[Info-vax] Changed Boot behavior on XP1000 since 2013

[Jan-Erik Soderholm]

Joukj wrote 2013-02-08 09:26:
> Hans Vlems wrote:
>> On 7 feb, 14:10, Volker Halle <volker_ha... at hotmail.com> wrote:
>>> HP has responded in the HP Support Forums and is actively collecting
>>> information about affected systems and thinking about a workaround in
>>> OpenVMS.
>>>
>>> If one of your systems is effected, please obtain the following
>>> information:
>>>
>>> 1. The Alpha system type (XP1000, AS1200, etc.)
>>>
>>> 2. P00>>> e toy:2 -w -n 3      (collect this after Power-Up or >>>
>>> INIT)
>>>
>>> Then log a call with HP (or maybe just post the information here or in
>>> the HP Support Forums):
>>>
>>> http://h30499.www3.hp.com/t5/Hardware/Changes-Boot-behavior-on-XP1000...
>>>
>>> Volker.
>>
>> The fix suggested there (by HP?):
>>
>> <HP>
>> The following code segment is used by OpenVMS:
>>
>> if .time_result [tim$w_year] GEQ .EXE$GL_TRANSITION_YEAR
>> then
>> time_result [tim$w_year] = .time_result [tim$w_year] + 1900;
>> else time_result [tim$w_year] = .time_result [tim$w_year] + 2000;
>>
>> OpenVMS engineering thought of the following fix / workaround for this
>> issue.  In the above code, change the hardcoded 1900 to a 1920 (1920 +
>> 93 = 2013).
>> </HP>
>>
>> Am I too critical here to consider that a kludge?
>> I'd rather fix the problem where it lies: the firmware. The next best
>> thing is to apply a fix BEFORE VMS gets booted.
>> My systems are far from production critical but I'd rather not run a
>> patched CPU$ROUTINES*.EXE for this problem.
>> Hans
>>
> 2 considerations;
>      1) OpenVMS should work with the hardware/firmware as it is. So if
>         the machine/firmware contains a bug OpenVMS should cope with it.
>      2) For patching the firmware HP will need:
>            a) people who know how firmware has been/is to be programmed.
>            b) machines to test it on
>          I wonder if HP has this "in-house"
>
>             Jouk

I see two realistic "solutions".

1. Fix/patch VMS to handle the brooken servers (as HP suggested).
2. Remove the brooken servers from the "supported" list in the SPD.

To ask for any change to servers that has gone
of support many years ago, is unrealistic.

The question is, has HP got a service request on this issue
from *any* paying customer yet?

Jan-Erik.