The current vnode layout is not smp-friendly by having frequently read data
avoidably sharing cachelines with very frequently modified fields. In
particular v_iflag inspected for VI_DOOMED can be found in the same line with
v_usecount. Instead make it available in the same cacheline as the v_op, v_data
and v_type which all get read all the time.
v_type is avoidably 4 bytes while the necessary data will easily fit in 1.
Shrinking it frees up 3 bytes, 2 of which get used here to introduce a new
flag field with a new value: VIRF_DOOMED.
Reviewed by: kib, jeff
Differential Revision: https://reviews.freebsd.org/D22715
When an NFSv4 server replies NFSERR_MINORVERSMISMATCH, it does not generate
a status result for the first operation in the compound. Without this
patch, this will result in a bogus EBADXDR error return.
Returning EBADXDR is relatively harmless, but a correct reply of
NFSERR_MINORVERSMISMATCH is needed by the pNFS client to select the correct
minor version to use for a File Layout DS now that there can be NFSv4.2
DS servers.
mount_nfs.c still needs to be fixed for this, although how the mount fails
is only useful to help sysadmins isolate why a mount fails.
Found during testing of the NFSv4.2 client and server.
MFC after: 2 weeks
This is a preliminary commit of NFSv4.2 definitions that will be used by
subsequent commits which adds NFSv4.2 support to the NFS client and server.
There will be a series of these preliminary commits that will prepare for
a major commit of the NFSv4.2 client/server changes currently found in
subversion under projects/nfsv42/sys.
Since r355472 added code which clears the XATTRSUPPORT bit for non-NFSv4.2
mounts, it is now safe to set it.
There will be a series of these preliminary commits that will prepare for
a major commit of the NFSv4.2 client/server changes currently found in
subversion under projects/nfsv42/sys.
This commit completes updates to nfsproto.h required by the NFSv4.2.
This patch adds code to macros to clear attribute bits not supported
by NFSv4.2. For now, these bits are never set anyhow, but this prepares
the code for the addition of NFSv4.2 support in a future commit.
There will be a series of these preliminary commits that will prepare for
a major commit of the NFSv4.2 client/server changes currently found in
subversion under projects/nfsv42/sys.
This is a preliminary commit of NFSv4.2 definitions that will be used by
subsequent commits which adds NFSv4.2 support to the NFS client and server.
There will be a series of these preliminary commits that will prepare for
a major commit of the NFSv4.2 client/server changes currently found in
subversion under projects/nfsv42/sys.
This is a preliminary commit of NFSv4.2 definitions that will be used by
subsequent commits which adds NFSv4.2 support to the NFS client and server.
There will be a series of these preliminary commits that will prepare for
a major commit of the NFSv4.2 client/server changes currently found in
subversion under projects/nfsv42/sys.
Generally, it's preferred that an application fork/setsid if it doesn't want
to keep its controlling TTY, but it could be that a debugger is trying to
steal it instead -- so it would hook in, drop the controlling TTY, then do
some magic to set things up again. In this case, TIOCNOTTY is quite handy
and still respected by at least OpenBSD, NetBSD, and Linux as far as I can
tell.
I've dropped the note about obsoletion, as I intend to support TIOCNOTTY as
long as it doesn't impose a major burden.
Reviewed by: bcr (manpages), kib
Differential Revision: https://reviews.freebsd.org/D22572
This allows bumping threadcount without taking the global devmtx lock.
In particular this eliminates contention on said lock while using bhyve
with multiple vms.
Reviewed by: kib
Tested by: markj
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D22548
We might race with reclaim, and then this is no longer a nfs vnode, in
which case we do not need to handle deferred vnode_pager_setsize()
either.
Reported by: rk@ronald.org
PR: 242184
Sponsored by: The FreeBSD Foundation
MFC after: 3 days
A crash was reported where the nr_client field was NULL during an upcall
to the nfsuserd daemon. Since nr_client == NULL only occurs when the
nfsuserd daemon is being shut down, it appeared to be caused by a race
between doing an upcall and the daemon shutting down.
By inspection two races were identified:
1 - The nfsrv_nfsuserd variable is used to indicate whether or not the
daemon is running. However it did not handle the intermediate phase
where the daemon is starting or stopping.
This was fixed by making nfsrv_nfsuserd tri-state and having the
functions that are called during start/stop to obey the intermediate
state.
2 - nfsrv_nfsuserd was checked to see that the daemon was running at
the beginning of an upcall, but nothing prevented the daemon from
being shut down while an upcall was still in progress.
This race probably caused the crash.
The patch fixes this by adding a count of upcalls in progress and
having the shut down function delay until this count goes to zero
before getting rid of nr_client and related data used by an upcall.
Tested by: avg (Panzura QA)
Reported by: avg
Reviewed by: avg
MFC after: 2 weeks
Differential Revision: https://reviews.freebsd.org/D22377
Top-level kern_renameat() increases the writecount on the mount point,
which, together with tmpfs unmount suspending the mount, already
ensures that unmount cannot proceed while rename unlocks and relocks
all operated vnodes.
Remove vfs_busy() call from tmpfs_rename() which was done while
holding a vnode lock, creating the deadlock. The only intent of the
busy operation seems to be the prevention of unmount, which is already
ensured.
Reported and tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
The pNFS server currently reports SpaceUsed (va_bytes) for the metadata
file. This in not correct, since the metadata file is always empty and,
as such, va_bytes is just the allocation for the empty file.
This patch adds va_bytes to the list of attributes acquired from the
DS for a file, so that it includes the allocated data size and is updated
when the file is written.
For files created on a pNFS server before this patch is applied, the
va_bytes value is estimated by rounding va_size up to a multiple of
BLKDEV_IOSIZE. Once the file is written after this patch has been
applied to the metadata server, the va_bytes returned for the file
will be correct.
This patch only affects a pNFS metadata server.
Found during testing of the NFSv4.2 pNFS server for the Allocate operation.
(Not yet in head/current.)
