Don't hold the scheduler lock while doing context switches. Instead we
unlock after selecting the new thread and switch within a spinlock
section leaving interrupts and preemption disabled to prevent local
concurrency. This means that mi_switch() is entered with the thread
locked but returns without. This dramatically simplifies scheduler
locking because we will not hold the schedlock while spinning on
blocked lock in switch.
This change has not been made to 4BSD but in principle it would be
more straightforward.
Discussed with: markj
Reviewed by: kib
Tested by: pho
Differential Revision: https://reviews.freebsd.org/D22778
Eliminate recursion from most thread_lock consumers. Return from
sched_add() without the thread_lock held. This eliminates unnecessary
atomics and lock word loads as well as reducing the hold time for
scheduler locks. This will eventually allow for lockless remote adds.
Discussed with: kib
Reviewed by: jhb
Tested by: pho
Differential Revision: https://reviews.freebsd.org/D22626
The ixl.4 manual page has documented that the threshold falsely detects
interrupt storms on 40Gbit NICs as long ago as 2015, and we have seen
similar false positives with the ioat(4) DMA device (which can push GB/s).
For example, synthetic load can be generated with tools/tools/ioat
'ioatcontrol 0 200 8192 1 1000' (allocate 200x8kB buffers, generate an
interrupt for each one, and do this for 1000 milliseconds). With
storm-detection disabled, the Broadwell-EP version of this device is capable
of generating ~350k real interrupts per second.
The following historical context comes from jhb@: Originally, the threshold
worked around incorrect routing of PCI INTx interrupts on single-CPU systems
which would end up in a hard hang during boot. Since the threshold was
added, our PCI interrupt routing was improved, most PCI interrupts use
edge-triggered MSI instead of level-triggered INTx, and typical systems have
multiple CPUs available to service interrupts.
On the off chance that the threshold is useful in the future, it remains
available as a tunable and sysctl.
Reviewed by: jhb
Sponsored by: Dell EMC Isilon
Differential Revision: https://reviews.freebsd.org/D20401
Bind the TCP pacer threads to NUMA domains and build per-domain
pacer-thread lookup tables. These tables allow us to use the
inpcb's NUMA domain information to match an inpcb with a pacer
thread on the same domain.
The motivation for this is to keep the TCP connection local to a
NUMA domain as much as possible.
Thanks to jhb for pre-reviewing an earlier version of the patch.
Reviewed by: rrs
Sponsored by: Netflix
Differential Revision: https://reviews.freebsd.org/D20134
There are only 19 bytes available for the name of an interrupt plus the
name(s) of handlers/drivers using it. There is a mechanism from the days of
shared interrupts that replaces some of the handler names with '+' when they
don't all fit into 19 bytes.
In modern times there is typically only one device on an interrupt, but long
device names are the norm, especially with embedded systems. Also, in systems
with multiple interrupt controllers, the names of the interrupts themselves
can be long. For example, 'gic0,s54: imx6_anatop0' doesn't fit, and
replacing the device driver name with a '+' provides no useful info at all.
When there is only one handler but its name was too long to fit, this
change truncates enough leading chars of the handler name (replacing them
with a '-' char to indicate that some chars are missing) to use all 19
bytes, preserving the unit number typically on the end of the name. Using
the prior example, this results in: 'gic0,s54:-6_anatop0' which provides
plenty of info to figure out which device is involved.
PR: 211946
Reviewed by: gonzo@ (prior version without the '-' char)
Differential Revision: https://reviews.freebsd.org/D19675
The goal of this change is to fix a problem with PCI shared interrupts
during suspend and resume.
I have observed a couple of variations of the following scenario.
Devices A and B are on the same PCI bus and share the same interrupt.
Device A's driver is suspended first and the device is powered down.
Device B generates an interrupt. Interrupt handlers of both drivers are
called. Device A's interrupt handler accesses registers of the powered
down device and gets back bogus values (I assume all 0xff). That data is
interpreted as interrupt status bits, etc. So, the interrupt handler
gets confused and may produce some noise or enter an infinite loop, etc.
This change affects only PCI devices. The pci(4) bus driver marks a
child's interrupt handler as suspended after the child's suspend method
is called and before the device is powered down. This is done only for
traditional PCI interrupts, because only they can be shared.
