emds: Revise rram related timing in emds #19275

Balaklaka · 2024-12-05T08:12:24Z

Updated estimated timing for rram storage using emds. Also updated documentation and test to a more proper indication of values for rram.

NordicBuilder · 2024-12-05T08:16:12Z

CI Information

^{To view the history of this post, clich the 'edited' button above}
Build number: 6

Inputs:

Sources:

sdk-nrf: PR head: 070328ff05fba447099208af1cd3700df1fb978e

more details

sdk-nrf:

PR head: 070328ff05fba447099208af1cd3700df1fb978e
merge base: bb2f537273f8ba6e6028c7496eb8ab6057f939dd
target head (main): f3ac9cfb784d163f4c1af11209976fcd5c403d5a
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 (4)

doc
│  ├── nrf
│  │  ├── libraries
│  │  │  ├── others
│  │  │  │  │ emds.rst
subsys
│  ├── emds
│  │  │ Kconfig
tests
│  ├── subsys
│  │  ├── emds
│  │  │  ├── emds_api
│  │  │  │  │ prj.conf
│  │  │  ├── emds_flash
│  │  │  │  ├── src
│  │  │  │  │  │ main.c

Outputs:

Toolchain

Version: b77d8c1312
Build docker image: docker-dtr.nordicsemi.no/sw-production/ncs-build:b77d8c1312_912848a074

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

◻️ Toolchain - Skipped: existing toolchain is used
✅ Build twister
- sdk-nrf test count: 57
✅ 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_libmodem-nrf
- test-fw-nrfconnect-nrf-iot_lwm2m
- test-fw-nrfconnect-nrf-iot_mosh
- test-fw-nrfconnect-nrf-iot_nrf_provisioning
- 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-ps
- test-fw-nrfconnect-rpc
- test-fw-nrfconnect-rs
- test-fw-nrfconnect-tfm
- test-fw-nrfconnect-thread
- test-fw-nrfconnect-zigbee
- test-low-level
- test-sdk-audio
- test-sdk-dfu
- test-sdk-find-my
- test-sdk-mcuboot
- test-sdk-pmic-samples
- test-sdk-sidewalk
- test-sdk-wifi
- test-secdom-samples-public

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

NordicBuilder · 2024-12-05T08:30:49Z

You can find the documentation preview for this PR at this link. It will be updated about 10 minutes after the documentation build succeeds.

Note: This comment is automatically posted by the Documentation Publish GitHub Action.

doc/nrf/libraries/others/emds.rst

subsys/emds/Kconfig

doc/nrf/libraries/others/emds.rst

tests/subsys/emds/emds_flash/src/main.c

ludvigsj

Tiny nitpicks left

doc/nrf/libraries/others/emds.rst

omkar3141 · 2024-12-07T12:40:46Z

doc/nrf/libraries/others/emds.rst

 If partial erase is not enabled or supported, the time of a complete sector erase has to be included in the :kconfig:option:`CONFIG_EMDS_FLASH_TIME_BASE_OVERHEAD_US`.
 When partial erase is enabled and supported by the hardware, include the time it takes for the scheduler to trigger, which is depending on the time defined in :kconfig:option:`CONFIG_SOC_FLASH_NRF_PARTIAL_ERASE_MS`.
 When changing the :kconfig:option:`CONFIG_EMDS_FLASH_TIME_BASE_OVERHEAD_US` option, it is important that the worst time is considered.
+When configuring these values using RRAM there is not a need to consider garbage collection in the same way as for flash, as it can be written to regardless of value. But the ERASEALL functionality


But the ERASEALL functionality should be avoided

How can you 'avoid' eraseall functionality? EMDS storage page erase anyway needed upon the next reboot right?

RRAMC provides functionality to protect the device against the erase all function.
See https://docs.nordicsemi.com/bundle/ps_nrf54L15/page/rramc.html#ariaid-title7

omkar3141 · 2024-12-07T12:41:33Z

doc/nrf/libraries/others/emds.rst

 These can be found by looking at datasheets, driver documentation, and the configuration of the application.
-:math:`s_\text{ate}` is the size of the allocation table entry used by the EMDS flash module, which is 8 B.
+:math:`s_\text{ate}` is the size of the allocation table entry used by the EMDS flash/RRAM module, which is 8 B.


