alu_dec.h File Reference

GE-130 packed/signed decimal ALU helpers. More...

#include <stdint.h>
#include "ge.h"
#include "alu_cc.h"

Go to the source code of this file.

Functions
void	alu_ap (struct ge *ge, uint16_t a, uint8_t alen, uint16_t b, uint8_t blen)
	AP 0xEA Add Packed: op1 = op1 + op2; CC set.
void	alu_sp (struct ge *ge, uint16_t a, uint8_t alen, uint16_t b, uint8_t blen)
	SP 0xEB Subtract Packed: op1 = op1 - op2; CC set.
void	alu_mp (struct ge *ge, uint16_t a, uint8_t alen, uint16_t b, uint8_t blen)
	MP 0xEC Multiply Packed: op1 = op1 * op2; CC set.
void	alu_dp (struct ge *ge, uint16_t a, uint8_t alen, uint16_t b, uint8_t blen)
	DP 0xED Divide Packed: op1[left L1-L2 chars] = quotient, op1[right L2+1 chars] = remainder; CC set.
void	alu_cmp (struct ge *ge, uint16_t a, uint8_t alen, uint16_t b, uint8_t blen)
	CMP 0xE9 Compare Packed (algebraic, no operand change); CC set.
void	alu_mvp (struct ge *ge, uint16_t a, uint8_t alen, uint16_t b, uint8_t blen)
	MVP 0xE8 Move Packed: op1 = op2 (sign preserved from op2); CC set.
void	alu_pk (struct ge *ge, uint16_t dst, uint8_t dlen, uint16_t src, uint8_t slen)
	PK 0xDA Pack: zoned op2 → packed op1 (no sign processing).
void	alu_upk (struct ge *ge, uint16_t dst, uint8_t dlen, uint16_t src, uint8_t slen)
	UPK 0xD8 Unpack: packed op2 → zoned op1 (no sign processing; zone of each result byte is taken from the pre-existing zone in ge->mem).
void	alu_pks (struct ge *ge, uint16_t dst, uint8_t dlen, uint16_t src, uint8_t slen)
	PKS 0xEE Pack with Sign: zoned op2 → packed op1; sign from zone of rightmost source byte (zone 0xA → negative, else positive).
void	alu_upks (struct ge *ge, uint16_t dst, uint8_t dlen, uint16_t src, uint8_t slen)
	UPKS 0xEF Unpack with Sign: packed op2 → zoned op1; zone always 0x4.
void	alu_edt (struct ge *ge, uint16_t pattern, uint8_t plen, uint16_t src)
	EDT 0xDE Edit packed source into pattern at op1.

Detailed Description

GE-130 packed/signed decimal ALU helpers.

Operand model (manual §2.3.5 / §5.3, verified lines ~3148-3169): PACKED format: two BCD digits per byte; the rightmost (highest-address) byte holds one digit in the high nibble and the SIGN in the low nibble. Addresses supplied to every function are the address of the RIGHTMOST (least-significant, highest-address) byte of the operand — matching the manual's "address of the rightmost position" rule. Length parameter L → field is L+1 bytes long; digits = 2L+1 (always odd).

Sign codes (manual ~line 3155): 0xB (1011) and 0xD (1101) → negative all others → positive Generated sign: 0xC (positive), 0xD (negative)

ZONED format (for PK/UPK/PKS/UPKS, §5.5.3.4-5.5.3.5, §5.6.2): One digit per byte; low nibble = digit, high nibble = zone (normally 0xF for EBCDIC-style, but 0x4 in GE notation; PKS interprets zone 0xA in the LAST/rightmost character as negative sign). UPKS always generates zone 0x4 on all result characters.

Condition code (§5.6.1.x, all packed arithmetic / compare): FA04 FA05 (= cc bit1 cc bit0) 0 0 overflow (or, for CMP: L1 < L2 overflow) 0 1 result < 0 1 0 result = 0 1 1 result > 0

For PKS / UPKS: 0 0 not possible 0 1 result < 0 1 0 result = 0 1 1 result > 0

EDT / PK / UPK: CC not used.

Definition in file alu_dec.h.

Function Documentation

alu_ap()

void alu_ap	(	struct ge *	ge,
		uint16_t	a,
		uint8_t	alen,
		uint16_t	b,
		uint8_t	blen
	)

AP 0xEA Add Packed: op1 = op1 + op2; CC set.

