The FSF would like to prepare some PR around this response an FSF supporter got from Intel customer care representative asking them about support for free BIOS on Intel hardware. Any thoughts from the coreboot community on this?
-------------------------- Thanks for your email.
Writing BIOS code is not like writing an OS device driver. Chipset specifications can vary not just between chipset models, but between steppings of the same chipset. Problems in chipset hardware and problems in BIOS code are hard to distinguish without specific hardware instrumentation. End user BIOS replacement with a third- party BIOS (whether free or not) on a commercial motherboard is not allowed by nearly all hardware vendors because of the potential for BIOS viruses and the risk of rendering the hardware useless through ill-advised modifications. For example, a laptop battery could explode if incorrect power management algorithms were applied.
BIOS is a part of the reliability and performance promise of the hardware. Chipset specifications at the level being discussed are commonly considered proprietary by all silicon vendors, not just Intel.
The open source firmware work that Intel *is* sponsoring could lead to a solution where proprietary low-level chipset initialization code from silicon vendors is made compatible with open source higher-level platform initialization and pre-boot management. If you are interested, we invite you to participate at www.tianocore.org.
Thanks once again for your interest in Intel.
Sincerely,
Intel Customer Support --------------------------
Thanks, Ward.
Quoting Ward Vandewege ward@gnu.org:
The FSF would like to prepare some PR around this response an FSF supporter got from Intel customer care representative asking them about support for free BIOS on Intel hardware. Any thoughts from the coreboot community on this?
Thanks for your email.
Writing BIOS code is not like writing an OS device driver. Chipset specifications can vary not just between chipset models, but between steppings of the same chipset. Problems in chipset hardware and problems in BIOS code are hard to distinguish without specific hardware instrumentation. End user BIOS replacement with a third- party BIOS (whether free or not) on a commercial motherboard is not allowed by nearly all hardware vendors because of the potential for BIOS viruses and the risk of rendering the hardware useless through ill-advised modifications. For example, a laptop battery could explode if incorrect power management algorithms were applied.
BIOS is a part of the reliability and performance promise of the hardware. Chipset specifications at the level being discussed are commonly considered proprietary by all silicon vendors, not just Intel.
The open source firmware work that Intel *is* sponsoring could lead to a solution where proprietary low-level chipset initialization code from silicon vendors is made compatible with open source higher-level platform initialization and pre-boot management. If you are interested, we invite you to participate at www.tianocore.org.
Thanks once again for your interest in Intel.
Sincerely,
Intel Customer Support
Thanks, Ward.
I don't know about that Ward, I think they are being a little silly on the matter. I have coreboot running on my Intel based board without exploding batteries. Sounds like some advertizing propoganda for tianocore to me:-)
Thanks - Joe
On Feb 11, 2008 1:07 PM, Ward Vandewege ward@gnu.org wrote:
BIOS is a part of the reliability and performance promise of the hardware. Chipset specifications at the level being discussed are commonly considered proprietary by all silicon vendors, not just Intel.
There's the real issue. This is pretty much the same line they've taken for the last 7 years. And, actually, not true. Just look at AMD or the ARM or the powerpc or ...
The open source firmware work that Intel *is* sponsoring could lead to a solution where proprietary low-level chipset initialization code from silicon vendors is made compatible with open source higher-level platform initialization and pre-boot management. If you are interested, we invite you to participate at www.tianocore.org.
yep, they've talked about this for 7 years too, and never once delivered. I even had a promise at several different times to have the code "in two weeks". I got the promise, but not the code.
Don't hold your breath.
ron
ron minnich wrote:
On Feb 11, 2008 1:07 PM, Ward Vandewege ward@gnu.org wrote:
BIOS is a part of the reliability and performance promise of the hardware. Chipset specifications at the level being discussed are commonly considered proprietary by all silicon vendors, not just Intel.
There's the real issue. This is pretty much the same line they've taken for the last 7 years. And, actually, not true. Just look at AMD or the ARM or the powerpc or ...
It's really just more FUD. If that was really the only reason they could release the specs and not offer support except for where they made errors in the data sheets. Problem solved.
-Bari
On Feb 11, 2008 4:07 PM, Ward Vandewege ward@gnu.org wrote:
The FSF would like to prepare some PR around this response an FSF supporter got from Intel customer care representative asking them about support for free BIOS on Intel hardware. Any thoughts from the coreboot community on this?
Thanks for your email.
Writing BIOS code is not like writing an OS device driver.
Yeah, nobody here knows what they're doing *eyeroll*
Chipset specifications can vary not just between chipset models, but between steppings of the same chipset. Problems in chipset hardware and problems in BIOS code are hard to distinguish without specific hardware instrumentation. End user BIOS replacement with a third- party BIOS (whether free or not) on a commercial motherboard is not allowed by nearly all hardware vendors because of the potential for BIOS viruses and the risk of rendering the hardware useless through ill-advised modifications.
Some (most?) hardware vendors say they only support MS Windows, the same logic could be applied. An open-source firmware provides a lot more assurance that the BIOS is free of malicious code.
For example, a laptop battery could explode if incorrect power management algorithms were applied.
And with the support of the vendor(s) and some better documentation, we can know how to avoid such issues.
BIOS is a part of the reliability and performance promise of the
hardware. Chipset specifications at the level being discussed are commonly considered proprietary by all silicon vendors, not just Intel.
Right. Because describing the boot process is somehow beneficial to the competition. I've seen a bunch of NDA'd docs from Via, and frankly, I'm completely lost as to how that would help anyone except a BIOS or driver developer.
