When safety requirements are met, it allows to avoid passing I/O requests
to GEOM g_up/g_down thread, executing them directly in the caller context.
That allows to avoid CPU bottlenecks in g_up/g_down threads, plus avoid
several context switches per I/O.
The defined now safety requirements are:
- caller should not hold any locks and should be reenterable;
- callee should not depend on GEOM dual-threaded concurency semantics;
- on the way down, if request is unmapped while callee doesn't support it,
the context should be sleepable;
- kernel thread stack usage should be below 50%.
To keep compatibility with GEOM classes not meeting above requirements
new provider and consumer flags added:
- G_CF_DIRECT_SEND -- consumer code meets caller requirements (request);
- G_CF_DIRECT_RECEIVE -- consumer code meets callee requirements (done);
- G_PF_DIRECT_SEND -- provider code meets caller requirements (done);
- G_PF_DIRECT_RECEIVE -- provider code meets callee requirements (request).
Capable GEOM class can set them, allowing direct dispatch in cases where
it is safe. If any of requirements are not met, request is queued to
g_up or g_down thread same as before.
Such GEOM classes were reviewed and updated to support direct dispatch:
CONCAT, DEV, DISK, GATE, MD, MIRROR, MULTIPATH, NOP, PART, RAID, STRIPE,
VFS, ZERO, ZFS::VDEV, ZFS::ZVOL, all classes based on g_slice KPI (LABEL,
MAP, FLASHMAP, etc).
To declare direct completion capability disk(9) KPI got new flag equivalent
to G_PF_DIRECT_SEND -- DISKFLAG_DIRECT_COMPLETION. da(4) and ada(4) disk
drivers got it set now thanks to earlier CAM locking work.
This change more then twice increases peak block storage performance on
systems with manu CPUs, together with earlier CAM locking changes reaching
more then 1 million IOPS (512 byte raw reads from 16 SATA SSDs on 4 HBAs to
256 user-level threads).
Sponsored by: iXsystems, Inc.
MFC after: 2 months
priority. If the write is requested by a system daemon, sleeping
there would starve resources and cause deadlock.
Reported and tested by: pho
Sponsored by: The FreeBSD Foundation
The flag was mandatory since r209792, where vm_page_grab(9) was
changed to only support the alloc retry semantic.
Suggested and reviewed by: alc
Sponsored by: The FreeBSD Foundation
Unify the 2 concept into a real, minimal, sxlock where the shared
acquisition represent the soft busy and the exclusive acquisition
represent the hard busy.
The old VPO_WANTED mechanism becames the hard-path for this new lock
and it becomes per-page rather than per-object.
The vm_object lock becames an interlock for this functionality:
it can be held in both read or write mode.
However, if the vm_object lock is held in read mode while acquiring
or releasing the busy state, the thread owner cannot make any
assumption on the busy state unless it is also busying it.
Also:
- Add a new flag to directly shared busy pages while vm_page_alloc
and vm_page_grab are being executed. This will be very helpful
once these functions happen under a read object lock.
- Move the swapping sleep into its own per-object flag
The KPI is heavilly changed this is why the version is bumped.
It is very likely that some VM ports users will need to change
their own code.
Sponsored by: EMC / Isilon storage division
Discussed with: alc
Reviewed by: jeff, kib
Tested by: gavin, bapt (older version)
Tested by: pho, scottl
Assign the rv variable a success code if the pager was not asked for
the page. Using an error code from the previous processed page caused
zeroing of the valid page, when e.g. the previous page was not
available in the pager.
Reported by: lstewart
Sponsored by: The FreeBSD Foundation
MFC after: 1 week
The vnode-backed md(4) has to map the unmapped bio because VOP_READ()
and VOP_WRITE() interfaces do not allow to pass unmapped requests to
the filesystem. Vnode-backed md(4) uses pbufs instead of relying on
the bio_transient_map, to avoid usual md deadlock.
Sponsored by: The FreeBSD Foundation
Tested by: pho, scottl
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
- The feature is dangerous because the kernel code didn't check
validity of the memory address provided from user space.
- It seems that mdconfig(8) never really supported attaching preloaded
memory disks.
- Preloaded memory disks are automatically attached during md(4)
initialization. Thus there shouldn't be much use for the feature.
PR: kern/169683
Discussed on: freebsd-hackers
kernel memory leakage to userspace. For the typical use, when a
filesystem put on the md disk, the change only results in CPU and
memory bandwidth spent to zero the page, since filsystems make sure
that user never see unwritten content. But if md disk is used as raw
device by userspace, the garbage is exposed.
Reported by: Paul Schenkeveld <freebsd@psconsult.nl>
MFC after: 2 weeks
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
to pull vm_param.h was removed. Other big dependency of vm_page.h on
vm_param.h are PA_LOCK* definitions, which are only needed for
in-kernel code, because modules use KBI-safe functions to lock the
pages.
Stop including vm_param.h into vm_page.h. Include vm_param.h
explicitely for the kernel code which needs it.
Suggested and reviewed by: alc
MFC after: 2 weeks
architectures (i386, for example) the virtual memory space may be
constrained enough that 2MB is a large chunk. Use 64K for arches
other than amd64 and ia64, with special handling for sparc64 due to
differing hardware.
Also commit the comment changes to kmem_init_zero_region() that I
missed due to not saving the file. (Darn the unfamiliar development
environment).
Arch maintainers, please feel free to adjust ZERO_REGION_SIZE as you
see fit.
