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 - use explicit f_mode checks instead of fsnotify_file_has_pre_content_watches() - fix compilation breakage with CONFIG_NOMMU - fixed up some changelogs v7->v8: - A bunch of work from Amir to cleanup the fast path for the common case of no watches, which cascades through the rest of th series to update the helpers and the hooks to use the new helpers. - A patch from Al to get rid of the __FMODE_NONOTIFY flag and cleanup the usage there, thanks Al! v6->v7: - As per Linus's suggestion, Amir added the file flag FMODE_NOTIFY_PERM that will be set at open time if the file has permission related watches (this is the original malware style permission watches and the new precontent watches). All of the VFS hooks and the page fault hooks use this flag to determine if they should generate a notification to allow for a much cheaper check in the common case. v5->v6: - Linus had problems with this and rejected Jan's PR (https://lore.kernel.org/linux-fsdevel/20240923110348.tbwihs42dxxltabc@quack3/ ), so I'm respinning this series to address his concerns. Hopefully this is more acceptable. - Change the page fault hooks to happen only in the case where we have to add a page, not where there exists pages already. - Amir added a hook to truncate. - We made the flag per SB instead of per fstype, Amir wanted this because of some potential issues with other file system specific work he's doing. - Dropped the bcachefs patch, there were some concerns that we were doing something wrong, and it's not a huge deal to not have this feature for now. - Unfortunately the xfs write fault path still has to do the page fault hook before we know if we have a page or not, this is because of the locking that's done before we get to the part where we know if we have a page already or not, so that's the path that is still the same from last iteration. - I've re-validated this series with btrfs, xfs, and ext4 to make sure I didn't break anything. v4->v5: - Cleaned up the various "I'll fix it on commit" notes that Jan made since I had to respin the series anyway. - Renamed the filemap pagefault helper for fsnotify per Christians suggestion. - Added a FS_ALLOW_HSM flag per Jan's comments, based on Amir's rough sketch. - Added a patch to disable btrfs defrag on pre-content watched files. - Added a patch to turn on FS_ALLOW_HSM for all the file systems that I tested. - Added two fstests (which will be posted separately) to validate everything, re-validated the series with btrfs, xfs, ext4, and bcachefs to make sure I didn't break anything. v3->v4: - Trying to send a final verson Friday at 5pm before you go on vacation is a recipe for silly mistakes, fixed the xfs handling yet again, per Christoph's review. - Reworked the file system helper so it's handling of fpin was a little less silly, per Chinner's suggestion. - Updated the return values to not or in VM_FAULT_RETRY, as we have a comment in filemap_fault that says if VM_FAULT_ERROR is set we won't have VM_FAULT_RETRY set. v2->v3: - Fix the pagefault path to do MAY_ACCESS instead, updated the perm handler to emit PRE_ACCESS in this case, so we can avoid the extraneous perm event as per Amir's suggestion. - Reworked the exported helper so the per-filesystem changes are much smaller, per Amir's suggestion. - Fixed the screwup for DAX writes per Chinner's suggestion. - Added Christian's reviewed-by's where appropriate. v1->v2: - reworked the page fault logic based on Jan's suggestion and turned it into a helper. - Added 3 patches per-fs where we need to call the fsnotify helper from their ->fault handlers. - Disabled readahead in the case that there's a pre-content watch in place. - Disabled huge faults when there's a pre-content watch in place (entirely because it's untested, theoretically it should be straightforward to do). - Updated the command numbers. - Addressed the random spelling/grammer mistakes that Jan pointed out. - Addressed the other random nits from Jan. --- Original email --- Hello, These are the patches for the bare bones pre-content fanotify support. The majority of this work is Amir's, my contribution to this has solely been around adding the page fault hooks, testing and validating everything. I'm sending it because Amir is traveling a bunch, and I touched it last so I'm going to take all the hate and he can take all the credit. There is a PoC that I've been using to validate this work, you can find the git repo here https://github.com/josefbacik/remote-fetch This consists of 3 different tools. 