

cvinayak force-pushed the github_ll_assert_dbg branch from 6fc003d to 89f2101 Compare October 9, 2025 13:00

Thalley previously approved these changes Oct 9, 2025

View reviewed changes

Bluetooth: Controller: Introduce LL_ASSERT_DBG/ERR

bb449fe

Introduce development and fatal assertion classification in the Controller implementation. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]>

cvinayak dismissed Thalley’s stale review via 67896b4 October 10, 2025 04:26

cvinayak force-pushed the github_ll_assert_dbg branch from 89f2101 to 67896b4 Compare October 10, 2025 04:26

cvinayak requested review from Thalley and Copilot October 10, 2025 05:37

Copilot AI reviewed Oct 10, 2025

View reviewed changes

cvinayak added 5 commits October 10, 2025 08:31

samples: Bluetooth: hci_uart: Make bt_ll_sw_split conf consistent

b989a92

Make overlay-all-bt_ll_sw_split configs consistent with the one used in hci_ipc for nRF5340 SoC. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]>

samples: Bluetooth: hci_uart: Reuse hci_vs_err_assert

acc5441

Reuse hci_vs_err_assert implementation to generate the HCI vendor-specific hardware error event. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]>

samples: Bluetooth: hci_ipc: Enable HCI vendor-specific h/w error event

949d5a0

Enable HCI vendor-specific h/w error event generation in samples and tests. Signed-off-by: Vinayak Kariappa Chettimada <[email protected]>

cvinayak force-pushed the github_ll_assert_dbg branch from 67896b4 to bc187a3 Compare October 10, 2025 06:32

cvinayak requested a review from Copilot October 10, 2025 06:32

Copilot AI reviewed Oct 10, 2025

View reviewed changes

Thalley approved these changes Oct 10, 2025

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Bluetooth: Controller: Reduce assertion check code size #90599

Bluetooth: Controller: Reduce assertion check code size #90599

cvinayak commented May 26, 2025 •

edited

Loading

Uh oh!

sonarqubecloud bot commented Jun 5, 2025

Uh oh!

This comment was marked as outdated.

Uh oh!

Copilot AI left a comment

Uh oh!

Thalley Oct 9, 2025

Uh oh!

cvinayak Oct 9, 2025

Uh oh!

Thalley Oct 9, 2025

Uh oh!

cvinayak Oct 9, 2025

Uh oh!

Thalley Oct 9, 2025

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI left a comment

Uh oh!

sonarqubecloud bot commented Oct 10, 2025

Uh oh!

Uh oh!

Bluetooth: Controller: Reduce assertion check code size #90599

Are you sure you want to change the base?

Bluetooth: Controller: Reduce assertion check code size #90599

Conversation

cvinayak commented May 26, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

sonarqubecloud bot commented Jun 5, 2025

Quality Gate passed

Uh oh!

This comment was marked as outdated.

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Uh oh!

Thalley Oct 9, 2025

Choose a reason for hiding this comment

Uh oh!

cvinayak Oct 9, 2025

Choose a reason for hiding this comment

Uh oh!

Thalley Oct 9, 2025

Choose a reason for hiding this comment

Uh oh!

cvinayak Oct 9, 2025

Choose a reason for hiding this comment

Uh oh!

Thalley Oct 9, 2025

Choose a reason for hiding this comment

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Uh oh!

sonarqubecloud bot commented Oct 10, 2025

Quality Gate passed

Uh oh!

Uh oh!

cvinayak commented May 26, 2025 •

edited

Loading