Foxconn owns up to dodgy BIOS crippling Linux

Summary: The other day I blogged about how the Foxconn G33M-S motherboard didn't play nicely with Linux because it broke ACPI support. Well, as it turns out the problem isn't a Foxconn one but it's down to American Megatrends (AMI) having shipped a defective BIOS. Also, as it turns out, other boards are also affected.

[UPDATE: The debug version on an updated BIOS is now available from Foxconn. Details here.] 

The other day I blogged about how the Foxconn G33M-S motherboard didn't play nicely with Linux because it broke ACPI support. Well, as it turns out the problem isn't a Foxconn one but it's down to American Megatrends (AMI) having shipped a defective BIOS. Also, as it turns out, other boards are also affected.

Here's the current state of play:

  • The issue has been reported to Ubuntu.
  • The buggy BIOS also allegedly affects the following motherboards:  - MSI P965 Platinum  - Asus P5K-E  - Asus P5E WS  - Asus P5E WS PRO
  • Foxconn say that an update will be released for the motherboard in the next few days.
  • No ETA for a fix from other vendors.
  • Nothing here points to any kind of conspiracy theory to lock Linux out of the game. The issue seems to be a bug, and this was made much worse than it needed to thanks to a poor tech support representative working for Foxconn. If tech support had logged the fault rather than argue that the board was Vista only, none of this would have been a problem as it would have been just another bug.

Thanks to Ryan for keeping me updated.

Topics: Hardware, Enterprise Software, Legal, Linux, Open Source, Operating Systems, Software

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

80 comments
Log in or register to join the discussion
  • About the treatment from Foxconn

    I've been apologized to for this multiple times, and I'm willing to give them the benefit of the doubt that they're committed to doing the right thing now.

    They've told me they reprimanded the employee responsible for those "Buy Vista" answers.

    Have they? I really wouldn't have even asked them to, their response to Linux inquiries now has been to ask the customer how they can help, instead of insulting them, so this satisfies me.

    Within the next week or so, this BIOS fix will get pumped out, much to Steve Ballmer's dismay ;) and we can all go back to doing what we do, being computer enthusiats without paying the kings ransom and giving up freedom to Microsoft.

    I still strongly believe Microsoft had a hand in this, but to what level is hard to say, had they bought off AMI it would be as unoffical and off the record as possible, they don't want the pesky DOJ tracing this to them, even though several million motherboards not capable of running Linux would delight them.

    The gloves are off now though, and it'll be harder for Foxconn to get by with the negligence excuse again. (and they make most motherboards for PC and Mac, so this is not a victory in vein.)
    TheAlmightyCthulhu
    • Well, at the very least, MS LIKES these kinds of problems, though, it

      appears they had nothing to do with it. On the bright side, about any motherboard company DOES want to be sure Linux works on their motherboards, and WILL quickly fix it when they are notified. It IS very simple to support Linux.
      DonnieBoy
  • Thanks for the clarification

    Thank you for helping clear this up, Adrian.

    This would never have happened with Windows, though. It's distressing that manufacturers let products slip through to market with such obvious, glaring defects in them.

    Doesn't anybody believe in testing, anymore?
    trentreviso
    • Sure it works.......

      "Microsoft said so!"

      Seriously though, Windows covers up most bad BIOS behavior, for disreputable companies that want to make a quick buck, the Vista seal is all they're wanting, screw you, you just BOUGHT the product, not like they need happy customers. ;)
      TheAlmightyCthulhu
    • Ditto

      Great job, Adrian! I appreciate the follow thru.

      Regarding testing, it's possible that they tested the first
      boards that made it past production, but didn't continue to
      test for Linux, and a bad batch got thru. The important thing
      is that there was an outcry and that Foxconn is dealing with
      the problem. Credit to Adrian and others for getting
      Foxconn's attention.
      Marcos El Malo
  • So it becomes their tech support's fault?

    Nice way to shift the blame. Well, if you guys had not pointed that wild, pathetic "conspiracy" finger at them in the first place, there might not have been so much baby crying after all.
    LBiege
    • Pass the buck

      No, Foxconn is not blaming their tech support. They're blaming American Megatrends, which wrote the BIOS. They're saying that they are excused from releasing a defective board because the defective part wasn't built by them.As if that excuses them from not testing their products before releasing them to market.

      When you're caught with a mistake, blame it on somebody else.

      And the conspiracy finger still applies. I still smell a rat here.
      trentreviso
      • they choose not to support the product for linux

        And frankly Linux is using some outdated portions of the BIOS. Looks like they are no keeping up with the changes.
        code_Warrior
        • Wrong!

          Outdated portions of the BIOS? More like AMI using tables to hand out parameters, depending on WHICH os is in use - because some OS's (Windows) don't implement ACPI correctly.

