Re: [RFC v3] fuse: use fs-iomap for better performance so we can containerize ext4

["Darrick J. Wong" <djwong@kernel.org>]

On Fri, Jul 18, 2025 at 01:55:48PM +0200, Amir Goldstein wrote:
> On Fri, Jul 18, 2025 at 10:54 AM Christian Brauner <brauner@kernel.org> wrote:
> >
> > On Thu, Jul 17, 2025 at 04:10:38PM -0700, Darrick J. Wong wrote:
> > > Hi everyone,
> > >
> > > DO NOT MERGE THIS, STILL!
> > >
> > > This is the third request for comments of a prototype to connect the
> > > Linux fuse driver to fs-iomap for regular file IO operations to and from
> > > files whose contents persist to locally attached storage devices.
> > >
> > > Why would you want to do that?  Most filesystem drivers are seriously
> > > vulnerable to metadata parsing attacks, as syzbot has shown repeatedly
> > > over almost a decade of its existence.  Faulty code can lead to total
> > > kernel compromise, and I think there's a very strong incentive to move
> > > all that parsing out to userspace where we can containerize the fuse
> > > server process.
> > >
> > > willy's folios conversion project (and to a certain degree RH's new
> > > mount API) have also demonstrated that treewide changes to the core
> > > mm/pagecache/fs code are very very difficult to pull off and take years
> > > because you have to understand every filesystem's bespoke use of that
> > > core code.  Eeeugh.
> > >
> > > The fuse command plumbing is very simple -- the ->iomap_begin,
> > > ->iomap_end, and iomap ->ioend calls within iomap are turned into
> > > upcalls to the fuse server via a trio of new fuse commands.  Pagecache
> > > writeback is now a directio write.  The fuse server is now able to
> > > upsert mappings into the kernel for cached access (== zero upcalls for
> > > rereads and pure overwrites!) and the iomap cache revalidation code
> > > works.
> > >
> > > With this RFC, I am able to show that it's possible to build a fuse
> > > server for a real filesystem (ext4) that runs entirely in userspace yet
> > > maintains most of its performance.  At this stage I still get about 95%
> > > of the kernel ext4 driver's streaming directio performance on streaming
> > > IO, and 110% of its streaming buffered IO performance.  Random buffered
> > > IO is about 85% as fast as the kernel.  Random direct IO is about 80% as
> > > fast as the kernel; see the cover letter for the fuse2fs iomap changes
> > > for more details.  Unwritten extent conversions on random direct writes
> > > are especially painful for fuse+iomap (~90% more overhead) due to upcall
> > > overhead.  And that's with debugging turned on!
> > >
> > > These items have been addressed since the first RFC:
> > >
> > > 1. The iomap cookie validation is now present, which avoids subtle races
> > > between pagecache zeroing and writeback on filesystems that support
> > > unwritten and delalloc mappings.
> > >
> > > 2. Mappings can be cached in the kernel for more speed.
> > >
> > > 3. iomap supports inline data.
> > >
> > > 4. I can now turn on fuse+iomap on a per-inode basis, which turned out
> > > to be as easy as creating a new ->getattr_iflags callback so that the
> > > fuse server can set fuse_attr::flags.
> > >
> > > 5. statx and syncfs work on iomap filesystems.
> > >
> > > 6. Timestamps and ACLs work the same way they do in ext4/xfs when iomap
> > > is enabled.
> > >
> > > 7. The ext4 shutdown ioctl is now supported.
> > >
> > > There are some major warts remaining:
> > >
> > > a. ext4 doesn't support out of place writes so I don't know if that
> > > actually works correctly.
> > >
> > > b. iomap is an inode-based service, not a file-based service.  This
> > > means that we /must/ push ext2's inode numbers into the kernel via
> > > FUSE_GETATTR so that it can report those same numbers back out through
> > > the FUSE_IOMAP_* calls.  However, the fuse kernel uses a separate nodeid
> > > to index its incore inode, so we have to pass those too so that
> > > notifications work properly.  This is related to #3 below:
> > >
> > > c. Hardlinks and iomap are not possible for upper-level libfuse clients
> > > because the upper level libfuse likes to abstract kernel nodeids with
> > > its own homebrew dirent/inode cache, which doesn't understand hardlinks.
> > > As a result, a hardlinked file results in two distinct struct inodes in
> > > the kernel, which completely breaks iomap's locking model.  I will have
> > > to rewrite fuse2fs for the lowlevel libfuse library to make this work,
> > > but on the plus side there will be far less path lookup overhead.
> > >
> > > d. There are too many changes to the IO manager in libext2fs because I
> > > built things needed to stage the direct/buffered IO paths separately.
> > > These are now unnecessary but I haven't pulled them out yet because
> > > they're sort of useful to verify that iomap file IO never goes through
> > > libext2fs except for inline data.
> > >
> > > e. If we're going to use fuse servers as "safe" replacements for kernel
> > > filesystem drivers, we need to be able to set PF_MEMALLOC_NOFS so that
> > > fuse2fs memory allocations (in the kernel) don't push pagecache reclaim.
> > > We also need to disable the OOM killer(s) for fuse servers because you
> > > don't want filesystems to unmount abruptly.
> > >
> > > f. How do we maximally contain the fuse server to have safe filesystem
> > > mounts?  It's very convenient to use systemd services to configure
> > > isolation declaratively, but fuse2fs still needs to be able to open
> > > /dev/fuse, the ext4 block device, and call mount() in the shared
> > > namespace.  This prevents us from using most of the stronger systemd
> >
> > I'm happy to help you here.
> >
> > First, I think using a character device for namespaced drivers is always
> > a mistake. FUSE predates all that ofc. They're incredibly terrible for
> > delegation because of devtmpfs not being namespaced as well as devices
> > in general. And having device nodes on anything other than tmpfs is just
> > wrong (TM).
> >
> > In systemd I ultimately want a bpf LSM program that prevents the
> > creation of device nodes outside of tmpfs. They don't belong on
> > persistent storage imho. But anyway, that's besides the point.
> >
> > Opening the block device should be done by systemd-mountfsd but I think
> > /dev/fuse should really be openable by the service itself.