MFC after: 2 weeks
Currently si_usecount is effectively a sum of usecounts from all associated
vnodes. This is maintained by special-casing for VCHR every time usecount is
modified. Apart from complicating the code a little bit, it has a scalability
impact since it forces a read from a cacheline shared with said count.
There are no consumers of the feature in the ports tree. In head there are only
2: revoke and devfs_close. Both can get away with a weaker requirement than the
exact usecount, namely just the count of active vnodes. Changing the meaning to
the latter means we only need to modify it on 0<->1 transitions, avoiding the
check plenty of times (and entirely in something like vrefact).
Reviewed by: kib, jeff
Tested by: pho
Differential Revision: https://reviews.freebsd.org/D22202
reudundant complicated checks and additional locking required only for
anonymous memory. Introduce vm_object_allocate_anon() to create these
objects. DEFAULT and SWAP objects now have the correct settings for
non-anonymous consumers and so individual consumers need not modify the
default flags to create super-pages and avoid ONEMAPPING/NOSPLIT.
Reviewed by: alc, dougm, kib, markj
Tested by: pho
Differential Revision: https://reviews.freebsd.org/D22119
flag and use the same system.
This enables further fault locking improvements by allowing more faults to
proceed with a shared lock.
Reviewed by: kib
Tested by: pho
Differential Revision: https://reviews.freebsd.org/D22116
Vast majority of uses the cache are just checking if there is an entry
present on process exit (and evicting it if so). Both checking and
eviction process are very expensive and put the lock protecting it high
up on the profile during poudriere -j 104.
Convert the linked list into a hash. This allows to almost always avoid
taking the lock in the first place (and consequently almost removes it
from the profile). Note only one lock is preserved as a split did not
meaningfully impact contention.
Should the cache be used for something it will still run into contention
issues. The code needs a rewrite, but should someone want to tidy it up
further the following can be done:
1) per-chain locks (or at least an array)
2) hashing by something else than just pid
Sponsored by: The FreeBSD Foundation
Make the nfsclient always call vnode_pager_setsize() with the vnode
exclusively locked. This ensures that page fault always can find the
backing page if the object size check succeeded. Set VV_VMSIZEVNLOCK
flag on NFS nodes.
The main offender breaking the interface in nfsclient is
nfs_loadattrcache(), which is used whenever server responded with
updated attributes, which can happen on non-changing operations as
well. Also, iod threads only have buffers locked (and even that is
LK_KERNPROC), but they still may call nfs_loadattrcache() on RPC
response.
Instead of immediately calling vnode_pager_setsize() if server
response indicated changed file size, but the vnode is not exclusively
locked, set a new node flag NVNSETSZSKIP. When the vnode exclusively
locked, or when we can temporary upgrade the lock to exclusive, call
vnode_pager_setsize(), by providing the nfsclient VOP_LOCK() implementation.
Tested by: pho
Discussed with: rmacklem
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
Differential revision: https://reviews.freebsd.org/D21883
Atomics are used for page busy and valid state when the shared busy is
held. The details of the locking protocol and valid and dirty
synchronization are in the updated vm_page.h comments.
Reviewed by: kib, markj
Tested by: pho
Sponsored by: Netflix, Intel
Differential Revision: https://reviews.freebsd.org/D21594
This is the first in a series of patches that promotes the page busy field
to a first class lock that no longer requires the object lock for
consistency.
Reviewed by: kib, markj
Tested by: pho
Sponsored by: Netflix, Intel
Differential Revision: https://reviews.freebsd.org/D21548
Node' cdp.si_name is the full path as provided by make_dev(9), it
should not be returned by VOP_VPTOCNP() when only the last component
is requested. Use the dirent entry instead.
With this note, handling of VDIR and VCHR nodes only differs in
handling of root vnode, which simplifies and unifies the logic.
Reported by: Li, Zhichao1 <Zhichao_Li1@Dell.com>
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
There are three more places in msdosfs_fat.c which might shift one
into the sign bit. While there, fix formatting of KASSERTs.
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
In case the implementation ever changes from using a chain of next pointers,
then changing the macro definition will be necessary, but changing all the
files that iterate over vm_map entries will not.
Drop a counter in vm_object.c that would have an effect only if the
vm_map entry count was wrong.
Discussed with: alc
Reviewed by: markj
Tested by: pho (earlier version)
Differential Revision: https://reviews.freebsd.org/D21882
vn_write already checks for vnode type to see if bwillwrite should be called.
This effectively reverts r244643.
Reviewed by: kib
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21905
During readdir() we guarantee that the tn_dir.tn_parent does not go
away, but it might be replaced by a parallel rename. Read tn_parent
only once, then use the cached value.
Reported and tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
To be consistent with replacing the mtx_lock()/mtx_unlock() calls on
the NFS node mutex (n_mtx) and ncl_iod_mutex, this patch replaces
all mtx_assert() calls on these mutexes with macros as well.
This will simplify changing these locks to sx locks in a future commit.
However, this change may be delayed indefinitely, since it appears there
is a deadlock when vnode_pager_setsize() is called to shrink the size
and the NFS node lock is held.
There is no semantic change as a result of this commit.
Suggested by: kib
MFC after: 1 week
Since the NFS node mutex needs to change to an sx lock so it can be held when
vnode_pager_setsize() is called and the iod lock is held when the NFS node lock
is acquired, the iod mutex will need to be changed to an sx lock as well.
To simply the future commit that changes both the NFS node lock and iod lock
to sx locks, this commit replaces all mtx_lock()/mtx_unlock() calls on the
iod lock with macros.
There is no semantic change as a result of this commit.
I don't know when the future commit will happen and be MFC'd, so I have
set the MFC on this commit to one week so that it can be MFC'd at the same
time.
Suggested by: kib
MFC after: 1 week
For a long time, some places in the NFS code have locked/unlocked the
NFS node lock with the macros NFSLOCKNODE()/NFSUNLOCKNODE() whereas
others have simply used mtx_lock()/mtx_unlock().