At the moment the change is only for x86.
Notable changes in core subsystems / interfaces:
- BUS_SUSPEND_INTR and BUS_RESUME_INTR methods are added to bus
interface along with convenience functions bus_suspend_intr and
bus_resume_intr;
- rman_set_irq_cookie and rman_get_irq_cookie functions are added to
provide a way to associate an interrupt resource with an interrupt
cookie;
- intr_event_suspend_handler and intr_event_resume_handler functions
are added to the MI interrupt handler interface.
I added two new interrupt handler flags, IH_SUSP and IH_CHANGED, to
implement the new intr_event functions. IH_SUSP marks a suspended
interrupt handler. IH_CHANGED is used to implement a barrier that
ensures that a change to the interrupt handler's state is visible
to future interrupts.
While there, I fixed some whitespace issues in comments and changed a
couple of logically boolean variables to be bool.
MFC after: 1 month (maybe)
Differential Revision: https://reviews.freebsd.org/D15755
given in random(4).
This includes updating of the relevant man pages, and no-longer-used
harvesting parameters.
Ensure that the pseudo-unit-test still does something useful, now also
with the "other" algorithm instead of Yarrow.
PR: 230870
Reviewed by: cem
Approved by: so(delphij,gtetlow)
Approved by: re(marius)
Differential Revision: https://reviews.freebsd.org/D16898
The code that iterates a list of interrupt handlers for a (shared)
interrupt, whether in the ISR context or in the context of an interrupt
thread, does so in a lock-free fashion. Thus, the routines that modify
the list need to take special steps to ensure that the iterating code
has a consistent view of the list. Previously, those routines tried to
play nice only with the code running in the ithread context. The
iteration in the ISR context was left to a chance.
After commit r336635 atomic operations and memory fences are used to
ensure that ie_handlers list is always safe to navigate with respect to
inserting and removal of list elements.
There is still a question of when it is safe to actually free a removed
element.
The idea of this change is somewhat similar to the idea of the epoch
based reclamation. There are some simplifications comparing to the
general epoch based reclamation. All writers are serialized using a
mutex, so we do not need to worry about concurrent modifications. Also,
all read accesses from the open context are serialized too.
So, we can get away just two epochs / phases. When a thread removes an
element it switches the global phase from the current phase to the other
and then drains the previous phase. Only after the draining the removed
element gets actually freed. The code that iterates the list in the ISR
context takes a snapshot of the global phase and then increments the use
count of that phase before iterating the list. The use count (in the
same phase) is decremented after the iteration. This should ensure that
there should be no iteration over the removed element when its gets
freed.
This commit also simplifies the coordination with the interrupt thread
context. Now we always schedule the interrupt thread when removing one
of handlers for its interrupt. This makes the code both simpler and
safer as the interrupt thread masks the interrupt thus ensuring that
there is no interaction with the ISR context.
P.S. This change matters only for shared interrupts and I realize that
those are becoming a thing of the past (and quickly). I also understand
that the problem that I am trying to solve is extremely rare.
PR: 229106
Reviewed by: cem
Discussed with: Samy Al Bahra
MFC after: 5 weeks
Differential Revision: https://reviews.freebsd.org/D15905
The primary reason for this commit is to separate mechanical and nearly
mechanical code changes from an upcoming fix for unsafe teardown of
shared interrupt handlers that have only filters (see D15905).
The technical rationale is that SLIST is sufficient. The only operation
that gets worse performance -- O(n) instead of O(1) is a removal of a
handler, but it is not a critical operation and the list is expected to
be rather short.
Additionally, it is easier to reason about SLIST when considering the
concurrent lock-free access to the list from the interrupt context and
the interrupt thread.
CK_SLIST is used because the upcoming change depends on the memory order
provided by CK_SLIST insert and the fact that CL_SLIST remove does not
trash the linkage in a removed element.
While here, I also fixed a couple of whitespace issues, made code under
ifdef notyet compilable, added a lock assertion to ithread_update() and
made intr_event_execute_handlers() static as it had no external callers.