The open source firmware work that Intel *is* sponsoring could lead to a
solution where proprietary low-level chipset initialization code from silicon vendors is made compatible with open source higher-level platform initialization and pre-boot management. If you are interested, we invite you to participate at www.tianocore.org.
Thanks once again for your interest in Intel.
Sincerely,
Intel Customer Support
I haven't the ambition to register with the site, is there any actual hardware init code there? And what's the point of starting another project, when coreboot can already load tiano (so I've heard)?
Thanks,
Ward.
-- Ward Vandewege ward@fsf.org Free Software Foundation - Senior System Administrator
Thanks for trying!
-Corey
On Feb 11, 2008 6:25 PM, Corey Osgood corey.osgood@gmail.com wrote:
Some (most?) hardware vendors say they only support MS Windows, the same logic could be applied. An open-source firmware provides a lot more assurance that the BIOS is free of malicious code.
The arguments against open source bios are almost exactly the arguments I heard years ago against open source OS.
FUD.
For example, a laptop battery could explode if incorrect power management algorithms were applied.
And with the support of the vendor(s) and some better documentation, we can know how to avoid such issues.
i want to know what laptops can explode if the software goes awry. In fact, I think the FAA wants to know. Laptops that can explode if software goes crazy should not be allowed on an airplane. Any chance intel will tell us more?
I haven't the ambition to register with the site, is there any actual hardware init code there? And what's the point of starting another project, when coreboot can already load tiano (so I've heard)?
coreboot has loaded tianocore.
I have registered at tianocore.org. There is no hardware init code there, nor will there ever be.
You can't build a working BIOS with the source code from tianocore.org -- unless you add coreboot to it.
ron
On Tuesday 12 February 2008, ron minnich wrote:
On Feb 11, 2008 6:25 PM, Corey Osgood corey.osgood@gmail.com wrote:
Some (most?) hardware vendors say they only support MS Windows, the same logic could be applied. An open-source firmware provides a lot more assurance that the BIOS is free of malicious code.
The arguments against open source bios are almost exactly the arguments I heard years ago against open source OS.
FUD.
True. But I also recall to never attribute to malovelence which can be explained by stupidity. That person must have been trained to answer this.
For example, a laptop battery could explode if incorrect power management algorithms were applied.
And besides the FAA problem, conclusively batteries cannot be charged when the main CPU is off -- cool design. Again, remember you are communicating with level 1 support.
coreboot has loaded tianocore.
Any pointers? Any hints?
As someone wrote (or quoted) on this list: EFI is an operating system still looking for a boot loader.
Torsten
ron minnich wrote:
On Feb 11, 2008 6:25 PM, Corey Osgood corey.osgood@gmail.com wrote:
For example, a laptop battery could explode if incorrect power management algorithms were applied.
And with the support of the vendor(s) and some better documentation, we can know how to avoid such issues.
Last time I checked all Lithium batteries had protection circuitry built into the pack to prevent any runaway conditions in the laptop charging circuits. The protection circuits override anything the laptop can throw at it short of a lightning bolt or exploding hard drives.
-Bari
On Feb 12, 2008 8:44 AM, bari bari@onelabs.com wrote:
Last time I checked all Lithium batteries had protection circuitry built into the pack to prevent any runaway conditions in the laptop charging circuits. The protection circuits override anything the laptop can throw at it short of a lightning bolt or exploding hard drives.
Yeah, that was the expected case. I don't know what the Intel person was claiming, but I hope it is wrong. Well, it is wrong :-)
ron
On 12.02.2008 18:55, ron minnich wrote:
On Feb 12, 2008 8:44 AM, bari bari@onelabs.com wrote:
Last time I checked all Lithium batteries had protection circuitry built into the pack to prevent any runaway conditions in the laptop charging circuits. The protection circuits override anything the laptop can throw at it short of a lightning bolt or exploding hard drives.
Yeah, that was the expected case. I don't know what the Intel person was claiming, but I hope it is wrong. Well, it is wrong :-)
If you control the embedded controller in charge of charging (sorry about the puns), you probably can fry a battery to the point where it starts burning. IIRC the OLPC EC code once had such problems. Then again, having a chipset spec will not help you at all when you try to mess with the EC.
Regards, Carl-Daniel
Carl-Daniel Hailfinger wrote:
If you control the embedded controller in charge of charging (sorry about the puns), you probably can fry a battery to the point where it starts burning. IIRC the OLPC EC code once had such problems. Then again, having a chipset spec will not help you at all when you try to mess with the EC.
That is why Li-Ion batteries have thermal protection circuitry. If they get too hot, the cells are open-circuited. There are several layers of protection in Li-Ion batteries to help protect them from ignition and explosion.
Here are a couple of links that get into more detail:
http://www.electronics-lab.com/articles/Li_Ion_reconstruct/
http://www.electronicproducts.com/ShowPage.asp?FileName=fa-nxp.mar2007.html
"Lithium will ignite and burn in oxygen when exposed to water or water vapors. It is the only metal that reacts with nitrogen at room temperature.
Lithium metal is flammable and potentially explosive when exposed to air and especially water, though it is far less dangerous than other alkali metals in this regard. The lithium-water reaction at normal temperatures is brisk but not violent. Lithium fires are difficult to extinguish, requiring special chemicals designed to smother them."
Great Google video here alkali+metals+water: http://video.google.com/videoplay?docid=1647759830306530002&q=alkali+met...
That is why car companies trying to launch vehicles that use Li-Ion batteries have spent so much effort on safety, since they may use thousands of cells in a single vehicle.
This is another reason why lithium iron phosphate or nanophosphate based cells are being used in new designs since they don't explode under extreme conditions.
KillaCycle - World’s Quickest Electric Motorcycle http://www.killacycle.com/
-Bari