bug#25860: 25.1; Double macro execution

[npostavs]

tags 25860 confirmed
quit

Tino Calancha <tino.calancha@gmail.com> writes:
> Thank you for the additional information.
> You are right.  This is a regression introduced in Emacs 25.1.
>
> I] Not observed with Emacs-24.5.
> II] I reproduce the issue with Emacs-25.1.
> *)  I see the issue with Emacs-25.1 even after loading kmacro.el
>     from Emacs-24.5.  That seems to point to changes in the
>     C source code; maybe some changes in src/macro.c?
>
> I am having some difficulties to do a bisect to track down
> changes in macro.c.  Maybe someone can do it much faster and find
> the commit causing this issue.

I traced the execution in gdb, the code removed in the following diff
seems to be the culprit:

--- i/src/keyboard.c
+++ w/src/keyboard.c
@@ -3029,14 +3029,6 @@ read_char (int commandflag, Lisp_Object map,
       Vunread_post_input_method_events
        = nconc2 (XCDR (tem), Vunread_post_input_method_events);
     }
-  /* When we consume events from the various unread-*-events lists, we
-     bypass the code that records input, so record these events now if
-     they were not recorded already.  */
-  if (!recorded)
-    {
-      record_char (c);
-      recorded = true;
-    }
 
  reread_first:

This was added in [1: 30a6b1f814].  Obviously just removing it will
reintroduce the bug it fixed, I'm not sure what the proper fix should
be.

1: 2015-08-10 21:15:03 +0300 30a6b1f81412044aa7dda5573b0142a0a03c4fd3
  Fix recording of events pushed onto unread-command-events



Slightly simpler reproduction: evaluate

    (defun insert-a ()
      (interactive)
      (execute-kbd-macro "a" 1))
    (define-key global-map [f12] 'insert-a)

then f3 f12 f4 C-x e