Hi Grant,
I was on that meeting too. Michal Simek (Xilinx).
Thanks, Michal
On 13.4.2018 21:48, Grant Likely wrote:
Here are the notes from yesterday’s meeting. Props to Daniel Thompson for taking notes!
Note: I didn’t capture the full list of attendees. If you were there, but aren’t in the list then let me know so I can add you to the official list. Eventually I’ll post these notes on a wiki for the project.
12 April 2018 Attendees
- Grant Likely (Arm)
- Ryan Harken (Linaro)
- Ruchika Gupta (NXP)
- Tom Rini (Konsulko)
- Peter Robinson (Red Hat)
- Alex Graf (SUSE)
- Daniel Thompson (Linaro)
- Ben Eckermann
(Incomplete list; Did not get full list of dial ins)
Agenda:
- Status and action item updates
- Other business
Notes: Status
- No progress on legal issues to get things shared for outside contributions
- No progress on converting EBBR to sphinx document
Devicetree
- Committee meeting will shrink scope to cover governanceissues (process, release process, etc).
- Will be starting a regular technical sync up call soon
AOB
- EBBR and different architectures
- Alexander Graf has started talking among u-boot team about extending linuxefi support more widely
- Udit K: What to do about architectures that are not yet in UEFI? # Grant: Not really in scope for EBBR, they should work with UEFI forum
- Grant: EBBR should be opt in (i.e. architecture representatives join us) rather then encompassing “everything”
- Udit K: What about big endian?
- Grant: Not UEFI… it merely looks like it.
- Tom: EBBR references other specifications, needs other specifications to take big endian before we move on it
- Udit K: How to handle devicetree updates?
- Grant: DT owned by platform is important, not discussed how to update it
- Grant: Should we create a DT specific section in EBBR?
- Udit K: Ideally, yes. We understand devicetree is owned by the platform but we have had better results using the devicetree in the kernel
- Peter: UEFI capsules?
- Alexander: Could use overlays to cope with difference between kernels
- Alexander: We cannot assume DTs will always be backwards compatible
- Grant: Historically have worked to ensure new kernels work with old devicetrees but not old kernels with new DTs
- Need to make sure firmware can always be recovered to a ‘safe’ state, and that DT updates don’t require reflashing the entire firmware.
Action: form sub team to draft DT update requirements.
When can others contribute?
- Expect to get things tidied up this week but the mailing list is open please discuss things here!
Arm.ebbr-discuss mailing list Arm.ebbr-discuss@arm.com