Requested by: alc
MFC after: 1 week
MFC with: r221853
sectors with all-zeroes.
The zeroes come from a static buffer; null(4) uses a dynamic buffer for
the same purpose (for /dev/zero). It might be a good idea to have a
static, shared, read-only all-zeroes page somewhere in the kernel that
md(4), null(4) and any other code that needs zeroes could use.
Reviewed by: kib
MFC after: 3 weeks
covering the whole page, free the page. Otherwise, clear the region and
mark it clean. Not marking the page dirty could reinstantiate cleared
data, but it is allowed by BIO_DELETE specification and saves unneeded
write to swap.
Reviewed by: alc
Tested by: pho
MFC after: 2 weeks
md(4) to using M_WAITOK malloc calls.
M_NOWAITOK allocations may fail when enough memory could be freed, but not
immediately. E.g. SU UFS becomes quite unhappy when metadata write return
error, that would happen for failed malloc() call.
Reported and tested by: pho
MFC after: 1 week
To protect against malicious software, we demand that the file name is at
a particular location (i.e. appended to the mdio structure) for it to be
treated as in-kernel.
Fix some wrong usages.
Note: this does not affect generated binaries as this argument is not used.
PR: 137213
Submitted by: Eygene Ryabinkin (initial version)
MFC after: 1 month
rlimit RLIMIT_SWAP that limits the amount of swap that may be reserved
for the uid.
The accounting information (charge) is associated with either map entry,
or vm object backing the entry, assuming the object is the first one
in the shadow chain and entry does not require COW. Charge is moved
from entry to object on allocation of the object, e.g. during the mmap,
assuming the object is allocated, or on the first page fault on the
entry. It moves back to the entry on forks due to COW setup.
The per-entry granularity of accounting makes the charge process fair
for processes that change uid during lifetime, and decrements charge
for proper uid when region is unmapped.
The interface of vm_pager_allocate(9) is extended by adding struct ucred *,
that is used to charge appropriate uid when allocation if performed by
kernel, e.g. md(4).
Several syscalls, among them is fork(2), may now return ENOMEM when
global or per-uid limits are enforced.
In collaboration with: pho
Reviewed by: alc
Approved by: re (kensmith)
possible future I-cache coherency operation can succeed. On ARM
for example the L1 cache can be (is) virtually mapped, which
means that any I/O that uses temporary mappings will not see the
I-cache made coherent. On ia64 a similar behaviour has been
observed. By flushing the D-cache, execution of binaries backed
by md(4) and/or NFS work reliably.
For Book-E (powerpc), execution over NFS exhibits SIGILL once in
a while as well, though cpu_flush_dcache() hasn't been implemented
yet.
Doing an explicit D-cache flush as part of the non-DMA based I/O
read operation eliminates the need to do it as part of the
I-cache coherency operation itself and as such avoids pessimizing
the DMA-based I/O read operations for which D-cache are already
flushed/invalidated. It also allows future optimizations whereby
the bcopy() followed by the D-cache flush can be integrated in a
single operation, which could be implemented using on-chips DMA
engines, by-passing the D-cache altogether.
filesystem supports additional operations using shared vnode locks.
Currently this is used to enable shared locks for open() and close() of
read-only file descriptors.
- When an ISOPEN namei() request is performed with LOCKSHARED, use a
shared vnode lock for the leaf vnode only if the mount point has the
extended shared flag set.
- Set LOCKSHARED in vn_open_cred() for requests that specify O_RDONLY but
not O_CREAT.
- Use a shared vnode lock around VOP_CLOSE() if the file was opened with
O_RDONLY and the mountpoint has the extended shared flag set.
- Adjust md(4) to upgrade the vnode lock on the vnode it gets back from
vn_open() since it now may only have a shared vnode lock.
- Don't enable shared vnode locks on FIFO vnodes in ZFS and UFS since
FIFO's require exclusive vnode locks for their open() and close()
routines. (My recent MPSAFE patches for UDF and cd9660 already included
this change.)
- Enable extended shared operations on UFS, cd9660, and UDF.
Submitted by: ups
Reviewed by: pjd (ZFS bits)
MFC after: 1 month
result in panic:
mdconfig -af blah.img -o force
mount /dev/md0 /mnt
mdconfig -du 0
Reviewed by: scottl
Approved by: rwatson (mentor)
Sponsored by: FreeBSD Foundation
Even though we got rid of device major numbers some time ago, device
drivers still need to provide unique device minor numbers to make_dev().
These numbers are only used inside the kernel. They are not related to
device major and minor numbers which are visible in devfs. These are
actually based on the inode number of the device.
It would eventually be nice to remove minor numbers entirely, but we
don't want to be too agressive here.
Because the 8-15 bits of the device number field (si_drv0) are still
reserved for the major number, there is no 1:1 mapping of the device
minor and unit numbers. Because this is now unused, remove the
restrictions on these numbers.
The MAXMAJOR definition was actually used for two purposes. It was used
to convert both the userspace and kernelspace device numbers to their
major/minor pair, which is why it is now named UMINORMASK.
minor2unit() and unit2minor() have now become useless. Both minor() and
dev2unit() now serve the same purpose. We should eventually remove some
of them, at least turning them into macro's. If devfs would become
completely minor number unaware, we could consider using si_drv0 directly,
just like si_drv1 and si_drv2.
Approved by: philip (mentor)
This fixes the panic which happens when mdcreate_vnode() calls vn_close()
and mddestroy() calls it again further down the error handling path.
Reviewed by: kris, kib
MFC after: 3 days
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>
