This implements the kernel glue needed (getc, putc, rxready).
This isn't a 16550 UART, even if the datasheet overview claims so.
The Linux ar933x support was used as a reference, however the uart code
is a reimplementation.
Attentive viewers will note that the uart code is based off of the ns8250
code and the UART bus code is a stubbed-out version of this. I'll be
replacing it with non-stubbed versions soon, making this a fully featured
driver.
Tested:
* AP121 reference board (AR933x), booting through the mountroot> prompt;
then doing some basic interactive tests in ddb.
0x3C: /* Per Intel document 325462-045US 01/2013. */
Add manpage to document all the goodness that is available in this
processor model.
Submitted by: hiren panchasara <hiren.panchasara@gmail.com>
Reviewed by: jimharris, sbruno
Obtained from: Yahoo! Inc.
MFC after: 2 weeks
later found to not be usable because the controller doesn't support the
same number of queues.
This is not the normal case, but does occur with the Chatham prototype
board.
Sponsored by: Intel
1. If we wanted to send exactly as many bytes as the socket buffer is
sized for, the inner loop of kern_sendfile() would see that the
socket is full before seeing that it had no more bytes left to send.
This would cause it to return EAGAIN to the caller instead of
success. Fix by changing the order that these conditions are tested.
2. Simplify the calculation for the bytes to send in each iteration of
the inner loop of kern_sendfile()
3. Fix some calls with bogus arguments to sf_buf_ext(). These would
only trigger on mbuf allocation failure, but would be hilariously
bad if they did trigger.
Submitted by: gibbs(3), andre(2)
Reviewed by: emax, andre
Obtained from: Netflix
MFC after: 1 week
unmapped I/O. That one exception is access to INQUIRY VPD request result.
Those requests are never unmapped now, but to be safe add respective check
there and allow unmapped I/O for the SIM by setting PIM_UNMAPPED flag.
before the vnode is vput() in vm_mmap_vnode(). Error return means
that there is no use reference on the vnode from the vm object
reference, and failing to restore v_writecount breaks the invariant
that v_writecount is less or equal to the usecount.
The situation observed when nfs client returns ESTALE for
VOP_GETATTR() after the open.
In collaboration with: pho
MFC after: 1 week
This was ported from the AR724x code and I think that also doesn't
quite work. I'll investigate that soon.
With this in place the system reset path works, so 'reset' from kdb
actually resets the SoC.
Tested:
* AP121 test board
Before this change they were just leaked. Fortunately USB sticks now use
only one CCB, and so leak was only 2KB per detach, while other bigger SIMs
with much more allocated CCBs are rarely detached.
MFC after: 2 weeks
This basically restores the spirit of r203535, which was partially reverted
in r205557, while we still map fixed amount to work around transient issues
we experienced with r203535.
Prodded by: avg
Tested by: avg
MFC after: 1 week
the thread reference on the vp->v_rdev and use the returned struct
cdev *dev instead of using vp->v_rdev. Call dev_strategy_csw()
instead of dev_strategy(), since we now own the reference.
Since the csw was already calculated, test d_flags to avoid mapping
the buffer if the driver supports unmapped requests [*].
Suggested by: kan [*]
Reviewed by: kan (previous version)
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
but assumes that a thread reference was already obtained on the passed
device. Use the function from physio(), to avoid two extra dev_mtx
lock and unlock. Note that physio() is always used as the cdevsw
method, or is called from a cdevsw method, and the caller already owns
the reference.
dev_strategy() is left to keep KPI intact, but now it is implemented
as a wrapper around dev_strategy_csw().
Do some style cleanup in physio().
Requested and reviewed by: kan (previous version)
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
maxbcache size fixed, the auto-tuned transient map is too small for
real-world load on i386.
Tested by: David Wolfskill
Sponsored by: The FreeBSD Foundation
buffer map size, auto-tuned on the 4GB machine. Having the maxbcache
bigger than the buffer map causes the transient bio map sizing logic
to assume that there is enough KVA to use approximately 90MB (buffer
map is sized to 110MB, and maxbcache is 200MB). The increase in the
KVA usage caused other big KVA consumers, like nvidia.ko, to fail the
initialization.
Change the definition for both PAE and non-PAE cases, since PAE is
even more KVA-starved.
Reported and tested by: David Wolfskill
Discussed with: alc
Sponsored by: The FreeBSD Foundation
which requires OVREF to be set to get proper playback volume, but which has
all zeroes in HDA controller subdevice IDs on PCI.
MFC after: 1 month
Sponsored by:
CPUs.
