Follow Slashdot blog updates by subscribing to our blog RSS feed

 


Forgot your password?
Close
typodupeerror
×
GNU is Not Unix Intel Power Hardware

Intel Announces a BIOS Implementation Test Suite 66

Posted by timothy from the gnu-way-to-look-at-things dept.
Josh Triplett writes "Intel announced the release of a BIOS Implementation Test Suite (BITS), a bootable pre-OS environment based on GNU GRUB2 that tests how well (or how badly) your BIOS has configured your platform hardware. BITS also includes Intel's official power management reference code, so you can override your BIOS's initialization with a known-good configuration. 'In addition to those changes to GRUB2 itself, BITS includes configuration files which build a menu exposing the various BITS functionality, including the test suites, hardware configuration, and exploratory tools. These scripts detect your system's CPU, and provide menu entries for all the available functionality on your hardware platform. You can also access all of the new commands we've added directly via the command line.'"
This discussion has been archived. No new comments can be posted.

Intel Announces a BIOS Implementation Test Suite More

Intel Announces a BIOS Implementation Test Suite

Comments Filter:

  • Nice one Intel (Score:1)

    by Anonymous Coward

    I don't often praise Intel (as their business practices seem overly shady on occasion) but I can definitely see this as being of use. Nice one guys.

    Now to try and integrate it into my pxe environment...

  • That's not gonna fly too well with overclockers (Score:5, Funny)

    by Rosco P. Coltrane ( 209368 ) on Friday February 25, 2011 @06:52AM (#35310638)

    a bootable pre-OS environment based on GNU GRUB2 that tests how well (or how badly) your BIOS has configured your platform hardware.


    !!INTEL BIOS WARNING!!

    We have just detect that you've configured your CPU in egg frying mode. Reverting to pansy mode. If you want a fast processor in pansy mode, please contact your nearest Intel dealer and open your wallet.

    • Re:That's not gonna fly too well with overclockers (Score:5, Insightful)

      by SpazmodeusG ( 1334705 ) on Friday February 25, 2011 @07:06AM (#35310700)

      Actually simply add a CPU and memory stress test and this would be great for overclocking!

      It's already got all the CPU identification stuff that overclockers need. It boots quick and off a USB drive too!

      The only thing it's missing is a CPU and memory stress tester. With that you'd be able to quickly change settings, reboot and test them without having to stuff around loading a full OS.

    • I fail to see in what way this wouldn't "fly too well" with overclockers. As far as I can see, there is nothing forcing you to use it. It is a tool that you can choose to use to investigate (and correct) how your BIOS initializes your hardware. Regardless, if you're such a daredevil 1337 tinkerer it should be fairly easy for you to remove this toolset if it somehow came preinstalled on your computer with a configuration that completely bypassed your interaction.

    • Re: (Score:1)

      by Anonymous Coward

      It's open source, so it would be trivial to remove that.

    • Re: (Score:2)

      by blueg3 ( 192743 )

      Why would they bother? Overclockers are the same people who buy Intel chips twice as often as everyone else.

  • Awesome (Score:5, Insightful)

    by Anonymous Coward on Friday February 25, 2011 @06:55AM (#35310646)

    Can we finally go to EFI or at least something that's not 20 years old now?

    • Re: (Score:1)

      by Anonymous Coward
      What the hell for? All you are going to do is use a microprocessor that's 40 years old with it.
    • A lot of mobos do use UEFI, switching when Sandy Bridge came out.

    • Re: (Score:1)

      by mpfife ( 655916 )
      I'm using UEFI on my new Sandybridge system right now - and MAN is it nicer. Booting on partitions larger than 2TB is probably one of the biggest wins now that 3TB drives are shipping (CMOS is limited to 2TB boot partitions). The GUI display with ability to use keyboard+mouse right away is very nice. It's also SO overdue that looking at older bios' makes me cringe. Really? Text setup with what looks like old BBS color schemes? Am I setting up by modem still? Sure they had the degrading SATA controller

  • Anybody knows if this can be used to enable VT-x on a cpu that the BIOS has disabled it on?
    I have HP machine with a Intel Core2 Duo E4400 and the BIOS does not have a switch to enable it for some reason.

  • Looks like it allows you to skip the BIOS and load a Linux kernel directly. Sort of like coreboot [coreboot.org].

    • Re: (Score:2)

      by qmaqdk ( 522323 )

      My bad. It's only pre-OS, not pre-BIOS. Looks like we'll have to waste 30 seconds on the BIOS after all.

  • So why need a BIOS in the first place? (Score:3, Interesting)

    by tropophobia ( 867619 ) on Friday February 25, 2011 @07:04AM (#35310688) Homepage
    BIOS does actually very little these days. The OS re-initializes most devices anyway on boot, using BIOS values only for reference. From first look, this release kind of makes BIOS obsolete. If it knows how to fix BIOS misconfiguration, then it can also configure it in the first place. The rest can be taken care by the OS.

    • Re: (Score:3, Funny)

      by Anonymous Coward

      Because some of us old-timers still run DOS (be it MS-DOS or FreeDOS) that relies solely on the BIOS for interfacing with the hardware so we can that ancient software that runs $$$$ worth of industrial equipment. Now get off my lawn.

      • so everyones computer should be designed with your legacy equipment in mind?

