Re: RSLVSP to QTEMP: why new MCH3601 on V5R2?

Although the following doesn't solve my original problem of resolving a
pointer to QTEMP, you might be interested to know that in V5R2 the RSLVSP's
extended 128-byte operand 2 template returns a system pointer to QTEMP in
positions 49-64, if the object named in operand 2 is found in QTEMP.

For example, create a data area named, TEST, in library QTEMP, then test
this little program, and at the breakpoint, system pointer "?CONTEXT" will
point to QTEMP.   Although this doesn't help me resolve to QTEMP from a
QTEMP name, it's interesting that here is at least instruction that
indirectly coughs up a QTEMP pointer---into the operand 2, but not into the
operand 1.

DCL SYSPTR ?OBJECT   AUTO;
DCL DD     NAME      AUTO CHAR(128) BDRY(16);
DCL SYSPTR ?CONTEXT  DEF(NAME) POS(49);

CPYBLAP  NAME, X"190A", X'00';
CPYBLAP  NAME(3:30), "NUTS", " ";
OR(S)    NAME(34:1), X"01";
RSLVSP   ?OBJECT, NAME, *, *;
BRK      "LOOK";
PEND;