virtio-blk and virtio-scsi invoke virtio_irqfd_notify() to send Used
Buffer Notifications from an IOThread. This involves an eventfd
write(2) syscall. Calling this repeatedly when completing multiple I/O
requests in a row is wasteful.
Use the defer_call() API to batch together virtio_irqfd_notify() calls
made during thread pool (aio=threads), Linux AIO (aio=native), and
io_uring (aio=io_uring) completion processing.
Behavior is unchanged for emulated devices that do not use
defer_call_begin()/defer_call_end() since defer_call() immediately
invokes the callback when called outside a
defer_call_begin()/defer_call_end() region.
fio rw=randread bs=4k iodepth=64 numjobs=8 IOPS increases by ~9% with a
single IOThread and 8 vCPUs. iodepth=1 decreases by ~1% but this could
be noise. Detailed performance data and configuration specifics are
available here:
https://gitlab.com/stefanha/virt-playbooks/-/tree/blk_io_plug-irqfd
This duplicates the BH that virtio-blk uses for batching. The next
commit will remove it.
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20230913200045.1024233-4-stefanha@redhat.com>
Reviewed-by: Michael S. Tsirkin <mst@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The synchronous bdrv_aio_cancel() function needs the acb's AioContext so
it can call aio_poll() to wait for cancellation.
It turns out that all users run under the BQL in the main AioContext, so
this callback is not needed.
Remove the callback, mark bdrv_aio_cancel() GLOBAL_STATE_CODE just like
its blk_aio_cancel() caller, and poll the main loop AioContext.
The purpose of this cleanup is to identify bdrv_aio_cancel() as an API
that does not work with the multi-queue block layer.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-ID: <20230912231037.826804-2-stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Klaus Jensen <k.jensen@samsung.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
thread_pool_free() might have been called on the `pool`, which would
be a reason for worker_thread() to quit. In this case,
`pool->request_cond` is been destroyed.
If worker_thread() didn't managed to signal `request_cond` before it
been destroyed by thread_pool_free(), we got:
util/qemu-thread-posix.c:198: qemu_cond_signal: Assertion `cond->initialized' failed.
One backtrace:
__GI___assert_fail (assertion=0x55555614abcb "cond->initialized", file=0x55555614ab88 "util/qemu-thread-posix.c", line=198,
function=0x55555614ad80 <__PRETTY_FUNCTION__.17104> "qemu_cond_signal") at assert.c:101
qemu_cond_signal (cond=0x7fffb800db30) at util/qemu-thread-posix.c:198
worker_thread (opaque=0x7fffb800dab0) at util/thread-pool.c:129
qemu_thread_start (args=0x7fffb8000b20) at util/qemu-thread-posix.c:505
start_thread (arg=<optimized out>) at pthread_create.c:486
Reported here:
https://lore.kernel.org/all/ZJwoK50FcnTSfFZ8@MacBook-Air-de-Roger.local/T/#u
To avoid issue, keep lock while sending a signal to `request_cond`.
Fixes: 900fa208f5 ("thread-pool: replace semaphore with condition variable")
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20230714152720.5077-1-anthony.perard@citrix.com>
Signed-off-by: Anthony PERARD <anthony.perard@citrix.com>
thread_pool_submit_aio() is always called on a pool taken from
qemu_get_current_aio_context(), and that is the only intended
use: each pool runs only in the same thread that is submitting
work to it, it can't run anywhere else.
Therefore simplify the thread_pool_submit* API and remove the
ThreadPool function parameter.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20230203131731.851116-5-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Use qemu_get_current_aio_context() where possible, since we always
submit work to the current thread anyways.
We want to also be sure that the thread submitting the work is
the same as the one processing the pool, to avoid adding
synchronization to the pool list.
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Message-Id: <20230203131731.851116-4-eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Just setting the max threads to 0 is enough to stop all workers.
Message-Id: <20220514065012.1149539-4-pbonzini@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Nicolas Saenz Julienne <nsaenzju@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Since commit f9fc8932b1 ("thread-posix: remove the posix semaphore
support", 2022-04-06) QemuSemaphore has its own mutex and condition
variable; this adds unnecessary overhead on I/O with small block sizes.