Suggested change

:math:`s_\text{ate}` is the size of the allocation table entry used by the EMDS flash/RRAM module, which is 8 B.

:math:`s_\text{ate}` is the size of the allocation table entry used by the EMDS flash/RRAM module, which is 8 B if NVS is used, else it is 16 B for ZMS.

The ate size is 8 regardless of ZMS or NVS. It is defined in emds_flash.c.

omkar3141 · 2024-12-07T12:42:13Z

doc/nrf/libraries/others/emds.rst

@@ -142,7 +144,7 @@ The following values can be inserted into the formula:
   This is the worst case overhead when a store is triggered in the middle of an erase on nRF52840 with :kconfig:option:`CONFIG_SOC_FLASH_NRF_PARTIAL_ERASE` enabled in the sample by default, and should be adjusted when using other configurations.
 *  Set :math:`t_\text{entry}` = 300 µs and :math:`t_\text{word}` = 41 µs. *Note: These values are valid only for this specific chip and configuration, and should be computed for the specific configuration whenever using EMDS.*
 *  The sample uses two entries, one for the RPL with 255 entries (:math:`s_i` = 2040 B) and one for the lightness state (:math:`s_i` = 3 B).
-*  The flash write block size :math:`s_\text{block}` in this case is 4 B, and the ATE size :math:`s_\text{ate}` is 8 B.
+*  The flash/RRAM write block size :math:`s_\text{block}` in this case is 4 B, and the ATE size :math:`s_\text{ate}` is 8 B.


Suggested change

* The flash/RRAM write block size :math:`s_\text{block}` in this case is 4 B, and the ATE size :math:`s_\text{ate}` is 8 B.

* The flash/RRAM write block size :math:`s_\text{block}` in this case is 4 B, and for NVS backend the ATE size :math:`s_\text{ate}` is 8 B.

The ate size is 8 regardless of ZMS or NVS. It is defined in emds_flash.c.

peknis · 2024-12-11T07:14:54Z

doc/nrf/libraries/others/emds.rst

 Overview
 ********


Suggested change

Overview

********

peknis · 2024-12-11T07:15:47Z

doc/nrf/libraries/others/emds.rst

 Implementation
 ==============
 The :kconfig:option:`CONFIG_EMDS` Kconfig option enables the emergency data storage.


Suggested change

Implementation

==============

The :kconfig:option:`CONFIG_EMDS` Kconfig option enables the emergency data storage.

Implementation

**************

The :kconfig:option:`CONFIG_EMDS` Kconfig option enables the emergency data storage.

peknis · 2024-12-18T06:36:20Z

subsys/emds/Kconfig

@@ -68,8 +73,7 @@ config EMDS_FLASH_TIME_BASE_OVERHEAD_US
 	   worst case scenario, before starting storing entries. This value
 	   is dependent on the chip used, and should be checked against the chip
 	   datasheet.
-	   For RRAM-based flash drivers, the time is defined by the erasing block size.
-	   The default erasing block size is 4096 bytes.
+	   For RRAM-based flash drivers, ERASEPROTECT should be used and we can disregard the SOC_FLASH_NRF_PARTIAL_ERASE_MS parameter.


Suggested change

For RRAM-based flash drivers, ERASEPROTECT should be used and we can disregard the SOC_FLASH_NRF_PARTIAL_ERASE_MS parameter.

For RRAM-based flash drivers, use ERASEPROTECT and disregard the SOC_FLASH_NRF_PARTIAL_ERASE_MS parameter.

