debug: coredump: add internal flash backend based on nrfx #21418

Damian-Nordic · 2025-04-04T11:44:01Z

Add Zephyr core dump backend that saves a core dump to the internal flash or RRAM partition named "coredump_partition".

This backend is an alternative to BACKEND_FLASH_PARTITION provided by Zephyr but it bypasses Zephyr flash device layer and uses nrfx directly, which offers the following benefits:

Bypasses synchronization primitives used by Zephyr flash or RRAM drivers. Currently, Zephyr flash drivers cannot be used to write to flash from a fault handler because of this.
Works with Partition Manager.
Minimizes the dependencies needed to successfully write a core dump, which is important as the core dump often needs to be written when the system is in the corrupted state.

Only flash and RRAM are supported for now (no MRAM support).

NordicBuilder · 2025-04-04T11:48:31Z

CI Information

^{To view the history of this post, clich the 'edited' button above}
Build number: 5

Inputs:

Sources:

sdk-nrf: PR head: 346ee03e6a264430ba88d3522a67cda38eb2a9b8

more details

sdk-nrf:

PR head: 346ee03e6a264430ba88d3522a67cda38eb2a9b8
merge base: 051cce44210010252b787bfd13c78f8a64d9342b
target head (main): 051cce44210010252b787bfd13c78f8a64d9342b
Diff

Github labels

Enabled	Name	Description
	ci-disabled	Disable the ci execution
	ci-all-test	Run all of ci, no test spec filtering will be done
	ci-force-downstream	Force execution of downstream even if twister fails
	ci-run-twister	Force run twister
	ci-run-zephyr-twister	Force run zephyr twister

List of changed files detected by CI (7)

CODEOWNERS
doc
│  ├── nrf
│  │  ├── releases_and_maturity
│  │  │  ├── releases
│  │  │  │  │ release-notes-changelog.rst
subsys
│  ├── debug
│  │  ├── CMakeLists.txt
│  │  ├── Kconfig
│  │  ├── coredump
│  │  │  ├── CMakeLists.txt
│  │  │  ├── Kconfig
│  │  │  │ coredump_backend_nrf_flash_partition.c

Outputs:

Toolchain

Version: 7cbc0036f4
Build docker image: docker-dtr.nordicsemi.no/sw-production/ncs-build:7cbc0036f4_8bf7ca4353

Test Spec & Results: ✅ Success; ❌ Failure; 🟠 Queued; 🟡 Progress; ◻️ Skipped; ⚠️ Quarantine

◻️ Toolchain - Skipped: existing toolchain is used
✅ Build twister
- sdk-nrf test count: 263
✅ Integration tests

Disabled integration tests

- desktop52_verification
- doc-internal
- test_ble_nrf_config
- test-fw-nrfconnect-apps
- test-fw-nrfconnect-ble_mesh
- test-fw-nrfconnect-ble_samples
- test-fw-nrfconnect-boot
- test-fw-nrfconnect-chip
- test-fw-nrfconnect-fem
- test-fw-nrfconnect-nfc
- test-fw-nrfconnect-nrf-iot_cloud
- test-fw-nrfconnect-nrf-iot_libmodem-nrf
- test-fw-nrfconnect-nrf-iot_mosh
- test-fw-nrfconnect-nrf-iot_positioning
- test-fw-nrfconnect-nrf-iot_samples
- test-fw-nrfconnect-nrf-iot_serial_lte_modem
- test-fw-nrfconnect-nrf-iot_thingy91
- test-fw-nrfconnect-nrf-iot_zephyr_lwm2m
- test-fw-nrfconnect-nrf_crypto
- test-fw-nrfconnect-proprietary_esb
- test-fw-nrfconnect-ps
- test-fw-nrfconnect-rpc
- test-fw-nrfconnect-rs
- test-fw-nrfconnect-tfm
- test-fw-nrfconnect-thread
- test-low-level
- test-sdk-audio
- test-sdk-dfu
- test-sdk-find-my
- test-sdk-mcuboot
- test-sdk-pmic-samples
- test-sdk-wifi
- test-secdom-samples-public