The AR933x is a mips24k based SoC with an AR9380 series SoC on board,
two gigabit ethernet interfaces and an internal 10/100mbit ethernet
switch. There's also the normal interfaces (USB, ethernet, uart, GPIO.)
The downside? There's a non-ns8250 UART device.
With a very basic UART driver (not in this commit) the SoC is initialised
and boots up. I'll commit the UART code soon and then link it into the
general setup path.
This code is a re-implementation based from the Linux kernel / openwrt
AR933x support.
TODO:
* UART (obviously)
* All of the ethernet, USB and wifi SoC glue, including ethernet PLL
programming.
data buffer for a ccb that is unmapped.
This case is currently not possible, since the SCI framework only
requests these pointers for doing SCSI/ATA translation of non-
READ/WRITE commands. The panic is more to protect against the
unlikely future scenario where additional commands could be unmapped.
Sponsored by: Intel
mechanism.
Now that all requests are timed, we are guaranteed to get a completion
notification, even if it is an abort status due to a timed out admin
command.
This has the effect of simplifying the controller and namespace setup
code, so that it reads straight through rather than broken up into
a bunch of different callback functions.
Sponsored by: Intel
Reviewed by: carl
start or reset. Also add a notifier for NVMe consumers for controller fail
conditions and plumb this notifier for nvd(4) to destroy the associated
GEOM disks when a failure occurs.
This requires a bit of work to cover the races when a consumer is sending
I/O requests to a controller that is transitioning to the failed state. To
help cover this condition, add a task to defer completion of I/Os submitted
to a failed controller, so that the consumer will still always receive its
completions in a different context than the submission.
Sponsored by: Intel
Reviewed by: carl
This is just as effective, and removes the need for a bunch of admin commands
to a controller that's going to be disabled shortly anyways.
Sponsored by: Intel
Reviewed by: carl
start process.
The spec indicates the OS driver should use Set Features (Software
Progress Marker) to set the pre-boot software load count to 0
after the OS driver has successfully been initialized. This allows
pre-boot software to determine if there have been any issues with the
OS loading.
Sponsored by: Intel
Reviewed by: carl
This flag was originally added to communicate to the sysctl code
which oids should be built, but there are easier ways to do this. This
needs to be cleaned up prior to adding new controller states - for example,
controller failure.
Sponsored by: Intel
Reviewed by: carl
The controller's IDENTIFY data contains MDTS (Max Data Transfer Size) to
allow the controller to specify the maximum I/O data transfer size. nvme(4)
already provides a default maximum, but make sure it does not exceed what
MDTS reports.
Sponsored by: Intel
Reviewed by: carl
that if a specific I/O repeatedly times out, we don't retry it indefinitely.
The default number of retries will be 4, but is adjusted using hw.nvme.retry_count.
Sponsored by: Intel
Reviewed by: carl
specified log page.
This satisfies the spec condition that future async events of the same type
will not be sent until the associated log page is fetched.
Sponsored by: Intel
Reviewed by: carl
NVMe error log entries include status, so breaking this out into
its own data structure allows it to be included in both the
nvme_completion data structure as well as error log entry data
structures.
While here, expose nvme_completion_is_error(), and change all of
the places that were explicitly looking at sc/sct bits to use this
macro instead.
Sponsored by: Intel
Reviewed by: carl
This protects against cases where a controller crashes with multiple
I/O outstanding, each timing out and requesting controller resets
simultaneously.
While here, remove a debugging printf from a previous commit, and add
more logging around I/O that need to be resubmitted after a controller
reset.
Sponsored by: Intel
Reviewed by: carl
While aborts are typically cleaner than a full controller reset, many times
an I/O timeout indicates other controller-level issues where aborts may not
work. NVMe drivers for other operating systems are also defaulting to
controller reset rather than aborts for timed out I/O.
Sponsored by: Intel
Reviewed by: carl
(Yes, the previous code temporarily broke EDMA TX. I'm sorry; I should've
actually setup ATH_BUF_FIFOEND on frames so txq->axq_fifo_depth was
cleared!)
This code implements a whole bunch of sorely needed EDMA TX improvements
along with CABQ TX support.
The specifics:
* When filling/refilling the FIFO, use the new TXQ staging queue
for FIFO frames
* Tag frames with ATH_BUF_FIFOPTR and ATH_BUF_FIFOEND correctly.
For now the non-CABQ transmit path pushes one frame into the TXQ
staging queue without setting up the intermediary link pointers
to chain them together, so draining frames from the txq staging
queue to the FIFO queue occurs AMPDU / MPDU at a time.
