diff options
| author | Eli Zaretskii | 2023-07-23 13:44:40 -0400 |
|---|---|---|
| committer | Eli Zaretskii | 2023-07-23 13:44:40 -0400 |
| commit | bbd91a05fa83b7b2a651d5386f4592db405bbcda (patch) | |
| tree | 9ee552f4ce2303c980ffe7ecc77201847edb17a5 /java/org | |
| parent | 865817633f6e4b548c9d138fdf792394d021e2f5 (diff) | |
| download | emacs-bbd91a05fa83b7b2a651d5386f4592db405bbcda.tar.gz emacs-bbd91a05fa83b7b2a651d5386f4592db405bbcda.zip | |
Fix rare aborts in CHECK_WINDOW_END
Those aborts happen because a window's window_end_vpos value
is inconsistent with the current matrix's number of rows, which
happens after resizing the mini-window.
* src/xdisp.c (mark_window_display_accurate_1): Don't validate the
window_end_valid flag if window_end_vpos is inconsistent with the
current_matrix's number of rows. This happens, e.g., when
Edebug wants to show a value in the mini-window that causes it to
resize, as the window above the mini-window was usually already
redisplayed, and we think its display is accurate.
Diffstat (limited to 'java/org')
0 files changed, 0 insertions, 0 deletions