Reviewed by: cem (earlier version)
MFC after: 4 weeks
Differential Revision: https://reviews.freebsd.org/D16016
It was introduced to the tree in r169320 and r169321 in May 2007.
It never got much use and never became a kernel default. The code
duplicates the default path quite a bit, with slight modifications. Just
yank out the cruft. Whatever goals were being aimed for can probably be met
within the existing framework, without a flag day option.
Mostly mechanical change: 'unifdef -m -UINTR_FILTER'.
Reviewed by: mmacy
Sponsored by: Dell EMC Isilon
Differential Revision: https://reviews.freebsd.org/D15546
Mainly focus on files that use BSD 2-Clause license, however the tool I
was using misidentified many licenses so this was mostly a manual - error
prone - task.
The Software Package Data Exchange (SPDX) group provides a specification
to make it easier for automated tools to detect and summarize well known
opensource licenses. We are gradually adopting the specification, noting
that the tags are considered only advisory and do not, in any way,
superceed or replace the license texts.
Add IRQ placement-only and ithread-only API variants. intr_event_bind
has been extended with sibling methods, as it has many more callsites in
existing code.
Reviewed by: kib@, adrian@ (earlier version)
Sponsored by: Dell EMC Isilon
Differential Revision: https://reviews.freebsd.org/D10586
in place. To do per-cpu stats, convert all fields that previously were
maintained in the vmmeters that sit in pcpus to counter(9).
- Since some vmmeter stats may be touched at very early stages of boot,
before we have set up UMA and we can do counter_u64_alloc(), provide an
early counter mechanism:
o Leave one spare uint64_t in struct pcpu, named pc_early_dummy_counter.
o Point counter(9) fields of vmmeter to pcpu[0].pc_early_dummy_counter,
so that at early stages of boot, before counters are allocated we already
point to a counter that can be safely written to.
o For sparc64 that required a whole dummy pcpu[MAXCPU] array.
Further related changes:
- Don't include vmmeter.h into pcpu.h.
- vm.stats.vm.v_swappgsout and vm.stats.vm.v_swappgsin changed to 64-bit,
to match kernel representation.
- struct vmmeter hidden under _KERNEL, and only vmstat(1) is an exclusion.
This is based on benno@'s 4-year old patch:
https://lists.freebsd.org/pipermail/freebsd-arch/2013-July/014471.html
Reviewed by: kib, gallatin, marius, lidl
Differential Revision: https://reviews.freebsd.org/D10156
it_need was wrong [*]. Restore the releases and add a comment
explaining why it is needed.
Noted by: alc [*]
Reviewed by: bde [*]
Sponsored by: The FreeBSD Foundation
Remove useless release semantic for some stores to it_need. For
stores where the release is needed, add a comment explaining why.
Fence after the atomic_cmpset() op on the it_need should be acquire
only, release is not needed (see above). The combination of
atomic_cmpset() + fence_acq() is better expressed there as
atomic_cmpset_acq().
Use atomic_cmpset() for swi' ih_need read and clear.
Discussed with: alc, bde
Reviewed by: bde
Comments wording provided by: bde
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
architectures. Atomic_cmpset_int(9) is a direct replacement, due to
loop. The change fixes arm, arm64, mips an sparc64, which lack
atomic_swap().
Suggested and reviewed by: alc
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
current value. It is believed that the change is the real fix for the
issue which was covered over by the r252683.
With the current code, if the interrupt handler sets it_need between
read and consequent reset, the update could be lost and
ithread_execute_handlers() would not be called in response to the lost
update.
The r252683 could have hide the issue since at the moment of commit,
atomic_load_acq_int() did locked cmpxchg on the variable, which puts
the cache line into the exclusive owned state and clears store
buffers. Then the immediate store of zero has very high chance of
reusing the exclusive state of the cache line and make the load and
store sequence operate as atomic swap.
For now, add the acq+rel fence immediately after the swap, to not
disturb current (but excessive) ordering. Acquire is needed for the
ih_need reads after the load, while release does not serve a useful
purpose [*].
Reviewed by: alc
Noted by: alc [*]
Discussed with: bde
Tested by: pho
Sponsored by: The FreeBSD Foundation
MFC after: 2 weeks
* GENERAL
- Update copyright.