Note: This message is automatically posted and updated by the CI

peknis · 2025-04-04T12:02:03Z

doc/nrf/releases_and_maturity/releases/release-notes-changelog.rst

@@ -938,7 +938,8 @@ Common Application Framework
 Debug libraries
 ---------------

-|no_changes_yet_note|
+* Added an experimental :ref:`Zephyr Core Dump <zephyr:coredump>` backend that writes a core dump to an internal flash or RRAM partition.
+  To enable this backend set the :kconfig:option:`CONFIG_DEBUG_COREDUMP_BACKEND_OTHER` and :kconfig:option:`CONFIG_DEBUG_COREDUMP_BACKEND_NRF_FLASH_PARTITION` Kconfig options.


Suggested change

To enable this backend set the :kconfig:option:`CONFIG_DEBUG_COREDUMP_BACKEND_OTHER` and :kconfig:option:`CONFIG_DEBUG_COREDUMP_BACKEND_NRF_FLASH_PARTITION` Kconfig options.

To enable this backend, set the :kconfig:option:`CONFIG_DEBUG_COREDUMP_BACKEND_OTHER` and :kconfig:option:`CONFIG_DEBUG_COREDUMP_BACKEND_NRF_FLASH_PARTITION` Kconfig options.

github-actions · 2025-04-04T12:10:36Z

You can find the documentation preview for this PR here.

Preview links for modified nRF Connect SDK documents:

https://ncsdoc.z6.web.core.windows.net/PR-21418/nrf/releases_and_maturity/releases/release-notes-changelog.html

de-nordic · 2025-04-08T08:44:21Z

subsys/debug/coredump/coredump_backend_nrf_flash_partition.c

+	if ((offset % FLASH_WRITE_SIZE) != 0) {
+		write_error = -EINVAL;
+		return;
+	}
+
+	if (!is_within_partition(offset, size)) {
+		write_error = -ENOMEM;
+		return;
+	}
+


I think these should be asserts. There is nothing we can do at this point, there is no recovery and returning error from here is pointless, because the problem is with the caller not properly aligning buffers.
Tests here are just adding to code, while this is the last stand driver that should have minimal code.

Regarding alignment, I agree, though do you think it is OK to trigger another fatal error (assertion) while processing the previous one? Won't it cause a core dump loop? Maybe I can just remove this check and let driver layer underneath fail if the alignment is wrong?

Regarding is_within_partition, I think we should keep it even when assertions are disabled - one may simply allocate to small partition (which is easy to misconfigure) and we don't want to do out of bounds writes in such a case. What do you think?

Regarding alignment, I agree, though do you think it is OK to trigger another fatal error (assertion) while processing the previous one? Won't it cause a core dump loop? Maybe I can just remove this check and let driver layer underneath fail if the alignment is wrong?

Yeah, I think we can just rely here on driver failing anyway.

Regarding is_within_partition, I think we should keep it even when assertions are disabled - one may simply allocate to small partition (which is easy to misconfigure) and we don't want to do out of bounds writes in such a case. What do you think?

Let it be as it as is.

The perfect solution here would be if we could run build in "test configuration" mode, where we could catch all problems that are misconfiguration at compile time. Because, notice that you will not have info that partition is too small until your core dump does not fit.

