STM32F303RE: Activate FLASHIAP

[LMESTM]

Description

As suggested in #10486 by @JanneKiiskila, we may activate FLASHIAP for F303RE.
Flash driver is already in place.

Pull request type

[ ] Fix
[ ] Refactor
[x ] Target update
[ ] Functionality change
[ ] Docs update
[ ] Test update
[ ] Breaking change

Reviewers

Release Notes

Tests results:
+-----------------------+---------------+-----------------------------+--------+--------------------+-------------+
| target                | platform_name | test suite                  | result | elapsed_time (sec) | copy_method |
+-----------------------+---------------+-----------------------------+--------+--------------------+-------------+
| NUCLEO_F303RE-GCC_ARM | NUCLEO_F303RE | tests-mbed_drivers-flashiap | OK     | 18.39              | default     |
+-----------------------+---------------+-----------------------------+--------+--------------------+-------------+
mbedgt: test suite results: 1 OK
mbedgt: test case report:
+-----------------------+---------------+-----------------------------+-----------------------------------+--------+--------+--------+--------                                                       ------------+
| target                | platform_name | test suite                  | test case                         | passed | failed | result | elapsed                                                       _time (sec) |
+-----------------------+---------------+-----------------------------+-----------------------------------+--------+--------+--------+--------                                                       ------------+
| NUCLEO_F303RE-GCC_ARM | NUCLEO_F303RE | tests-mbed_drivers-flashiap | FlashIAP - init                   | 1      | 0      | OK     | 0.09                                                                      |
| NUCLEO_F303RE-GCC_ARM | NUCLEO_F303RE | tests-mbed_drivers-flashiap | FlashIAP - program                | 1      | 0      | OK     | 0.22                                                                      |
| NUCLEO_F303RE-GCC_ARM | NUCLEO_F303RE | tests-mbed_drivers-flashiap | FlashIAP - program across sectors | 1      | 0      | OK     | 0.16                                                                      |
| NUCLEO_F303RE-GCC_ARM | NUCLEO_F303RE | tests-mbed_drivers-flashiap | FlashIAP - program errors         | 1      | 0      | OK     | 0.1                                                                       |
| NUCLEO_F303RE-GCC_ARM | NUCLEO_F303RE | tests-mbed_drivers-flashiap | FlashIAP - timing                 | 1      | 0      | OK     | 1.01                                                                      |
+-----------------------+---------------+-----------------------------+-----------------------------------+--------+--------+--------+--------                                                       ------------+
mbedgt: test case results: 5 OK
mbedgt: completed in 30.58 sec

[JanneKiiskila]

LGTM, we need this to next Mbed OS 5.12.x patch release, please!
@adbridge

[ciarmcom]

@LMESTM, thank you for your changes.
@ARMmbed/mbed-os-maintainers please review.

[JanneKiiskila]

@0xc0170 - merge time!

[adbridge]

@0xc0170 - merge time!

@JanneKiiskila I think more like CI time first...

[adbridge]

CI started

[jeromecoutant]

Hi

FLASHIAP tests becomes OK with NUCLEO_F303RE.

But FLASHIAP enables also device_key tests, which are FAIL...

@JanneKiiskila could you check this ?

mbed test -t ARM -m NUCLEO_F303RE -v -n features-device_key-tests-device_key-functionality
==> KVMap init issue ?

[JanneKiiskila]

@davidsaada - could you give it a look please?

[davidsaada]

@JanneKiiskila @LMESTM I don't have the board here, nor see the failure log, so it will be hard for me to tell why this fails.
That said, the test results posted here are not relevant for the change. They only test the FlashIAP driver, but the change adds the FlashIAP block device. This is best tested with the general block device test.
i.e. run:

mbed test -v -t GCC_ARM -m NUCLEO_F303RE --profile debug -n features-storage-tests-blockdevice-general_block_device

This may give more clues regarding the failure.

[davidsaada]

@jeromecoutant Is the "KVMap init issue?" question based on a speculation or on log results you see?
Any logs (preferably with debug profile) will be welcomed.

[jeromecoutant]

KVMap init issue ?

This is the first debug step :-)
I added a printf at the beginning of some KVMap functions.

• With DISCO_L475 (test OK), I get the init printf in the test log.

[1556540133.44][CONN][RXD] >>> Running case #1: 'Device Key - long consistency test'...
[1556540133.51][CONN][INF] found KV pair in stream: {{__testcase_start;Device Key - long consistency test}}, queued...
[1556540133.52][CONN][RXD] KVMap init
[1556540133.54][CONN][RXD] KVMap attach
[1556540133.55][CONN][RXD] KVMap config_lookup
...

• With NUCLEO_F3, I couldn't and test is failing:

[1556539842.21][CONN][RXD] >>> Running case #1: 'Device Key - long consistency test'...
[1556539842.27][CONN][INF] found KV pair in stream: {{__testcase_start;Device Key - long consistency test}}, queued...
[1556539842.29][CONN][RXD] KVMap config_lookup
mbedgt: :95::FAIL: Expected Not-Equal

[davidsaada]

Thanks @jeromecoutant. It seems like there's a problem with using the FlashIAP block device (test fails on getting the internal block device).
Can you run the general block device test as following and post the results?

mbed test -v -t GCC_ARM -m NUCLEO_F303RE --profile debug -n features-storage-tests-blockdevice-general_block_device

[davidsaada]

@jeromecoutant BTW, it could well be that the internal flash of this board is small, not enough for the allocation of KVStore, hence the failure in device key test. Do you have the sector map of this board?
In addition, can you rerun the the FlashIAP test (whose results are posted in the description). It may well be that few of tests there also sense that there's not enough room in the flash for testing. But in this case, the test is skipped and you should see prints telling that.

[JanneKiiskila]

@davidsaada - whare are the space requirements? It has the same amount of flash as Nucleo F411RE, but in a more sensible approach (2 kilobyte blocks, 256 pieces of them).

[davidsaada]

@JanneKiiskila Took it offline with @jeromecoutant. Seems like it's not related to space requirements, but to something else we're trying to figure out now.

[davidsaada]

@yossi2le

[yossi2le]

I see that this device is missing "device_has": [ "TRNG" ]. DeviceKey needs TRNG in order to work correctly or any other support which enables the MBEDTLS_ENTROPY_NV_SEED macro.

[JanneKiiskila]

No, it does not have TRNG. It is similar to Nucleo F411, we need to inject the entropy externally.

[yossi2le]

and are you injecting it from outside?

[yossi2le]

@JanneKiiskila I don't think it is related to the TRNG and injection of the device key. The test should create a dummy ROT if there is no TRNG support. I have to run now but I will be happy to help debugging the issue tomorrow morning.
The only problem I do not have a board here. Do we have a remote one?

[JanneKiiskila]

Yes, it is available in RaaS.

[yossi2le]

Excellent, can you tell me on which RaaS. Please send me the details offline?

[mbed-ci]

Test run: SUCCESS

Summary: 11 of 11 test jobs passed
Build number : 1
Build artifacts

[0xc0170]

@yossi2le @jeromecoutant What is the status here? I set it to needs work until we get confirmation the tests are fixed

[yossi2le]

I have submitted a PR #10516 which should solve the issue of device key

[jeromecoutant]

Approved as FLASHIAP tests are OK with this PR.

DEVICE_KEY tests will then be OK with #10516

[JanneKiiskila]

Excellent, let's get them in and enable stuff. 👍