| Age | Commit message (Collapse) | Author | Lines |
|---|
|
Signed-off-by: Bernd Schubert <bschubert@ddn.com>
|
|
Add support for filesystem passthrough read/write of files.
When the FUSE_PASSTHROUGH capability is enabled, the FUSE server may
decide, while handling the "open" or "create" requests, if the given
file can be accessed by that process in "passthrough" mode, meaning that
all the further read and write operations would be forwarded by the
kernel directly to the backing file rather than to the FUSE server.
All requests other than read or write are still handled by the server.
This allows for an improved performance on reads and writes, especially
in the case of reads at random offsets, for which no (readahead)
caching mechanism would help, reducing the performance gap between FUSE
and native filesystem access.
Extend also the passthrough_hp example with the new passthrough feature.
This example opens a kernel backing file per FUSE inode on the first
FUSE file open of that inode and closes the backing file on the release
of the last FUSE file on that inode.
All opens of the same inode passthrough to the same backing file.
A combination of fi->direct_io and fi->passthrough is allowed.
It means that read/write operations go directly to the server, but mmap
is done on the backing file.
This allows to open some fds of the inode in passthrough mode and some
fd of the same inode in direct_io/passthrough_mmap mode.
Signed-off-by: Alessio Balsini <balsini@android.com>
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
|
|
Add in the rename of FUSE_DIRECT_IO_RELAX to
FUSE_DIRECT_IO_ALLOW_MMAP.
|
|
|
|
This syncs fuse_kernel.h to <linux-6.3>/include/uapi/linux/fuse.h
Special handling is done for setxattr as in linux commit
52a4c95f4d24b struct fuse_setxattr_in was extended. Extended
struct is only used when FUSE_SETXATTR_EXT is passed in FUSE_INIT
reply.
|
|
https://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse.git/commit/?h=for-next&id=53e949edb7692dce02220eba926c9d75ecbb47f7"
This reverts commit 7f430a39db5a30979d75a906af891a38ebce1a3c because it
breaks setxattr().
Fixes: #730
|
|
Right now fuse kernel serializes direct writes on the
same file. This serialization is good for such FUSE
implementations which rely on the inode lock to
avoid any data inconsistency issues but it hurts badly
such FUSE implementations which have their own mechanism
of dealing with cache/data integrity and can handle
parallel direct writes on the same file.
This patch allows parallel direct writes on the same file to be
enabled with the help of a flag FOPEN_PARALLEL_DIRECT_WRITES.
FUSE implementations which want to use this feature can
set this flag during fuse init. Default behaviour remains
same i.e no parallel direct writes on the same file.
Corresponding fuse kernel patch(Merged).
https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?h=v6.2&id=153524053bbb0d27bb2e0be36d1b46862e9ce74c
|
|
https://git.kernel.org/pub/scm/linux/kernel/git/mszeredi/fuse.git/commit/?h=for-next&id=53e949edb7692dce02220eba926c9d75ecbb47f7
|
|
|
|
Just a further sync with the in-kernel flags.
|
|
In fuse kernel, 'commit 53db28933e95 ("fuse: extend init flags")'
made the changes to handle flags going beyond 32 bits but i think
changes were not done in libfuse to handle the same.
This patch prepares the ground in libfuse for incoming FUSE kernel
patches (Atomic open + lookup) where flags went beyond 32 bits.
It makes struct same as in fuse kernel resulting in name change of
few fields.
|
|
Allow requesting from kernel to avoid flush on close at file open
time. If kernel does not support FOPEN_NOFLUSH flag, the request
will be ignored.
For passthrough_hp example, request to avoid flush on close when
writeback cache is disabled and file is opened O_RDONLY.
Signed-off-by: Amir Goldstein <amir73il@gmail.com>
|
|
They are illegal in C89/90.
|
|
|
|
|
|
This constant is not defined in the kernel, so it will be lost when
fuse_kernel.h is not synchronized. Instead, the kernel just passes a
flag value of "1", so for now we also use a literal in userspace.
|
|
Commit 24b35c3d97ffdbf0a1f8e8b4e94ed892343603a6 had a simple mistake in
its changelog entry.
|
|
|
|
Add support for the relatively new copy_file_range() syscall. Backend
filesystems can now implement an efficient way of cloning/duplicating
data ranges within files. See 'man 2 copy_file_range' for more details.
|
|
Taken from Linux kernel commit 3b7008b226f3.
|
|
Taken from Linux kernel commit 27bcd37.
