Terje Malmedal reports that a fuse filesystem with 32 million inodes
on a machine with lots of memory can take up to 30 minutes to process
FORGET requests when all those inodes are evicted from the icache.
To solve this, create a BATCH_FORGET request that allows up to about
8000 FORGET requests to be sent in a single message.
This request is only sent if userspace supports interface version 7.16
or later, otherwise fall back to sending individual FORGET messages.
Reported-by: Terje Malmedal <terje.malmedal@usit.uio.no>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Terje Malmedal reports that a fuse filesystem with 32 million inodes
on a machine with lots of memory can go unresponsive for up to 30
minutes when all those inodes are evicted from the icache.
The reason is that FORGET messages, sent when the inode is evicted,
are queued up together with regular filesystem requests, and while the
huge queue of FORGET messages are processed no other filesystem
operation can proceed.
Since a full fuse request structure is allocated for each inode, these
take up quite a bit of memory as well.
To solve these issues, create a slim 'fuse_forget_link' structure
containing just the minimum of information required to send the FORGET
request and chain these on a separate queue.
When userspace is asking for a request make sure that FORGET and
non-FORGET requests are selected fairly: for each 8 non-FORGET allow
16 FORGET requests. This will make sure FORGETs do not pile up, yet
other requests are also allowed to proceed while the queued FORGETs
are processed.
Reported-by: Terje Malmedal <terje.malmedal@usit.uio.no>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Userspace filesystem can request data to be retrieved from the inode's
mapping. This request is synchronous and the retrieved data is queued
as a new request. If the write to the fuse device returns an error
then the retrieve request was not completed and a reply will not be
sent.
Only present pages are returned in the retrieve reply. Retrieving
stops when it finds a non-present page and only data prior to that is
returned.
This request doesn't change the dirty state of pages.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Userspace filesystem can request data to be stored in the inode's
mapping. This request is synchronous and has no reply. If the write
to the fuse device returns an error then the store request was not
fully completed (but may have updated some pages).
If the stored data overflows the current file size, then the size is
extended, similarly to a write(2) on the filesystem.
Pages which have been completely stored are marked uptodate.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse:
mm: export generic_pipe_buf_*() to modules
fuse: support splice() reading from fuse device
fuse: allow splice to move pages
mm: export remove_from_page_cache() to modules
mm: export lru_cache_add_*() to modules
fuse: support splice() writing to fuse device
fuse: get page reference for readpages
fuse: use get_user_pages_fast()
fuse: remove unneeded variable
When splicing buffers to the fuse device with SPLICE_F_MOVE, try to
move pages from the pipe buffer into the page cache. This allows
populating the fuse filesystem's cache without ever touching the page
contents, i.e. zero copy read capability.
The following steps are performed when trying to move a page into the
page cache:
- buf->ops->confirm() to make sure the new page is uptodate
- buf->ops->steal() to try to remove the new page from it's previous place
- remove_from_page_cache() on the old page
- add_to_page_cache_locked() on the new page
If any of the above steps fail (non fatally) then the code falls back
to copying the page. In particular ->steal() will fail if there are
external references (other than the page cache and the pipe buffer) to
the page.
Also since the remove_from_page_cache() + add_to_page_cache_locked()
are non-atomic it is possible that the page cache is repopulated in
between the two and add_to_page_cache_locked() will fail. This could
be fixed by creating a new atomic replace_page_cache_page() function.
fuse_readpages_end() needed to be reworked so it works even if
page->mapping is NULL for some or all pages which can happen if the
add_to_page_cache_locked() failed.
A number of sanity checks were added to make sure the stolen pages
don't have weird flags set, etc... These could be moved into generic
splice/steal code.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Update some fs code to make use of new helper functions introduced
in the previous patch. Should be no significant change in behaviour
(except CIFS now calls send_sig under i_lock, via inode_newsize_ok).
Reviewed-by: Christoph Hellwig <hch@lst.de>
Acked-by: Miklos Szeredi <miklos@szeredi.hu>
Cc: linux-nfs@vger.kernel.org
Cc: Trond.Myklebust@netapp.com
Cc: linux-cifs-client@lists.samba.org
Cc: sfrench@samba.org
Signed-off-by: Nick Piggin <npiggin@suse.de>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Make the max_background and congestion_threshold parameters of a FUSE
mount tunable at runtime by adding the respective knobs to its directory
within the fusectl filesystem.
