Announcement

Collapse
No announcement yet.

A Dream Come True: Running Coreboot On A Modern, Retail Desktop Motherboard

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • pietrushnic
    replied
    Originally posted by ezst036 View Post
    I really wish we could convince one of the motherboard manufacturers to dump the old style BIOS and go with something like an open source Coreboot as factory equipment.
    ezst036 key question is how to convince them? From my experience it is close to impossible, the only way you can do that is to create so strong open-source firmware environment that they cannot ignore benefits of it. What essentially means open-source firmware ecosystem have to provide significant business value with minimal risk. For now risk is too high and that's why we have no mainstream hardware with coreboot out-of-the-box. On the other side I think it may be hard to have coreboot out-of-the-box big business needs someone to call and ask to solve the problems. This is not something you get from open-source community, but from companies operating in this community (hence Linux and Red Hat).

    Leave a comment:


  • ezst036
    replied
    I really wish we could convince one of the motherboard manufacturers to dump the old style BIOS and go with something like an open source Coreboot as factory equipment.

    Leave a comment:


  • pietrushnic
    replied
    For those who still watching this thread. We would like to let you know we working on some improvements to Dasharo compatible with MSI PRO Z690-A. Details are in Dasharo Github project. Please let us know what do you think. You can always reach us out on Dasharo Matrix.

    Leave a comment:


  • pal666
    replied
    pietrushnic i wish you success in producing x670e port

    Leave a comment:


  • pietrushnic
    replied
    Originally posted by arQon View Post

    > Also I would like to understand better why you consider this disaster

    I think you ran into a language barrier there. It means "Only if <something UNLIKELY>".
    arQon ok, it was definitely misunderstanding on my side. Thank you for your support and comments.

    Leave a comment:


  • miczyg
    replied
    Originally posted by plan-g View Post
    Any chance someone could post the iommu groups? There is a script here to dump the info: https://www.reddit.com/r/VFIO/wiki/boardiommu/

    This could help figure out if the board could assign a GPU or other devices using PCI passthru and KVM. Having an open firmware and jailing proprietary operating systems would be nice.
    Here are the groups: https://pastebin.com/FugAd8Qg

    Leave a comment:


  • arQon
    replied
    Originally posted by pietrushnic View Post
    Pilot attempt was mad by 3mdeb without any MSI support or knowledge.
    My bad.

    > We tried to reach MSI Poland but they were not interested

    Well, crap.

    > Also I would like to understand better why you consider this disaster

    I think you ran into a language barrier there. It means "Only if <something UNLIKELY>".

    Leave a comment:


  • plan-g
    replied
    Any chance someone could post the iommu groups? There is a script here to dump the info: https://www.reddit.com/r/VFIO/wiki/boardiommu/

    This could help figure out if the board could assign a GPU or other devices using PCI passthru and KVM. Having an open firmware and jailing proprietary operating systems would be nice.

    Leave a comment:


  • pietrushnic
    replied
    Originally posted by arQon View Post
    "Incomplete" is potentially just a minor inconvenience. "Unable to use XMP" though is quite a significant hurdle (given how half-assed JEDEC is), as is not being able to boost the CPU.
    arQon agree and that's why there are plans to enable XMP. Feel free to vote for that feature.

    Originally posted by arQon View Post
    Still, this is *huge* progress for Coreboot, and now that MSI has gone down this road at all, I think there's a strong likelihood that their plan is to migrate as much of their product to it as possible, unless this "pilot" attempt turns out to be disastrous.
    Pilot attempt was mad by 3mdeb without any MSI support or knowledge. We tried to reach MSI Poland but they were not interested, so it is very unlikely that MSI will consider open-source firmware for their products yet. Also I would like to understand better why you consider this disaster, is it because of performance results and the fact that we follow standards and silicon vendor guidlines? Would you change mind if we would deliver firmware that beats vendor BIOS performance?

    Leave a comment:


  • zir_blazer
    replied
    Originally posted by arQon View Post
    "Incomplete" is potentially just a minor inconvenience. "Unable to use XMP" though is quite a significant hurdle (given how half-assed JEDEC is), as is not being able to boost the CPU.
    I actually consider than XMP is the half-assed one. I tend to eat popcorn reading about people having stability issues after enabling it, since you typically also require higher Memory Controller voltages for memory speeds above what is officially supported by the Processor, a limit that most people seems to not take into account at all. Some Motherboards apply higher voltages automatically when enabling XMP to take care of possible stability issues and make it user friendly, but at that point you're overclocking the CPU without explicitly being told that you are doing so.
    Thus XMP is more than On/Off. And you need a rather wide array of manual controls in case it doesn't works on the first try.


    Originally posted by arQon View Post
    Still, this is *huge* progress for Coreboot, and now that MSI has gone down this road at all, I think there's a strong likelihood that their plan is to migrate as much of their product to it as possible, unless this "pilot" attempt turns out to be disastrous.
    MSI is not involved in any way. The job was entirely done by 3mdeb on its own with no vendor support other than Intel for the FSP.

    Leave a comment:

Working...
X