- Make kernel options for RANDOM_YARROW and RANDOM_DUMMY. Set
neither to ON, which means we want Fortuna
- If there is no 'device random' in the kernel, there will be NO
random(4) device in the kernel, and the KERN_ARND sysctl will
return nothing. With RANDOM_DUMMY there will be a random(4) that
always blocks.
- Repair kern.arandom (KERN_ARND sysctl). The old version went
through arc4random(9) and was a bit weird.
- Adjust arc4random stirring a bit - the existing code looks a little
suspect.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Redo read_random(9) so as to duplicate random(4)'s read internals.
This makes it a first-class citizen rather than a hack.
- Move stuff out of locked regions when it does not need to be
there.
- Trim RANDOM_DEBUG printfs. Some are excess to requirement, some
behind boot verbose.
- Use SYSINIT to sequence the startup.
- Fix init/deinit sysctl stuff.
- Make relevant sysctls also tunables.
- Add different harvesting "styles" to allow for different requirements
(direct, queue, fast).
- Add harvesting of FFS atime events. This needs to be checked for
weighing down the FS code.
- Add harvesting of slab allocator events. This needs to be checked for
weighing down the allocator code.
- Fix the random(9) manpage.
- Loadable modules are not present for now. These will be re-engineered
when the dust settles.
- Use macros for locks.
- Fix comments.
* src/share/man/...
- Update the man pages.
* src/etc/...
- The startup/shutdown work is done in D2924.
* src/UPDATING
- Add UPDATING announcement.
* src/sys/dev/random/build.sh
- Add copyright.
- Add libz for unit tests.
* src/sys/dev/random/dummy.c
- Remove; no longer needed. Functionality incorporated into randomdev.*.
* live_entropy_sources.c live_entropy_sources.h
- Remove; content moved.
- move content to randomdev.[ch] and optimise.
* src/sys/dev/random/random_adaptors.c src/sys/dev/random/random_adaptors.h
- Remove; plugability is no longer used. Compile-time algorithm
selection is the way to go.
* src/sys/dev/random/random_harvestq.c src/sys/dev/random/random_harvestq.h
- Add early (re)boot-time randomness caching.
* src/sys/dev/random/randomdev_soft.c src/sys/dev/random/randomdev_soft.h
- Remove; no longer needed.
* src/sys/dev/random/uint128.h
- Provide a fake uint128_t; if a real one ever arrived, we can use
that instead. All that is needed here is N=0, N++, N==0, and some
localised trickery is used to manufacture a 128-bit 0ULLL.
* src/sys/dev/random/unit_test.c src/sys/dev/random/unit_test.h
- Improve unit tests; previously the testing human needed clairvoyance;
now the test will do a basic check of compressibility. Clairvoyant
talent is still a good idea.
- This is still a long way off a proper unit test.
* src/sys/dev/random/fortuna.c src/sys/dev/random/fortuna.h
- Improve messy union to just uint128_t.
- Remove unneeded 'static struct fortuna_start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
* src/sys/dev/random/yarrow.c src/sys/dev/random/yarrow.h
- Improve messy union to just uint128_t.
- Remove unneeded 'staic struct start_cache'.
- Tighten up up arithmetic.
- Provide a method to allow eternal junk to be introduced; harden
it against blatant by compress/hashing.
- Assert that locks are held correctly.
- Fix the nasty pre- and post-read overloading by providing explictit
functions to do these tasks.
- Turn into self-sufficient module (no longer requires randomdev_soft.[ch])
- Fix some magic numbers elsewhere used as FAST and SLOW.
Differential Revision: https://reviews.freebsd.org/D2025
Reviewed by: vsevolod,delphij,rwatson,trasz,jmg
Approved by: so (delphij)
remains. Xen is planning to phase out support for PV upstream since it
is harder to maintain and has more overhead. Modern x86 CPUs include
virtualization extensions that support HVM guests instead of PV guests.
In addition, the PV code was i386 only and not as well maintained recently
as the HVM code.
- Remove the i386-only NATIVE option that was used to disable certain
components for PV kernels. These components are now standard as they
are on amd64.
- Remove !XENHVM bits from PV drivers.