* In the CABQ case, manually tag the list with ATH_BUF_FIFOPTR and
ATH_BUF_FIFOEND so a chain of frames is pushed into the FIFO
at once.
* Now that frames are in a FIFO pending queue, we can top up the
FIFO after completing a single frame. This means we can keep
it filled rather than waiting for it drain and _then_ adding
more frames.
* The EDMA restart routine now walks the FIFO queue in the TXQ
rather than the pending queue and re-initialises the FIFO with
that.
* When restarting EDMA, we may have partially completed sending
a list. So stamp the first frame that we see in a list with
ATH_BUF_FIFOPTR and push _that_ into the hardware.
* When completing frames, only check those on the FIFO queue.
We should never ever queue frames from the pending queue
direct to the hardware, so there's no point in checking.
* Until I figure out what's going on, make sure if the TXSTATUS
for an empty queue pops up, complain loudly and continue.
This will stop the panics that people are seeing. I'll add
some code later which will assist in ensuring I'm populating
each descriptor with the correct queue ID.
* When considering whether to queue frames to the hardware queue
directly or software queue frames, make sure the depth of
the FIFO is taken into account now.
* When completing frames, tag them with ATH_BUF_BUSY if they're
not the final frame in a FIFO list. The same holding descriptor
behaviour is required when handling descriptors linked together
with a link pointer as the hardware will re-read the previous
descriptor to refresh the link pointer before contiuning.
* .. and if we complete the FIFO list (ie, the buffer has
ATH_BUF_FIFOEND set), then we don't need the holding buffer
any longer. Thus, free it.
Tested:
* AR9380/AR9580, STA and hostap
* AR9280, STA/hostap
TODO:
* I don't yet trust that the EDMA restart routine is totally correct
in all circumstances. I'll continue to thrash this out under heavy
multiple-TXQ traffic load and fix whatever pops up.
On any I/O timeout, check for csts.cfs==1. If set, the controller
is reporting fatal status and we reset the controller immediately,
rather than trying to abort the timed out command.
This changeset also includes deferring the controller start portion
of the reset to a separate task. This ensures we are always performing
a controller start operation from a consistent context.
Sponsored by: Intel
Reviewed by: carl
invoke it from nvmecontrol(8).
Controller reset will be performed in cases where I/O are repeatedly
timing out, the controller reports an unrecoverable condition, or
when explicitly requested via IOCTL or an nvme consumer. Since the
controller may be in such a state where it cannot even process queue
deletion requests, we will perform a controller reset without trying
to clean up anything on the controller first.
Sponsored by: Intel
Reviewed by: carl
Each set of frames pushed into a FIFO is represented by a list of
ath_bufs - the first ath_buf in the FIFO list is marked with
ATH_BUF_FIFOPTR; the last ath_buf in the FIFO list is marked with
ATH_BUF_FIFOEND.
Multiple lists of frames are just glued together in the TAILQ as per
normal - except that at the end of a FIFO list, the descriptor link
pointer will be NULL and it'll be tagged with ATH_BUF_FIFOEND.
For non-EDMA chipsets this is a no-op - the ath_txq frame list (axq_q)
stays the same and is treated the same.
For EDMA chipsets the frames are pushed into axq_q and then when
the FIFO is to be (re) filled, frames will be moved onto the FIFO
queue and then pushed into the FIFO.
So:
* Add a new queue in each hardware TXQ (ath_txq) for staging FIFO frame
lists. It's a TAILQ (like the normal hardware frame queue) rather than
the ath9k list-of-lists to represent FIFO entries.
* Add new ath_buf flags - ATH_TX_FIFOPTR and ATH_TX_FIFOEND.
* When allocating ath_buf entries, clear out the flag value before
returning it or it'll end up having stale flags.
* When cloning ath_buf entries, only clone ATH_BUF_MGMT. Don't clone
the FIFO related flags.
* Extend ath_tx_draintxq() to first drain the FIFO staging queue, _then_
drain the normal hardware queue.
Tested:
* AR9280, hostap
* AR9280, STA
* AR9380/AR9580 - hostap
TODO:
* Test on other chipsets, just to be thorough.
Also add logic to clean up all outstanding asynchronous event requests
when resetting or shutting down the controller, since these requests
will not be explicitly completed by the controller itself.
Sponsored by: Intel
function.
This allows for completions outside the normal completion path, for example
when an ABORT command fails due to the controller reporting the targeted
command does not exist. This is mainly for protection against a faulty
controller, but we need to clean up our internal request nonetheless.
