Problem installing MVS 3.8: Machine check due to host error

[mark_davidson]

If there's an answer to this question posted elsewhere, I'd love to be pointed in that direction. :)

I'm trying to install MVS 3.8 on my G4 PowerBook using Hercules 3.06, using the instructions and images from Jay Moseley's web site.  

I am able to run the mvsdasd script to create the DASD files, and I have un-tar'd the archives from Jay's site in a manner which (I believe) matches the layout he describes.

However, when I run Hercules and attempt to run the ibcdadsi step, I run into problems.  At some point during the job, I get this on the Hercules console:

HHCTT001W Timer thread set priority -20 failed: Permission denied                              
HHCTT002I Timer thread started: tid=F0407000, pid=540, priority=15                              
HHCPN001I Control panel thread started: tid=A0C45820, pid=540                                  
HHCAO001I Hercules Automatic Operator thread started;                                          
          tid=F070D000, pri=16, pid=540                                                        
ipl 0280                                                                                        
HHCTE009I Client 127.0.0.1 connected to 1052 device 0:0009                                      
HHCCP017I CPU0000: Machine check due to host error: User defined signal 1                      
PSW=01040000 80000B44 INST=47800B54     BC    8,2900(0,0)            branch_on_condition        
R:00000B54:K:06=9D004000 47200B54 47F00C04 980F0220  .. ......0..q...                          
GR00=00000000  GR01=0000000A  GR02=00000360  GR03=00004570                                      
GR04=00000149  GR05=0000E720  GR06=00000000  GR07=01DC0005                                      
GR08=0000002D  GR09=80003602  GR10=00001000  GR11=00002000                                      
GR12=00003000  GR13=00004000  GR14=9000289A  GR15=A00035C2                                      
HHCCP011I CPU0000: Disabled wait state                                                          
          PSW=00020000 800000E2                      

I've tried the mvsdasd/IBCDASDI steps several times, and this happens at different times.  The telnet connection window doesn't show any errors:

Hercules version 3.06 built on Jan  9 2009 20:19:50
running on mark-davidsons-powe (Darwin-9.8.0.Darwin Kernel Version 9.8.0: Wed Jul 15 16:57:01  Power Macintosh UP)
Connected to device 0:0009

IBC105A  DEFINE INPUT DEVICE. DASDI  7.80
HHCTE006A Enter input for console device 0009
input=1442,00c
                    DASDI  7.80
SMP001 JOB 'INITIALIZE SMP001 3350 VOLUME'
       MSG TODEV=1052,TOADDR=009
       DADEF TODEV=3350,TOADDR=148,IPL=NO,VOLID=SMP001,BYPASS=YES
       VLD NEWVOLID=SMP001,OWNERID=HERCULES
       VTOCD STRTADR=1,EXTENT=15
       END
IBC163A  END OF JOB.
                    DASDI  7.80
WORK02 JOB 'INITIALIZE WORK02 3350 VOLUME'
       MSG TODEV=1052,TOADDR=009
       DADEF TODEV=3350,TOADDR=149,IPL=NO,VOLID=WORK02,BYPASS=YES
       VLD NEWVOLID=WORK02,OWNERID=HERCULES
       VTOCD STRTADR=1,EXTENT=15
       END

If anyone can tell me what's going on, it would be much appreciated.

Thanks.

Mark Davidson

[kerravon86]

--- In H390-MVS@..., "mark_davidson" <mdavidson1963@...> wrote:
>
> If there's an answer to this question posted elsewhere,
> I'd love to be pointed in that direction. :)

I don't know the answer to your question, but the
lack of answers from anyone else highlights the
fact that you're using an older package that isn't
widely used anymore. Most people use TK3 or a
derivative of that, and most common problems are
already known and have answers available.

BFN.  Paul.

[mark_davidson]

--- In H390-MVS@..., "kerravon86" <kerravon86@...> wrote:

>
> --- In H390-MVS@..., "mark_davidson" <mdavidson1963@> wrote:
> >
> > If there's an answer to this question posted elsewhere,
> > I'd love to be pointed in that direction. :)
>
> I don't know the answer to your question, but the
> lack of answers from anyone else highlights the
> fact that you're using an older package that isn't
> widely used anymore. Most people use TK3 or a
> derivative of that, and most common problems are
> already known and have answers available.
>
> BFN.  Paul.
>

Paul---

Thanks for the reply!  I shall attempt to move to TK3 if that is what other people are using.  I tried this same process on an Intel-based Mac and had no problems, so I thought it might be prudent to point this out to someone.

Thanks again.

Mark

[kerravon86]

--- In H390-MVS@..., "mark_davidson" <mdavidson1963@...> wrote:

>
> > I don't know the answer to your question, but the
> > lack of answers from anyone else highlights the
> > fact that you're using an older package that isn't
> > widely used anymore. Most people use TK3 or a
> > derivative of that, and most common problems are
> > already known and have answers available.
>
> Thanks for the reply!  I shall attempt to move to TK3
> if that is what other people are using.  I tried this
> same process on an Intel-based Mac and had no problems,
> so I thought it might be prudent to point this out to someone.

