On Thu, Nov 21, 2024 at 12:22 PM Jan Kara <jack@xxxxxxx> wrote: > > I'm posting here the series I'm currently carrying in my tree [1]. The changes > from v8 Josef posted are not huge but big enough that I think it's worth a > repost. Unless somebody speaks up, the plan is to merge into fsnotify branch > after the merge window closes. > > [1] https://git.kernel.org/pub/scm/linux/kernel/git/jack/linux-fs.git/log/?h=fsnotify_hsm > > v8: > https://lore.kernel.org/all/cover.1731684329.git.josef@xxxxxxxxxxxxxx > v7: > https://lore.kernel.org/linux-fsdevel/cover.1731433903.git.josef@xxxxxxxxxxxxxx/ > v6: > https://lore.kernel.org/linux-fsdevel/cover.1731355931.git.josef@xxxxxxxxxxxxxx/ > v5: > https://lore.kernel.org/linux-fsdevel/cover.1725481503.git.josef@xxxxxxxxxxxxxx/ > v4: > https://lore.kernel.org/linux-fsdevel/cover.1723670362.git.josef@xxxxxxxxxxxxxx/ > v3: > https://lore.kernel.org/linux-fsdevel/cover.1723228772.git.josef@xxxxxxxxxxxxxx/ > v2: > https://lore.kernel.org/linux-fsdevel/cover.1723144881.git.josef@xxxxxxxxxxxxxx/ > v1: > https://lore.kernel.org/linux-fsdevel/cover.1721931241.git.josef@xxxxxxxxxxxxxx/ > > v8->v9: > - fix DAX fault handling for ext4 & xfs > - rework FMODE_ constants a bit to keep FMODE_NONOTIFY a single bit > - move file_set_fsnotify_mode() out of line as it's quite big > - fold fsnotify_file_object_watched() into the single caller I tested this with my new test cases on LTP branch fan_hsm. One test broke, I posted a suggested fix for patch 3/19. > - use explicit f_mode checks instead of fsnotify_file_has_pre_content_watches() > - fix compilation breakage with CONFIG_NOMMU > - fixed up some changelogs > Other than patch 3, all looks good to me. Thanks, Amir.