Signed-off-by: Csaba Henk <csaba@gluster.com>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
The practical values for these limits depend on the design of the
filesystem server so let userspace set them at initialization time.
Signed-off-by: Csaba Henk <csaba@gluster.com>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Add notification messages that allow the filesystem to invalidate VFS
caches.
Two notifications are added:
1) inode invalidation
- invalidate cached attributes
- invalidate a range of pages in the page cache (this is optional)
2) dentry invalidation
- try to invalidate a subtree in the dentry cache
Care must be taken while accessing the 'struct super_block' for the
mount, as it can go away while an invalidation is in progress. To
prevent this, introduce a rw-semaphore, that is taken for read during
the invalidation and taken for write in the ->kill_sb callback.
Cc: Csaba Henk <csaba@gluster.com>
Cc: Anand Avati <avati@zresearch.com>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
This patch lets filesystems handle masking the file mode on creation.
This is needed if filesystem is using ACLs.
- The CREATE, MKDIR and MKNOD requests are extended with a "umask"
parameter.
- A new FUSE_DONT_MASK flag is added to the INIT request/reply. With
this the filesystem may request that the create mode is not masked.
CC: Jean-Pierre André <jean-pierre.andre@wanadoo.fr>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Update fuse_conn_init() such that it doesn't take @sb and move bdi
registration into a separate function. Also separate out
fuse_conn_kill() from fuse_put_super().
These will be used to implement cuse.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Make fuse_sync_release() a generic helper function that doesn't need a
struct inode pointer. This makes it suitable for use by CUSE.
Change return value of fuse_release_common() from int to void.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Move setting ff->fh, ff->nodeid and file->private_data outside
fuse_finish_open(). Add ->open_flags member to struct fuse_file.
This simplifies the argument passing to fuse_finish_open() and
fuse_release_fill(), and paves the way for creating an open helper
that doesn't need an inode pointer.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Use ff->fc and ff->nodeid instead of passing down the inode.
This prepares this function for use by CUSE, where the inode is not
owned by a fuse filesystem.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Add new members ->fc and ->nodeid to struct fuse_file. This will aid
in converting functions for use by CUSE, where the inode is not owned
by a fuse filesystem.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Add fuse_conn->release() so that fuse_conn can be embedded in other
structures.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Separate out fuse_conn_init() from new_conn() and while at it
initialize fuse_conn->entry during conn initialization.
This will be used by CUSE.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Add fuse_ prefix to request_send*() and get_root_inode() as some of
those functions will be exported for CUSE. With or without CUSE
export, having the function names scoped is a good idea for
debuggability.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Implement poll support. Polled files are indexed using kh in a RB
tree rooted at fuse_conn->polled_files.
Client should send FUSE_NOTIFY_POLL notification once after processing
FUSE_POLL which has FUSE_POLL_SCHEDULE_NOTIFY set. Sending
notification unconditionally after the latest poll or everytime file
content might have changed is inefficient but won't cause malfunction.
fuse_file_poll() can sleep and requires patches from the following
thread which allows f_op->poll() to sleep.
http://thread.gmane.org/gmane.linux.kernel/726176
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
The file handle, fuse_file->fh, is opaque value supplied by userland
FUSE server and uniqueness is not guaranteed. Add file kernel handle,
fuse_file->kh, which is allocated by the kernel on file allocation and
guaranteed to be unique.
This will be used by poll to match notification to the respective file
but can be used for other purposes where unique file handle is
necessary.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Add include protectors to include/linux/fuse.h and fs/fuse/fuse_i.h.
Signed-off-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Implement the get_parent export operation by sending a LOOKUP request with
".." as the name.
Implement looking up an inode by node ID after it has been evicted from
the cache. This is done by seding a LOOKUP request with "." as the name
(for all file types, not just directories).
The filesystem can set the FUSE_EXPORT_SUPPORT flag in the INIT reply, to
indicate that it supports these special lookups.
Thanks to John Muir for the original implementation of this feature.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Cc: "J. Bruce Fields" <bfields@fieldses.org>
Cc: Trond Myklebust <trond.myklebust@fys.uio.no>
Cc: Matthew Wilcox <matthew@wil.cx>
Cc: David Teigland <teigland@redhat.com>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Implement export_operations, to allow fuse filesystems to be exported to
NFS. This feature has been in the out-of-tree fuse module, and is widely
used and tested.