- Remove various shims required for XEN (e.g. PT_UPDATES_FLUSH, LOAD_CR3,
etc.)
- Remove duplicate copy of <xen/features.h>.
- Remove unused, i386-only xenstored.h.
Differential Revision: https://reviews.freebsd.org/D2362
Reviewed by: royger
Tested by: royger (i386/amd64 HVM domU and amd64 PVH dom0)
Relnotes: yes
This code has had an extensive rewrite and a good series of reviews, both by the author and other parties. This means a lot of code has been simplified. Pluggable structures for high-rate entropy generators are available, and it is most definitely not the case that /dev/random can be driven by only a hardware souce any more. This has been designed out of the device. Hardware sources are stirred into the CSPRNG (Yarrow, Fortuna) like any other entropy source. Pluggable modules may be written by third parties for additional sources.
The harvesting structures and consequently the locking have been simplified. Entropy harvesting is done in a more general way (the documentation for this will follow). There is some GREAT entropy to be had in the UMA allocator, but it is disabled for now as messing with that is likely to annoy many people.
The venerable (but effective) Yarrow algorithm, which is no longer supported by its authors now has an alternative, Fortuna. For now, Yarrow is retained as the default algorithm, but this may be changed using a kernel option. It is intended to make Fortuna the default algorithm for 11.0. Interested parties are encouraged to read ISBN 978-0-470-47424-2 "Cryptography Engineering" By Ferguson, Schneier and Kohno for Fortuna's gory details. Heck, read it anyway.
Many thanks to Arthur Mesh who did early grunt work, and who got caught in the crossfire rather more than he deserved to.
My thanks also to folks who helped me thresh this out on whiteboards and in the odd "Hallway track", or otherwise.
My Nomex pants are on. Let the feedback commence!
Reviewed by: trasz,des(partial),imp(partial?),rwatson(partial?)
Approved by: so(des)
interrupts and report the largest value seen as sysctl
debug.max_kstack_used. Useful to estimate how close the kernel stack
size is to overflow.
In collaboration with: Larry Baird <lab@gta.com>
Sponsored by: The FreeBSD Foundation (kib)
MFC after: 1 week
than u_char.
Migrate post_filter to use an int for a CPU rather than u_char.
Change intr_event_bind() to use an int for CPU rather than u_char.
It touches the ppc, sparc64, arm and mips machdep code but it should
(hah!) be a no-op.
Tested:
* i386, AMD64 laptops
Reviewed by: jhb
These changes prevent sysctl(8) from returning proper output,
such as:
1) no output from sysctl(8)
2) erroneously returning ENOMEM with tools like truss(1)
or uname(1)
truss: can not get etype: Cannot allocate memory
there is an environment variable which shall initialize the SYSCTL
during early boot. This works for all SYSCTL types both statically and
dynamically created ones, except for the SYSCTL NODE type and SYSCTLs
which belong to VNETs. A new flag, CTLFLAG_NOFETCH, has been added to
be used in the case a tunable sysctl has a custom initialisation
function allowing the sysctl to still be marked as a tunable. The
kernel SYSCTL API is mostly the same, with a few exceptions for some
special operations like iterating childrens of a static/extern SYSCTL
node. This operation should probably be made into a factored out
common macro, hence some device drivers use this. The reason for
changing the SYSCTL API was the need for a SYSCTL parent OID pointer
and not only the SYSCTL parent OID list pointer in order to quickly
generate the sysctl path. The motivation behind this patch is to avoid
parameter loading cludges inside the OFED driver subsystem. Instead of
adding special code to the OFED driver subsystem to post-load tunables
into dynamically created sysctls, we generalize this in the kernel.
Other changes:
- Corrected a possibly incorrect sysctl name from "hw.cbb.intr_mask"
to "hw.pcic.intr_mask".
- Removed redundant TUNABLE statements throughout the kernel.
- Some minor code rewrites in connection to removing not needed
TUNABLE statements.
- Added a missing SYSCTL_DECL().
- Wrapped two very long lines.
- Avoid malloc()/free() inside sysctl string handling, in case it is
called to initialize a sysctl from a tunable, hence malloc()/free() is
not ready when sysctls from the sysctl dataset are registered.
