bug#69083: Emacs's keyboard hook state is not reset on session lock (Windows)

[Raffael Stocker]

Eli Zaretskii <eliz@gnu.org> writes:

>> Is there a better place where the remove_w32_kbdhook call could go such
>> that cleanup can always happen?
>
> I think that place is term_ntproc.

I have added a call to ‘remove_w32_kbdhook’ there, but also left it with
the WM_DESTROY message.  The cleanup now always happens and the
setup/remove calls are correctly nested when using multiple frames.

>> >> Unfortunately, this didn't work for me.  I tried calling
>> >> ‘EnumWindows(find_child_console, ...)’ with a ‘child_process’ instance
>> >> containing the current process id as returned by ‘GetCurrentProcessId’,
>> >> but I don't seem to get a useful window handle.
>> >
>> > What do you mean?  What is the result of using find_child_console?
>> > does the condition in find_child_console, which looks at the
>> > process_id of all windows, never match the process ID of the Emacs
>> > session running with -nw?  Or what else goes wrong?

I figured this out now.  The ‘find_child_console’ function looks for a
‘ConsoleWindowClass’ window, but the Emacs process itself only has
‘Emacs Clipboard’ and ‘IME’ windows.  The latter seems to be the one I
need.  I added a function ‘find_ime_window’ in ‘w32console.c’ that
checks for this window when called through ‘EnumThreadWindows’.  I can
now register for session notifications using this window.

To handle the notifications, I added some code to ‘drain_message_queue’,
calling ‘reset_w32_kbdhook_state’ from there.  This now correctly resets
the hook state in console Emacs.

I added ‘WINDOWSNT’ #ifdefs around calls to and the definition of
‘reset_w32_kbdhook_state’, as with the setup/remove functions.  I also
cleaned up the ‘remove_w32_kbdhook’ function to use the previously
obtained function pointer.

I believe the attached version should now tick all the important boxes.

Regards,
Raffael

reset_kbdhook_2.patch
Description: reset_kbdhook_state patch