Sponsored by: Intel
the submit action assuming the qpair lock has already been acquired.
Also change nvme_qpair_submit_request to just lock/unlock the mutex
around a call to this new function.
This fixes a recursive mutex acquisition in the retry path.
Sponsored by: Intel
using vm_radix_node_page() == NULL, the compiler is able to generate one
less conditional branch when vm_radix_isleaf() is used. More use cases
involving the inner loops of vm_radix_insert(), vm_radix_lookup{,_ge,_le}(),
and vm_radix_remove() will follow.
Reviewed by: attilio
Sponsored by: EMC / Isilon Storage Division
instead of axq_link.
This (among a bunch of uncommitted work) is required for EDMA chips
to correctly transmit frames on the CABQ.
Tested:
* AR9280, hostap mode
* AR9380/AR9580, hostap mode (staggered beacons)
TODO:
* This code only really gets called when burst beacons are used;
it glues multiple CABQ queues together when sending to the hardware.
* More thorough bursted beacon testing! (first requires some work with
the beacon queue code for bursted beacons, as that currently uses the
link pointer and will fail on EDMA chips.)
the descriptor link pointer, rather than directly.
This is needed on AR9380 and later (ie, EDMA) NICs so the multicast queue
has a chance in hell of being put together right.
Tested:
* AR9380, AR9580 in hostap mode, CABQ traffic (but with other patches..)
In physio, check if device can handle unmapped IO and pass an
appropriately mapped buffer to the driver strategy routine. The
only driver in the tree that can handle unmapped buffers is one
exposed by GEOM, so mark it as such with the new flag in the
driver cdevsw structure.
This fixes insta-panics on hosts, running dconschat, as /dev/fwmem
is an example of the driver that makes use of physio routine, but
bypasses the g_down thread, where the buffer gets mapped normally.
Discussed with: kib (earlier version)
Merge change from illumos:
1694 Add type-aware print() action
This is a very nice feature implemented in upstream Dtrace.
A complete description is available here:
http://dtrace.org/blogs/eschrock/2011/10/26/your-mdb-fell-into-my-dtrace/
This change bumps the DT_VERS_* number to 1.9.0 in
accordance to what is done in illumos.
While here also include some minor cleanups to ease further merging
and appease clang with a fix by Fabian Keil.
Illumos Revisions: 13501:c3a7090dbc16
13483:f413e6c5d297
Reference:
https://www.illumos.org/issues/1560https://www.illumos.org/issues/1694
Tested by: Fabian Keil
Obtained from: Illumos
MFC after: 1 month
Merge changes from illumos:
1451 DTrace needs toupper()/tolower() subroutines
1457 lltostr() D subroutine should take an optional base
This change bumps the DT_VERS_* number to 1.8.1 in
accordance to what is done in illumos.
The test suite we currently include is outdated and
doesnt support some updates in tst.subr.d which had to
be left out for now.
Illumos Revisions: r13458 5e394d8db762
r13459 c3454574dd1a
Reference:
https://www.illumos.org/issues/1451https://www.illumos.org/issues/1457
Tested by: Fabian Keil
Obtained from: Illumos
MFC after: 1 month
It is already done in SSIF interface code.
This reduces contention/spinning reported by many users.
PR: kern/172166
Submitted by: Eric van Gyzen <eric at vangyzen.net>
MFC after: 2 weeks
for migrating callouts to new CPU. This value is passed to
callout_cc_add() in order to update properly precision field in case of
rescheduling/migration.
Reviewed by: mav
extra read from PxCI/PxSACT registers. If only NCQ commands are running, we
don't really need PxCI. If only non-NCQ commands are running we don't need
PxSACT. Mixed set may happen only on controllers with FIS-based switching
when port multiplier is attached, and then we have to read both registers.
MFC after: 1 month
- Replace single done mutex with per-disk ones. On system with several
disks on several HBAs that removes small, but measurable lock congestion.
- Modify disk destruction process to not destroy the mutex prematurely.
- Remove some extra pointer derefences.
Merge change from illumos:
1455 DTrace tracemem() should take an optional size argument
Our local enhancements to dt_print_bytes were equivalent to
those in illumos but we made it match the illumos version
to ease further code merges.
For now leave out tst.smallsize.d and tst.smallsize.d.out
since those don't seem to work cleanly on FreeBSD.
This change bumps the DT_VERS_* number to 1.7.1 in accordance
to what is done in illumos.