Definition at line 253 of file alu_dec.c.

alu_cmp()

void alu_cmp	(	struct ge *	ge,
		uint16_t	a,
		uint8_t	alen,
		uint16_t	b,
		uint8_t	blen
	)

CMP 0xE9 Compare Packed (algebraic, no operand change); CC set.

Overflow if L1 < L2.

Definition at line 559 of file alu_dec.c.

alu_dp()

void alu_dp	(	struct ge *	ge,
		uint16_t	a,
		uint8_t	alen,
		uint16_t	b,
		uint8_t	blen
	)

DP 0xED Divide Packed: op1[left L1-L2 chars] = quotient, op1[right L2+1 chars] = remainder; CC set.

Overflow if: L1 <= L2, L2 > 7, divisor == 0, or result doesn't fit. On overflow, operation is NOT performed.

Definition at line 459 of file alu_dec.c.

alu_edt()

void alu_edt	(	struct ge *	ge,
		uint16_t	pattern,
		uint8_t	plen,
		uint16_t	src
	)

EDT 0xDE Edit packed source into pattern at op1.

pattern: address of leftmost byte of pattern; plen = pattern length (bytes). src: address of RIGHTMOST byte of packed source field. Pattern control chars: 0x20 = SST (digit substitute + zero suppress), 0x21 = TSZ (digit substitute, ends zero suppress), 0x22 = RSZ (reset zero suppress start), other = insertion char (kept as-is). First byte of pattern is the fill character. CC: 1 0 = zero suppression still active (result = 0); 1 1 = active off. (manual §5.5.3.6 / §5.5.3.7 CC table; note "not possible" for 0/1.)

Definition at line 868 of file alu_dec.c.

alu_mp()

void alu_mp	(	struct ge *	ge,
		uint16_t	a,
		uint8_t	alen,
		uint16_t	b,
		uint8_t	blen
	)

MP 0xEC Multiply Packed: op1 = op1 * op2; CC set.

Constraints: blen <= 8, blen < alen; op1 must have ≥blen+1 leading zero digits; otherwise overflow (CC=0, operation not performed).

Definition at line 358 of file alu_dec.c.

alu_mvp()

void alu_mvp	(	struct ge *	ge,
		uint16_t	a,
		uint8_t	alen,
		uint16_t	b,
		uint8_t	blen
	)

MVP 0xE8 Move Packed: op1 = op2 (sign preserved from op2); CC set.

Definition at line 614 of file alu_dec.c.

alu_pk()

void alu_pk	(	struct ge *	ge,
		uint16_t	dst,
		uint8_t	dlen,
		uint16_t	src,
		uint8_t	slen
	)

PK 0xDA Pack: zoned op2 → packed op1 (no sign processing).

Definition at line 682 of file alu_dec.c.

alu_pks()

void alu_pks	(	struct ge *	ge,
		uint16_t	dst,
		uint8_t	dlen,
		uint16_t	src,
		uint8_t	slen
	)

PKS 0xEE Pack with Sign: zoned op2 → packed op1; sign from zone of rightmost source byte (zone 0xA → negative, else positive).

CC set.

Definition at line 761 of file alu_dec.c.

alu_sp()

void alu_sp	(	struct ge *	ge,
		uint16_t	a,
		uint8_t	alen,
		uint16_t	b,
		uint8_t	blen
	)

SP 0xEB Subtract Packed: op1 = op1 - op2; CC set.

Definition at line 324 of file alu_dec.c.

alu_upk()

void alu_upk	(	struct ge *	ge,
		uint16_t	dst,
		uint8_t	dlen,
		uint16_t	src,
		uint8_t	slen
	)

UPK 0xD8 Unpack: packed op2 → zoned op1 (no sign processing; zone of each result byte is taken from the pre-existing zone in ge->mem).

Definition at line 728 of file alu_dec.c.

alu_upks()

void alu_upks	(	struct ge *	ge,
		uint16_t	dst,
		uint8_t	dlen,
		uint16_t	src,
		uint8_t	slen
	)

UPKS 0xEF Unpack with Sign: packed op2 → zoned op1; zone always 0x4.

CC set (reflects sign/value of packed source).

Definition at line 804 of file alu_dec.c.