Since the NFS node mutex needs to change to an sx lock so it can be held when
vnode_pager_setsize() is called, replace all occurrences of mtx_lock/mtx_unlock
with the macros to simply making the change to an sx lock in future commit.
There is no semantic change as a result of this commit.
I am not sure if the change to an sx lock will be MFC'd soon, so I put
an MFC of 1 week on this commit so that it could be MFC'd with that commit.
Suggested by: kib
MFC after: 1 week
When a file is unlinked, the denode is not reclaimed until the last
reference is dropped, but the directory entry is immediately up for reuse.
This is a problem later when createde goes to grab a denode for the newly
created entry -- we search the hash and find a dead denode, then return that
without even bumping the reference count and the data later gets truncated
when the the last reference to the unlinked file is dropped.
This manifested itself as a broken in-place strip(1) on msdosfs. elfcopy
will do a sequence incredibly roughly like this:
open("/mnt/foo", ...) => fd 3
mmap()
unlink("/mnt/foo")
open("/mnt/foo", ...) => fd 4
write(4, ...)
close(4)
close(3)
and the resulting file would be truncated, but the write succeeded, as long
as a reference to the unlinked file had not been closed.
Some archaeology indicates that this bug has likely existed since msdosfs
was converted to use vfs_hash instead of a home rolled hash implementation
in r143570. Prior to that point, the hashget implementation would do a
refcnt check while searching and explicitly only return a denode with
de_refcnt != 0. vfs_hash did not yet have the callback that it does today,
so this slipped away and did not come back when it later grew that
functionality.
The comment indicating that we want to skip these denodes has been updated
to reflect where this is actually done. My repo-diving session seems to
indicate that the refcnt check was likely never actually below the comment,
to be pedantic, but instead a detail wrapped up in the hashget
implementation since the beginning of its inclusion into FreeBSD.
This bug was the cause behind the issue addressed in r352557.
Reported by: jhibbits
Reviewed by: kib
MFC after: 1 week
Differential Revision: https://reviews.freebsd.org/D21731
node lock when shrinking.
This is similar to r252528, applied to the above commit.
Apparently there is a race which makes necessary at least to keep the
n_size and pager size consistent when extending. Current suspect is
that iod threads perform vnode_pager_setsize() without taking the
vnode lock, which corrupts the file content.
Reported and tested by: Masachika ISHIZUKA <ish@amail.plala.or.jp>
Discussed with: rmacklem (related issues)
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
There are 3 counters modified all the time in this structure - one for
keeping the structure alive, one for preventing unmount and one for
tracking active writers. Exact values of these counters are very rarely
needed, which makes them a prime candidate for conversion to a per-cpu
scheme, resulting in much better performance.
Sample benchmark performing fstatfs (modifying 2 out of 3 counters) on
a 104-way 2 socket Skylake system:
before: 852393 ops/s
after: 76682077 ops/s
Reviewed by: kib, jeff
Tested by: pho
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21637
That revision addressed a Coverity CID that could lead to a buffer overflow,
but it had an off-by-one error in the buffer size check.
Reported by: Coverity
Coverity CID: 1405530
MFC after: 3 days
MFC-With: 351961
Sponsored by: The FreeBSD Foundation
* When unparenting a vnode, actually clear the flag. AFAIK this is basically
a no-op because we only unparent a vnode when reclaiming it or when
unlinking.
* There's no need to call fuse_vnode_setparent during reclaim, because we're
about to free the vnode data anyway.
Reviewed by: emaste
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21630
vnode_pager_setsize() under the node mutex.
r248567 moved some calls of vnode_pager_setsize() after the node lock
is unlocked, do the rest now.
Reported and tested by: peterj
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
of the reverse.
This fixes Linux sysctl(8) binary - it assumes the first two
directory entries are always "." and "..". There might be other
Linux apps affected by this.
NB it might be a good idea to rewrite it using queue(3).
Reviewed by: kib
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21550
Setting the B_INVALONERR flag before a synchronous write causes the buf
cache to forcibly invalidate contents if the write fails (BIO_ERROR).
This is intended to be used to allow layers above the buffer cache to make
more informed decisions about when discarding dirty buffers without
successful write is acceptable.
As a proof of concept, use in msdosfs to handle failures to mark the on-disk
'dirty' bit during rw mount or ro->rw update.
Extending this to other filesystems is left as future work.
PR: 210316
Reviewed by: kib (with objections)
Sponsored by: Dell EMC Isilon
Differential Revision: https://reviews.freebsd.org/D21539
When communicating with a FUSE server that implements version 7.8 (or older)
of the FUSE protocol, the FUSE_WRITE request structure is 16 bytes shorter
than normal. The protocol version check wasn't applied universally, leading
to an extra 16 bytes being sent to such servers. The extra bytes were
allocated and bzero()d, so there was no information disclosure.
Reviewed by: emaste
MFC after: 3 days
MFC-With: r350665
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21557
There are several mechanisms by which a vm_page reference is held,
preventing the page from being freed back to the page allocator. In
particular, holding the page's object lock is sufficient to prevent the
page from being freed; holding the busy lock or a wiring is sufficent as
well. These references are protected by the page lock, which must
therefore be acquired for many per-page operations. This results in
false sharing since the page locks are external to the vm_page
structures themselves and each lock protects multiple structures.
Transition to using an atomically updated per-page reference counter.
The object's reference is counted using a flag bit in the counter. A
second flag bit is used to atomically block new references via
pmap_extract_and_hold() while removing managed mappings of a page.
Thus, the reference count of a page is guaranteed not to increase if the
page is unbusied, unmapped, and the object's write lock is held. As
a consequence of this, the page lock no longer protects a page's
identity; operations which move pages between objects are now
synchronized solely by the objects' locks.