1. populate. This just creates all the stub files in the directory from the source directory. Just run ./populate ~/linux ~/hsm-linux and it'll recursively create all of the stub files and directories. 2. remote-fetch. This is the actual PoC, you just point it at the source and destination directory and then you can do whatever. ./remote-fetch ~/linux ~/hsm-linux. 3. mmap-validate. This was to validate the pagefault thing, this is likely what will be turned into the selftest with remote-fetch. It creates a file and then you can validate the file matches the right pattern with both normal reads and mmap. Normally I do something like ./mmap-validate create ~/src/foo ./populate ~/src ~/dst ./rmeote-fetch ~/src ~/dst ./mmap-validate validate ~/dst/foo I did a bunch of testing, I also got some performance numbers. I copied a kernel tree, and then did remote-fetch, and then make -j4 Normal real 9m49.709s user 28m11.372s sys 4m57.304s HSM real 10m6.454s user 29m10.517s sys 5m2.617s So ~17 seconds more to build with HSM. I then did a make mrproper on both trees to see the size [root@fedora ~]# du -hs /src/linux 1.6G /src/linux [root@fedora ~]# du -hs dst 125M dst This mirrors the sort of savings we've seen in production. Meta has had these patches (minus the page fault patch) deployed in production for almost a year with our own utility for doing on-demand package fetching. The savings from this has been pretty significant. The page-fault hooks are necessary for the last thing we need, which is on-demand range fetching of executables. Some of our binaries are several gigs large, having the ability to remote fetch them on demand is a huge win for us not only with space savings, but with startup time of containers. There will be tests for this going into LTP once we're satisfied with the patches and they're on their way upstream. Thanks, Josef Al Viro (1): fs: get rid of __FMODE_NONOTIFY kludge Amir Goldstein (11): fsnotify: opt-in for permission events at file open time fsnotify: check if file is actually being watched for pre-content events on open fanotify: don't skip extra event info if no info_mode is set fanotify: rename a misnamed constant fanotify: reserve event bit of deprecated FAN_DIR_MODIFY fsnotify: introduce pre-content permission events fsnotify: pass optional file access range in pre-content event fsnotify: generate pre-content permission event on truncate fanotify: introduce FAN_PRE_ACCESS permission event fanotify: report file range info with pre-content events fanotify: allow to set errno in FAN_DENY permission response Jan Kara (1): ext4: add pre-content fsnotify hook for DAX faults Josef Bacik (6): fanotify: disable readahead if we have pre-content watches mm: don't allow huge faults for files with pre content watches fsnotify: generate pre-content permission event on page fault xfs: add pre-content fsnotify hook for DAX faults btrfs: disable defrag on pre-content watched files fs: enable pre-content events on supported file systems fs/btrfs/ioctl.c | 9 +++ fs/btrfs/super.c | 2 +- fs/ext4/file.c | 3 + fs/ext4/super.c | 3 + fs/fcntl.c | 4 +- fs/notify/fanotify/fanotify.c | 31 +++++++-- fs/notify/fanotify/fanotify.h | 15 ++++ fs/notify/fanotify/fanotify_user.c | 137 +++++++++++++++++++++++++++++-------- fs/notify/fsnotify.c | 83 +++++++++++++++++++++- fs/open.c | 62 +++++++++++++---- fs/xfs/xfs_file.c | 14 ++++ fs/xfs/xfs_super.c | 2 +- include/linux/fanotify.h | 18 +++-- include/linux/fs.h | 44 +++++++++--- include/linux/fsnotify.h | 78 ++++++++++++++++----- include/linux/fsnotify_backend.h | 53 +++++++++++++- include/linux/mm.h | 1 + include/uapi/asm-generic/fcntl.h | 1 - include/uapi/linux/fanotify.h | 18 +++++ mm/filemap.c | 90 ++++++++++++++++++++++++ mm/memory.c | 22 ++++++ mm/nommu.c | 7 ++ mm/readahead.c | 14 ++++ security/selinux/hooks.c | 3 +- 24 files changed, 624 insertions(+), 90 deletions(-)
