Mainframe Calculator

PSW Interpretation Basics

The Program Status Word (PSW) is one of the first things to look at in a dump because it captures the current execution state of the processor and the address of the next instruction to be executed.

In practical debugging, the PSW helps answer a few immediate questions: where execution was headed, whether the task was waiting, whether certain interrupts were enabled, and what addressing mode or control state clues might matter before you even inspect the failing instruction.

What the PSW tells you first

The PSW reflects general system state, including whether the machine is waiting or processing, whether I/O interrupts are enabled, and the address of the next instruction to execute.

• The next instruction address tells you where control was about to go.
• The wait indication helps distinguish a running loop from a task that is simply parked.
• Interrupt-enable and control bits can explain why expected work is not progressing.
• Condition code and state clues help you interpret surrounding instructions and branching logic.

Key fields in practice

• Instruction address — usually the first field to inspect. It points to the code region that needs attention.
• Condition code — useful when the failing area involves compares, arithmetic, or branch-on-condition logic.
• Wait state bit — identifies whether the system or task is paused rather than actively failing.
• Interrupt masks — check these when external, I/O, or related work is not being serviced.
• Addressing and control bits — important when you suspect a mode mismatch, storage-protection issue, or unexpected control transition.

Using the instruction address

The fastest practical step is to take the PSW next-instruction address and map it to a module, CSECT, load module listing, or disassembly. If the PSW keeps showing the same address or a small repeating series of addresses across multiple snapshots, the program is likely looping.

In an abend the address tells you where to start reading. In a suspected loop, repeated PSW snapshots can reveal the repeating path before you fully decode register contents.

What to be careful about

The PSW shows processor state around the current execution point, but you still need registers, surrounding instructions, storage contents, and calling context to understand why control arrived there. It is easy to confuse the address of the next instruction with the instruction that actually caused trouble — depending on the failure type, the address shown may be one instruction past the problem.

Address Alignment Basics

Mainframe data structures generally require their starting address to be on a specific boundary. Misalignment can cause program checks or unexpected behavior, so recognizing boundary codes quickly is useful during dump reading.

The three boundaries

• Halfword (H) — address is divisible by 2. Required for halfword instructions and two-byte data fields.
• Fullword (F) — address is divisible by 4. Required for fullword data, most register-based instructions.
• Doubleword (D) — address is divisible by 8. Required for doubleword data, floating-point, and some control blocks.

Reading the alignment code in the calculator

Enter any hex address into the Mainframe Calculator. The small code column between Hex and Notes displays H, F, or D automatically, or leaves it blank for an odd address. This lets you verify alignment for any value you are working with without doing the mental arithmetic yourself.

Reading Hex Offsets in Dumps

Much of hands-on mainframe debugging involves computing where a field is relative to a known base address. The Mainframe Calculator is built specifically for this kind of work.

Typical workflow

• Enter a base address — a register value or a known control block start — in the BASE row.
• Enter a DSECT or listing displacement in one of the OFFSET rows.
• Set the offset sign to plus to add, or minus to subtract.
• Read the TOTAL row to see the resulting storage address in hex, decimal, and octal.
• Use the alignment code on the TOTAL row to confirm the result lands on the expected boundary.

If you need to compare several alternative offsets, enter each one in a different OFFSET row and toggle them on or off with the plus/minus button. The SaveArea lets you park an intermediate result so you can continue working without losing it.

EBCDIC Quick Reference

EBCDIC (Extended Binary Coded Decimal Interchange Code) is the character encoding used on IBM mainframes. Unlike ASCII, EBCDIC does not place digits and letters in a contiguous range, which matters when you are examining character data in a dump.

Ranges to know

• Digits 0–9 — hex F0 through F9.
• Uppercase A–I — hex C1 through C9.
• Uppercase J–R — hex D1 through D9.
• Uppercase S–Z — hex E2 through E9.
• Lowercase a–i — hex 81 through 89.
• Space — hex 40.
• Period — hex 4B.

Notice the gap between I (C9) and J (D1) and between R (D9) and S (E2) — these are not contiguous ranges the way ASCII A–Z is. A quick range check in a dump storage display helps you confirm whether bytes contain printable characters or binary data.