          The truth is ACPI is broken under windows. Linux works just fine with the published ACPI spec.
          SpikeyMike
          • ACPI was co-authored by Microsoft ....

            and Intel. It is fully implimented in Windows. The truth is it is Linux that is behind in implimenting ACPI anot Windows.
            ShadeTree
          • So? They are the sole authors of OOXML...

            ... and they can't implement that correctly either!
            SpikeyMike
          • Works fine with office.

            What is your issue?
            ShadeTree
          • He's referring...

            ...to the ISO implementation of OOXML.

            Not that it's relevant here.
            AndyCee
          • And Linux's ACPI is written by Intel.

            At least, the Maintainer for Linux's ACPI implementation is Len Brown, and he signs patches using his Intel email address. He's been doing this job since at least 2003 as well.
            Zogg
          • What is your point?

            One person from Intel working on Linux kernel really isn't the same thing as a staff of people from Microsoft implimenting ACPI now is it?
            ShadeTree
          • One person??? Try learning how Linux is developed first!

            Len Brown is the [b]Maintainer[/b], which means that he's "[u]in charge[/u]". It doesn't mean that he's the only person doing it!?!! It doesn't even mean that he's the only person from Intel!

            For a complete list of [i]everyone[/i] who has worked on Linux ACPI, I suggest you consult the mailing list archives:

            http://marc.info/?l=linux-acpi&r=1&w=2

            http://www.lesswatts.org/projects/acpi/mailing-lists.php

            And as for "a staff of people from Microsoft implimenting(sic) ACPI" for Windows, how do [u]you[/u] know how many people Microsoft has assigned to this task? Has Microsoft told you? Granted, it's [b]probably[/b] more than one, but that's still no more than a guess.

            So we have ACPI, co-written by Microsoft and Intel, and with Intel driving Linux's ACPI support.
            Zogg
          • Don't be obtuse!

            The title of the poster's post was Intel wrote the ACPI for Linux and gave one person from Intel as proof. Still you have the nerve to attack my post. My role as the engineer in charge of OS support for one of the top three OEMs allows me close contact with Microsoft including the team that works on Power Management support. It is not a quess! You guys don't have a clue!
            ShadeTree
          • Arrogant bluster, masking poor reading comprehension!

            [i]"The title of the poster's post was Intel wrote the ACPI for Linux"[/i]

            Correct up to here, but then falls down [b]miserably[/b] when you carry on to say:

            [i]" and gave one person from Intel as proof."[/i]

            My [i]actual[/i] statement was:

            [i]"[T]he Maintainer for Linux's ACPI implementation is Len Brown, and he signs patches using his Intel email address."[/i]

            So [b]your[/b] problem is that [b]you[/b] have no idea what "the Maintainer" of a Linux subsystem does. Which in turn implies that your "role as the engineer in charge of OS support for one of the top three OEMs" doesn't [i]even remotely[/i] cover any Linux-based OS.

            [i]"Still you have the nerve to attack my post."[/i]

            Oh my! [b]Now[/b] I'm scared! You really shouldn't advertise your insecurities like this in a public forum.
            Zogg
          • Let me simplify this so even you can understand.

            ACPI is implimented in the BIOS or EFI depending on the motherboard. Intel does not write their own BIOS or EFI. Claiming that Intel wrote ACPI support for Linux is luidicrous. They don't even write it for themselves. AMI does! Is that clear enough? Your arguements as were the original poster's are specious.

            Your attempts to use misdirection are futile. This is not about the Linux development model which I am very aware of. That is a red herring! What we are talking about is the ridiculous claim that Intel is behind ACPI support for Linux. Take your own insecurities some where they are appreciated. You are coming off as uniformed and hysterical!
            ShadeTree
          • Just calm down and learn to read.

            [i]"Claiming that Intel wrote ACPI support for Linux is luidicrous(sic)."[/i]

            If you look back, you'll see that I [b]actually[/b] said (and repeated):

            "[i][T]he Maintainer for Linux's ACPI implementation is Len Brown, and he signs patches using his Intel email address.[/i]"

            Intel [b]is[/b] actively involved in writing the Linux kernel's ACPI support. That's a demonstrable [b]fact[/b].

            [i]"Your attempts to use misdirection are futile. This is not about the Linux development model which I am very aware of."[/i]

            Except that you clearly [b]don't[/b], because you obviously don't know what the Maintainer of a Linux subsystem does.

            [i]"What we are talking about is the ridiculous claim that Intel is behind ACPI support for Linux."[/i]

            What on Earth is so ridiculous about Intel, co-author of ACPI, helping Linux support ACPI? Have you misunderstood me?

            And the purpose behind this thread was to show that when you said:

            "[i]The truth is it is Linux that is behind in implimenting(sic) ACPI[/i]"

            that you were spreading FUD. The [i]real[/i] truth is that one ACPI co-author writes Windows' ACPI support (all by itself), and the other leads the development of Linux's ACPI support.

            "[i]You are coming off as uniformed and hysterical![/i]"

            Err, no. That's [u]definitely[/u] you.
            Zogg