Illumos Revision: 13457:571b0355c2e3
Reference:
https://www.illumos.org/issues/1455
Tested by: Fabian Keil
Obtained from: Illumos
MFC after: 1 month
that could never be reached in vm_radix_insert(). (If the pointer being
checked by the panic call were ever NULL, the immmediately preceding loop
would have already crashed on a NULL pointer dereference.)
Reviewed by: attilio (an earlier version)
Sponsored by: EMC / Isilon Storage Division
Use destroy_dev_sched_cb() to not wait for device destruction while holding
GEOM topology lock (that actually caused deadlock). Use request counting
protected by mutex to properly wait for outstanding requests completion in
cases of device closing and geom destruction. Unlike r227009, this code
does not block taskqueue thread for indefinite time, waiting for completion.
more topology change done that may require its attention. Add few missing
g_do_wither() calls in respective places to signal it.
This fixes potential infinite loop here when some provider is withered, but
still opened or connected for some reason and so can not be destroyed. For
example, see r227009 and r227510.
related issues.
Moving the TX locking under one lock made things easier to progress on
but it had one important side-effect - it increased the latency when
handling CABQ setup when sending beacons.
This commit introduces a bunch of new changes and a few unrelated changs
that are just easier to lump in here.
The aim is to have the CABQ locking separate from other locking.
The CABQ transmit path in the beacon process thus doesn't have to grab
the general TX lock, reducing lock contention/latency and making it
more likely that we'll make the beacon TX timing.
The second half of this commit is the CABQ related setup changes needed
for sane looking EDMA CABQ support. Right now the EDMA TX code naively
assumes that only one frame (MPDU or A-MPDU) is being pushed into each
FIFO slot. For the CABQ this isn't true - a whole list of frames is
being pushed in - and thus CABQ handling breaks very quickly.
The aim here is to setup the CABQ list and then push _that list_ to
the hardware for transmission. I can then extend the EDMA TX code
to stamp that list as being "one" FIFO entry (likely by tagging the
last buffer in that list as "FIFO END") so the EDMA TX completion code
correctly tracks things.
Major:
* Migrate the per-TXQ add/removal locking back to per-TXQ, rather than
a single lock.
* Leave the software queue side of things under the ATH_TX_LOCK lock,
(continuing) to serialise things as they are.
* Add a new function which is called whenever there's a beacon miss,
to print out some debugging. This is primarily designed to help
me figure out if the beacon miss events are due to a noisy environment,
issues with the PHY/MAC, or other.
* Move the CABQ setup/enable to occur _after_ all the VAPs have been
looked at. This means that for multiple VAPS in bursted mode, the
CABQ gets primed once all VAPs are checked, rather than being primed
on the first VAP and then having frames appended after this.
Minor:
* Add a (disabled) twiddle to let me enable/disable cabq traffic.
It's primarily there to let me easily debug what's going on with beacon
and CABQ setup/traffic; there's some DMA engine hangs which I'm finally
trying to trace down.
* Clear bf_next when flushing frames; it should quieten some warnings
that show up when a node goes away.
Tested:
* AR9280, STA/hostap, up to 4 vaps (staggered)
* AR5416, STA/hostap, up to 4 vaps (staggered)
TODO:
* (Lots) more AR9380 and later testing, as I may have missed something here.
* Leverage this to fix CABQ hanling for AR9380 and later chips.
* Force bursted beaconing on the chips that default to staggered beacons and
ensure the CABQ stuff is all sane (eg, the MORE bits that aren't being
correctly set when chaining descriptors.)
to stuck beacons.
* Set the cabq readytime (ie, how long to burst for) to 50% of the total
beacon interval time
* fix the cabq adjustment calculation based on how the beacon offset is
calculated (the SWBA/DBA time offset.)
This is all still a bit magic voodoo but it does seem to have further
quietened issues with missed/stuck beacons under my local testing.
In any case, it better matches what the reference HAL implements.
Obtained from: Qualcomm Atheros
held. The ttm_buffer_object_transfer() does not need the mutex locked
at all, except for the call to the driver sync_obj_ref() method.
Reported and tested by: dumbbell
MFC after: 2 weeks
With some recent busdma refactoring, sometimes it happens that a sync
op gets called when bus_dmamap_load() never got called, which results
in a spurious warning about a map mismatch when no sync operations will
actually happen anyway. Now the check is done only if a sync operation
is actually performed, and the result of the check is a panic, not just
a printf.