The vm_page_wire() and vm_page_unwire() KPIs are changed. The former
requires that either the object lock or the busy lock is held. The
latter no longer has a return value and may free the page if it releases
the last reference to that page. vm_page_unwire_noq() behaves the same
as before; the caller is responsible for checking its return value and
freeing or enqueuing the page as appropriate. vm_page_wire_mapped() is
introduced for use in pmap_extract_and_hold(). It fails if the page is
concurrently being unmapped, typically triggering a fallback to the
fault handler. vm_page_wire() no longer requires the page lock and
vm_page_unwire() now internally acquires the page lock when releasing
the last wiring of a page (since the page lock still protects a page's
queue state). In particular, synchronization details are no longer
leaked into the caller.
The change excises the page lock from several frequently executed code
paths. In particular, vm_object_terminate() no longer bounces between
page locks as it releases an object's pages, and direct I/O and
sendfile(SF_NOCACHE) completions no longer require the page lock. In
these latter cases we now get linear scalability in the common scenario
where different threads are operating on different files.
__FreeBSD_version is bumped. The DRM ports have been updated to
accomodate the KPI changes.
Reviewed by: jeff (earlier version)
Tested by: gallatin (earlier version), pho
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D20486
CID 1404532 fixes a signed vs unsigned comparison error in fuse_vnop_bmap.
It could potentially have resulted in VOP_BMAP reporting too many
consecutive blocks.
CID 1404364 is much worse. It was an array access by an untrusted,
user-provided variable. It could potentially have resulted in a malicious
file system crashing the kernel or worse.
Reported by: Coverity
Reviewed by: emaste
MFC after: 3 days
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21466
The "nd" argument for nfsrv_proxyds() is no longer used by the function.
This patch deletes it. This allows a subsequent patch to delete the "nd"
argument from nfsvno_getattr(), since it's only use of "nd" was to pass it
to nfsrv_proxyds().
Getting rid of the "nd" argument from nfsvno_getattr() avoids confusion
over why it might need "nd".
This patch is trivial and does not have any semantic effect.
These were fully neutered in r177676 (2008), but not removed at the time for
unclear reasons. They're totally dead code, so go ahead and yank them now.
No functional change.
After r294954, it is an invariant that bread returns non-NULL bp if and only
if the routine succeeded. On error, it handles any buffer cleanup
internally. So the brelse(NULL) here was just redundant.
No functional change.
Discussed with: kib (extracted from a larger differential)
The "nd" argument for nfsrv_checkdsattr() is no longer used by the function.
This patch deletes it. This allows subsequent patches to delete the "nd"
argument from nfsrv_proxyds(), since it's only use of "nd" was to pass it
to nfsrv_checkdsattr(). The same will then be true for nfsvno_getattr(),
which passes "nd" to nfsrv_proxyds().
Getting rid of the "nd" argument from nfsvno_getattr() avoids confusion
over why it might need "nd".
This patch is trivial and does not have any semantic effect.
Found by inspection while working on the NFSv4.2 server.
Specifically, the following was broken:
$ mount -t procfs procfs /proc
$ ls -l /proc
r351741 reworked readdir slightly to avoid pfs_node/pidhash LOR, but
inadvertently regressed pid == NO_PID; new pfs_lookup_proc() fails for the
obvious reasons, and later pfs_visible_proc doesn't capture the
pid == NO_PID -> return 1 aspect of pfs_visible. We can infact skip this
whole block if we're operating on a directory w/ NO_PID, as it's always
visible.
Reported by: trasz
Reviewed by: mjg
Differential Revision: https://reviews.freebsd.org/D21518
Similarly to the other routine stop taking the interlock for the lower
vnode. The interlock for nullfs vnode is still taken to ensure
stability of ->v_data.
Reviewed by: kib
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21480
Current implementation of vnode_create_vobject() and
vnode_destroy_vobject() is written so that it prepared to handle the
vm object destruction for live vnode. Practically, no filesystems use
this, except for some remnants that were present in UFS till today.
One of the consequences of that model is that each filesystem must
call vnode_destroy_vobject() in VOP_RECLAIM() or earlier, as result
all of them get rid of the v_object in reclaim.
Move the call to vnode_destroy_vobject() to vgonel() before
VOP_RECLAIM(). This makes v_object stable: either the object is NULL,
or it is valid vm object till the vnode reclamation. Remove code from
vnode_create_vobject() to handle races with the parallel destruction.
Reviewed by: markj
Tested by: pho
Sponsored by: The FreeBSD Foundation
Differential revision: https://reviews.freebsd.org/D21412
vnode usecount drops to 0 all the time (e.g. for directories during path lookup).
When that happens the kernel would always lock the exclusive lock for the vnode
in order to call vinactive(). This blocks other threads who want to use the vnode
for looukp.
vinactive is very rarely needed and can be tested for without the vnode lock held.
This patch gives filesytems an opportunity to do it, sample total wait time for
tmpfs over 500 minutes of poudriere -j 104:
before: 557563641706 (lockmgr:tmpfs)
after: 46309603301 (lockmgr:tmpfs)
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21371
* A small error in r338152 let to the returned size always being exactly
eight bytes too large.
* The FUSE_LISTXATTR operation works like Linux's listxattr(2): if the
caller does not provide enough space, then the server should return ERANGE
rather than return a truncated list. That's true even though in FUSE's
case the kernel doesn't provide space to the client at all; it simply
requests a maximum size for the list. We previously weren't handling the
case where the server returns ERANGE even though the kernel requested as
much size as the server had told us it needs; that can happen due to a
race.
* We also need to ensure that a pathological server that always returns
ERANGE no matter what size we request in FUSE_LISTXATTR won't cause an
infinite loop in the kernel. As of this commit, it will instead cause an
infinite loop that exits and enters the kernel on each iteration, allowing
signals to be processed.
Reviewed by: cem
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21287
This is part of the preparation to remove flags argument from VOP_UNLOCK.
Also has a side effect of fixing stacking on top of nullfs broken by r351472.
