On Sun, Jun 1, 2025 at 1:27 AM Eli Zaretskii <eliz@gnu.org> wrote:

> Cc: ben@bensimms.moe
> From: Ship Mints <shipmints@gmail.com>
> Date: Sat, 31 May 2025 17:46:11 -0400
>
> I'm thinking it might have been this (I cc'd Ben) commit 9cbbdcee132588a11dc03c3da371176aaa13927c it's
> the only one that looks relevant.  That said, it is possible I'm tickling a bug that's been there longer.
>
> I pulled a master today and recompiled from clean and it still happens.  It's a bit spurious, but regular over
> the last few days.  There are no stipples used in the buffer in question, so this looks like a side effect of that
> change, I'm guessing.  There are specified spaces in the body of the buffer and also in the header-line that
> are "stretch" glyphs.
>
> This comes up as I go through a battery of tests for the revised vtable that has image content in a test jig
> and with multiple vtables sharing the same buffer (I doubt that has any bearing).
>
> I will try to see if I can reproduce without images if that indicates anything. 
>
> $ uname -a
>
> Darwin host.local 21.6.0 Darwin Kernel Version 21.6.0: Mon Jun 24 00:56:10 PDT 2024;
> root:xnu-8020.240.18.709.2~1/RELEASE_X86_64 x86_64
>
> * thread #1, queue = 'com.apple.main-thread', stop reason = EXC_BAD_ACCESS (code=1, address=0xa8)
>   * frame #0: 0x00000001000e79bb emacs`prepare_face_for_display(f=0x0000000102f41b70,
> face=0x0000000000000000) at xfaces.c:4657:16 [opt]
>     frame #1: 0x000000010025d599 emacs`ns_draw_stretch_glyph_string(s=0x00007ff7bfef9eb0) at
> nsterm.m:4176:8 [opt]
>     frame #2: 0x000000010025a532 emacs`ns_draw_glyph_string(s=0x00007ff7bfef9eb0) at
> nsterm.m:4571:7 [opt]

This code:

          if (s->row->mouse_face_p
              && cursor_in_mouse_face_p (s->w))
            {
              face = FACE_FROM_ID_OR_NULL (s->f,
                                           MOUSE_HL_INFO (s->f)->mouse_face_face_id);

              if (!s->face)
                face = FACE_FROM_ID (s->f, MOUSE_FACE_ID);
              prepare_face_for_display (s->f, face);

seems to have a thinko: it should say this instead:

          if (s->row->mouse_face_p
              && cursor_in_mouse_face_p (s->w))
            {
              face = FACE_FROM_ID_OR_NULL (s->f,
                                           MOUSE_HL_INFO (s->f)->mouse_face_face_id);

              if (!face)  <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
                face = FACE_FROM_ID (s->f, MOUSE_FACE_ID);
              prepare_face_for_display (s->f, face);

Can you try that change?

Making a normal build now.  Will test today/tomorrow.