Foxconn BIOS breaks linux acpi.

[Quiksilver]

Link.

I personally don't use linux or a foxconn motherboard, but this might be important to someone.

 

[AmberClad]

I don't recall having any issues with any Dells or my old Foxconn board in Linux.

I wouldn't put it past them to have a buggy BIOS though - my Foxconn G965 somehow permanently lost the ability to run TRAS higher than 15 after I filled it up with 4 dimms. So it effectively could no longer run JEDEC spec (luckily my RAM didn't have any issues with the slightly tighter timings).

EDIT: For clarification, I haven't owned the specific board they're referring to, and I use Slackware, not Ubuntu. So YMMV.

The Microsoft conspiracy theory is sort of amusing to read, in any case.

 

[Soulkeeper]

that's why i stick to more name brand, well know manufacturers like Abit and Asus with a history of supporting all OS's
not worth the risk, i mean c'mon there are like 20+ board makers out there, i'll stick to the top 5 for my choices

 

[toadeater]

I don't think MS was doing anything malicious here, I think they were just caught being negligent. I wonder if they even bothered to test the board, or do they just hand out WHQL certificates to anyone who bribes them? You know, it's not like Nvidia and ATI's WHQL drivers have been bug-free, so what does WHQL mean anymore? It does not guarantee anything really.

 

[VirtualLarry]

Originally posted by: toadeater
I don't think MS was doing anything malicious here, I think they were just caught being negligent. I wonder if they even bothered to test the board, or do they just hand out WHQL certificates to anyone who bribes them? You know, it's not like Nvidia and ATI's WHQL drivers have been bug-free, so what does WHQL mean anymore? It does not guarantee anything really.

According to the dissassembly, it appears that the ACPI code is apparently thus:
if (OS = Microsoft)
working_ACPI_table();
else if (OS = Linux)
subltely_broken_ACPI_table();

Btw, does anyone know if the Foxconn 780G mobo is afflicted with this issue? I was debating picking one up, they are $67 shipped at NewEgg.

 

[toadeater]

Originally posted by: VirtualLarry

Originally posted by: toadeater
I don't think MS was doing anything malicious here, I think they were just caught being negligent. I wonder if they even bothered to test the board, or do they just hand out WHQL certificates to anyone who bribes them? You know, it's not like Nvidia and ATI's WHQL drivers have been bug-free, so what does WHQL mean anymore? It does not guarantee anything really.

According to the dissassembly, it appears that the ACPI code is apparently thus:
if (OS = Microsoft)
working_ACPI_table();
else if (OS = Linux)
subltely_broken_ACPI_table();

Yes, but was it broken intentionally, or was it just badly made?

Btw, apparently there is an Intel ACPI spec and an MS Vista ACPI spec. Anyone know more about this?

 

[taltamir]

well, appernatly redirecting linux to the vista table fixes a lot of the bugs, but not all of them, he said there were 7 other hidden checks inside there to see if you run linux that cause issues, and he changed them as well to hide it further...
It is also not implemeting ACPI, instead it detects what OS you have (win95, 98, NT, 2000/XP, vista, or linux are the options it detects) and then sends each one to a table that makes it "sorta work" with it... if they had just implemented ACPI with the full intel spec it would work with ALL oses without any issue as long as the OS supports ACPI...

It seems to be a combination of sabotage AND poor coding. Or they could have INTENDED to support linux, and then left it half finished (which does more damage then linking it to the vista table) and that caused it to break.

Either option is feasable.

BTW, no way in hell this board will work with windows 7 without a bios update... so again... crappy coding, but why are there 7 hidden windows check within the windows table itself? he says the only reason would be to deter someone from pointing linux at it for a quick fix.
He is also pissed at the responses they gave him which were "we don't support linux, buy windows vista" and "sure it is written to ACPI spec, microsoft would not have given us their WHQL certification if it wasn't"

 

[NXIL]

The Microsoft conspiracy theory is sort of amusing to read, in any case.

http://hehe2.net/thedarkside/m...n-the-foxconn-debacle/

After looking through the disassembled BIOS for the last several hours, rebooting it, and tweaking it more, I?d say this is very intentional, I?ve found redundant checks to make sure it?s really running on Windows, regardless what the OS tells it it is, and then of course fatal errors that will kernel panic FreeBSD or Linux, scattered all over the place, even in the table path for Windows 9x, NT, 2000, XP, and Vista, and had to correct them (Well, at least divert them off into a segment of RAM I hope to god I?m sure about)

No, this looks extremely calculated, it?s like they knew someone would probably go tearing it apart eventually and so tried to scatter landmines out so as to where you?d probably hit one eventually.

