Failure of TCPIP machine under CMS 7, service level 111

[Melinda Varian]

I have recently brought TCP/IP up on a P/370 (a System/370 workstation)
running VM/ESA 370 Feature with CMS 7 at the 111 service level. I am
using exactly the same version of TCP/IP as on our 3090 running VM/ESA
1.0 ESA Feature with CMS 7 at the 104 service level (as well as on our
4381 running HPO 4.2 and CMS 5).

I was unable to get the TCPIP service machine up at that level of CMS.
After I moved a service level 104 CMS nucleus to the P/370, TCPIP ran
flawlessly, so I believe that the problem is some service applied to
CMS between levels 104 and 111.

Here are the symptoms:

PCCA3 initializing:
Device NEWELC:
Type: LCS, Status: Not started
Envelope queue size: 0
Address: 0660
DMSITP142T Specification exception occurred at D302C0 in routine TCPIP
during SPIE exit routine
DMSITP144T Operation exception occurred at C1D3F2 in routine TCPIP while
UFDBUSY = 02; re-IPL CMS
DMSABE2047I Dump started; please wait
DMSABE1297I Dump has been taken
DMKDSP450W CP entered; disabled wait PSW '00020000 40E6CB10'

My tracing shows a BALR to an area that is all zeros. Setting RELPAGE
off changes the symptoms significantly (but results ultimately in a
message about being out of free storage). I think the error is likely
to be in free storage management and is probably specific to OS simula-
tion. (Nothing else I've run on this system has had any problem.)

A search of IBMLink has turned up nothing, so I thought I would ask if
anybody on the list recognizes the problem.

Melinda Varian,
Princeton University

[Martha McConaghy]

Melinda,

I think you are right in your assesment. I'm currently running TCP/IP
under CMS 7 service level 114 and don't see your problem at all. I've
looked through the PUT memo for 9304 (114), but didn't see anything that
specifically mentions TCP/IP. If you want, I can send you a copy.

Martha