summaryrefslogtreecommitdiff
path: root/lisp/gnus/gnus-kill.el
diff options
context:
space:
mode:
authorTassilo Horn <tsdh@gnu.org>2021-08-08 16:45:50 +0200
committerTassilo Horn <tsdh@gnu.org>2021-08-08 21:06:17 +0200
commit61677ac3e4685d8f81c3b90eb751d9b5e8a3732d (patch)
tree77141521ecc1fce031ec4513c6241eb72a66ec2a /lisp/gnus/gnus-kill.el
parent051434fdefd6418bf1f0cd28c087b31cb3921f48 (diff)
downloademacs-61677ac3e4685d8f81c3b90eb751d9b5e8a3732d.tar.gz
emacs-61677ac3e4685d8f81c3b90eb751d9b5e8a3732d.tar.bz2
emacs-61677ac3e4685d8f81c3b90eb751d9b5e8a3732d.zip
Un-obsolete dired-in-this-tree-p and use it again in certain places.
This reverts parts of b425966b07, and 7b50ed553f, i.e. it reverts the obsoletion of dired-in-this-tree-p and switches some new callers of file-in-directory-p back to using dired-in-this-tree-p. It turned out that using file-in-directory-p can be a major performance regression in case you have a dired buffer of a remote directory which has become (almost) inaccessible. Any attempt to open a new dired buffer is checking if a buffer for that directory already exists (in terms of dired-buffers-for-dir) which meant that file-in-directory-p was called with the directory of any existing dired buffer including the inaccessible one where the file-truename call in file-in-directory-p could block for seconds or even minutes. * lisp/dired.el (dired-in-this-tree-p): Undo obsoletion. (dired-buffers-for-dir): Use dired-in-this-tree-p as before. * lisp/dired-aux.el (dired-kill-tree,dired-tree-down): Une dired-in-this-tree-p as before.
Diffstat (limited to 'lisp/gnus/gnus-kill.el')
0 files changed, 0 insertions, 0 deletions