future further optimizations where the vm_object lock will be held
in read mode most of the time the page cache resident pool of pages
are accessed for reading purposes.
The change is mostly mechanical but few notes are reported:
* The KPI changes as follow:
- VM_OBJECT_LOCK() -> VM_OBJECT_WLOCK()
- VM_OBJECT_TRYLOCK() -> VM_OBJECT_TRYWLOCK()
- VM_OBJECT_UNLOCK() -> VM_OBJECT_WUNLOCK()
- VM_OBJECT_LOCK_ASSERT(MA_OWNED) -> VM_OBJECT_ASSERT_WLOCKED()
(in order to avoid visibility of implementation details)
- The read-mode operations are added:
VM_OBJECT_RLOCK(), VM_OBJECT_TRYRLOCK(), VM_OBJECT_RUNLOCK(),
VM_OBJECT_ASSERT_RLOCKED(), VM_OBJECT_ASSERT_LOCKED()
* The vm/vm_pager.h namespace pollution avoidance (forcing requiring
sys/mutex.h in consumers directly to cater its inlining functions
using VM_OBJECT_LOCK()) imposes that all the vm/vm_pager.h
consumers now must include also sys/rwlock.h.
* zfs requires a quite convoluted fix to include FreeBSD rwlocks into
the compat layer because the name clash between FreeBSD and solaris
versions must be avoided.
At this purpose zfs redefines the vm_object locking functions
directly, isolating the FreeBSD components in specific compat stubs.
The KPI results heavilly broken by this commit. Thirdy part ports must
be updated accordingly (I can think off-hand of VirtualBox, for example).
Sponsored by: EMC / Isilon storage division
Reviewed by: jeff
Reviewed by: pjd (ZFS specific review)
Discussed with: alc
Tested by: pho
In particular, do not lock Giant conditionally when calling into the
filesystem module, remove the VFS_LOCK_GIANT() and related
macros. Stop handling buffers belonging to non-mpsafe filesystems.
The VFS_VERSION is bumped to indicate the interface change which does
not result in the interface signatures changes.
Conducted and reviewed by: attilio
Tested by: pho
If you have a binary on a filesystem which is also mounted over by
nullfs, you could execute the binary from the lower filesystem, or
from the nullfs mount. When executed from lower filesystem, the lower
vnode gets VV_TEXT flag set, and the file cannot be modified while the
binary is active. But, if executed as the nullfs alias, only the
nullfs vnode gets VV_TEXT set, and you still can open the lower vnode
for write.
Add a set of VOPs for the VV_TEXT query, set and clear operations,
which are correctly bypassed to lower vnode.
Tested by: pho (previous version)
MFC after: 2 weeks
the full length from offset is being flushed. Note that for now VOP_FSYNC
does not support offset and length parameters so we still do the same full
VOP_FSYNC. This issue was reported at FreeNAS support site as FreeNAS
ticket #1096.
Submitted by: "ceckerle" <ce.freenas eckerle net>
Prodded by: gcooper
Reviewed by: rmacklem
MFC after: 2 weeks
NFS server and reuse it for writes as well to allow writes to the backing
store to be clustered.
- Use a prime number for the size of the heuristic table (1017 is not
prime).
- Move the logic to locate a heuristic entry from the table and compute
the sequential count out of VOP_READ() and into a separate routine.
- Use the logic from sequential_heuristic() in vfs_vnops.c to update the
seqcount when a sequential access is performed rather than just
increasing seqcount by 1. This lets the clustering count ramp up
faster.
- Allow for some reordering of RPCs and if it is detected leave the current
seqcount as-is rather than dropping back to a seqcount of 1. Also,
when out of order access is encountered, cut seqcount in half rather than
dropping it all the way back to 1 to further aid with reordering.
- Fix the new NFS server to properly update the next offset after a
successful VOP_READ() so that the readahead actually works.
Some of these changes came from an earlier patch by Bjorn Gronwall that was
forwarded to me by bde@.
