Yes it is right direction.  Yes it is sub_dp.  I can dig out examples on
Monday.

Also, with some planning,  sub tasks can also have sub tasks
On Jul 5, 2014 1:21 PM, "Tony Thigpen" <[email protected]> wrote:

> Thanks, that means that I my thinking is in the right direction.
>
> For the pipi call, do you remember if it was the CEEPIPI(init_sub_dp)
> function?
>
> Tony Thigpen
>
> -----Original Message -----
>  From: Sam Siegel
>  Sent: 07/05/2014 01:51 PM
>
>> Tony. In z/os I've done the following:
>> 1) root task attaches non-le asm subtaks.
>> 2) each sub tasks issues pipi with parms for non- main program
>> 3) each sub task saves and reuses its own le token.
>> 4) each sub task call LE program via pipi and passes parms as documented.
>>
>> I'm not at office, so I can't give more details or name of manual
>>
>> Sam
>> On Jul 5, 2014 9:32 AM, "Tony Thigpen" <[email protected]> wrote:
>>
>>  I am working in z/VSE, but I know you z/OS guys/gals have been involved
>>> with LE for a longer period so maybe you have been down this road before.
>>>
>>> In one vendor product I maintain, I have a mixture of LE/C,
>>> Non-LE/Assembler and LE/enabled Assembler.
>>>
>>> I am in the situation where I need to attach additional subtasks to my
>>> main task and those subtasks need to utilize LE services.
>>>
>>> I have been unable to find any documentation on running multiple LE tasks
>>> in a single address space. I am not even sure if it is supported in
>>> z/VSE,
>>> or even in z/OS.
>>>
>>> I believe I will need to ATTACH the subtask and have the subtask issue
>>> it's own PIPI call.
>>>
>>> I have used PIPI before. I have created subtasks before. I just never
>>> tried to do both together.
>>>
>>> If anyone has tried or done anything like this on ether z/OS or z/VSE,
>>> can
>>> you post something about your experience and what you found worked or did
>>> not work?
>>>
>>> Thanks,
>>>
>>> Tony Thigpen
>>>
>>>
>>
>>

Reply via email to