Neither bug nor necessarily a mistake. Just a little incomplete. What the
book means is that the full 13 bits of the return address is included in the
stack, so your subroutine need not know from whence it was called. However,
once the subroutine has returned PCLATH is still set the way it was to
access the subroutine so if you do further work which is PCLATH dependent
you must set it to the current page after the subroutine is called.
The verbage in the data book which addresses this is: 'PCLATH is not
affected by a PUSH or POP operation.' (section 4.3.2 in my 1996 DS30390D
document, p. 41)

Brian
-----Original Message-----
From: Bob Smith x31105 <rsmith@KODAK.COM>
To: PICLIST@MITVMA.MIT.EDU <PICLIST@MITVMA.MIT.EDU>
Date: Sunday, March 08, 1998 10:37 PM
Subject: PCLATH and P16C74A ??


>Hello All,
>
>Writing my first code for a PC16C74A - the data book indicates that
>I need to take care of PCLATH<3> when using CALL or GOTO beyond the
>2k boundary, and further states that I don't need to worry about it
>on RETURN...  I've found that to be incorrect, at least when running
>MPSIM...  is this an MPSIM bug?  or a mistake in the data book?
>
>Thanks!
>
>Bob