|
|
This creates a separate device file descriptor for each processing thread,
which might improve performance.
|
|
|
|
See renameat2() system call in linux-3.15 and later kernels.
|
|
|
|
This allows the filesystem to specify the time granularity it
supports when the kernel is responsible for updating times
("writeback_cache" option).
|
|
Asynchronous direct I/O is supported by linux kernels 3.13 and
later, writeback caching is supported by 3.14 and later.
|
|
Check for __KERNEL__ instead of __linux__ and use the standard int
types instead of the linux specific ones.
|
|
This switches the -o no_readdirplus option to a tristate
string: -o readdirplus=(yes|no|auto)
Telling the kernel to always use readdirplus is beneficial to
filesystems (e.g. GlusterFS) where the cost to perform readdir
and readdirplus are identical.
The default remains "auto" (if supported).
|
|
Make requested poll events available to the filesystem. If the requested
eventsare not available, then this field is zero.
|
|
This patch implements readdirplus support in FUSE usersapce. It adds
a new fuse lowlevel operations fuse_lowleve_ops::readdir_plus,
corespoding mount options and helper functions to maintain buffer.
[From: Eric Wong <normalperson@yhbt.net>]
This makes our terminology consistent with NFS and
our kernel module, as well as reducing user/developer
confusion in the command-line.
Note: I'm keeping "fuse_add_direntry_plus" since that is
less standardized in its use than "readdirplus" for now.
Signed-off-by: Feng Shuo <steve.shuo.feng@gmail.com>
|
|
Use <linux/types.h> for linux and define types used for other operating systems
using <stdint.h> types.
|
|
Several caching logic changes have been made on the kernel side
to better support network-based fuse filesystems. These include
kernel side mtime checking and read path cache revalidation. The
new caching logic is enabled through the FUSE_AUTO_INVAL_DATA
init flag. Export this to the user via the '[no_]auto_inval_data'
mount option.
Signed-off-by: Feng Shuo <steve.shuo.feng@gmail.com>
|
|
Add missing flags that userspace derived from the protocol version number. This
makes the protocol more flexible.
|
|
fallocate filesystem operation preallocates media space for the given file.
If fallocate returns success then any subsequent write to the given range
never fails with 'not enough space' error.
|
|
Use the ISO C standard compliant form instead of the gcc extension in the
interface definition.
|
|
This partially reverts commit 4b2157c44e6ad7e692fcffb7450143e83151d36b.
Remove mmap/munmap suppor as this missed the interface changes for Linux-3.3
(API version 7.18).
Only revert the mmap/munmap bits and leave the retrieve_reply API fix in place
as well as the optimization in fuse_send_data_iov_fallback().
|
|
libfuse part to allow a FUSE file-system to tell the kernel when a
file or directory is deleted. If the specified dentry has the
specified inode number, the kernel will unhash it.
Signed-off-by: John Muir <john@jmuir.com>
Signed-off-by: Miklos Szeredi <mszeredi@suse.cz>
|
|
Currently this is only useful for CUSE. Also update retrieve_reply()
method.
|
|
Reported by Antonio SJ Musumeci
|
|
This fixes problems with emulating flock() with POSIX locking.
Reported by Sebastian Pipping.
As with lock/setlk/getlk most filesystems don't need to implement
this, as the kernel takes care of file locking. The only reason to
implement locking operations is for network filesystems which want
file locking to work between clients.
|
|
Fix the ambiguity of ioctl ABI on the kernel/userspace boundary
for 32bit vs. 64bit userspace
|
|
This allows forget requests to be processed faster and doesn't require
a modification to fuse filesystems. Reported by Terje Malmedal
|
|
Retrieve data stored in the kernel buffers for a given inode.
|
|
Request data to be stored in the kernel buffers for a given inode.
|
|
|
|
threshold tunable.
|
|
kernel and userspace. Notably libfuse didn't correctly handle the
case when the supported major versions didn't match
|
|
for mknod, mkdir and create requests by linux kernel version
2.6.31 or later, otherwise the umask is set to zero. Also
introduce a new feature flag: FUSE_CAP_DONT_MASK. If the kernel
supports this feature, then this flag will be set in conn->capable
in the ->init() method. If the filesystem sets this flag in in
conn->want, then the create modes will not be masked.
* Add low level interfaces for lookup cache and attribute
invalidation. This feature is available in linux kernels 2.6.31
or later. Patch by John Muir
* Kernel interface version is now 7.12
|
|
|
