[lkml]   [2010]   [Dec]   [9]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
    SubjectRe: [PATCH 1/3] X86: Revamp reboot behaviour to match Windows more closely
    On Thu, Dec 09, 2010 at 02:53:35PM -0800, H. Peter Anvin wrote:
    > On 12/09/2010 02:51 PM, Matthew Garrett wrote:
    >> I'm afraid I don't understand your argument. The date cutoff would be on
    >> the order of 2001 (anything after this will have been tested with XP).
    >> The spec that defines this behaviour only came into existence in August
    >> 2000, and any older hardware will be missing the flag that indicates
    >> that this feature is supported. It doesn't seem realistic to believe
    >> that there's any real body of hardware that sets the flag but otherwise
    >> has a broken implementation.
    > 2001 is probably a good date, then.
    > It's pretty safe you'll see the bit being set on systems which are older
    > than that, even if it was not defined at the time it was created -- just
    > being garbage. That's par for the course in BIOS land.

    There's a revision field in the FADT. They'd need to simultaneously
    provide an incorrect revision *and* by pure luck set the 10th bit of a
    32-bit register. Is it possible? Yes. Is it likely? No, and I don't see
    a benefit in adding extra code to force hardware into a less-tested

    Matthew Garrett |

     \ /
      Last update: 2010-12-10 00:01    [W:0.021 / U:2.952 seconds]
    ©2003-2016 Jasper Spaans. hosted at Digital OceanAdvertise on this site