Discussed with: bde, rmacklem, fs@
Submitted by: Bjorn Gronwall (1, 4)
MFC after: 2 weeks
which requires that ni_strictrelative be set to 0, post-r224810.
Tested by: swills (earlier version), geo dot liaskos at gmail.com
Approved by: re (kib)
PMC/SYSV/...).
No FreeBSD version bump, the userland application to query the features will
be committed last and can serve as an indication of the availablility if
needed.
Sponsored by: Google Summer of Code 2010
Submitted by: kibab
Reviewed by: arch@ (parts by rwatson, trasz, jhb)
X-MFC after: to be determined in last commit with code from this project
incorrectly calling vm_object_page_clean(). They are passing the length of
the range rather than the ending offset of the range.
Perform the OFF_TO_IDX() conversion in vm_object_page_clean() rather than the
callers.
Reviewed by: kib
MFC after: 3 weeks
so turn off UFS-specific hack that assumes so in ZFS case.
Before the change we can miss returning some directory entries to a
NFS client.
I believe that the hack should be moved to ufs_readdir(), but until we find
somebody who will do it, turn it off for ZFS in NFS server code.
Submitted by: rmacklem
Discussed with: rmacklem, mckusick
MFC after: 3 days
busy. This fixes a race where we can pass invalid mount point to VFS_VGET()
via vp->v_mount when exported file system was forcibly unmounted between
nfsrv_fhtovp() and VFS_VGET().
Reviewed by: kib
MFC after: 5 days
this is the only place that is different from all the other nfsrv_fhtovp()
consumers.
This simplifies nfsrv_fhtovp() a bit and also eliminates one
vn_lock/VOP_UNLOCK() cycle in case of NFSv3.
- Implement NFSRV_FLAG_BUSY flag for nfsrv_fhtovp() that tells it to leave
mount point busy.
Reviewed by: kib
MFC after: 5 days
only for !usevget case. If VFS_VGET is working, the vnode shared lock
is obtained recursively and vput() shall be done, not vunref().
Submitted by: rmacklem
Tested by: Josh Carroll <josh.carroll gmail com>
MFC after: 3 days
not support VFS_VGET, like msdosfs, do not call VOP_LOOKUP() for
dotdot on the root directory. Our filesystems expect that VFS handles
dotdot lookups on root on its own.
Reported and tested by: kevlo
MFC after: 2 weeks
- We are fine by only share-locking the vnode.
- Remove assertion that doesn't hold for ZFS where we cross mount points
boundaries by going into .zfs/snapshot/<name>/.
Reviewed by: rmacklem
MFC after: 1 month
Previously the failing operation would allocate an mbuf and construct an
error reply, but because the function did not return 0, the NFS server
assumed it had failed to generate a reply and would leak the reply mbuf as
well as not sending the reply to the NFS client.
PR: kern/140853
Submitted by: Ted Faber faber at isi edu (remove)
Reviewed by: rmacklem (remove)
MFC after: 1 week
This completes the fix from r185586.
PR: kern/139059
Reported by: Daniel Braniss <danny@cs.huji.ac.il>
Submitted by: Jaakko Heinonen <jh@saunalahti.fi>
Tested by: Daniel Braniss <danny@cs.huji.ac.il>
MFC after: 3 days
the VFS. Now all the VFS_* functions and relating parts don't want the
context as long as it always refers to curthread.
In some points, in particular when dealing with VOPs and functions living
in the same namespace (eg. vflush) which still need to be converted,
pass curthread explicitly in order to retain the old behaviour.
Such loose ends will be fixed ASAP.
While here fix a bug: now, UFS_EXTATTR can be compiled alone without the
UFS_EXTATTR_AUTOSTART option.
VFS KPI is heavilly changed by this commit so thirdy parts modules needs
to be recompiled. Bump __FreeBSD_version in order to signal such
situation.
validation code on ZFS.
