Every thing should be as simple as possible but no simpler. A non-zero wait
count in the current RB of a task is only one of the reasons that it may not
eligible to run.
Address space ineligible
TCB nondispatchable
Current RB has wait count > 0
--
Shmuel (Seymour J.) Metz
On Sun, 16 Jul 2023 11:08:51 +, Seymour J Metz wrote:
>With ATTACH, you need to play games to prevent two tasks from running
>concurrently on two CPUs. With coroutines, you have multiples contexts within
>a single thread; there is no need for explicit interlocking.
>
Il a les défauts de ses
Oh I see!
Thanks. Yes, that would make sense, although from my trivial understanding
it might make coroutines less useful for my kind of work.
Roops
On Sun, 16 Jul 2023, 12:08 Seymour J Metz, wrote:
> With ATTACH, you need to play games to prevent two tasks from running
> concurrently on two
I must have missed the point of something, because on first reading, it's
analogous to what we could do with ATTACH, ECB and WAIT in assembly under
z/OS and MVS, or the equivalents in PL/I and COBOL (I assume) where we have
a subtask which can wait for an event and then resume operation from its
With ATTACH, you need to play games to prevent two tasks from running
concurrently on two CPUs. With coroutines, you have multiples contexts within a
single thread; there is no need for explicit interlocking.
--
Shmuel (Seymour J.) Metz
http://mason.gmu.edu/~smetz3
Coroutines were never intended to replace tasks, despite what wiki implies in
the lead. See Conway, Melvin E. (July 1963). "Design of a Separable
Transition-diagram Compiler" (PDF). Communications of the ACM. ACM. 6 (7):
396–408. doi:10.1145/33.366704. at
It's not expected that a SLIP dump is required to debug an application error
(whether it's a segment fault or anything else). Quite the reverse.
So in what way was the CEEDUMP and/or transaction dump not adequate? That
should be brought to the attention of the LE team.
Peter Relson
z/OS Core
Think generators and iterators.
From: IBM Mainframe Discussion List on behalf of
Rupert Reynolds
Sent: Sunday, July 16, 2023 2:03 PM
To: IBM-MAIN@LISTSERV.UA.EDU
Subject: Re: C++ coroutines are recent, and difficult?
Oh I see!
Thanks. Yes, that would
Peter,
If you can run GTF on the Zxplore system and capture the output, you might
want to try to use the DEBUG option of the SLIP SET command:
DEBUG
For a SLIP SET trap, allows you to determine why a trap that you set is
not working as you expected by indicating which of the
The Zxplore admin just advised me that the definitions they had set up to
provide DASD space for an SVCDUMP may have expired at the last IPL they did a
week ago. I will report back when he finds an answer to that question.
Peter
From: IBM Mainframe Discussion List On Behalf Of
Farley, Peter
It seems to me that the Zxplore Python is badly broken and has major issues
with codepage translation. we’ve seen this before when you opened an issue for
pyzfile. I’m going to update the ticket to that effect so the maintainers
don’t waste time chasing shadows.
> On 17 Jul 2023, at 3:57 am,
> On 17 Jul 2023, at 12:44 pm, Farley, Peter
> <031df298a9da-dmarc-requ...@listserv.ua.edu> wrote:
>
> Dave,
>
> It may well be that the python V3.9.2 on the Zxplore system is “broken” as
> you suggest, but getting a dump that proves it to the Zxplore admins still
> could help goad them
Mike,
I suspect I do not have the authority to run GTF on the Zxplore system, not to
mention I have no knowledge of how to run it in the first place. My userid on
Zxplore is just like any of the other “students” registered to learn there,
with deliberately limited authority.
It would fall to
This is the latest SLIP setup from today which did not catch the abend when the
python script is executed from an ssh login to z/OS on the Zxplore system:
SLIP
SET,ID=PJF1,JOBNAME=Z02446*,COMP=0C4,ACTION=SVCD,AL=(H,P,S),SDATA=(ALLNUC,PSA,RGN,LPA,CSA,SQA,TRT,GRSQ,SUM),END
Peter
From: IBM
Peter,
A CEEDUMP was in fact supplied to the ibmdb team on github in the initial error
report that I filed there. I generate a CEEDUMP in my $HOME directory every
time I execute the example python script.
I don't think any of the ibmdb team has reviewed the dump that I submitted as
of yet,
Dave,
It may well be that the python V3.9.2 on the Zxplore system is “broken” as you
suggest, but getting a dump that proves it to the Zxplore admins still could
help goad them to more rapidly upgrade the python that is available there. I
don’t see that as a useless exercise or as wasting
Here are our LE settings. It seems DYNDUMP is significant, you can specify this
as a runtime option https://www.ibm.com/docs/en/zos/2.4.0?topic=ulero-dyndump.
CEEDUMP(60,SYSOUT=*,FREE=END,
SPIN=UNALLOC)
DYNDUMP(*USERID.CEEDUMP,DYNAMIC,TDUMP)
ENVAR("TZ=GMT-8")
Everything in a wait is not "eligible" to run. That's pretty much the
definition of wait. :)
Brian
--
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to lists...@listserv.ua.edu with the message:
18 matches
Mail list logo