Reported by: cy
Sponsored by: The FreeBSD Foundation
Some places only take the interlock to hold the vnode, which was a requiremnt
before they started being manipulated with atomics. Use the newly introduced
vholdnz to bump the count.
Reviewed by: kib
Tested by: pho
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21358
Use pointer arithmetic (as now done in makefs, and in NetBSD) instead of
taking the address of array element. No functional change, but this
makes it easier to compare different versions of this file.
Reviewed by: kib
MFC after: 1 week
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21365
null_nodeget which follows almost always finds the target vnode in the hash,
avoiding insmntque1 altogether. Should it be needed, it already checks if the
lock needs to be upgraded.
Reviewed by: kib
Tested by: pho
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D20244
There is no need to duplicate this file when it can be trivially
shared (just exposing sections previously under #ifdef _KERNEL).
MFC with: r351273
Differential Revision: The FreeBSD Foundation
There is no reason to duplicate this file when it can be trivially
shared (just exposing one section previously under #ifdef _KERNEL).
Reviewed by: imp, cem
MFC with: r351273
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21346
Suppose that a binary was executed from tmpfs mount, and the text
vnode was reclaimed while the binary was still running. It is
possible during even the normal operations since tmpfs vnode'
vm_object has swap type, and no references on the vnode is held. Also
assume that the text vnode was revived for some reason. Then, on the
process exit or exec, unmapping of the text mapping tries to remove
the text reference from the vnode, but since it went from
recycle/instantiation cycle, there is no reference kept, and assertion
in VOP_UNSET_TEXT_CHECKED() triggers.
Fix this by keeping a use reference on the tmpfs vnode for each exec
reference. This prevents the vnode reclamation while executable map
entry is active.
Do it by adding per-mount flag MNTK_TEXT_REFS that directs
vop_stdset_text() to add use ref on first vnode text use, and
per-vnode VI_TEXT_REF flag, to record the need on unref in
vop_stdunset_text() on last vnode text use going away. Set
MNTK_TEXT_REFS for tmpfs mounts.
Reported by: bdrewery
Tested by: sbruno, pho (previous version)
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
The FUSE_LISTXATTR operation always returns the full list of a file's
extended attributes, in all namespaces. There's no way to filter the list
server-side. However, currently FreeBSD's fusefs driver sends a namespace
string with the FUSE_LISTXATTR request. That behavior was probably copied
from fuse_vnop_getextattr, which has an attribute name argument. It's
been there ever since extended attribute support was added in r324620. This
commit removes it.
Reviewed by: cem
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21280
In FUSE protocol 7.9, the size of the FUSE_GETATTR request has increased.
However, the fusefs driver is currently not sending the additional fields.
In our implementation, the additional fields are always zero, so I there
haven't been any test failures until now. But fusefs-lkl requires the
request's length to be correct.
Fix this bug, and also enhance the test suite to catch similar bugs.
PR: 239830
MFC after: 2 weeks
MFC-With: 350665
Sponsored by: The FreeBSD Foundation
This commit imports the new fusefs driver. It raises the protocol level
from 7.8 to 7.23, fixes many bugs, adds a test suite for the driver, and
adds many new features. New features include:
* Optional kernel-side permissions checks (-o default_permissions)
* Implement VOP_MKNOD, VOP_BMAP, and VOP_ADVLOCK
* Allow interrupting FUSE operations
* Support named pipes and unix-domain sockets in fusefs file systems
* Forward UTIME_NOW during utimensat(2) to the daemon
* kqueue support for /dev/fuse
* Allow updating mounts with "mount -u"
* Allow exporting fusefs file systems over NFS
* Server-initiated invalidation of the name cache or data cache
* Respect RLIMIT_FSIZE
* Try to support servers as old as protocol 7.4
Performance enhancements include:
* Implement FUSE's FOPEN_KEEP_CACHE and FUSE_ASYNC_READ flags
* Cache file attributes
* Cache lookup entries, both positive and negative
* Server-selectable cache modes: writethrough, writeback, or uncached
* Write clustering
* Readahead
* Use counter(9) for statistical reporting
PR: 199934 216391 233783 234581 235773 235774 235775
PR: 236226 236231 236236 236291 236329 236381 236405
PR: 236327 236466 236472 236473 236474 236530 236557
PR: 236560 236844 237052 237181 237588 238565
Reviewed by: bcr (man pages)
Reviewed by: cem, ngie, rpokala, glebius, kib, bde, emaste (post-commit
review on project branch)
MFC after: 3 weeks
Relnotes: yes
Sponsored by: The FreeBSD Foundation
Pull Request: https://reviews.freebsd.org/D21110
- Provide unionfs_add_writecount() which passes the writecount to the
lower or upper vnode as appropriate.
- In unionfs VOP_RECLAIM() implementation, annulate unionfs
writecounts from upper or lower vnode. It is not clear that it is
always correct to remove the all references from either lower or
upper vnode, but we currently do not track which vnode get how many
refs anyway.
Reported and tested by: t_uemura@macome.co.jp
MFC after: 1 week
Sponsored by: The FreeBSD Foundation
When a fusefs file system is mounted using the writeback cache, the cache
may still be bypassed by opening a file with O_DIRECT. When writing with
O_DIRECT, the cache must be invalidated for the affected portion of the
file. Fix some panics caused by inadvertently invalidating too much.
Sponsored by: The FreeBSD Foundation
FUSE file systems can optionally support interrupting outstanding
operations. However, the file system does not identify to the kernel at
mount time whether it's capable of doing that. Instead it signals its
noncapability by returning ENOSYS to the first FUSE_INTERRUPT operation it
receives. That's a problem for reliable signal delivery, because the kernel
must choose which thread should get a signal before it knows whether the
FUSE server can handle interrupts. The problem is even worse because the
FUSE protocol allows a file system to simply ignore all FUSE_INTERRUPT
operations.
Fix the signal delivery logic by making interruptibility an opt-in mount
option. This will require a corresponding change to libfuse, but not to
most file systems that link to libfuse.
Bump __FreeBSD_version due to the new mount option.