/me slaps his head and remembers that fsopen/fsconfig/fsmount exist.
Can you pass an fsopen fd to an unprivileged process and have that
second process call fsmount?

If so, then it would be more convenient if mount.safe/systemd-mountfsd
could pass open fds for /dev/fuse fsopen then the fuse server wouldn't
need any special /dev access at all.  I think then the fuse server's
service could have:

DynamicUser=true
ProtectSystem=true
ProtectHome=true
PrivateTmp=true
PrivateDevices=true
DevicePolicy=strict

(I think most of those are redundant with DynamicUser=true but a lot of
my systemd-fu is paged out ATM.)

My goal here is extreme containment -- the code doing the fs metadata
parsing has no privileges, no write access except to the fds it was
given, no network access, and no ability to read anything outside the
root filesystem.  Then I can get back to writing buffer
overflows^W^Whigh quality filesystem code in peace.

> > So we can try and allowlist /dev/fuse in vfs_mknod() similar to
> > whiteouts. That means you can do mknod() in the container to create
> > /dev/fuse (Personally, I would even restrict this to tmpfs right off the
> > bat so that containers can only do this on their private tmpfs mount at
> > /dev.)
> >
> > The downside of this would be to give unprivileged containers access to
> > FUSE by default. I don't think that's a problem per se but it is a uapi
> > change.

Yeah, that is a new risk.  It's still better than metadata parsing
within the kernel address space ... though who knows how thoroughly fuse
has been fuzzed by syzbot :P

> > Let me think a bit about alternatives. I have one crazy idea but I'm not
> > sure enough about it to spill it.

Please do share, #f is my crazy unbaked idea. :)

> I don't think there is a hard requirement for the fuse fd to be opened from
> a device driver.
> With fuse io_uring communication, the open fd doesn't even need to do io.
> 
> > > protections because they tend to run in a private mount namespace with
> > > various parts of the filesystem either hidden or readonly.
> > >
> > > In theory one could design a socket protocol to pass mount options,
> > > block device paths, fds, and responsibility for the mount() call between
> > > a mount helper and a service:
> >
> > This isn't a problem really. This should just be an extension to
> > systemd-mountfsd.

I suppose mount.safe could very well call systemd-mount to go do all the
systemd-related service setup, and that would take care of udisks as
well.

> This is relevant not only to systemd env.
> 
> I have been experimenting with this mount helper service to mount fuse fs
> inside an unprivileged kubernetes container, where opening of /dev/fuse
> is restricted by LSM policy:
> 
> https://github.com/pfnet-research/meta-fuse-csi-plugin?tab=readme-ov-file#fusermount3-proxy-modified-fusermount3-approach

That sounds similar to what I was thinking about, though there are a lot
of TLAs that I don't understand.

--D