Reviewed by: cognet (who prevented me from donning a point hat)
This particular scenario was easily reproduced using a NFS export. When the
first 'zfs unmount' occurred, it returned EBUSY via this path, while
vflush() had flushed references on the filesystem's root vnode, which in
turn caused its v_interlock to be destroyed. The next time 'zfs unmount'
was called, vflush() tried to obtain this lock, which caused this panic.
Since vflush() on FreeBSD is a definitive call, there is no need to check
vfsp->vfs_count after it completes. Simply #ifdef sun this check.
Submitted by: avg
Reviewed by: avg
Approved by: ken (mentor)
MFC after: 1 month
The scope of these callbacks is primarily to support actions that affect the
taskqueue's thread environments. They are entirely optional, and
consequently are introduced as a new API: taskqueue_set_callback().
This interface allows the caller to specify that a taskqueue requires a
callback and optional context pointer for a given callback type.
The callback types included in this commit can be used to register a
constructor and destructor for thread-local storage using osd(9). This
allows a particular taskqueue to define that its threads require a specific
type of TLS, without the need for a specially-orchestrated task-based
mechanism for startup and shutdown in order to accomplish it.
Two callback types are supported at this point:
- TASKQUEUE_CALLBACK_TYPE_INIT, called by every thread when it starts, prior
to processing any tasks.
- TASKQUEUE_CALLBACK_TYPE_SHUTDOWN, called by every thread when it exits,
after it has processed its last task but before the taskqueue is
reclaimed.
While I'm here:
- Add two new macros, TQ_ASSERT_LOCKED and TQ_ASSERT_UNLOCKED, and use them
in appropriate locations.
- Fix taskqueue.9 to mention taskqueue_start_threads(), which is a required
interface for all consumers of taskqueue(9).
Reviewed by: kib (all), eadler (taskqueue.9), brd (taskqueue.9)
Approved by: ken (mentor)
Sponsored by: Spectra Logic
MFC after: 1 month
not every time an intermediate root (including the first devfs) is
mounted.
This is also consistent with waking up via root_mount_complete.
Reviewed by: jhb
MFC after: 13 days
This issue would be silent most of the time, but if the requested memory
is a multiple of a page size, then accessing one element beyond the end
would lead to a kernel page fault.
Otherwise, the unlucky last type would just be inaccessible.
Reported by: glebius
Tested by: glebius
MFC after: 6 days
running time for a full fsck. It also reduces the random access time
for large files and speeds the traversal time for directory tree walks.
The key idea is to reserve a small area in each cylinder group
immediately following the inode blocks for the use of metadata,
specifically indirect blocks and directory contents. The new policy
is to preferentially place metadata in the metadata area and
everything else in the blocks that follow the metadata area.
The size of this area can be set when creating a filesystem using
newfs(8) or changed in an existing filesystem using tunefs(8).
Both utilities use the `-k held-for-metadata-blocks' option to
specify the amount of space to be held for metadata blocks in each
cylinder group. By default, newfs(8) sets this area to half of
minfree (typically 4% of the data area).
This work was inspired by a paper presented at Usenix's FAST '13:
www.usenix.org/conference/fast13/ffsck-fast-file-system-checker
Details of this implementation appears in the April 2013 of ;login:
www.usenix.org/publications/login/april-2013-volume-38-number-2.
A copy of the April 2013 ;login: paper can also be downloaded
from: www.mckusick.com/publications/faster_fsck.pdf.
Reviewed by: kib
Tested by: Peter Holm
MFC after: 4 weeks
u_long. Before this change it was of type int for syscalls, but prototypes
in sys/stat.h and documentation for chflags(2) and fchflags(2) (but not
for lchflags(2)) stated that it was u_long. Now some related functions
use u_long type for flags (strtofflags(3), fflagstostr(3)).
- Make path argument of type 'const char *' for consistency.
Discussed on: arch
Sponsored by: The FreeBSD Foundation
UMTX_PROFILING should really analyze the distribution of locks as they
index entries in the umtxq_chains hash-table.
However, the current implementation does add/dec the length counters
for *every* thread insert/removal, measuring at all really userland
contention and not the hash distribution.
Fix this by correctly add/dec the length counters in the points where
it is really needed.
Please note that this bug brought us questioning in the past the quality
of the umtx hash table distribution.
To date with all the benchmarks I could try I was not able to reproduce
any issue about the hash distribution on umtx.
Sponsored by: EMC / Isilon storage division
Reviewed by: jeff, davide
MFC after: 2 weeks
original 2us are indeed not enough, 3us are working quite well on my tests.
