A padded move of a shorter field, and a numbered indicator tested alone #643
Loading…
Reference in a new issue
No description provided.
Delete branch "rpg-9-ah-padded-move-20261026"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
A padded move of a shorter field, and a numbered indicator tested alone
An assignment of a shorter character field to a longer one (
aout = ain,char(10) and char(5)) compiles to a typed store of the value at byte 0 and
a blank fill of the rest: the padded store that the source emitter already
knows for literals. The statement reader wrote the store and the fill as
two lines, which had no rule, so the statement failed. When the fill
completes a proved padded store (
rpg_source_parts.padded) that covers thewhole field, the store line
F.0 = Vnow stands alone, as it does for aliteral, and the source emitter writes the assignment
F = V.A numbered indicator tested alone (
if *in03) compares the indicator withthe compiler's own '1'. The emitter turned
F = '1'into the testFforindicator fields and for the indicator array (
*in(N)) only, and wroteif *in03 = '1', which compiles to other instructions (its '1' is aliteral of the source). The rule now covers
*inNN.r7d-plain-in03is the public reproduction: its emit and rebuilt displayfile round-trip to a match except the file identity.
r7h-prt-oflemitsif *in90;as its source does.Gates
1026f9630181bef6d15b