      • Re: (Score:2)

        by Rich0 ( 548339 )

        Ok, so we'll leave the BIOS as a standard item in PCs that come equipped with ISA or MFM drive controller cards. :)

      • Commission some legacy stuff and leave us alone!
        We do not have to pay for your obsolescence.

      • ooh! I see a business opportunity that'll require you to splash out $$$$ for a new version of exactly the same thing.

        ka-ching, ka-ching, ka-ching.

        sure, you'll be pissed, but you know, someone's got to stimulate the economy and the rest of us have been playing this legacy upgrade dance for years.

    • As far as I know the BIOS is unfortunately still involved with anything related to power management through ACPI - suspend/resume etc.

    • To make my Aspire One hang once in a while :( Not even the Linux SysReq keys work anymore, so I expect it is the BIOS that does it.

    • Re:So why need a BIOS in the first place? (Score:5, Insightful)

      by Thelasko ( 1196535 ) on Friday February 25, 2011 @10:40AM (#35311864) Journal

      BIOS does actually very little these days.

      Then why does it take so frigging long to load?

      • It's waiting for you to press the 'Any' key to continue.

      • Then why does it take so frigging long to load?

        BIOS does very little, but one of the things it does is a basic test of all the memory in a machine. That takes a while to work. Besides, if it were much faster, your hard disk wouldn't have a chance to spin up.

    • Re:So why need a BIOS in the first place? (Score:5, Interesting)

      by Cyrano de Maniac ( 60961 ) on Friday February 25, 2011 @12:56PM (#35313262)

      BIOS does actually very little these days. The OS re-initializes most devices anyway on boot

      Well, actually being a BIOS developer, I can state with absolute confidence that you're wrong about BIOS doing very little these days.

      The BIOS these days takes care of an incredible amount of work, such as detecting, training, testing, and configuring RAM, initializing the CPU state on many cores, configuring the interconnect between processors (QPI on some recent Intel processors, HT on AMD), setting up system memory maps, probing and setting up the entire IO fabric, building tables (e.g. ACPI) that fully describe every nitty-gritty aspect of the system to the OS, make your USB keyboard and mouse functional for ancient OSes, work around problems in hardware, have small drivers for accessing myriad devices for reading blocks from boot devices, in the case of EFI/UEFI manage options for boot ordering as well as bazillions of basic system settings, actually implementing each and every one of those bazillion settings, handle all sorts of hardware abstractions in the form of BIOS/EFI calls, manage and configure IO BARs, provide code to handle all sorts of potential correctable (and sadly sometimes uncorrectable) hardware errors, in some cases provide disaster fallback paths if you manage to corrupt the main BIOS image, in the case of EFI provide a runtime environment for pre-OS applications, etc. -- and do all of this with absolutely nothing underneath it other than hardware. If you think this is "very little", I'd encourage you to find a job developing BIOS code, and I think you'd be overwhelmed by the sheer bulk of the codebase in a modern BIOS. Just the source code trees these days push a fair bit over the 100MB level. Seriously.

      Having also worked on OSes and kernel-level device drivers, it is true that the OS re-initializes a fair bit of the hardware, but not nearly the level of hardware the BIOS initializes (have fun trying to re-train RAM or reconfigure the CPU interconnect, for example). If anything the trend has been toward the BIOS taking on greater and greater responsibility for device initialization and provision of runtime services to make the OS less aware of "quirks" in the hardware. That's not to say there isn't a ton of work the OS still has to do, but your statement vastly over-trivializes the role of the BIOS in modern machines.

      • Re: (Score:2)

        by Nick Ives ( 317 )

        It's surprising how common the idea that "BIOS does very little" is, even amongst developers.

        The coreboot guys did a Google tech talk [youtube.com] about all the fun they had during their initial LinuxBIOS days as they began to discover just how much work a BIOS does. It's an hour long but quite fascinating!

      • but your statement vastly over-trivializes the role of the BIOS in modern machines.

        Hey, he doesn't understand it so he assigns it minimal value. There's a Dilbert where the PHB assigns Dilbert 3 minutes to design a world-wide client server architecture.

      • This may sound ignorant, but what exactly what type of re-training does
        RAM need with every boot?

        Sounds a bit odd to have to retrain memory at all, let alone with each boot....

        • Re: (Score:3)

          by Agripa ( 139780 )

          DDR2, DDR3, and GDDR5 require skew compensation (and perhaps equalization) for various signals because of manufacturing variations in the signal environment (motherboard, sockets, DIMMs, Number of occupied sockets, DIMM or chip loading, etc.) and in some cases because of the design (DDR3 chains some signals from chip to chip) in order to meet setup and hold requirements. GDDR5 is sensitive enough to require retraining even with temperature variations.

  • the Year of HURD [gnu.org] Desktop is near?

  • Pendrive Linux? (Score:3)

    by jginspace ( 678908 ) <.jginspace. .at. .yahoo.com.> on Friday February 25, 2011 @09:48AM (#35311358) Homepage Journal
    As it's Grub-compatible, I hope it's going to be easy to add to a multi-boot usb toolkit. Along the lines of: http://www.pendrivelinux.com/boot-multiple-iso-from-usb-multiboot-usb/ [pendrivelinux.com]
  • Microsoft already snatched the "BITS" acronym to refer to the "background intelligent transfer service" used by Windows Update and other services. Hopefully there won't be any unnecessary confusion or ambiguity.

Slashdot Top Deals

E = MC ** 2 +- 3db

Close