If you want someone to point things out to, with a view
to get a product improvement, your best bet is to use
the tk3qstart program in the files area of:

http://tech.groups.yahoo.com/group/hercules-os380/

and let it do the default install of MVS/380.

I just reported some things I would like to see
improved in the installation, so there might be a
new version soon (the author is responsive to
suggestions, and appreciates any feedback he can
get).

The current version is fine as-is though. Just
some minor things I noticed during the install.

BFN.  Paul.

[halfmeg]

> mark_davidson wrote:

><snip>

>  I tried this same process on an Intel-based Mac and had no problems,
> so I thought it might be prudent to point this out to someone.

Your encounter of a problem on native Mac versus an Intel-based Mac would more likely be a Hercules problem, not MVS or other guest OS.

Mac is Big-Endian while Intel-base Mac is Little-Endian.  That may not be the problem however as Hercules should be Endian independant.

The earlier message in your post:

"HHCTT001W Timer thread set priority -20 failed: Permission denied"

may have more bearing on a problem which occurs at "different times" and might have more feedback on the hercules forum than here on the MVS one.

Phil

[Jay Maynard-2]

On Mon, Aug 24, 2009 at 02:14:51AM -0000, halfmeg wrote:
> Your encounter of a problem on native Mac versus an Intel-based Mac would
> more likely be a Hercules problem, not MVS or other guest OS.

Nitpick: All Macs are native, both Intel and PowerPC.

> Mac is Big-Endian while Intel-base Mac is Little-Endian.  That may not be
> the problem however as Hercules should be Endian independant.

It is. It goes to some great lengths to handle either host endianness
properly.

> The earlier message in your post:
> "HHCTT001W Timer thread set priority -20 failed: Permission denied"
> may have more bearing on a problem which occurs at "different times" and
> might have more feedback on the hercules forum than here on the MVS one.

This is a standard message on OS X, be it Intel or PowerPC. The mechanism
Hercules uses to set different thread priorities on Linux doesn't work on OS
X. I've had a to-do for years to fix that.
--
Jay Maynard, K5ZC                   http://www.conmicro.com
http://jmaynard.livejournal.com      http://www.tronguy.net
http://www.hercules-390.org               (Yes, that's me!)
Buy Hercules stuff at http://www.cafepress.com/hercules-390

[mark_davidson]

On Sun, Aug 23, 2009 at 7:14 PM, halfmeg<opplr@...> wrote:

>
>
>> mark_davidson wrote:
>
>><snip>
>
>> I tried this same process on an Intel-based Mac and had no problems,
>> so I thought it might be prudent to point this out to someone.
>
> Your encounter of a problem on native Mac versus an Intel-based Mac would
> more likely be a Hercules problem, not MVS or other guest OS.

Oh, I completely agree (and certainly didn't mean to imply this was a
problem with MVS, especially since I've had better luck on my
Intel-based Mac).

> Mac is Big-Endian while Intel-base Mac is Little-Endian. That may not be the
> problem however as Hercules should be Endian independant.
>
> The earlier message in your post:
>
> "HHCTT001W Timer thread set priority -20 failed: Permission denied"
>
> may have more bearing on a problem which occurs at "different times" and
> might have more feedback on the hercules forum than here on the MVS one.

Jay Moseley and I have been discussing this off-list, and he has been
very helpful.  I will probably try to send this information to someone
who works on the Hercules core to see if they have any insights.

Mark

[halfmeg]

> Jay Maynard wrote:

> > halfmeg wrote:
> > Your encounter of a problem on native Mac versus an Intel-based
> > Mac would more likely be a Hercules problem, not MVS or other
> > guest OS.

> Nitpick: All Macs are native, both Intel and PowerPC.
 
><snip>

Last "MAC" I had was a LISA, so don't know much about them at all.

I was thinking more along the lines of something similar to when Hercules on a "IBM PC" had to have AMD instead of i586 specified ( or something similar ) during compilation or the executable would include non-existent instructions for the CPU in the binaries and promptly failed when it attempted to execute one.

Intel and PowerPC based MACs may be able to run binaries compiled on the other, regardless of compile options, without problems.

Phil

[au1john]

<snip>
Last "MAC" I had was a LISA, so don't know much about them at all.

My memory says that Lisa was pre-Macintosh, both of which used a Motorola
68000 processor. Gee, we are almost back in the MVS 3.8 days :) (1983)

[Jay Maynard-2]

On Mon, Aug 24, 2009 at 01:52:33PM -0000, halfmeg wrote:
> I was thinking more along the lines of something similar to when Hercules
> on a "IBM PC" had to have AMD instead of i586 specified ( or something
> similar ) during compilation or the executable would include non-existent
> instructions for the CPU in the binaries and promptly failed when it
> attempted to execute one.