It has not been originally merged into mainline, because doing the NFS
export in userspace was thought to be a cleaner and more efficient way of
doing it, than through the kernel.
While that is true, it would also have involved a lot of duplicated effort
at reimplementing NFS exporting (all the different versions of the
protocol). This effort was unfortunately not undertaken by anyone, so we
are left with doing it the easy but less efficient way.
If this feature goes in, the out-of-tree fuse module can go away,
which would have several advantages:
- not having to maintain two versions
- less confusion for users
- no bugs due to kernel API changes
Comment from hch:
- Use the same fh_type values as XFS, since we use the same fh encoding.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Prior to 2.6.26 fuse only supported single page write requests. In theory all
fuse filesystem should be able support bigger than 4k writes, as there's
nothing in the API to prevent it. Unfortunately there's a known case in
NTFS-3G where big writes cause filesystem corruption. There could also be
other filesystems, where the lack of testing with big write requests would
result in bugs.
To prevent such problems on a kernel upgrade, disable big writes by default,
but let filesystems set a flag to turn it on.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Cc: Szabolcs Szakacsits <szaka@ntfs-3g.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Node ID is 64bit but it is passed as unsigned long to some functions. This
breakage wasn't noticed, because libfuse uses unsigned long too.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
If the READ request returned a short count, then either
- cached size is incorrect
- filesystem is buggy, as short reads are only allowed on EOF
So assume that the size is wrong and refresh it, so that cached read() doesn't
zero fill the missing chunk.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Quoting Linus (3 years ago, FUSE inclusion discussions):
"User-space filesystems are hard to get right. I'd claim that they
are almost impossible, unless you limit them somehow (shared
writable mappings are the nastiest part - if you don't have those,
you can reasonably limit your problems by limiting the number of
dirty pages you accept through normal "write()" calls)."
Instead of attempting the impossible, I've just waited for the dirty page
accounting infrastructure to materialize (thanks to Peter Zijlstra and
others). This nicely solved the biggest problem: limiting the number of pages
used for write caching.
Some small details remained, however, which this largish patch attempts to
address. It provides a page writeback implementation for fuse, which is
completely safe against VM related deadlocks. Performance may not be very
good for certain usage patterns, but generally it should be acceptable.
It has been tested extensively with fsx-linux and bash-shared-mapping.
Fuse page writeback design
--------------------------
fuse_writepage() allocates a new temporary page with GFP_NOFS|__GFP_HIGHMEM.
It copies the contents of the original page, and queues a WRITE request to the
userspace filesystem using this temp page.
The writeback is finished instantly from the MM's point of view: the page is
removed from the radix trees, and the PageDirty and PageWriteback flags are
cleared.
For the duration of the actual write, the NR_WRITEBACK_TEMP counter is
incremented. The per-bdi writeback count is not decremented until the actual
write completes.
On dirtying the page, fuse waits for a previous write to finish before
proceeding. This makes sure, there can only be one temporary page used at a
time for one cached page.
This approach is wasteful in both memory and CPU bandwidth, so why is this
complication needed?
The basic problem is that there can be no guarantee about the time in which
the userspace filesystem will complete a write. It may be buggy or even
malicious, and fail to complete WRITE requests. We don't want unrelated parts
of the system to grind to a halt in such cases.
Also a filesystem may need additional resources (particularly memory) to
complete a WRITE request. There's a great danger of a deadlock if that
allocation may wait for the writepage to finish.
Currently there are several cases where the kernel can block on page
writeback:
- allocation order is larger than PAGE_ALLOC_COSTLY_ORDER
- page migration
- throttle_vm_writeout (through NR_WRITEBACK)
- sync(2)
Of course in some cases (fsync, msync) we explicitly want to allow blocking.
So for these cases new code has to be added to fuse, since the VM is not
tracking writeback pages for us any more.
As an extra safetly measure, the maximum dirty ratio allocated to a single
fuse filesystem is set to 1% by default. This way one (or several) buggy or
malicious fuse filesystems cannot slow down the rest of the system by hogging
dirty memory.
With appropriate privileges, this limit can be raised through
'/sys/class/bdi/<bdi>/max_ratio'.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Register FUSE's backing_dev_info under sysfs with the name "fuse-MAJOR:MINOR"
Make the fuse control filesystem use s_dev instead of a fuse specific ID.