So if it is a mistake, or incompetence, then it?s the most meticulous, targeted, and dare I say, anal retentive incompetence I?ve seen.

Bill Gates said in an email:

One thing I find myself wondering about is whether we shouldn?t try to make the ?ACPI? extensions somehow Windows specific.

It seems unfortunate if we do this work and get our partners to do the work and the result is that Linux works great without having to do the work.

Maybe there is no way to avoid this problem but it does bother me.

Maybe we could define the API?s so that they work well with NT and not the others even if they are open.

Or maybe we could patent something related to this.

Maybe not a "conspiracy" in the "24" or CSI, but you may recall the saying "DOS isn't done till Lotus won't run"....Wordperfect going belly up because Office/Word ran better on Windows using undocumented OS calls and hooks......the Microsoft monopoly trial where they swore under oath that Internet Explorer had to be part of the operating system....(hence Netscape's demise).

Interesting that Asus is embracing Linux, as are other manufacturers....makes sense, for low cost hardware i.e 3-400$ laptop: not good for margins to have to pay Microsoft tax of 50-100$....

NXIL

 

[tcsenter]

After looking through the disassembled BIOS for the last several hours, rebooting it, and tweaking it more, I?d say this is very intentional, I?ve found redundant checks to make sure it?s really running on Windows, regardless what the OS tells it it is, and then of course fatal errors that will kernel panic FreeBSD or Linux, scattered all over the place, even in the table path for Windows 9x, NT, 2000, XP, and Vista, and had to correct them (Well, at least divert them off into a segment of RAM I hope to god I?m sure about).

I thought I would give an update on the conclusion of this saga after a "real" programmer with Linux kernel and BIOS expertise delved into the matter.

It appears that almost every "problem" detailed by Ryan was in fact due to bugs in the Linux kernel, not Foxconn's BIOS. Foxconn's BIOS did have one trivial bug that was not central to any of the serious problems Ryan reported. Garrett tested the latest Linux release candidate on a Foxconn G33M board running the same "defective" Foxconn BIOS version. Other than the restart issue, no problems were noted or encountered.

The OS 'sniffing' code in Foxconn's BIOS is in reality completely incapable of detecting whether it is running any _OSI supporting Linux version released in the past four years, treating it identically to Windows XP, 2003, or Vista.

In response, Ryan dismisses Garrett's contribution as nothing more than a personal attack against him. Ever better, Ryan goes on to suggest that, instead of making poor little Ryan look like an idiot by exposing his understanding of code as barely rank amateur, Garrett's time and energies would be better spent helping Foxconn and AMI write better BIOS code, and getting the bugs fixed in Linux.


Cliffs:

- "Ryan" a.k.a "TheAlmightyCthulhu" accuses Foxconn of deliberately sabotaging Linux, presumably at the behest of Microsoft

- His claims are Digged and Slashdotted to death, even reported or blurbed uncritically by supposedly reputable tech-centric webzines

- Ryan becomes an instant celebrity, an angry mob grows behind him calling for boycotts of Foxconn, complaints to government agencies, investigations of Microsoft

- Foxconn responds as one might expect any company to under such negative publicity, it apologizes and pledges to fix the problems in a forth-coming BIOS

- A "real" coder with Linux kernel and BIOS expertise examines Ryan's analysis of Foxconn's BIOS, then weighs in on the merits

- Ryan is proven to be fantastically wrong on almost every count, dismisses Garrett's expert rebuttal as a personal attack

- Ryan's exposure as a complete idiot and Foxconn's exoneration receives 1/1000th as much publicity or attention

 

[Fox5]

Originally posted by: VirtualLarry

Originally posted by: toadeater
I don't think MS was doing anything malicious here, I think they were just caught being negligent. I wonder if they even bothered to test the board, or do they just hand out WHQL certificates to anyone who bribes them? You know, it's not like Nvidia and ATI's WHQL drivers have been bug-free, so what does WHQL mean anymore? It does not guarantee anything really.

According to the dissassembly, it appears that the ACPI code is apparently thus:
if (OS = Microsoft)
working_ACPI_table();
else if (OS = Linux)
subltely_broken_ACPI_table();

Btw, does anyone know if the Foxconn 780G mobo is afflicted with this issue? I was debating picking one up, they are $67 shipped at NewEgg.

It does have the issue too.

Also, these bios are bought straight from a company and then customized for the hardware. It likely had preset tables for different OSes (perhaps even working around bugs within windows) and foxconn never bothered to handle anything other than vista/xp.