Sponsored by: The FreeBSD Foundation
fticket_wait_answer would spin if it received an unhandled signal whose
default disposition is to terminate. The reason is because msleep(9) would
return EINTR even for a masked signal. One reason is when the thread is
stopped, which happens for example during sigexit(). Fix this bug by
returning immediately if fticket_wait_answer ever gets interrupted a second
time, for any reason.
Sponsored by: The FreeBSD Foundation
1) Don't explicitly not mask SIGKILL. kern_sigprocmask won't allow it to be
masked, anyway.
2) Fix an infinite loop bug. If a process received both a maskable signal
lower than 9 (like SIGINT) and then received SIGKILL,
fticket_wait_answer would spin. msleep would immediately return EINTR,
but cursig would return SIGINT, so the sleep would get retried. Fix it
by explicitly checking whether SIGKILL has been received.
3) Abandon the sig_isfatal optimization introduced by r346357. That
optimization would cause fticket_wait_answer to return immediately,
without waiting for a response from the server, if the process were going
to exit anyway. However, it's vulnerable to a race:
1) fatal signal is received while fticket_wait_answer is sleeping.
2) fticket_wait_answer sends the FUSE_INTERRUPT operation.
3) fticket_wait_answer determines that the signal was fatal and returns
without waiting for a response.
4) Another thread changes the signal to non-fatal.
5) The first thread returns to userspace. Instead of exiting, the
process continues.
6) The application receives EINTR, wrongly believes that the operation
was successfully interrupted, and restarts it. This could cause
problems for non-idempotent operations like FUSE_RENAME.
Reported by: kib (the race part)
Sponsored by: The FreeBSD Foundation
* Fix the kernel build with gcc by removing a redundant extern declaration
* In the tests, fix a printf format specifier that assumed LP64
Sponsored by: The FreeBSD Foundation
These fields will not be equal only in case if bigalloc filesystem feature is turned on.
This feature is not supported for now.
Reported by: Christopher Krah, Thomas Barabosch, and Jan-Niclas Hilgert of Fraunhofer FKIE
Reported as: FS-27-EXT2-12: Denial of Service in openat-0 (vm_fault_hold/ext2_clusteracct)
MFC after: 2 weeks
The ext2fs fragments are different from ufs fragments.
In case of ext2fs the fragment should be equal or more then block size.
The values more than block size are used only in case of bigalloc feature, which is does not supported for now.
Reported by: Christopher Krah, Thomas Barabosch, and Jan-Niclas Hilgert of Fraunhofer FKIE
Reported as: FS-22-EXT2-9: Denial of service in ftruncate-0 (ext2_balloc)
MFC after: 2 weeks
Reported by: Christopher Krah, Thomas Barabosch, and Jan-Niclas Hilgert of Fraunhofer FKIE
Reported as: FS-11-EXT2-6: Denial Of Service in write-1 (ext2_balloc)
MFC after: 2 weeks
We were leaking the fuse ticket if the original operation completed before
the daemon received the INTERRUPT operation. Fixing this was easier than I
expected.
Sponsored by: The FreeBSD Foundation
Previously fusefs would never recycle vnodes. After VOP_INACTIVE, they'd
linger around until unmount or the vnlru reclaimed them. This commit
essentially actives and inlines the old reclaim_revoked sysctl, and fixes
some issues dealing with the attribute cache and multiply linked files.
Sponsored by: The FreeBSD Foundation
counter(9) is more performant than using atomic instructions to update
sysctls that just report statistics to userland.
Sponsored by: The FreeBSD Foundation
Fix memory leaks relating to FUSE_BMAP and FUSE_CREATE. There are still
leaks relating to FUSE_INTERRUPT, but they'll be harder to fix since the
server is legally allowed to never respond to a FUSE_INTERRUPT operation.
Sponsored by: The FreeBSD Foundation
As of protocol 7.23, fuse file systems can specify their cache behavior on a
per-mountpoint basis. If they set FUSE_WRITEBACK_CACHE in
fuse_init_out.flags, then they'll get the writeback cache. If not, then
they'll get the writethrough cache. If they set FOPEN_DIRECT_IO in every
FUSE_OPEN response, then they'll get no cache at all.
The old vfs.fusefs.data_cache_mode sysctl is ignored for servers that use
protocol 7.23 or later. However, it's retained for older servers,
especially for those running in jails that lack access to the new protocol.
This commit also fixes two other minor test bugs:
* WriteCluster:SetUp was using an uninitialized variable.
* Read.direct_io_pread wasn't verifying that the cache was actually
bypassed.
Sponsored by: The FreeBSD Foundation
The fusefs kernel module allegedly supported no_attrcache, no_readahed,
no_datacache, no_namecache, and no_mmap mount options, but the mount_fusefs
binary never did. So there was no way to ever activate these options.
Delete them. Some of them have alternatives:
no_attrcache: set the attr_valid time to 0 in FUSE_LOOKUP and FUSE_GETATTR
responses.
no_readahed: set max_readahead to 0 in the FUSE_INIT response.
no_datacache: set the vfs.fusefs.data_cache_mode sysctl to 0, or (coming
soon) set the attr_valid time to 0 and set FUSE_AUTO_INVAL_DATA in
the FUSE_INIT response.
no_namecache: set entry_valid time to 0 in FUSE_LOOKUP and FUSE_GETATTR
responses.
Sponsored by: The FreeBSD Foundation
If a server supports a timestamp granularity other than 1ns, it can tell the
client this as of protocol 7.23. The client will use that granularity when
updating its cached timestamps during write. This way the timestamps won't
appear to change following flush.
Sponsored by: The FreeBSD Foundation
As of r349396 the kernel will internally update the mtime and ctime of files
on write. It will also flush the mtime should a SETATTR happen before the
data cache gets flushed. Now it will flush the ctime too, if the server is
using protocol 7.23 or higher.
This is the only case in which the kernel will explicitly set a file's
ctime, since neither utimensat(2) nor any other user interfaces allow it.