peknis · 2024-12-18T06:37:48Z

doc/nrf/libraries/others/emds.rst

@@ -128,9 +128,9 @@ The :c:func:`emds_store_time_get` function estimates the required worst-case tim
   t_\text{store} = t_\text{base} + \sum_{i = 1}^n \left(t_\text{entry} + t_\text{word}\left(\left\lceil\frac{s_\text{ate}}{s_\text{block}}\right\rceil + \left\lceil\frac{s_i}{s_\text{block}}\right\rceil \right)\right)

 where :math:`t_\text{base}` is the value specified by :kconfig:option:`CONFIG_EMDS_FLASH_TIME_BASE_OVERHEAD_US`, :math:`t_\text{entry}` is the value specified by :kconfig:option:`CONFIG_EMDS_FLASH_TIME_ENTRY_OVERHEAD_US` and :math:`t_\text{word}` is the value specified by :kconfig:option:`CONFIG_EMDS_FLASH_TIME_WRITE_ONE_WORD_US`.
-:math:`s_i` is the size of the :math:`i`\ th entry in bytes and :math:`s_\text{block}` is the number of bytes in one word of flash.
+:math:`s_i` is the size of the :math:`i`\ th entry in bytes and :math:`s_\text{block}` is the number of bytes in one word of flash, while for RRAM the write-buffer size i number of 128-bit words (16 bytes).


Suggested change

:math:`s_i` is the size of the :math:`i`\ th entry in bytes and :math:`s_\text{block}` is the number of bytes in one word of flash, while for RRAM the write-buffer size i number of 128-bit words (16 bytes).

:math:`s_i` is the size of the :math:`i`\ th entry in bytes and :math:`s_\text{block}` is the number of bytes in one word of flash, while for RRAM the write-buffer size ``i`` number of 128-bit words (16 bytes).

peknis · 2024-12-18T06:38:43Z

doc/nrf/libraries/others/emds.rst

@@ -64,7 +64,7 @@ The :c:func:`emds_is_ready` function can be called to check if EMDS is prepared
 Once the data storage has completed, a callback is called if provided in :c:func:`emds_init`.
 This callback notifies the application that the data storage has completed, and can be used to reboot the CPU or execute another function that is needed.

-After completion of :c:func:`emds_store`, the :c:func:`emds_is_ready` function call will return error, since it can no longer guarantee that the data will fit into the flash area.
+After completion of :c:func:`emds_store`, the :c:func:`emds_is_ready` function call will return error, since it can no longer guarantee that the data will fit into the persistent memory area.


Suggested change

After completion of :c:func:`emds_store`, the :c:func:`emds_is_ready` function call will return error, since it can no longer guarantee that the data will fit into the persistent memory area.

After completion of :c:func:`emds_store`, the :c:func:`emds_is_ready` function call will return an error, because it can no longer guarantee that the data will fit into the persistent memory area.

peknis · 2024-12-18T06:40:43Z

doc/nrf/libraries/others/emds.rst

 If partial erase is not enabled or supported, the time of a complete sector erase has to be included in the :kconfig:option:`CONFIG_EMDS_FLASH_TIME_BASE_OVERHEAD_US`.
 When partial erase is enabled and supported by the hardware, include the time it takes for the scheduler to trigger, which is depending on the time defined in :kconfig:option:`CONFIG_SOC_FLASH_NRF_PARTIAL_ERASE_MS`.
 When changing the :kconfig:option:`CONFIG_EMDS_FLASH_TIME_BASE_OVERHEAD_US` option, it is important that the worst time is considered.
+When configuring these values using RRAM there is not a need to consider garbage collection in the same way as for flash, as it can be written to regardless of value. But the ERASEALL functionality


Suggested change

When configuring these values using RRAM there is not a need to consider garbage collection in the same way as for flash, as it can be written to regardless of value. But the ERASEALL functionality

When configuring these values using RRAM, you do not need to consider garbage collection in the same way as for flash, as it can be written to regardless of the value.

However, avoid the ERASEALL functionality,

peknis · 2024-12-18T06:41:21Z

doc/nrf/libraries/others/emds.rst

 If partial erase is not enabled or supported, the time of a complete sector erase has to be included in the :kconfig:option:`CONFIG_EMDS_FLASH_TIME_BASE_OVERHEAD_US`.
 When partial erase is enabled and supported by the hardware, include the time it takes for the scheduler to trigger, which is depending on the time defined in :kconfig:option:`CONFIG_SOC_FLASH_NRF_PARTIAL_ERASE_MS`.
 When changing the :kconfig:option:`CONFIG_EMDS_FLASH_TIME_BASE_OVERHEAD_US` option, it is important that the worst time is considered.
+When configuring these values using RRAM there is not a need to consider garbage collection in the same way as for flash, as it can be written to regardless of value. But the ERASEALL functionality
+should be avoided as that can increase the time before the EMDS store functions are called.


