Handling *Entry parms from command (using MAX(8)) in a subproc

I have a new RPGLE application that gets its parameters from a command.  The
command is defined as:

  Parm       Members     Type( *char )  Len( 10 )  Max( 8 )
  Parm       MembrTypes  Type( *char )  Len( 10 )  Max( 8 )
  Parm       Libraries   Type( *char )  Len( 10 )  Max( 8 )
  Parm       OmitMTypes  Type( *char )  Len( 10 )  Max( 8 )

By virtue of the MAX (Maximum values allowed) parameter, set at eight, the
value issued to the CPP is a structure, starting with a 2-byte binary
number-of-values, followed by the values themselves.  The nature of the
beast is that it is a variable-length structure, and "unreliable results"
follow the last entry.

In the initial development of the RPG program, I was able to process the
*Entry parms in the mainline.  Essentially, this process takes the parameter
values and plants them in display file variables for display.  I defined a
data structure for each parm, i.e.:
   d pMbrs_ds     ds     82
   d  pMbrs_c             5i 0
   d  pMbrs_a            10a   Dim( 8 ) Overlay( pMbrs_ds : 3)

and processed as:
      For  x = 1 to pMbrs_c
        Sel_Mbrs( x ) = pMbrs_a( x )
      Endfor

(Sel_Mbrs is an array which overlays the eight display fields.)

In the main procedure, this works fine.  However...

I would like to move this to a subproc.  I have only included a snippet of
the logic, and there is more I need to do with these parameter values that I
would really REALLY like to encapsulate in a subproc.  Unfortunately, I
haven't found the right incantation for passing *Entry parms to a subproc.
I already tried putting the *Entry Plist in the subproc, but that isn't
allowed.

What's interesting is that a dump shows the subproc getting the same values
as when I had this in the main proc, but once I use the value, I get
garbage.  Here is a brief snippet of that subproc, preceded by the prototype
and the CallP in the main proc, with more comments below.  (Note: code is
"compressed" and not in proper columns.)

   d ProcessParms pr
   d   @Mbrs             82
   d   @MbrTypes         82
   d   @Libraries        82
   d   @OmitMTypes       82

   c             CallP     ProcessParms( pMbrs
   c                                   : pMbrTypes
   c                                   : pLibraries
   c                                   : pOmitMTypes )

   c             Dump(a)
   c             Eval      *inLR = *On

   c   *Entry    plist
   c             parm           pMbrs            82
   c             parm           pMbrTypes        82
   c             parm           pLibraries       82
   c             parm           pOmitMTypes      82

   p ProcessParms b
    *
   d ProcessParms pi
   d   pMbrs             82
   d   pMbrTypes         82
   d   pLibraries        82
   d   pOmitMTypes       82

   d pMbrs_ds     ds     82
   d  pMbrs_c             5i 0
   d  pMbrs_a            10a   Dim( 8 ) Overlay( pMbrs_ds : 3)


   d Sel_Mbrs_ds  ds     80    Inz
   d  Sel_Mbr1           10a   Overlay( Sel_Mbrs_ds : 1 )
   d  Sel_Mbr2           10a   Overlay( Sel_Mbrs_ds : *next )
   d  Sel_Mbr3           10a   Overlay( Sel_Mbrs_ds : *next )
   d  Sel_Mbr4           10a   Overlay( Sel_Mbrs_ds : *next )
   d  Sel_Mbr5           10a   Overlay( Sel_Mbrs_ds : *next )
   d  Sel_Mbr6           10a   Overlay( Sel_Mbrs_ds : *next )
   d  Sel_Mbr7           10a   Overlay( Sel_Mbrs_ds : *next )
   d  Sel_Mbr8           10a   Overlay( Sel_Mbrs_ds : *next )
   d Sel_Mbrs            10    Dim( 8 )  Overlay( Sel_Mbrs_ds : 1 )

   d x            s       5i 0

    /free
      pMbrs_ds       = pMbrs         ;
      pMbrTypes_ds   = pMbrTypes     ;
      pLibraries_ds  = pLibraries    ;
      pOmitMTypes_ds = pOmitMTypes   ;

      For  x = 1 to pMbrs_c          ;
        Sel_Mbrs( x ) = pMbrs_a( x ) ;
      Endfor                         ;
    /end-free

   p ProcessParms    e

In the DUMP output (called from the main proc after this subproc is called),
I am looking at the section titled "Local variables for subprocedure :
PROCESSPARMS".  The value for pMbrs is as I expected, but the value for
pMbrs_ds is garbage.

Is it an issue with *Entry parms?  Or command parameters of variable
lengths?  Or something I missed in TFM?

TIA,
Dan