Sponsored by: The FreeBSD Foundation
Writing should implicitly update a file's mtime and ctime. For fuse, the
server is supposed to do that. But the client needs to do it too, because
the FUSE_WRITE response does not include time attributes, and it's not
desirable to issue a GETATTR after every WRITE. When using the writeback
cache, there's another hitch: the kernel should ignore the mtime and ctime
fields in any GETATTR response for files with a dirty write cache.
Sponsored by: The FreeBSD Foundation
Writes that extend a file should update the file's size. r344185 restricted
that behavior for fusefs to only happen when the data cache was enabled.
That probably made sense at the time because the attribute cache wasn't
fully baked yet. Now that it is, we should always update the cached file
size during write.
Sponsored by: The FreeBSD Foundation
Use the standard facilities for getpages and putpages instead of bespoke
implementations that don't work well with the writeback cache. This has
several corollaries:
* Change the way we handle short reads _again_. vfs_bio_getpages doesn't
provide any way to handle unexpected short reads. Plus, I found some more
lock-order problems. So now when the short read is detected we'll just
clear the vnode's attribute cache, forcing the file size to be requeried
the next time it's needed. VOP_GETPAGES doesn't have any way to indicate
a short read to the "caller", so we just bzero the rest of the page
whenever a short read happens.
* Change the way we decide when to set the FUSE_WRITE_CACHE bit. We now set
it for clustered writes even when the writeback cache is not in use.
Sponsored by: The FreeBSD Foundation
The final server unref should be done by the server thread to prevent
deadlock in the client cdevpriv destructor, which cannot destroy
itself.
MFC after: 1 week
Sponsored by: Mellanox Technologies
Remove a lingering use of splbio().
The buffer must be locked by the caller. No functional change
intended.
Reviewed by: kib
MFC after: 1 week
Sponsored by: The FreeBSD Foundation
NANDFS has been broken for years. Remove it. The NAND drivers that
remain are for ancient parts that are no longer relevant. They are
polled, have terrible performance and just for ancient arm
hardware. NAND parts have evolved significantly from this early work
and little to none of it would be relevant should someone need to
update to support raw nand. This code has been off by default for
years and has violated the vnode protocol leading to panics since it
was committed.
Numerous posts to arch@ and other locations have found no actual users
for this software.
Relnotes: Yes
No Objection From: arch@
Differential Revision: https://reviews.freebsd.org/D20745
b_fsprivate1 needs to be initialized even for write operations, probably
because a buffer can be used to read, write, and read again with the final
read serviced by cache.
Sponsored by: The FreeBSD Foundation
VOP_GETPAGES intentionally tries to read beyond EOF, so fuse_read_biobackend
can't rely on bp->b_resid > 0 indicating a short read. And adjusting
bp->b_count after a short read seems to cause some sort of resource leak.
Instead, store the shortfall in the bp->b_fsprivate1 field.
Sponsored by: The FreeBSD Foundation
Even if a short read is caused by EOF, it's still necessary to bzero the
remaining buffer, because that buffer could become valid as a result of a
future ftruncate or pwrite operation.
Reported by: fsx
Sponsored by: The FreeBSD Foundation
A fuse server may return a short read for three reasons:
* The file is opened with FOPEN_DIRECT_IO. In this case, the short read
should be returned directly to userland. We already handled this case
correctly.
* The file was truncated server-side, and the read hit EOF. In this case,
the kernel should update the file size. Fixed in the case of VOP_READ.
Fixing this for VOP_GETPAGES is TODO.
* The file is opened in writeback mode, there are dirty buffers past what
the server thinks is the file's EOF, and the read hit what the server
thinks is the file's EOF. In this case, the client is trying to read a
hole, and should zero-fill it. We already handled this case, and I added
a test for it.
Sponsored by: The FreeBSD Foundation
None of the new features are implemented yet. This commit just adds the new
protocol definitions and adds backwards-compatibility code for pre 7.23
servers.
Sponsored by: The FreeBSD Foundation
Jumping from protocol 7.15 to 7.21 adds several new features. While they're
all potentially useful, they're also all optional, and I'm not implementing
any right now because my highest priority lies in a later version.
Sponsored by: The FreeBSD Foundation
fuse_kernel.h is based on Linux's fuse.h. In r349250 I modified
fuse_kernel.h by generating a diff of two versions of Linux's fuse.h and
applying it to our tree. patch succeeded, but it put one chunk in the wrong
location. This commit fixes that. No functional changes.
Sponsored by: The FreeBSD Foundation
This protocol level adds two new features: the ability for the server to
store or retrieve data into/from the client's cache. But the messages
aren't defined soundly since they identify the file only by its inode,
without the generation number. So it's possible for them to modify the
wrong file's cache. Also, I don't know of any file systems in ports that
use these messages. So I'm not implementing them. I did add a (disabled)
test for the store message, however.
Sponsored by: The FreeBSD Foundation
VOP_READ and VOP_WRITE take the seqcount in blocks in a 16-bit field.
However, fcntl allows you to set the seqcount in bytes to any nonnegative
31-bit value. The result can be a 16-bit overflow, which will be
sign-extended in functions like ffs_read. Fix this by sanitizing the
argument in kern_fcntl. As a matter of policy, limit to IO_SEQMAX rather
than INT16_MAX.
Also, fifos have overloaded the f_seqcount field for a completely different
purpose ever since r238936. Formalize that by using a union type.
Reviewed by: cem
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D20710
Previously we allowed servers as old as 7.1 to connect (there never was a
7.0). However, we wrongly assumed a few things about protocols older than
7.8. This commit attempts to support servers as old as 7.4 but no older. I
added no new tests because I'm not sure there actually _are_ any servers
this old in the wild.
Sponsored by: The FreeBSD Foundation
This protocol version adds one new feature: the ability for the server to
set the maximum number of background requests and a "congestion threshold"
with ill-defined properties. I don't know of any fuse file systems in ports
that use this feature, so I'm not implementing it.