Problem: when opening file with O_CREAT|O_EXCL NFS has to jump through
extra hoops to ensure O_EXCL semantics. Namely, client supplies of 8
bytes (NFSX_V3CREATEVERF) bytes of verification data to uniquely
identify this create request. Server then creates a new file with access
mode 0, copies received 8 bytes into va_atime member of struct vattr and
attempt to set the atime on file using VOP_SETATTR. If that succeeds, it
fetches file attributes with VOP_GETATTR and verifies that atime
timestamps match. If timestamps do not match, NFS server concludes it
has probbaly lost the race to another process creating the file with the
same name and bails with EEXIST.
This scheme works OK when exported FS is FFS, but if underlying
filesystem is ZFS _and_ server is running 64bit kernel, it breaks down
due to sanity checking in zfs_setattr function, which refuses to accept
any timestamps which have tv_sec that cannot be represented as 32bit
int. Since struct timespec fields are 64 bit integers on 64bit platforms
and server just copies NFSX_V3CREATEVERF bytes info va_atime, all eight
bytes supplied by client end up in va_atime.tv_sec, forcing it out of
valid 32bit range.
The solution this change implements is simple: it treats
NFSX_V3CREATEVERF as two 32bit integers and unpacks them separately into
va_atime.tv_sec and va_atime.tv_nsec respectively, thus guaranteeing
that tv_sec remains in 32 bit range and ZFS remains happy.
Reviewed by: kib
and server. This replaces the RPC implementation of the NFS client and
server with the newer RPC implementation originally developed
(actually ported from the userland sunrpc code) to support the NFS
Lock Manager. I have tested this code extensively and I believe it is
stable and that performance is at least equal to the legacy RPC
implementation.
The NFS code currently contains support for both the new RPC
implementation and the older legacy implementation inherited from the
original NFS codebase. The default is to use the new implementation -
add the NFS_LEGACYRPC option to fall back to the old code. When I
merge this support back to RELENG_7, I will probably change this so
that users have to 'opt in' to get the new code.
To use RPCSEC_GSS on either client or server, you must build a kernel
which includes the KGSSAPI option and the crypto device. On the
userland side, you must build at least a new libc, mountd, mount_nfs
and gssd. You must install new versions of /etc/rc.d/gssd and
/etc/rc.d/nfsd and add 'gssd_enable=YES' to /etc/rc.conf.
As long as gssd is running, you should be able to mount an NFS
filesystem from a server that requires RPCSEC_GSS authentication. The
mount itself can happen without any kerberos credentials but all
access to the filesystem will be denied unless the accessing user has
a valid ticket file in the standard place (/tmp/krb5cc_<uid>). There
is currently no support for situations where the ticket file is in a
different place, such as when the user logged in via SSH and has
delegated credentials from that login. This restriction is also
present in Solaris and Linux. In theory, we could improve this in
future, possibly using Brooks Davis' implementation of variant
symlinks.
Supporting RPCSEC_GSS on a server is nearly as simple. You must create
service creds for the server in the form 'nfs/<fqdn>@<REALM>' and
install them in /etc/krb5.keytab. The standard heimdal utility ktutil
makes this fairly easy. After the service creds have been created, you
can add a '-sec=krb5' option to /etc/exports and restart both mountd
and nfsd.
The only other difference an administrator should notice is that nfsd
doesn't fork to create service threads any more. In normal operation,
there will be two nfsd processes, one in userland waiting for TCP
connections and one in the kernel handling requests. The latter
process will create as many kthreads as required - these should be
visible via 'top -H'. The code has some support for varying the number
of service threads according to load but initially at least, nfsd uses
a fixed number of threads according to the value supplied to its '-n'
option.
Sponsored by: Isilon Systems
MFC after: 1 month
to add more V* constants, and the variables changed by this patch were often
being assigned to mode_t variables, which is 16 bit.
Approved by: rwatson (mentor)
years by the priv_check(9) interface and just very few places are left.
Note that compatibility stub with older FreeBSD version
(all above the 8 limit though) are left in order to reduce diffs against
old versions. It is responsibility of the maintainers for any module, if
they think it is the case, to axe out such cases.
This patch breaks KPI so __FreeBSD_version will be bumped into a later
commit.
