topic Processor type vs kernel in Operating System - HP-UX

Processor type vs kernel

Jun Zhang_4 — Wed, 01 Oct 2003 09:22:03 GMT

At my D370, the command
grep $(model | sed 's;.*/;;') /usr/sam/lib/mo/sched.models
tells me
D370 2.0 PA8000
and the command
file /stand/vmunix
tells me
/stand/vmunix: PA-RISC1.1 executable -not stripped
The PA-RISC versions are different.
If I do the same two commands on my A400, the PA-RISC versions agree with each other.
What the problem with the D370?

Jun

Re: Processor type vs kernel

Stefan Stechemesser — Wed, 01 Oct 2003 09:35:14 GMT

sounds like you have a 32 bit kernel on a 64 bit machine. This is a good idea if you have EISA cards on the D370, because there are no 64 bit drivers for EISA cards. D370 can run both PA-RISC 1.1 and 2.0 code.

Re: Processor type vs kernel

Jeff Schussele — Wed, 01 Oct 2003 09:45:43 GMT

Hi Jun,

Nothing wrong with running 32-bit OS on a D-class as it's a 32/64 capable machine.
The PA8000 was the first PA-RISC2.0 capable processor.
The processor in the A-class is also a PA-RISC2.0 CPU, but that system cannot run a 32-bit kernel. It's strictly a 64-bit system.
AFAIK only the D & K-class systems could run either kernels. Everything before, only 32-bit. Everything after, only 64-bit.

Rgds,
Jeff

Re: Processor type vs kernel

Jun Zhang_4 — Wed, 01 Oct 2003 11:47:53 GMT

I currently run an application built for PA-RISC 2 on this D370, and it runs well but have noticeable performance issues. The application is called jBASE, an 32-bit application, but claimed by the software builder for PA-RISC 2.
In your context I feel the association of 32-bit with PA-RISC 1, and 64-bit with PA-RISC 2, my question will be,
How to understand a 32-bit app built for pa2?
Is the combination of kernel/cpu type/bitness to be blamed for the poor performace of jBASE?

Jun

Re: Processor type vs kernel

Joshua Scott — Wed, 01 Oct 2003 11:52:55 GMT

Jeff, to add to your list, the R-class (essentially the same as the D's) and the T600 could also run either. :)