NonStop S-Series Server Description Manual (G06.24+)

Instruction Processing Environments

HP NonStop S-Series Server Description Manual—520331-003

5-2

Native and TNS Programs and Processes

The RISC processor that is the instruction processing unit (IPU) of each NonStop

S-series processor executes only RISC instructions. However, to accommodate the

large installed base of TNS programs, the processor has been designed to accept and

execute TNS programs in addition to programs that are compiled directly to RISC

instructions.

Applications that are compiled directly to RISC instructions are in the form of native

object code. Such code, when executed, becomes a native process, illustrated in

the leftmost sequence in Figure 5-1. The RISC processor reads RISC instructions

from memory (backed up by the object code file) and executes those instructions

directly, using RISC stack conventions. RISC stack conventions are described in

Section 7, Native Execution Mode.

Applications that have been compiled as TNS instructions require the TNS instructions

to be converted to RISC instructions. There are two methods of performing this

conversion, differing in how and when the instruction conversion is performed.

In the case of an accelerated program (middle sequence in Figure 5-1), the TNS

code is converted to RISC instructions by an Accelerator program prior to run time.

The result of this preprocessing is that accelerated code (translated sequences of

RISC instructions) exists in the same code file as the original TNS object code.

However, because it is possible that some TNS coding cannot be translated

unambiguously before run time, and because P-relative data is not accelerated, the

TNS code remains available and will be accessed in those circumstances.

In the case of a nonaccelerated TNS program (right sequence in Figure 5-1), the

process, when executing, causes resident millicode to read and interpret TNS

instructions from memory. Thus the process appears to be executing TNS instructions.

In both of the latter two cases, the process uses TNS stack conventions (described in

Section 6, TNS Execution Modes) and is therefore categorized as a TNS process.