Comment 60 for bug 43745

Sergio Callegari (callegar) wrote :

Addendum... can it be related to this?
I am just quoting from debian bug 277298 and sites quoted therein...

I saw this:

http://lists.us.dell.com/pipermail/linux-poweredge/2005-February/042501.html

which suggests that the RTC hang issue can be solved by setting
CONFIG_HPET_RTC_IRQ=y
I have confirmed that on a dell 1850, which used to hang when hwclock
ran, the above configuration setting allows a normal boot. I suspect,
but have not yet confirmed, that disabling HPET may also prevent hangs.

and also...

Found by Googling for "8208CA", and then reading Google's
    "cached" copy. (This stuff isn't easy to find - the original
    has evidently been taken down:

    sunsite.rediris.es/sites/download.intel.nl/
design/chipsets/specupdt/290739.htm
                    ^^ Estonia??)

    This document may or may not explain 2.8.1 real-time clock
    problem.

Problem:
    Under certain conditions, a CPU generated I/O read to RTC
    (Real Time Clock) registers 0-9 may return an incorrect
    value. The issue occurs on the read path from the RTC
    registers and the RTC value in the registers is not impacted.
    Should the certain conditions occur, one or more of the bits
    read from the RTC registers may be incorrect. The issue has
    only been found using a synthetic test and has not been seen
    using commercially available software. [ <= N.B. ]

   "Implication:
    An operating system or software applications which
    synchronizes the time/date value with the RTC <registers may
    get an incorrect value.

   "Workaround:
    BIOS workaround is available in the Intel ICH3-S BIOS
    Specification Update Rev 2.01. The workaround does take into
    account multiple CPUs and Hyper Threading. The workaround uses
    timers to zero-in on the window where invalide RTC I/O read
    data could be returned. Using the software SMI, HPET timers
    and port 71 traps, the BIOS will ensure that there are no
    accesses to the RTC during this timing window.