Indeed rather a "broad brush" when the problem was only with the last item, the "PI" record. Especially when SLIP was mentioned and I knew that SLIP's record does contain the high halves.
For a person as experienced as the OP, the data presented really was enough (albeit I guess not in the way he was thinking). Sure, it would be nice (even expected) that the PI entry show 64-bit GRs. Ask for it formally, please. No one has thought that important enough in 20 years to bother asking for. Such new support would make sense to me. Given that this was a PIC 38 (and would also have been true for PICs 39, 3A, 3B), "VPH" and "VPA" form the 8-byte translation exception address. VPH..... 7FFFF000 VPA..... 7FFFF800 So for this specific PIC, we have the offending address (the high half in VPH, the low half without bits 52-63 in VPA, the full low half in the slot of whichever register it turned out to be, in this case reg 15). Peter Relson z/OS Core Technology Design ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
