Attention is currently required from: Felix Singer, Nico Huber, Michał Żygowski, Paul Menzel, Michael Niewöhner.

View Change

1 comment:

• Patchset:

  • Patch Set #4:

    Maybe they are unique. However I have this feeling that DMI checks are more reliable. […]

    We could use both DMI and SVID/SDID checks:

    • DMI good, SVID/SDID good ---> [1]
    • DMI good, SVID/SDID bad ---> [2]
    • DMI bad, SVID/SDID good ---> [3]
    • DMI bad, SVID/SDID bad ---> [4]

    [1]: full match, proceed as usual.
    [2]: unexpected (DMI is more specific than SVID/SDID), abort unless forced
    [3]: mention that things could work, but abort unless forced
    [4]: unknown machine, abort unless forced

    Running into [2] would be very unusual. I'd explain the situation (DMI strings match, but SVID/SDID do not; could be a boot firmware bug or an unsupported system) and warn against continuing unless one is certain that the mismatched SVID/SDID is spurious (not something the mailing list would know).

    For [3], I'd make flashrom say:

    • What happened: SVID/SDID match, but DMI strings do not
    • What the above implies: the machine might just be a rebranded system
    • What to do: if one can risk a brick, try forcing and see what happens. if it works, then add the DMI strings to flashrom and submit a patch

    In both cases, log the known DMI strings and SVID/SDID values (print both expected and actual SVID/SDID in [2]).

    Thoughts?

To view, visit change 55715. To unsubscribe, or for help writing mail filters, visit settings.