*Attendees*: Riku, Anton, Glen, Shebu, Ben C, Joanna, Janos, Matteo, Don,
Xinyu
*Minutes*:
- Glen: Begin w/ TFC Kanban Board
- Glen: H/W Status
- Working thru getting Chromebooks imported. In progress
- Cypress boards flashing and booting, starting on jobs work. In a
week ready for boards to be shipped to Cambridge.
- NXP board is in the queue.
- MBedTLS:
- Glen: Have advanced as far as can. Need AMI files from Dean A.
Blocked awaiting AMI files.
- Janos: Progressed since last week, but working on resolving
internal issues. No estimate yet from Dean.
- Glen: Plan to move to expect script efforts with team approval
- Matteo: Agree Expect Scripts are the next item to work. Would like
completed, so good direction to go.
- Code Coverage:
- A sync w/ Leonardo and Joanna to discuss next steps/Current status
on CC shall be planned. Glen to set up
- Glen: Performance/Fixes needed?
- Riku: Nightly jobs failing due to too many parallel checkouts of
same repo/same version. TFC-20. Need to decide how to implement
- For TF-M, looks currently like 1 week work to modify CI to reduce
parallel checkouts, TF-A still looking to get an estimate.
- Ben: TF-A may be a bit larger task, but still need to look.
- Ben: Other solutions?
- Glen: Could Arm TF-A / TF-M team do this to offload the work?
- Not currently
- Ben: Need a short-term and long-term solution. Short-term - added
a server, long-term, change the configuration changes.
- Can use Staging Server to test it out.
- Git checkout taking lots of the build time.
- Shebu: Failing every night?
- Not sure.
- Riku: Last week, 1 success all other failed.
- Ben: The new server increasing capacity has caused this issue to
show itself.
- Work around is to potentially limit the number of parallel jobs.
- Next Steps:
- Leonardo scope TF-A.
- Create two subtasks for TF-A and TF-M
- Riku: Put in work-around to limit number of parallel jobs
- Need a TFC ticket here?
<end>
Attendees: Janos Fallath, Matteo, Joanna, Don, Glen, Xinyu, Ben C
Minutes:
- Cancel next week instance - folks are out
Platform enablement
- Chromebook: Arthur wrapping up the work, when receive the platforms
will be ready.
- Cypress: Got the info and moving forward again.
- NXP Platform: Some back ground work to see best way to integrate
- Focus back in Cypress
Other
- expect scripts. in holding pattern with focus on mbedTLS
- MC: Two months away from end a FY. Would like expect scripts are
finalized. Would like as a background task when things get blocked
- MC: Would like to at a minimum know what needs to be done and what
help Leonardo may need. LAVA help for example? Anything that
can be done
in parallel. So what is left and who needs to do which task?
- MBedTLS
- JF: Don't have AMI's yet, working to get them out. So blocked.
Can't set up environment.
- GV: Jumped forward to M3's for now. Will need to get to having
this up on a real system.
- GV: By end of the week, M3 should be done, having the AMI files
will be key at that point to move on.
- JF: Originally, scripts / files were to be pulled in. Now thinking
of a restricted repo and sharing it - may not have that option. What can
be used beyond a public repo?
- JF: Can Linaro see the private Arm instance?
- Glen: Have to check if can see the private repo *ACTION Glen*.
- Glen: Set up a sync to review repo sharing options *ACTION Glen*
- Movng to gitlab?
- JF: Not at this point
- JF: Could move the test repo to gitlab, but same issue.
- SC Approved:
- Code Coverage tool
- Joanna: A new ticket raised TFC-160. Can we disable code coverage
for now until the gitlab.arm.com is back up. Without it, jobs are
failing.
- MC: Should we wait?
- Joanna: TF-A has been down since Friday.
- MC: Can we see what Dean does to resolve?
- Plan to see if can disable code coverage until resolved.
- MC: Can we be explicit in the ticket on what Linaro is to do. i.e
request disable code coverage on TF-A
Attendees: Janos Fallath, Don, Glen, Xinyu, Ben, Joanna
Minutes:
- Server is ready to bring online, Ben just prioritizing time to get it in
now.
- TFC-110: Licenses available in October - pushed out till then
- TFC-92 Complete - Can now create docker containers for FVP models
- TFC-5 - Add additional FVP models
Is this needed still or will developers add their own?
- MBedTLS
Need to add Janos to the Jira
Need ANI files
Forward email to Janos
- Boards
May start NXP board since Chromebook and Cypress platforms blocked
Chromebooks in shipping
- Expect scripts
Still need some modifications, but not working on - MBedTLS higher
priority
Joanna - Will these get worked on if gaps in board come up?
Glen - yes this could happen
- Code coverage: The bigger aggregated reports are still open
Attendees: Don, Anton, Joanna, Ben, Dave R, Xinyu
FVP: TFC-92 updates?
Ben: Working in staging, needs Docs, tested and moved to staging. Leonardo
back and to work this. Subtasks are TFC-107,108,115
Anton - Planned to complete TF-M release this week. Within the next two
days.
After the release Ben to bring up the new Server. Already set up and
waiting. Should it be all production, partial staging? Plan to start out
as all Production. Can adjust if still need more staging.
Anton: Is it possible to connect two CI's to one LAVA backend?
Ben: It's possible with correct priorities. LAVA has its own Scheduler.
Anton has a private build from an internal CI that he will send binaries to
be tested in LAVA.
Dave R: Notes on MBeb TLS
- There have been some changes to how MBed TLS tests are run. Will make
this public soon. No impacts, but should be easier to use the public repo
versus a private copy
- Unable to host private repos on Github - the impact is to migrate
private repos to GitLab. Expect this to change CI triggering. May require
a detailed call on impacts.
- Dave is here Wed Thursday next week, then out for 4 weeks. Dave will
provide a delegate.