RE: IEA304W SYSTEM WAIT STATE CODE 80009064 DURING IEAVNP10 INITIALIZATION
| Newsgroups | gmane.comp.emulators.hercules390.general |
|---|---|
| Message-ID | <[email protected]> |
rixhawg wrote: > I am not sure why, but on my older PC it worked fine > with no limit on the number of zIIPS per GP. However, > on my newer PC, it appears that z/OS is adhering to > IBM's marketing decision to only allow one zIIP engine > per two GP processors. Interesting! (*) The only possible explanation is IBM's code behaves differently depending on Hercules calculated "Capability" values returned in the SYSIB 1.1.1 (Basic-machine configuration) and/or SYSIB 1.2.2 (Basic-machine All CPUs) system information blocks returned by the B27D STSI (Store System Information) instruction (which are calculated based on the HOST'S real CPU speed; see the "MIPSreal()", "stsi_capability()" and "DEF_INST(store_system_information)" functions in source member control.c). Since your newer PC is presumably faster than your older PC, it makes sense that Hercules's calculated "Capability" values would be different, and it is likely this difference in the "Capability" of each system that is causing IBM's code to behave differently. > I now have my system configured for 5 GP engines > and 3 zIIP engines and all appears to be working fine > running it under Hercules 4.1. Whatever works! :) I'm just happy you figured it out. > I want to thank all of you for your suggestions and assistance. > This issue was driving me crazy! Me too! :) But I'm satisfied with classifying this as not being a bug in Hercules nor even in z/OS, but rather just an unfortunate side effect of running certain guest operating systems (e.g. z/OS) under an emulator instead of on real iron. -- "Fish" (David B. Trout) Software Development Laboratories http://www.softdevlabs.com mail: [email protected] (*) https://www.youtube.com/watch?v=LEudTOJQ8xo