Suggested change

should be avoided as that can increase the time before the EMDS store functions are called.

because that can increase the time before the EMDS store functions are called.

peknis · 2024-12-18T10:12:18Z

How about a changelog entry? Is this important enough to add one?

doc/nrf/libraries/others/emds.rst

Updated estimated timing for rram storage using emds. Also updated documentation and test to a more proper indication of values for rram. Signed-off-by: Ingar Kulbrandstad <[email protected]>

Balaklaka · 2024-12-18T15:40:21Z

How about a changelog entry? Is this important enough to add one?

This is not something new, so I think it is not important enough to add a changelog entry.

alxelax

LGTM, just a minor question.

alxelax · 2025-01-02T12:09:04Z

subsys/emds/Kconfig

@@ -68,8 +73,7 @@ config EMDS_FLASH_TIME_BASE_OVERHEAD_US
 	   worst case scenario, before starting storing entries. This value
 	   is dependent on the chip used, and should be checked against the chip
 	   datasheet.
-	   For RRAM-based flash drivers, the time is defined by the erasing block size.
-	   The default erasing block size is 4096 bytes.
+	   For RRAM-based persistent memory driver, use ERASEPROTECT and disregard the SOC_FLASH_NRF_PARTIAL_ERASE_MS parameter.


use ERASEPROTECT what does it mean?

See https://docs.nordicsemi.com/bundle/ps_nrf54L15/page/rramc.html#ariaid-title7

Hm... it is not clear why is ERASE PROTECT required for emds? Seems this is up to the application whether it uses or doesn't.

Balaklaka requested review from a team as code owners December 5, 2024 08:12

github-actions bot added doc-required PR must not be merged without tech writer approval. changelog-entry-required Update changelog before merge. Remove label if entry is not needed or already added. labels Dec 5, 2024

Balaklaka removed the changelog-entry-required Update changelog before merge. Remove label if entry is not needed or already added. label Dec 5, 2024

ludvigsj requested changes Dec 5, 2024

View reviewed changes

Balaklaka force-pushed the develop/revise_rram_time_emds branch from 9d09608 to 780cf30 Compare December 5, 2024 19:07

github-actions bot added the changelog-entry-required Update changelog before merge. Remove label if entry is not needed or already added. label Dec 5, 2024

Balaklaka force-pushed the develop/revise_rram_time_emds branch from 780cf30 to 0eeab6e Compare December 5, 2024 21:59

ludvigsj requested changes Dec 6, 2024

View reviewed changes

doc/nrf/libraries/others/emds.rst Outdated Show resolved Hide resolved

doc/nrf/libraries/others/emds.rst Outdated Show resolved Hide resolved

omkar3141 reviewed Dec 7, 2024

View reviewed changes

peknis requested changes Dec 11, 2024

View reviewed changes

Balaklaka force-pushed the develop/revise_rram_time_emds branch from 0eeab6e to a646365 Compare December 17, 2024 14:55

Balaklaka requested review from ludvigsj, peknis and omkar3141 December 17, 2024 14:59

peknis requested changes Dec 18, 2024

View reviewed changes

Balaklaka force-pushed the develop/revise_rram_time_emds branch from a646365 to e216aee Compare December 18, 2024 09:34

Balaklaka requested a review from peknis December 18, 2024 09:35

peknis approved these changes Dec 18, 2024

View reviewed changes

ludvigsj requested changes Dec 18, 2024

View reviewed changes

doc/nrf/libraries/others/emds.rst Outdated Show resolved Hide resolved

emds: Revise rram related timing in emds

070328f

Updated estimated timing for rram storage using emds. Also updated documentation and test to a more proper indication of values for rram. Signed-off-by: Ingar Kulbrandstad <[email protected]>