Those days are pretty much gone...

> Intel and PowerPC based MACs may be able to run binaries compiled on the
> other, regardless of compile options, without problems.

Intel-based Macs can run PowerPC OS X binaries through a technology called
Rosetta, which is a PowerPC emulator. PowerPC-based Macs can't run Intel OS
X binaries at all.

Mac OS X since the introduction of Intel support in 10.4 has been delivered
in two flavors, and since 10.5 has been delivered with both 32- and 64-bit
versions of the programs it contains.

Mac binaries, these days, are typically delivered as a universal binary
image, which contains both the Intel and PowerPC versions in one bundle. OS
X will run the correct binary out of the file. The development software can
cross-compile, not only between Intel and PPC, but also between 32- and
64-bit versions (thus, a universal binary bundle can contain 4 separate
images, if it's desired to wring every alst drop of performance out of a
64-bit host; Hercules is delivered this way). A 32-bit PPC or Intel Mac can
run 32-bit binaries; a 64-bit PPC or Intel Mac can run 64-bit or 32-bit
binaries with no problems.
--
Jay Maynard, K5ZC                   http://www.conmicro.com
http://jmaynard.livejournal.com      http://www.tronguy.net
http://www.hercules-390.org               (Yes, that's me!)
Buy Hercules stuff at http://www.cafepress.com/hercules-390

[Mike Schwab]

http://www.macgeek.org/museum/applelisa1/

On Mon, Aug 24, 2009 at 9:22 AM, au1john<au1john@...> wrote:
> <snip>
> Last "MAC" I had was a LISA, so don't know much about them at all.
>
> My memory says that Lisa was pre-Macintosh, both of which used a Motorola
> 68000 processor. Gee, we are almost back in the MVS 3.8 days :) (1983)
--
Mike A Schwab, Springfield IL USA
Where do Forest Rangers go to get away from it all?

[King Khan-2]

Hello Group,

Today we got the alert as " ARC0909E MCDS CONTROL DATA SET IS ABOUT 090% FULL  " in mainframe. I understand that the MCDS volume is about to reach its threshold value. Can anyone advice me the procedure to monitor its volume via SDSF.

Your advice in this issue would be very useful to me.


      Get your new Email address!
Grab the Email name you've always wanted before someone else does!
http://mail.promotions.yahoo.com/newdomains/aa/

[chbirr01]

--- In H390-MVS@..., King Khan <aquarius_big@...> wrote:
>
>
> Hello Group,
>
> Today we got the alert as " ARC0909E MCDS CONTROL DATA SET IS ABOUT 090% FULL  " in mainframe. I understand that the MCDS volume is about to reach its threshold value. Can anyone advice me the procedure to monitor its volume via SDSF.
>
> Your advice in this issue would be very useful to me.
>
>
Hello,
 
F your.hsm.procname,BACKVOL CDS

will be your friend

Christian

[Enrico Sorichetti-2]

--- In H390-MVS@..., King Khan <aquarius_big@...> wrote:

>
>
> Hello Group,
>
> Today we got the alert as " ARC0909E MCDS CONTROL DATA SET IS ABOUT 090% FULL  " in mainframe. I understand that the MCDS volume is about to reach its threshold value. Can anyone advice me the procedure to monitor its volume via SDSF.
>
> Your advice in this issue would be very useful to me.
>
>
>       Get your new Email address!
> Grab the Email name you've always wanted before someone else does!
> http://mail.promotions.yahoo.com/newdomains/aa/
>

what about asking Your storage support group ??
anyway, SDSF is not an HSM related tool

[King Khan-2]

Hello Group,

Thanks for your answer,. Well in which Parmlib we can find the MCDS, BCDS information.




________________________________
From: chbirr01 <TSAdmin@...>
To: H390-MVS@...
Sent: Monday, August 24, 2009 22:13:40
Subject: [H390-MVS] Re: ARC0909E MCDS CONTROL DATA SET IS ABOUT 090% FULL

 
--- In H390-MVS@yahoogroup s.com, King Khan <aquarius_big@ ...> wrote:
>
>
> Hello Group,
>
> Today we got the alert as " ARC0909E MCDS CONTROL DATA SET IS ABOUT 090% FULL  " in mainframe. I understand that the MCDS volume is about to reach its threshold value. Can anyone advice me the procedure to monitor its volume via SDSF.
>
> Your advice in this issue would be very useful to me.
>
>
Hello,

F your.hsm.procname, BACKVOL CDS

will be your friend

Christian





      New Email names for you!
Get the Email name you've always wanted on the new @ymail and @rocketmail.
Hurry before someone else does!
http://mail.promotions.yahoo.com/newdomains/aa/