- Bumped FreeBSD version to indicate SYSCTL API change.
MFC after: 2 weeks
Sponsored by: Mellanox Technologies
binding their threads to particular CPU.
Changing ithread cpu mask is now performed by special cpuset_setithread().
It creates additional cpuset root group on first bind invocation.
No objection: jhb
Tested by: hiren
MFC after: 2 weeks
Sponsored by: Yandex LLC
Contains:
* Refactor the hardware RNG CPU instruction sources to feed into
the software mixer. This is unfinished. The actual harvesting needs
to be sorted out. Modified by me (see below).
* Remove 'frac' parameter from random_harvest(). This was never
used and adds extra code for no good reason.
* Remove device write entropy harvesting. This provided a weak
attack vector, was not very good at bootstrapping the device. To
follow will be a replacement explicit reseed knob.
* Separate out all the RANDOM_PURE sources into separate harvest
entities. This adds some secuity in the case where more than one
is present.
* Review all the code and fix anything obviously messy or inconsistent.
Address som review concerns while I'm here, like rename the pseudo-rng
to 'dummy'.
Submitted by: Arthur Mesh <arthurmesh@gmail.com> (the first item)
in the ithread code where we could lose ithread interrupts if
intr_event_schedule_thread() is called while the ithread is already
running. Effectively memory writes could be ordered incorrectly
such that the unatomic write of 0 to ithd->it_need (in ithread_loop)
could be delayed until after it was set to be triggered again by
intr_event_schedule_thread().
This was particularly a big problem for CAM because CAM optimizes
scheduling of ithreads such that it only signals camisr() when it
queues to an empty queue. This means that additional completion
events would not unstick CAM and quickly lead to a complete lockup
of the CAM stack.
To fix this use atomics in all places where ithd->it_need is accessed.
Submitted by: delphij, mav
Obtained from: TrueOS, iXsystems
MFC After: 1 week
the handler address. Add a mark to distinguish between filter and
handler.
Note that the arguments for both filter and handler are same.
Sponsored by: The FreeBSD Foundation
Reviewed by: jhb
MFC after: 1 week
Make loadavg calculation callout direct. There are several reasons for it:
- it is very simple and doesn't worth context switch to SWI;
- since SWI is no longer used here, we can remove twelve years old hack,
excluding this SWI from from the loadavg statistics;
- it fixes problem when eventtimer (HPET) shares interrupt with some other
device, and that interrupt thread counted as permanent loadavg of 1; now
loadavg accounted before that interrupt thread is scheduled.
Sponsored by: Google Summer of Code 2012, iXsystems inc.
Tested by: flo, marius, ian, Fabian Keil, markj
cpuset mask for the associated interrupt thread.
The text used above is verbatim from r195249 and the code should now be
in line with the intent of that commit.
as controlled by kern.random.sys.harvest.swi. SWI harvesting feeds into
the interrupt FIFO and each event is estimated as providing a single bit of
entropy.
Reviewed by: markm, obrien
MFC after: 2 weeks
In rare event when fast and ithread interrupts share the same vector
and the fast handler was registered first, we can end up trying to
schedule the ithread that is not created yet. The kernel built with
INVARIANTS then triggers an assertion.
Change the order to create the ithread first and only then add the
handler that needs it to the interrupt event handlers list.
Reviewed by: jhb
the cached name used for KTR_SCHED traces when a thread's name changes.
This way KTR_SCHED traces (and thus schedgraph) will notice when a thread's
name changes, most commonly via execve().
MFC after: 2 weeks
It seems strchr() and strrchr() are used more often than index() and
rindex(). Therefore, simply migrate all kernel code to use it.
For the XFS code, remove an empty line to make the code identical to
the code in the Linux kernel.
sintrcnt/sintrnames which are symbols containing the size of the 2
tables.
- For amd64/i386 remove the storage of intr* stuff from assembly files.
This area can be widely improved by applying the same to other
architectures and likely finding an unified approach among them and
move the whole code to be MI. More work in this area is expected to
happen fairly soon.
No MFC is previewed for this patch.
Tested by: pluknet
Reviewed by: jhb
Approved by: re (kib)