To be more safe set minimal period to 5us and to be even more safe replicate
here from HPET mechanism of rereading counter after programming comparator.
This change allows to handle 30K of short nanosleep() calls per second on
Raspberry Pi instead of just 8K before.
Discussed with: gonzo
SIGSTOP if stop signals are currently deferred. This can occur if a
process is stopped via SIGSTOP while a thread is running or runnable
but before it has set TDF_SBDRY.
Tested by: pho
Reviewed by: kib
MFC after: 1 week
satisfy some alignment restrictions. Do not set TW_OSLI_REQ_FLAGS_CCB
flag for mapped data, pass the csio->data_ptr in the req->data.
Do not put the ccb pointer into req->data ever, ccb is stored in
req->orig_req already.
Submitted by: Shuichi KITAGUCHI <ki@hh.iij4u.or.jp>
PR: kern/177020
Previously TRIM processing was very bursty. This was made worse by the fact
that TRIM requests on SSD's are typically much slower than reads or writes.
This often resulted in stalls while large numbers of TRIM's where processed.
In addition due to the way the TRIM thread was only woken by writes, deletes
could stall in the queue for extensive periods of time.
This patch adds a number of controls to how often the TRIM thread for each
SPA processes its outstanding delete requests.
vfs.zfs.trim.timeout: Delay TRIMs by up to this many seconds
vfs.zfs.trim.txg_delay: Delay TRIMs by up to this many TXGs (reduced to 32)
vfs.zfs.vdev.trim_max_bytes: Maximum pending TRIM bytes for a vdev
vfs.zfs.vdev.trim_max_pending: Maximum pending TRIM segments for a vdev
vfs.zfs.trim.max_interval: Maximum interval between TRIM queue processing
(seconds)
Given the most common TRIM implementation is ATA TRIM the current defaults
are targeted at that.
Reviewed by: pjd (mentor)
Approved by: pjd (mentor)
MFC after: 2 weeks
Currently, the trim module uses the same algorithm for data and cache
devices when deciding to issue TRIM requests, based on how far in the
past the TXG is.
Unfortunately, this is not ideal for cache devices, because the L2ARC
doesn't use the concept of TXGs at all. In fact, when using a pool for
reading only, the L2ARC is written but the TXG counter doesn't
increase, and so no new TRIM requests are issued to the cache device.
This patch fixes the issue by using time instead of the TXG number as
the criteria for trimming on cache devices. The basic delay principle
stays the same, but parameters are expressed in seconds instead of
TXGs. The new parameters are named trim_l2arc_limit and
trim_l2arc_batch, and both default to 30 second.
Reviewed by: pjd (mentor)
Approved by: pjd (mentor)
Obtained from: 17122c31ac
MFC after: 2 weeks
This patch adds some improvements to the way the trim module considers
TXGs:
- Free ZIOs are registered with the TXG from the ZIO itself, not the
current SPA syncing TXG (which may be out of date);
- L2ARC are registered with a zero TXG number, as L2ARC has no concept
of TXGs;
- The TXG limit for issuing TRIMs is now computed from the last synced
TXG, not the currently syncing TXG. Indeed, under extremely unlikely
race conditions, there is a risk we could trim blocks which have been
freed in a TXG that has not finished syncing, resulting in potential
data corruption in case of a crash.
Reviewed by: pjd (mentor)
Approved by: pjd (mentor)
Obtained from: 5b46ad40d9
MFC after: 2 weeks
The trim map inflight writes tree assumes non-conflicting writes, i.e.
that there will never be two simultaneous write I/Os to the same range
on the same vdev. This seemed like a sane assumption; however, in
actual testing, it appears that repair I/Os can very well conflict
with "normal" writes.
I'm not quite sure if these conflicting writes are supposed to happen
or not, but in the mean time, let's ignore repair writes for now. This
should be safe considering that, by definition, we never repair blocks
that are freed.
Reviewed by: pjd (mentor)
Approved by: pjd (mentor)
Obtained from: Source: 6a3cebaf7c
This adds TRIM support to cache vdevs. When ARC buffers are removed
from the L2ARC in arc_hdr_destroy(), arc_release() or l2arc_evict(),
the size previously occupied by the buffer gets scheduled for TRIMming.
As always, actual TRIMs are only issued to the L2ARC after
txg_trim_limit.
Reviewed by: pjd (mentor)
Approved by: pjd (mentor)
Obtained from: 31aae37399
MFC after: 2 weeks
includes MFV 238590, 238592, 247580
MFV 238590, 238592:
In the first zfs ioctl restructuring phase, the libzfs_core library was
introduced. It is a new thin library that wraps around kernel ioctl's.
