MCU-Driver-HAL and CI Combined WG Meeting 2021-07-01

Date

Jul 1, 2021

Participants

  • James

  • Saheer Babu

  • Evelyne Donnaes

  • Eric Finco

  • Bill Fletcher

  • Ian Fyall

  • Kumar Gala

  • Chokri Gammoudi

  • Erwan Gouriou

  • Haley Guo

  • Martin Kojtal

  • Laurent Meunier

  • François Ozog

  • Andy Powers

  • Kenney Qu

  • Marcelo Salazar

  • Qinghao Shi

  • Paul Sokolovskyy

  • Bechir JABRI

  • Filip Jagodzinski

  • jerome coutant

  • Konstantinos Chatzis

  • Mark Hicken

  • Wilfried

Slides

See https://linaro.atlassian.net/wiki/spaces/HAL/pages/28517859701

Notes

Proposal on contribution guidelines

KG: What’s the basis for maintainership. Will there be e.g. a UART (specific) maintainer. Do we have enough people?

ED: Need to understand who will come forward for the role. Maybe one person can take a number of roles

KG: Right now have 2 repos. If someone contributes an API change - do they need to make the change to the vendor repos. What’s the workflow?

ED: Need to bring in this workflow. From the API repo each vendor/partner can decide at what point they want to take on the change.

KG: Think that devalues the project. What’s the value proposition if can’t say that eg the I2C driver interface works across all vendors. Same case as different vendor APIs today vs different versions with this API

ED: Need to discuss this. The idea is that all of the driver implementation matches the API.

Going to put this in a PR.

EF: Project is under the Linaro umbrella. Is it consistent with Linaro rules? Also what about DCO vs CLA etc?

BF: Believe this scope is something that needs to be aligned with open source best practises specifically for code management (PRs etc). Linaro aims for that and other projects TF, Zephyr etc are similarly aligned. Don’t think there are Linaro specifics in this case (relating to Linaro organisation/reporting/infrastructure)

KG: Yes, aligned with open source way of working. We explicitly checked the DCO vs CLA and generally recommend DCO.

Legacy/New Documentation

EF: For 198 initiated by ST, what will happen with this kind of material?

ED: Was not going to bring over the recording

EF: What about the slide set?

ED: Can bring it over and put in the Linaro archive.

EF: Think there’s also a slideset from NXP.

BJ: For layering (slides showing the layers and lower peripherals and discussion about handling physical vs virtual peripherals) there’s something from Arm also.

LM: What’s the deadline for approval of open issues

ED: Is next week possible?

EF/LM: Think next week is a good idea - either approval or comments.

Status Update

LM: What the goal of the work on Ticker API. Not sure this has been approved. Is it a valid API?

MS: This is focussed on MbedOS API 6.xx to enable the tests. It’s the existing API and we are cleaning up the tests.

ED: It’s about testing the code we have which is currently the Mbed HAL.

EF: It’s a bit misleading that it’s ended up with an MCU-Driver-HAL flag. Some other work that we discussed has not ended up here. In general how do you see the process of (re-)introducing changes in GitHub.

ED: Need to discuss how to evolve whilst keeping a working implementation.

CI and Tests

EF: Will there be specific CI documentation for this project or will it be the general LAVA view.

ED: The intention is to have something specific for this project.

KG: Is there some reason why the GitHub action isn’t being used.

Recording