This patch needs to be credited 50-50 with rwatson@ as he found time to
explain me how the priv_check() works in detail and to review patches.
Tested by: Giovanni Trematerra <giovanni dot trematerra at gmail dot com>
Reviewed by: rwatson
Now, only some few places still require thread passing (mostly the ones which
access to VOP_* functions) and will be fixed once the primitive also will be.
Tested by: Giovanni Trematerra <giovanni dot trematerra at gmail dot com>
Removed dead code that assumed that M_TRYWAIT can return NULL; it's not true
since the advent of MBUMA.
Reviewed by: arch
There are ongoing disputes as to whether we want to switch to directly using
UMA flags M_WAITOK/M_NOWAIT for mbuf(9) allocation.
BO_LOCK/UNLOCK/MTX when manipulating the bufobj.
- Create a new lock in the bufobj to lock bufobj fields independently.
This leaves the vnode interlock as an 'identity' lock while the bufobj
is an io lock. The bufobj lock is ordered before the vnode interlock
and also before the mnt ilock.
- Exploit this new lock order to simplify softdep_check_suspend().
- A few sync related functions are marked with a new XXX to note that
we may not properly interlock against a non-zero bv_cnt when
attempting to sync all vnodes on a mountlist. I do not believe this
race is important. If I'm wrong this will make these locations easier
to find.
Reviewed by: kib (earlier diff)
Tested by: kris, pho (earlier diff)
conjuction with 'thread' argument passing which is always curthread.
Remove the unuseful extra-argument and pass explicitly curthread to lower
layer functions, when necessary.
KPI results broken by this change, which should affect several ports, so
version bumping and manpage update will be further committed.
Tested by: kris, pho, Diego Sardina <siarodx at gmail dot com>
Remove this argument and pass curthread directly to underlying
VOP_LOCK1() VFS method. This modify makes the code cleaner and in
particular remove an annoying dependence helping next lockmgr() cleanup.
KPI results, obviously, changed.
Manpage and FreeBSD_version will be updated through further commits.
As a side note, would be valuable to say that next commits will address
a similar cleanup about VFS methods, in particular vop_lock1 and
vop_unlock.
Tested by: Diego Sardina <siarodx at gmail dot com>,
Andrea Di Pasquale <whyx dot it at gmail dot com>
Only ops which used namei still remained.
- Implement a scheme for reducing the overhead of tracking which vops
require giant by constantly reducing the number of recursive giant
acquires to one, leaving us with only one vfslocked variable.
- Remove all NFSD lock acquisition and release from the individual nfs
ops. Careful examination has shown that they are not required. This
greatly simplifies the code.
Sponsored by: Isilon Systems, Inc.
Discussed with: rwatson
Tested by: kkenn
Approved by: re
This way we may support multiple structures in v_data vnode field within
one file system without using black magic.
Vnode-to-file-handle should be VOP in the first place, but was made VFS
operation to keep interface as compatible as possible with SUN's VFS.
BTW. Now Solaris also implements vnode-to-file-handle as VOP operation.
VFS_VPTOFH() was left for API backward compatibility, but is marked for
removal before 8.0-RELEASE.
Approved by: mckusick
Discussed with: many (on IRC)
Tested with: ufs, msdosfs, cd9660, nullfs and zfs
forward cases by converting from unconditional acquisition of Giant
around vnode operations to conditional acquisition:
- Remove nfsrv_access_withgiant(), and cause nfsrv_access() to now
assert that Giant will be held if it is required for the vnode.
- Add nfsrv_fhtovp_locked(), which will drop the NFS server lock if
required, and modify nfsrv_fhtovp() to conditionally acquire
Giant if required.
- In the VOP's not dealing with more than one vnode at a time (i.e.,
not involving a lookup), conditionally acquire Giant.
This removes Giant use for MPSAFE file systems for a number of quite
important RPCs, including getattr, read, write. It leaves
unconditional Giant acquisitions in vnode operations that interact
with the name space or more than one vnode at a time as these
require further work.
Tested by: kris
Reviewed by: kib
