A static pointer with INZ(*NULL), and the cycle proofs under OPTION(*SRCSTMT) #636
Loading…
Reference in a new issue
No description provided.
Delete branch "rpg-9-z-pointer-inz-20261025"
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 static pointer with INZ(*NULL), and the cycle proofs under OPTION(*SRCSTMT)
A static pointer declared
inz(*null)compiles to a 16-byte fill with the byte0 in the static initialization, where the pointer fields of the compiler are a
null-pointer store. The declaration rule knew the zero fill only as the empty
value of a varying-length field, so the program failed at its first use of the
pointer ("the initial value has no rule"). The rule now writes
pointer inz(*null)for a 16-byte pointer filled with zero. Theinit-static-characterand
init-static-packedprobes are the public reproduction; their rebuiltobjects still differ, because the source folds
%len(tag)into the literalthat the object holds.
With
OPTION(*SRCSTMT)the compiler statements are 2100000001 and 2100000002.An origin rule marked both partial whenever they were recovered, because no
proof of the compiler effects existed then. The cycle start template and the
return at the end of the main procedure prove them now, as they prove the
statements 1000001 and 1000002 of a program without the option, so a proved
statement keeps its recovered status; an unproved one stays partial, and the
statements keep their compiler kind and their counts.
SC-ILE/DSPLYand thed4-srcstmt-dspfprobe are the public reproduction; theTSTRIP-STRIPPEDgolden text follows (two status lines).
Gates
8c98db2e6f5d925475b5