The idea is to provide a forward-compatible way of dealing with new
features. Arguments are passed in nvlists and not random zfs_cmd fields,
new-style ioctls are logged to pool history using a new method of
history logging.
http://blog.delphix.com/matt/2012/01/17/the-future-of-libzfs/
MFV 247580 [1]:
To address issues of several deadlocks and race conditions the locking
code around dsl_dataset was rewritten and the interface to synctasks
was changed.
User-Visible Changes:
"zfs snapshot" can create more arbitrary snapshots at once (atomically)
"zfs destroy" destroys multiple snapshots at once
"zfs recv" has improved performance
Backward Compatibility:
I have extended the compatibility layer to support full backward
compatibility by remapping or rewriting the responsible ioctl arguments.
Old utilities are fully supported by the new kernel module.
Forward Compatibility:
New utilities work with old kernels with the following restrictions:
- creating, destroying, holding and releasing of multiple snapshots
at once is not supported, this includes recursive (-r) commands
Illumos ZFS issues:
2882 implement libzfs_core
2900 "zfs snapshot" should be able to create multiple,
arbitrary snapshots at once
3464 zfs synctask code needs restructuring
References:
https://www.illumos.org/issues/2882https://www.illumos.org/issues/2900https://www.illumos.org/issues/3464 [1]
MFC after: 1 month
Sponsored by: Hybrid Logic Inc. [1]
locked. vnode_pager_setsize() might sleep waiting for the page after
EOF be unbusied.
Call vnode_pager_setsize() both for the regular and directory vnodes.
Reported by: mich
Reviewed by: rmacklem
Discussed with: avg, jhb
MFC after: 2 weeks
bufobj counter of the writes in progress is incremented. Other thread
inspecting the bufobj would consider it clean.
For the regular vnodes, the vnode lock is typically held both by the
thread performing the bufwrite() and an other thread doing syncing,
which prevents the situation. On the other hand, writes to the VCHR
vnodes are done without holding vnode lock.
Increment the write ref counter for the buffer object before calling
bundirty().
Sponsored by: The FreeBSD Foundation
Tested by: pho
MFC after: 2 weeks
becoming too low, the softdep flush thread processes the workitems,
which frees the space in journal, and then unsuspends the fs. The
softdep_flush() and other workitem processing functions busy the
filesystem before iterating over the worklist, to prevent the parallel
unmount from freeing the mount data. The vfs_busy() is called with
MBF_NOWAIT flag.
Now, if the unmount is already started and the filesystem is suspended
due to low journal space, the journal is never flushed and filesystem
is never unsuspended, because vfs_busy(MBF_NOWAIT) call cannot succeed
for the unmounting fs, and softdep_flush() does not process the
workitems. Unmount needs to write metadata, where it hangs in the
"suspfs" state.
Move the vn_start_write() call in the dounmount() before setting the
MNTK_UNMOUNT flag. This practically ensures that softdep_flush()
processed the pending journal writes by making dounmount() wait for
the lift of the suspension.
Sponsored by: The FreeBSD Foundation
Reported and tested by: pho
MFC after: 2 weeks
we need to call ufs_checkpath() to walk from our new location to
the root of the filesystem to ensure that we do not encounter
ourselves along the way. Until now, we accomplished this by reading
the ".." entries of each directory in our path until we reached
the root (or encountered an error). This change tries to avoid the
I/O of reading the ".." entries by first looking them up in the
name cache and only doing the I/O when the name cache lookup fails.
Reviewed by: kib
Tested by: Peter Holm
MFC after: 4 weeks
Setting DSCP support is done via O_SETDSCP which works for both
IPv4 and IPv6 packets. Fast checksum recalculation (RFC 1624) is done for IPv4.
Dscp can be specified by name (AFXY, CSX, BE, EF), by value
(0..63) or via tablearg.
Matching DSCP is done via another opcode (O_DSCP) which accepts several
classes at once (af11,af22,be). Classes are stored in bitmask (2 u32 words).
Many people made their variants of this patch, the ones I'm aware of are
(in alphabetic order):
Dmitrii Tejblum
Marcelo Araujo
Roman Bogorodskiy (novel)
Sergey Matveichuk (sem)
Sergey Ryabin
PR: kern/102471, kern/121122
MFC after: 2 weeks
auto-correction. This change makes re(4) establish a link with
a system using non-crossover UTP cable.
Tested by: Michael BlackHeart < amdmiek <> gmail dot com >