Damian-Nordic · 2025-04-09T09:02:09Z

@nordic-krch @nordicjm could you please review?

subsys/debug/coredump/Kconfig

Add Zephyr core dump backend that saves a core dump to the internal flash or RRAM partition named "coredump_partition". This backend is an alternative to BACKEND_FLASH_PARTITION provided by Zephyr but it bypasses Zephyr flash device layer and uses nrfx directly, which offers the following benefits: 1. Bypasses synchronization primitives used by Zephyr flash or RRAM drivers. Currently, Zephyr flash drivers cannot be used to write to flash from a fault handler because of this. 2. Works with Partition Manager. 3. Minimizes the dependencies needed to successfully write a core dump, which is important as the core dump often needs to be written when the system is in the corrupted state. Only flash and RRAM are supported for now (no MRAM support). Signed-off-by: Damian Krolik <damian.krolik@nordicsemi.no>

Damian-Nordic requested review from a team and nordic-krch as code owners April 4, 2025 11:44

github-actions bot added the doc-required PR must not be merged without tech writer approval. label Apr 4, 2025

Damian-Nordic requested review from nvlsianpu and de-nordic April 4, 2025 11:46

Damian-Nordic force-pushed the coredump_nrfx branch from e609e22 to 905e41a Compare April 4, 2025 11:50

peknis requested changes Apr 4, 2025

View reviewed changes

Damian-Nordic force-pushed the coredump_nrfx branch from 905e41a to 5e677de Compare April 6, 2025 21:35

Damian-Nordic requested a review from peknis April 7, 2025 07:50

Damian-Nordic mentioned this pull request Apr 7, 2025

log_rpc: replace crash log with crash dump #21420

Draft

peknis approved these changes Apr 7, 2025

View reviewed changes

de-nordic reviewed Apr 8, 2025

View reviewed changes

Damian-Nordic force-pushed the coredump_nrfx branch from 5e677de to 262beaa Compare April 8, 2025 09:59

Damian-Nordic requested a review from de-nordic April 8, 2025 10:38

nordicjm reviewed Apr 9, 2025

View reviewed changes

subsys/debug/coredump/Kconfig Outdated Show resolved Hide resolved

nordic-krch approved these changes Apr 9, 2025

View reviewed changes

Damian-Nordic force-pushed the coredump_nrfx branch from 262beaa to 346ee03 Compare April 9, 2025 14:31

Damian-Nordic requested a review from nordicjm April 9, 2025 14:31

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

debug: coredump: add internal flash backend based on nrfx #21418

debug: coredump: add internal flash backend based on nrfx #21418

Damian-Nordic commented Apr 4, 2025

NordicBuilder commented Apr 4, 2025 •

edited

Loading

peknis Apr 4, 2025

github-actions bot commented Apr 4, 2025

de-nordic Apr 8, 2025

Damian-Nordic Apr 8, 2025

de-nordic Apr 8, 2025

Damian-Nordic commented Apr 9, 2025

	To enable this backend set the :kconfig:option:`CONFIG_DEBUG_COREDUMP_BACKEND_OTHER` and :kconfig:option:`CONFIG_DEBUG_COREDUMP_BACKEND_NRF_FLASH_PARTITION` Kconfig options.
	To enable this backend, set the :kconfig:option:`CONFIG_DEBUG_COREDUMP_BACKEND_OTHER` and :kconfig:option:`CONFIG_DEBUG_COREDUMP_BACKEND_NRF_FLASH_PARTITION` Kconfig options.

debug: coredump: add internal flash backend based on nrfx #21418

Are you sure you want to change the base?

debug: coredump: add internal flash backend based on nrfx #21418

Conversation

Damian-Nordic commented Apr 4, 2025

NordicBuilder commented Apr 4, 2025 • edited Loading

CI Information

Inputs:

Sources:

Github labels

Outputs:

Toolchain

Test Spec & Results: ✅ Success; ❌ Failure; 🟠 Queued; 🟡 Progress; ◻️ Skipped; ⚠️ Quarantine

peknis Apr 4, 2025

Choose a reason for hiding this comment

github-actions bot commented Apr 4, 2025

de-nordic Apr 8, 2025

Choose a reason for hiding this comment

Damian-Nordic Apr 8, 2025

Choose a reason for hiding this comment

de-nordic Apr 8, 2025

Choose a reason for hiding this comment

Damian-Nordic commented Apr 9, 2025

NordicBuilder commented Apr 4, 2025 •

edited

Loading