The previously used quiesce_all_cpus walks all CPUs and waits until curthread
can run on them. Even on contemporary machines this becomes a significant
problem under load when it can literally take minutes for the operation to
complete. With the patch the stall is normally less than 1 second.
Reviewed by: kib, jeff (previous version)
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21740
A variant of this facility was already used by rmlocks where IPIs would
enforce ordering.
This allows to elide fences where they are rarely needed and the cost of
IPI (should it be necessary) is cheaper.
Reviewed by: kib, jeff (previous version)
Sponsored by: The FreeBSD Foundation
Differential Revision: https://reviews.freebsd.org/D21740
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
The former is frequently accessed (e.g., in vfs_cache_lookup) and shares the
cacheline with v_usecount, avoidably adding to cache misses during concurrent
lookup. The latter is almost unused and probably can get garbage-collected.
The struct does not change in size despite enum vs char * discrepancy.
On 64-bit archs there used to be 4 bytes padding after v_type giving 480 bytes
in total.
broken in r355082. Reduce some locking in nearby related object type
checks.
Reviewed by: kib, markj
Differential Revision: https://reviews.freebsd.org/D22565
unset until the object is recycled so this check is stable. Now that we
can acquire the ref without a lock it is not necessary to group these
operations and we can avoid it entirely in many cases.
Reviewed by: kib, markj
Differential Revision: https://reviews.freebsd.org/D22565
in ofw_gpiobus_probe() return BUS_PROBE_DEFAULT rather than 0; we are not
the only possible driver to handle this device, we're just slightly better
than the base gpiobus (which probes at BUS_PROBE_GENERIC).
In the time since this code was first written, the gpio controller bindings
aquired the concept of a "hog" node which could be used to preset one or
more gpio pins as input or output at a specified level. This change doesn't
fully implement the hogging concept, it just filters out hog nodes when
instantiating child devices by scanning for child nodes in the fdt data.
The whole concept of having child nodes under the controller node is not
supported by the standard bindings, and appears to be a freebsd extension,
probably left over from the days when we had no support for cross-tree
phandle references in the fdt data.
Remove the duplicate macros that defined a subset of the VM_MEMATTR values.
While here use VM_MEMATTR macros when filling in the MAIR register.
Reviewed by: alc, markj
Sponsored by: DARPA, AFRL
Differential Revision: https://reviews.freebsd.org/D22241
We already assert the lock is held later during tty_rel_free(), but it is
arguably good form to clarify locking expectations here as well at the
top-level that other drivers use.
vn_open_cred() assumes that it is called from the top-level of a VFS
syscall. Writers must call bwillwrite() before locking any VFS
resource to wait for cleanup of dirty buffers.
ZFS getextattr() and setextattr() VOPs do call vn_open_cred(), which
results in wait for unrelated buffers while owning ZFS vnode lock (and
ZFS does not use buffer cache). VN_OPEN_INVFS allows caller to skip
bwillwrite.
Note that ZFS is still incorrect there, because it starts write on an
mp and locks a vnode while holding another vnode lock.
Reported by: Willem Jan Withagen <wjw@digiware.nl>
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
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
Since version 2.11.0, QEMU became bug-compatible with
PowerVM's vty implementation, by inserting a \0 after
every \r going to the guest. Guests are expected to
workaround this issue by removing every \0 immediately
following a \r.
Reviewed by: jhibbits
Differential Revision: https://reviews.freebsd.org/D22171
The use of the uma trash procedures is automatic, there's no need to
pass them explicitly here.
Reviewed by: markj
Sponsored by: Dell EMC Isilon
Differential Revision: https://reviews.freebsd.org/D22582
tty_pts.c relies on sys/tty.h for sys/mutex.h. Include it directly instead
of relying on this pollution to ease the diff for anyone that wants to try
converting the tty lock to anything other than a mutex.
(e.g. root->left = NULL) to affect the behavior of that function. This
change stops that data manipulation, and instead calls a pair of
functions, one for the left direction and the other for the right,
with the function called depending whether or not we currently null
the root child in that direction to control the behavior of
vm_map_splay_merge.
Reviewed by: kib
Tested by: pho
Differential Revision: https://reviews.freebsd.org/D22589
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
In case L2ARC read failed, l2arc_read_done() creates _different_ ZIO
to read data from the original storage device. Unfortunately pointer
to the failed ZIO remains in hdr->b_l1hdr.b_acb->acb_zio_head, and if
some other read try to bump the ZIO priority, it will crash.
The problem is reproducible by corrupting L2ARC content and reading
some data with prefetch if l2arc_noprefetch tunable is changed to 0.
With the default setting the issue is probably not reproducible now.
MFC after: 2 weeks
Sponsored by: iXsystems, Inc.
as requiresd by the socket API specification.
Thanks to Inaki Baz Castillo, who found this bug running the userland
stack with valgrind and reported the issue in
https://github.com/sctplab/usrsctp/issues/408
MFC after: 1 week
union members in vm_page.h to store the zone and slab. Remove some nearby
dead code.
Reviewed by: markj
Differential Revision: https://reviews.freebsd.org/D22564
I have no good explanation why it happens, but I found that in B2B mode
at least Xeon v4 NTB leaks accesses to its configuration memory at BAR0
originated from the link side to its host side. DMAR predictably blocks
those, making access to remote scratchpad registers in B2B mode impossible.
This change creates identity mapping in DMAR covering the BAR0 addresses,
making the NTB work fine with DMAR enabled. It seems like allowing single
4KB range at 32KB offset may be enough, but I don't see a reason to be so
specific.
MFC after: 1 week
Sponsored by: iXsystems, Inc.
The kernel RPCSEC_GSS code sets the credential (called a client) lifetime
to the lifetime of the Kerberos ticket, which is typically several hours.
As such, when a user's credentials change such as being added to a new group,
it can take several hours for this change to be recognized by the NFS server.
This patch adds a sysctl called kern.rpc.gss.lifetime_max which can be set
by a sysadmin to put a cap on the time to expire for the credentials, so that
a sysadmin can reduce the timeout.
It also fixes a bug, where time_uptime is added twice when GSS_C_INDEFINITE
is returned for a lifetime. This has no effect in practice, sine Kerberos
never does this.
Tested by: pen@lysator.liu.se
PR: 242132
Submitted by: pen@lysator.liu.se
MFC after: 2 weeks
because nothing ever changes this field for read-only mounts and we want
to verify that it is still 0 when we unmount.
Reviewed by: mckusick
Approved by: mckusick (mentor)
Sponsored by: Netflix
more statistcs than are exported via the ABI stable vmstat interface.
Rename uz_count to uz_bucket_size because even I was confused by the
name after returning to the source years later.
Reviewed by: rlibby
Differential Revision: https://reviews.freebsd.org/D22554
processor in pmap_invalidate_{all,page,range}(). These functions are using
an instruction that broadcasts the TLB invalidation to every processor, so
even if a thread migrates in the middle of one of these functions every
processor will still perform the required TLB invalidations.
Reviewed by: andrew, markj
MFC after: 10 days
Differential Revision: https://reviews.freebsd.org/D22502
This was inherited from iwlwifi, which drives devices supported by both
iwn(4) and iwm(4) in FreeBSD. In iwm(4) _mvm is meaningless, so remove
it. OpenBSD made the same change a long time ago. No functional change
intended.
MFC after: 2 weeks
Sponsored by: The FreeBSD Foundation
FreeBSD-only function should live in the O/S specific source file.
This essentially reverts r349929 Now that ipftest and ipfreplay are
disabled in FreeBSD 11-stable.
MFC after: 3 days
