Foxconn bioses

[Anony]

found this post on the ubuntu forums
http://ubuntu-virginia.ubuntuforums.org ... p?t=869249
any thoughts?

[Fitzcarraldo]

The following explains it more succinctly:

http://www.custompc.co.uk/news/604521/f ... page1.html

but I suspect it's more likely a cock-up than a conspiracy.

[chickpea]

Yeah, I get the feeling that this is just extraordinary stupidity coupled with laziness rather than mendacity.

[Fitzcarraldo]

I've been reading the two Ubuntu threads about this in more detail, plus the umpteen blog posts and other forum posts that are scattered across the Web about this issue. It's actually quite interesting because it highlights how poor ACPI support in BIOSs and some Linux kernels can be. For example, I happened to notice one person commenting that if they turned ACPI off their DVD drive started to work!:

BEGIN QUOTE
But brings up another point, I set up a Lenovo laptop for a lady, but there is obviously a BIOS issue. One of two things happen, either it ignores the CD/DVD hardware or It won't suspend resume.

The two are mutually exculsive. I can enable ACPI, and I get suspend resume, but the OS in unable to see the CD/DVD harware, it's as if it doesn't exist. Or I can turn ACPI off, I get no suspend resume, and in fact I get no auto poweroff either, but the OS can see the CD /DVD hardware, and it mounts a CD or DVD fine. This has got to Faulty BIOS. The question this article raises for me is "is this deliberate?"
END QUOTE

Now, I expect the optical drive in the lady's Lenovo laptop is ATA, ATAPI or whatever. But that's not always the case. For example the internal optical drive on my Acer laptop appears to be USB-connected (diagnostics under Linux and Windows XP report it as a USB drive, and it runs slower than I would expect for an ATA or ATAPI device), so the following text from another Web site entitled "Various notes about Acer Travelmate 3004WTMi with Linux" (different model to mine) is also interesting:

BEGIN QUOTE
Boot with “pci=noacpi” or even “acpi=off”. You won’t have ACPI power saving, but the external Firewire DVD-drive and network will work. On at least installed system, “linux pci=noacpi” seems to be enough.
END QUOTE

This Foxconn story has highlighted ACPI support in firmware/software. Even if the guy who first raised this issue, one "Ryan", is wrong about Foxconn deliberately sabotaging a BIOS against Linux, it looks like there might now be some much-needed focus on improving ACPI support for Linux in some BIOSes and some Linux kernels. The SL Forum, and many other Linux distros' forums, have numerous threads relating to ACPI problems. So perhaps this guy has done Linux a favour, even if some of his language comes across as rather obnoxious.

[chickpea]

Unfortunately, would have to term his tone as "petulant", which doesn't help his cause. But I do feel like BIOS/motherboard manufacturers are REALLY ignoring the ACPI spec, either intentionally, or sloppily, or both. In either event, it causes a lot of headaches. My HP dv6000 series laptop for example. I have had CONSTANT unending problems with the real time clock. For whatever reason, the kernel just can't control it. 2.6.25 for the most part solved this and the real time clock would be reliably polled to set the system time. But since upgrading to 2.6.26, It has stopped working. I suspect that the BIOS and DSDT table is at fault. Although this is an HP/AMD laptop (so both companies are known linux supporters) it just fails to be painfree. Under 2.6.21 kernels, the system used to kernel panic on boot and not boot up because of this issue. It has been a real pain that has come and gone over the life of this machine.

[Fitzcarraldo]

Yes, an HP dv6000 series laptop was specifically mentioned in one of the threads that I read regarding the Foxconn BIOS -- maybe the Ubuntu thread -- as suffering from this buggy BIOS problem with Linux as well. To be honest, having read about a hardware problem on some models of HP laptops (HP Pavilion dv6000 and dv9000) and now reading that the HP BIOS is buggy w.r.t. ACPI, has put me off buying HP laptops. Not that HP BIOSs are the only suffers: see the two Ubuntu threads on the Foxconn BIOS saga for evidence of the same bug or similar bugs in other manufacturers BIOSs.

[chickpea]

Yes, my BIOS is notorious as well. After a REALLY bad experience with HP customer service (unrelated to this laptop) I will no longer be buying any of their products. I got this laptop as a replacement from Bestbuy due to unfixable warranty issues with an old sony laptop that I had, so not like I chose it ( wouldn't with it's chipset Buggy BIOS/Broadcom wireless UGH!)

[Stupot]

Update: Apparently it affects a whole group of motherboards using MSI BIOS.

http://izanbardprince.wordpress.com/200 ... -the-code/

[Fitzcarraldo]

If MSI fix this bug in AMI and Award BIOSs then a large number of Linux users are likely to benefit by better ACPI support. Since buggy ACPI support is a perennial topic in Linux forums, this could be a significant advance for the OS.

[mdmarmer]

http://ubuntuforums.org/showthread.php?t=877721

Mike