recent changes to display code

recent changes to display code

[Glenn Morris]

Something in your 2007-11-04 change to the src/ directory of the
EMACS_22_BASE branch has fairly badly messed up Emacs for me. The
ChangeLog entry is also missing.

./src/emacs -Q
  -> mode-line and scroll-bar are "missing" (in the sense of being
  empty and just looking like the normal buffer background).

The scroll bar appears if I type something into scratch.

This is on GNU/Linux, configured with:

--with-x-toolkit=athena --without-toolkit-scroll-bars

Re: recent changes to display code

[Chong Yidong]

Glenn Morris <rgm@gnu.org> writes:

> Something in your 2007-11-04 change to the src/ directory of the
> EMACS_22_BASE branch has fairly badly messed up Emacs for me. The
> ChangeLog entry is also missing.
>
> ./src/emacs -Q
>   -> mode-line and scroll-bar are "missing" (in the sense of being
>   empty and just looking like the normal buffer background).
>
> The scroll bar appears if I type something into scratch.

I see the problem too.  What bug is this redisplay change supposed to
fix, and why is it in EMACS_22_BASE as opposed to the trunk?

Re: recent changes to display code

[Richard Stallman]

    I see the problem too.  What bug is this redisplay change supposed to
    fix, and why is it in EMACS_22_BASE as opposed to the trunk?

It is meant to fix the problem that Joe Wells reported with overlay
properties.  I put it in Emacs 22 because these bugs were reported for
Emacs 22.  But I agree we should move it into the trunk.  Would
you please do that?

I'm surprised that this change affected any behavior.  It was only
supposed to make the argument OVERLAY available to
face_for_overlay_string.  I will take another look to see what might
have caused such a problem.

I wonder if there is some hidden constraint on the order of fields in
`struct it'.

Re: recent changes to display code

[Chong Yidong]

Richard Stallman <rms@gnu.org> writes:

> I'm surprised that this change affected any behavior.  It was only
> supposed to make the argument OVERLAY available to
> face_for_overlay_string.  I will take another look to see what might
> have caused such a problem.
>
> I wonder if there is some hidden constraint on the order of fields in
> `struct it'.

No, the bug arises because of a rather strange (presumable) typo in
your definition of from_overlay in handle_face_prop:

      Lisp_Object from_overlay
      = (it->current.overlay_string_index >= 0
         ? it->string_overlays[it->current.overlay_string_index]
            : from_overlay);
             ^^^^^^^^^^^^

I've fixed this in EMACS_22_BASE.