Check the QTAILQ directly instead of adding the indirection of a
semaphore's count. Using a semaphore has not been necessary since
qemu_cond_timedwait was introduced; the new code has to be careful about
spurious wakeups but it is simpler, for example thread_pool_cancel does
not have to worry about synchronizing the semaphore count with the number
of elements of pool->request_list.
Note that the return value of qemu_cond_timedwait (0 for timeout, 1 for
signal or spurious wakeup) is different from that of qemu_sem_timedwait
(-1 for timeout, 0 for success).
Reported-by: Lukáš Doktor <ldoktor@redhat.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Nicolas Saenz Julienne <nsaenzju@redhat.com>
Message-Id: <20220514065012.1149539-3-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The completion bottom half was scheduled within the pool->lock
critical section. That actually results in worse performance,
because the worker thread can run its own small critical section
and go to sleep before the bottom half starts running.
Note that this simple change does not produce an improvement without
changing the thread pool QemuSemaphore to a condition variable.
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Nicolas Saenz Julienne <nsaenzju@redhat.com>
Message-Id: <20220514065012.1149539-2-pbonzini@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The thread pool regulates itself: when idle, it kills threads until
empty, when in demand, it creates new threads until full. This behaviour
doesn't play well with latency sensitive workloads where the price of
creating a new thread is too high. For example, when paired with qemu's
'-mlock', or using safety features like SafeStack, creating a new thread
has been measured take multiple milliseconds.
In order to mitigate this let's introduce a new 'EventLoopBase'
property to set the thread pool size. The threads will be created during
the pool's initialization or upon updating the property's value, remain
available during its lifetime regardless of demand, and destroyed upon
freeing it. A properly characterized workload will then be able to
configure the pool to avoid any latency spikes.
Signed-off-by: Nicolas Saenz Julienne <nsaenzju@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Acked-by: Markus Armbruster <armbru@redhat.com>
Message-id: 20220425075723.20019-4-nsaenzju@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
- ran regexp "qemu_mutex_lock\(.*\).*\n.*if" to find targets
- replaced result with QEMU_LOCK_GUARD if all unlocks at function end
- replaced result with WITH_QEMU_LOCK_GUARD if unlock not at end
Signed-off-by: Daniel Brodsky <dnbrdsky@gmail.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-id: 20200404042108.389635-3-dnbrdsky@gmail.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
No header includes qemu-common.h after this commit, as prescribed by
qemu-common.h's file comment.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-Id: <20190523143508.25387-5-armbru@redhat.com>
[Rebased with conflicts resolved automatically, except for
include/hw/arm/xlnx-zynqmp.h hw/arm/nrf51_soc.c hw/arm/msf2-soc.c
block/qcow2-refcount.c block/qcow2-cluster.c block/qcow2-cache.c
target/arm/cpu.h target/lm32/cpu.h target/m68k/cpu.h target/mips/cpu.h
target/moxie/cpu.h target/nios2/cpu.h target/openrisc/cpu.h
target/riscv/cpu.h target/tilegx/cpu.h target/tricore/cpu.h
target/unicore32/cpu.h target/xtensa/cpu.h; bsd-user/main.c and
net/tap-bsd.c fixed up]
commit 3c80ca15 fixed a deadlock scenarion with nested aio_poll invocations.
However, the rescheduling of the completion BH introcuded unnecessary spinning
in the main-loop. On very fast file backends this can even lead to the
"WARNING: I/O thread spun for 1000 iterations" message popping up.
Callgrind reports about 3-4% less instructions with this patch running
qemu-img bench on a ramdisk based VMDK file.
Fixes: 3c80ca158c
Cc: qemu-stable@nongnu.org
Signed-off-by: Peter Lieven <pl@kamp.de>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 20170213135235.12274-16-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Message-id: 20170213135235.12274-15-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
AioContext is fairly self contained, the only dependency is QEMUTimer but
that in turn doesn't need anything else. So move them out of block-obj-y
to avoid introducing a dependency from io/ to block-obj-y.
main-loop and its dependency iohandler also need to be moved, because
later in this series io/ will call iohandler_get_aio_context.
[Changed copyright "the QEMU team" to "other QEMU contributors" as
suggested by Daniel Berrange and agreed by Paolo.
--Stefan]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Message-id: 20170213135235.12274-2-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