Sponsored by: The FreeBSD Foundation
If the fuse daemon supports FUSE_BMAP, then use that for the block mapping.
Otherwise, use the same technique used by vop_stdbmap. Report large values
for runp and runb in order to maximize read clustering and minimize upcalls,
even if we don't know the true layout.
The major result of this change is that sequential reads to FUSE files will
now usually happen 128KB at a time instead of 64KB.
Sponsored by: The FreeBSD Foundation
* Don't always write the last page synchronously. That's not actually
required. It was probably just masking another bug that I fixed later,
possibly in r349021.
* Enable the NotifyWriteback tests now that Writeback cache is working.
* Add a test to ensure that the write cache isn't flushed synchronously when
in writeback mode.
Sponsored by: The FreeBSD Foundation
fusefs will now use cluster_read. This allows readahead of more than one
cache block. However, it won't yet actually cluster the reads because that
requires VOP_BMAP, which fusefs does not yet implement.
Sponsored by: The FreeBSD Foundation
rename the source to gsb_crc32.c.
This is a prerequisite of unifying kernel zlib instances.
PR: 229763
Submitted by: Yoshihiro Ota <ota at j.email.ne.jp>
Differential Revision: https://reviews.freebsd.org/D20193
fusefs will now read ahead at most one cache block at a time (usually 64
KB). Clustered reads are still TODO. Individual file systems may disable
read ahead by setting fuse_init_out.max_readahead=0 during initialization.
Sponsored by: The FreeBSD Foundation
Our fusefs(5) module supports three cache modes: uncached, write-through,
and write-back. However, the write-through mode (which is the default) has
never actually worked as its name suggests. Rather, it's always been more
like "write-around". It wrote directly, bypassing the cache. The cache
would only be populated by a subsequent read of the same data.
This commit fixes that problem. Now the write-through mode works as one
would expect: write(2) immediately adds data to the cache and then blocks
while the daemon processes the write operation.
A side effect of this change is that non-cache-block-aligned writes will now
incur a read-modify-write cycle of the cache block. The old behavior
(bypassing write cache entirely) can still be achieved by opening a file
with O_DIRECT.
PR: 237588
Sponsored by: The FreeBSD Foundation
Enable write clustering in fusefs whenever cache mode is set to writeback
and the "async" mount option is used. With default values for MAXPHYS,
DFLTPHYS, and the fuse max_write mount parameter, that means sequential
writes will now be written 128KB at a time instead of 64KB.
Also, add a regression test for PR 238565, a panic during unmount that
probably affects UFS, ext2, and msdosfs as well as fusefs.
PR: 238565
Sponsored by: The FreeBSD Foundation
An errant vfs_bio_clrbuf snuck in in r348931. Surprisingly, it doesn't have
any effect most of the time. But under some circumstances it cause the
buffer to behave in a write-only fashion.
Sponsored by: The FreeBSD Foundation
The current "writeback" cache mode, selected by the
vfs.fusefs.data_cache_mode sysctl, doesn't do writeback cacheing at all. It
merely goes through the motions of using buf(9), but then writes every
buffer synchronously. This commit:
* Enables delayed writes when the sysctl is set to writeback cacheing
* Fixes a cache-coherency problem when extending a file whose last page has
just been written.
* Removes the "sync" mount option, which had been set unconditionally.
* Adjusts some SDT probes
* Adds several new tests that mimic what fsx does but with more control and
without a real file system. As I discover failures with fsx, I add
regression tests to this file.
* Adds a test that ensures we can append to a file without reading any data
from it.
This change is still incomplete. Clustered writing is not yet supported,
and there are frequent "panic: vm_fault_hold: fault on nofault entry" panics
that I need to fix.
Sponsored by: The FreeBSD Foundation
fusefs's I/O methods were originally copy/pasted from nfsclient. This
commit removes some irrelevant parts, like stuff involving B_NEEDCOMMIT.
Sponsored by: The FreeBSD Foundation
Both filesystems do no use vnode_pager_dealloc() which would handle
this case otherwise. Nullfs because vnode vm_object handle never
points to nullfs vnode. Tmpfs because its vm_object is never vnode
object at all.
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
In r348560 I thought that FUSE_EXPORT_SUPPORT was required for cases where
the node to be invalidated (or the parent of the entry to be invalidated)
wasn't cached. But I realize now that that's not the case. During entry
invalidation, if the parent isn't in the vfs hash table, then it must've
been reclaimed. And since fuse_vnop_reclaim does a cache_purge, that means
the entry to be invalidated has already been removed from the namecache.
And during inode invalidation, if the inode to be invalidated isn't in the
vfs hash table, then it too must've been reclaimed. In that case it will
have no buffer cache to invalidate.
Sponsored by: The FreeBSD Foundation
Protocol 7.12 adds a way for the server to notify the client that it should
invalidate an inode's data cache and/or attributes. This commit implements
that mechanism. Unlike Linux's implementation, ours requires that the file
system also supports FUSE_EXPORT_SUPPORT (NFS-style lookups). Otherwise the
invalidation operation will return EINVAL.
Sponsored by: The FreeBSD Foundation
Protocol 7.12 adds a way for the server to notify the client that it should
invalidate an entry from its name cache. This commit implements that
mechanism.
Sponsored by: The FreeBSD Foundation
FUSE allows entries to be cached for a limited amount of time. fusefs's
vnop_lookup method already implements that using the timeout functionality
of cache_lookup/cache_enter_time. However, lookups for the NFS server go
through a separate path: vfs_vget. That path can't use the same timeout
functionality because cache_lookup/cache_enter_time only work on pathnames,
whereas vfs_vget works by inode number.
This commit adds entry timeout information to the fuse vnode structure, and
checks it during vfs_vget. This allows the NFS server to take advantage of
cached entries. It's also the same path that FUSE's asynchronous cache
invalidation operations will use.
Sponsored by: The FreeBSD Foundation