Balaklaka force-pushed the develop/revise_rram_time_emds branch from e216aee to 070328f Compare December 18, 2024 13:24

ludvigsj approved these changes Dec 18, 2024

View reviewed changes

Balaklaka removed the changelog-entry-required Update changelog before merge. Remove label if entry is not needed or already added. label Dec 18, 2024

alxelax approved these changes Jan 2, 2025

View reviewed changes

nordicjm approved these changes Jan 6, 2025

View reviewed changes

nordicjm merged commit cac49ac into nrfconnect:main Jan 6, 2025
14 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

emds: Revise rram related timing in emds #19275

emds: Revise rram related timing in emds #19275

Balaklaka commented Dec 5, 2024

NordicBuilder commented Dec 5, 2024 •

edited

Loading

NordicBuilder commented Dec 5, 2024

ludvigsj left a comment

omkar3141 Dec 7, 2024

Balaklaka Dec 17, 2024

omkar3141 Dec 7, 2024

Balaklaka Dec 17, 2024

omkar3141 Dec 7, 2024

Balaklaka Dec 17, 2024

peknis Dec 11, 2024

peknis Dec 11, 2024

peknis Dec 18, 2024

peknis Dec 18, 2024

peknis Dec 18, 2024

peknis Dec 18, 2024

peknis Dec 18, 2024

peknis commented Dec 18, 2024

Balaklaka commented Dec 18, 2024

alxelax left a comment

alxelax Jan 2, 2025

Balaklaka Jan 3, 2025

alxelax Jan 3, 2025

	:math:`s_\text{ate}` is the size of the allocation table entry used by the EMDS flash/RRAM module, which is 8 B.
	:math:`s_\text{ate}` is the size of the allocation table entry used by the EMDS flash/RRAM module, which is 8 B if NVS is used, else it is 16 B for ZMS.

	* The flash/RRAM write block size :math:`s_\text{block}` in this case is 4 B, and the ATE size :math:`s_\text{ate}` is 8 B.
	* The flash/RRAM write block size :math:`s_\text{block}` in this case is 4 B, and for NVS backend the ATE size :math:`s_\text{ate}` is 8 B.

	For RRAM-based flash drivers, ERASEPROTECT should be used and we can disregard the SOC_FLASH_NRF_PARTIAL_ERASE_MS parameter.
	For RRAM-based flash drivers, use ERASEPROTECT and disregard the SOC_FLASH_NRF_PARTIAL_ERASE_MS parameter.

	:math:`s_i` is the size of the :math:`i`\ th entry in bytes and :math:`s_\text{block}` is the number of bytes in one word of flash, while for RRAM the write-buffer size i number of 128-bit words (16 bytes).
	:math:`s_i` is the size of the :math:`i`\ th entry in bytes and :math:`s_\text{block}` is the number of bytes in one word of flash, while for RRAM the write-buffer size ``i`` number of 128-bit words (16 bytes).

	After completion of :c:func:`emds_store`, the :c:func:`emds_is_ready` function call will return error, since it can no longer guarantee that the data will fit into the persistent memory area.
	After completion of :c:func:`emds_store`, the :c:func:`emds_is_ready` function call will return an error, because it can no longer guarantee that the data will fit into the persistent memory area.

	When configuring these values using RRAM there is not a need to consider garbage collection in the same way as for flash, as it can be written to regardless of value. But the ERASEALL functionality
	When configuring these values using RRAM, you do not need to consider garbage collection in the same way as for flash, as it can be written to regardless of the value.
	However, avoid the ERASEALL functionality,

	should be avoided as that can increase the time before the EMDS store functions are called.
	because that can increase the time before the EMDS store functions are called.

emds: Revise rram related timing in emds #19275

emds: Revise rram related timing in emds #19275

Conversation

Balaklaka commented Dec 5, 2024

NordicBuilder commented Dec 5, 2024 • edited Loading

CI Information

Inputs:

Sources:

Github labels

Outputs:

Toolchain

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

NordicBuilder commented Dec 5, 2024

ludvigsj left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

peknis commented Dec 18, 2024

Balaklaka commented Dec 18, 2024

alxelax left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

NordicBuilder commented Dec 5, 2024 •

edited

Loading