From fb086fe264291111d3171e9acde475fabaacb550 Mon Sep 17 00:00:00 2001 From: Oskar Roesler Date: Wed, 9 Jul 2025 17:54:58 +0200 Subject: [PATCH] gh-81520: Document unexpected `os.path.ismount` behaviour with btrfs subvolumes (GH-136058) (cherry picked from commit 591abcc01fcf1c65c7fdfaca7274f5d3f9f022da) Co-authored-by: Oskar Roesler --- Doc/library/os.path.rst | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/Doc/library/os.path.rst b/Doc/library/os.path.rst index cd2c5dfbb3f122..32a2970d2d3a2c 100644 --- a/Doc/library/os.path.rst +++ b/Doc/library/os.path.rst @@ -298,9 +298,10 @@ the :mod:`glob` module.) device than *path*, or whether :file:`{path}/..` and *path* point to the same i-node on the same device --- this should detect mount points for all Unix and POSIX variants. It is not able to reliably detect bind mounts on the - same filesystem. On Windows, a drive letter root and a share UNC are - always mount points, and for any other path ``GetVolumePathName`` is called - to see if it is different from the input path. + same filesystem. On Linux systems, it will always return ``True`` for btrfs + subvolumes, even if they aren't mount points. On Windows, a drive letter root + and a share UNC are always mount points, and for any other path + ``GetVolumePathName`` is called to see if it is different from the input path. .. versionchanged:: 3.4 Added support for detecting non-root mount points on Windows.