add-variable-watcher usage in lisp/frame.el

add-variable-watcher usage in lisp/frame.el

[Robert Pluim]

lisp/frame.el has:

(mapc (lambda (var)
        (add-variable-watcher var (symbol-function 'set-buffer-redisplay)))

which causes (describe-variable 'truncate-lines) to produce output like

  Automatically becomes permanently buffer-local when set.
  Calls these functions when changed: (#<subr set-buffer-redisplay>)

Note the "#<subr... there.

Not using symbol-function makes this output (set-buffer-redisplay)
instead, which I think looks better, Iʼm wondering why symbol-function
was used?

Thanks

Robert

Re: add-variable-watcher usage in lisp/frame.el

[Noam Postavsky]

On 1 June 2018 at 05:18, Robert Pluim <rpluim@gmail.com> wrote:
> lisp/frame.el has:
>
> (mapc (lambda (var)
>         (add-variable-watcher var (symbol-function 'set-buffer-redisplay)))

> Not using symbol-function makes this output (set-buffer-redisplay)
> instead, which I think looks better, Iʼm wondering why symbol-function
> was used?

Passing a subr object tells the variable watcher mechanism to call the
C function directly (without a possible GC call), which is closer to
what happened in Emacs 25 (see diff of [1: d3faef9bae]). I don't know
that it's really needed in this case though.

[1: d3faef9bae]: 2016-12-02 20:25:15 -0500
  Ensure redisplay using variable watcher
  https://git.savannah.gnu.org/cgit/emacs.git/commit/?id=d3faef9baedadc9eaec46814ba9bbe5168048328

Re: add-variable-watcher usage in lisp/frame.el

[Robert Pluim]

Noam Postavsky <npostavs@gmail.com> writes:

> On 1 June 2018 at 05:18, Robert Pluim <rpluim@gmail.com> wrote:
>> lisp/frame.el has:
>>
>> (mapc (lambda (var)
>>         (add-variable-watcher var (symbol-function 'set-buffer-redisplay)))
>
>> Not using symbol-function makes this output (set-buffer-redisplay)
>> instead, which I think looks better, Iʼm wondering why symbol-function
>> was used?
>
> Passing a subr object tells the variable watcher mechanism to call the
> C function directly (without a possible GC call), which is closer to
> what happened in Emacs 25 (see diff of [1: d3faef9bae]). I don't know
> that it's really needed in this case though.

Thanks for the explanation. By all means lets avoid adding potential GC to code
thatʼs not expecting it.

Robert