This makes it easier to match directories under /sys/fs/fuse/connections/ with
directories under /sys/class/bdi, and with actual mounts.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Cc: Peter Zijlstra <a.p.zijlstra@chello.nl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Libfuse basically creates a new thread for each new request. This is fine for
synchronous requests, which are naturally limited. However background
requests (especially writepage) can cause a thread creation storm.
To avoid this, limit the number of background requests available to userspace.
This is done by introducing another queue for background requests, and a
counter for the number of "active" requests, which are currently available for
userspace.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Move the fields 'dentry' and 'vfsmount' into the request specific union, since
these are only used for the RELEASE request.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Some open flags (O_APPEND, O_DIRECT) can be changed with fcntl(F_SETFL, ...)
after open, but fuse currently only sends the flags to userspace in open.
To make it possible to correcly handle changing flags, send the
current value to userspace in each read and write.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Currently reading a fuse file will stop at cached i_size and return
EOF, even though the file might have grown since the attributes were
last updated.
So detect if trying to read past EOF, and refresh the attributes
before continuing with the read.
Thanks to mpb for the report.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
For mandatory locking the userspace filesystem needs to know the lock
ownership for read, write and truncate operations.
This patch adds the necessary fields to the protocol.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch adds a new helper function fuse_write_fill() which makes it
possible to send WRITE requests asynchronously.
A new flag for WRITE requests is also added which indicates that this a write
from the page cache, and not a "normal" file write.
This patch is in preparation for writable mmap support.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Each WRITE request must carry a valid file descriptor. When a page is written
back from a memory mapping, the file through which the page was dirtied is not
available, so a new mechananism is needed to find a suitable file in
->writepage(s).
A list of fuse_files is added to fuse_inode. The file is removed from the
list in fuse_release().
This patch is in preparation for writable mmap support.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This patch allows fuse filesystems to implement open(..., O_TRUNC) as a single
request, instead of separate truncate and open requests.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Getattr and lookup operations can be running in parallel to attribute changing
operations, such as write and setattr.
This means, that if for example getattr was slower than a write, the cached
size attribute could be set to a stale value.
To prevent this race, introduce a per-filesystem attribute version counter.
This counter is incremented whenever cached attributes are modified, and the
incremented value stored in the inode.
Before storing new attributes in the cache, getattr and lookup check, using
the version number, whether the attributes have been modified during the
request's lifetime. If so, the returned attributes are not cached, because
they might be stale.
Thanks to Jakub Bogusz for the bug report and test program.
[akpm@linux-foundation.org: coding-style fixes]
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Cc: Jakub Bogusz <jakub.bogusz@gemius.pl>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The following operation didn't check if sending the request was allowed:
setattr
listxattr
statfs
Some other operations don't explicitly do the check, but VFS calls
->permission() which checks this.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The VFS checks sticky bits on the parent directory even if the filesystem
defines it's own ->permission(). In some situations (sshfs, mountlo, etc) the
user does have permission to delete a file even if the attribute based
checking would not allow it.
So work around this by storing the permission bits separately and returning
them in stat(), but cutting the permission bits off from inode->i_mode.
This is slightly hackish, but it's probably not worth it to add new
infrastructure in VFS and a slight performance penalty for all filesystems,
just for the sake of fuse.
[Jan Engelhardt] cosmetic fixes
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Cc: Jan Engelhardt <jengelh@linux01.gwdg.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
fuse_permission() didn't refresh inode attributes before using them, even if
the validity has already expired.
Thanks to Junjiro Okajima for spotting this.
Also remove some old code to unconditionally refresh the attributes on the
root inode.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Make lifetime of 'struct fuse_file' independent from 'struct file' by adding a
reference counter and destructor.
This will enable asynchronous page writeback, where it cannot be guaranteed,
that the file is not released while a request with this file handle is being
served.
The actual RELEASE request is only sent when there are no more references to
the fuse_file.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Use wake_up_all instead of wake_up in put_reserved_req(), otherwise it is
possible that the right task is not woken up.
Also create a separate reserved_req_waitq in addition to the blocked_waitq,
since they fulfill totally separate functions.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Set the read and write congestion state if the request queue is close to
blocking, and clear it when it's not.
This prevents unnecessary blocking in readahead and (when writable mmaps are
allowed) writeback.
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
