Split icount stuff from system/cpu-timers.h.
There are 17 files which only require icount.h, 7 that only
require cpu-timers.h, and 7 that require both.
Reviewed-by: Pierrick Bouvier <pierrick.bouvier@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
Adaptive polling has a big problem: It doesn't consider that an event
loop can wait for many different events that may have very different
typical latencies.
For example, think of a guest that tends to send a new I/O request soon
after the previous I/O request completes, but the storage on the host is
rather slow. In this case, getting the new request from guest quickly
means that polling is enabled, but the next thing is performing the I/O
request on the backend, which is slow and disables polling again for the
next guest request. This means that in such a scenario, polling could
help for every other event, but is only ever enabled when it can't
succeed.
In order to fix this, keep a separate AioPolledEvent for each
AioHandler. We will then know that the backend file descriptor always
has a high latency and isn't worth polling for, but we also know that
the guest is always fast and we should poll for it. This solves at least
half of the problem, we can now keep polling for those cases where it
makes sense and get the improved performance from it.
Since the event loop doesn't know which event will be next, we still do
some unnecessary polling while we're waiting for the slow disk. I made
some attempts to be more clever than just randomly growing and shrinking
the polling time, and even to let callers be explicit about when they
expect a new event, but so far this hasn't resulted in improved
performance or even caused performance regressions. For now, let's just
fix the part that is easy enough to fix, we can revisit the rest later.
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20250307221634.71951-6-kwolf@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Headers in include/sysemu/ are not only related to system
*emulation*, they are also used by virtualization. Rename
as system/ which is clearer.
Files renamed manually then mechanical change using sed tool.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Tested-by: Lei Yang <leiyang@redhat.com>
Message-Id: <20241203172445.28576-1-philmd@linaro.org>
Currently the QemuLockCnt data structure and associated functions are
in the include/qemu/thread.h header. Move them to their own
qemu/lockcnt.h. The main reason for doing this is that it means we
can autogenerate the documentation comments into the docs/devel
documentation.
The copyright/author in the new header is drawn from lockcnt.c,
since the header changes were added in the same commit as
lockcnt.c; since neither thread.h nor lockcnt.c state an explicit
license, the standard default of GPL-2-or-later applies.
We include the new header (and the .c file, which was accidentally
omitted previously) in the "RCU" part of MAINTAINERS, since that
is where the lockcnt.rst documentation is categorized.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Acked-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 20240816132212.3602106-7-peter.maydell@linaro.org
aio_context_set_thread_pool_params() takes two int64_t arguments to
set the minimum and maximum number of threads in the pool. We do
some bounds checking on these, but we don't catch the case where the
inputs are negative. This means that later in the function when we
assign these inputs to the AioContext::thread_pool_min and
::thread_pool_max fields, which are of type int, the values might
overflow the smaller type.
A negative number of threads is meaningless, so make
aio_context_set_thread_pool_params() return an error if either min or
max are negative.
Resolves: Coverity CID 1547605
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Message-id: 20240723150927.1396456-1-peter.maydell@linaro.org
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The Big QEMU Lock (BQL) has many names and they are confusing. The
actual QemuMutex variable is called qemu_global_mutex but it's commonly
referred to as the BQL in discussions and some code comments. The
locking APIs, however, are called qemu_mutex_lock_iothread() and
qemu_mutex_unlock_iothread().
The "iothread" name is historic and comes from when the main thread was
split into into KVM vcpu threads and the "iothread" (now called the main
loop thread). I have contributed to the confusion myself by introducing
a separate --object iothread, a separate concept unrelated to the BQL.
The "iothread" name is no longer appropriate for the BQL. Rename the
locking APIs to:
- void bql_lock(void)
- void bql_unlock(void)
- bool bql_locked(void)
There are more APIs with "iothread" in their names. Subsequent patches
will rename them. There are also comments and documentation that will be
updated in later patches.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Acked-by: Fabiano Rosas <farosas@suse.de>
Acked-by: David Woodhouse <dwmw@amazon.co.uk>
Reviewed-by: Cédric Le Goater <clg@kaod.org>
Acked-by: Peter Xu <peterx@redhat.com>
Acked-by: Eric Farman <farman@linux.ibm.com>
Reviewed-by: Harsh Prateek Bora <harshpb@linux.ibm.com>
Acked-by: Hyman Huang <yong.huang@smartx.com>
Reviewed-by: Akihiko Odaki <akihiko.odaki@daynix.com>
Message-id: 20240102153529.486531-2-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Delete these functions because nothing calls these functions anymore.
I introduced these APIs in commit 98563fc3ec ("aio: add
aio_context_acquire() and aio_context_release()") in 2014. It's with a
sigh of relief that I delete these APIs almost 10 years later.
Thanks to Paolo Bonzini's vision for multi-queue QEMU, we got an
understanding of where the code needed to go in order to remove the
limitations that the original dataplane and the IOThread/AioContext
approach that followed it.
Emanuele Giuseppe Esposito had the splendid determination to convert
large parts of the codebase so that they no longer needed the AioContext
lock. This was a painstaking process, both in the actual code changes
required and the iterations of code review that Emanuele eked out of
Kevin and me over many months.
Kevin Wolf tackled multitudes of graph locking conversions to protect
in-flight I/O from run-time changes to the block graph as well as the
clang Thread Safety Analysis annotations that allow the compiler to
check whether the graph lock is being used correctly.
And me, well, I'm just here to add some pizzazz to the QEMU multi-queue
block layer :). Thank you to everyone who helped with this effort,
including Eric Blake, code reviewer extraordinaire, and others who I've
forgotten to mention.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-ID: <20231205182011.1976568-11-stefanha@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
This is the big patch that removes
aio_context_acquire()/aio_context_release() from the block layer and
affected block layer users.
There isn't a clean way to split this patch and the reviewers are likely
the same group of people, so I decided to do it in one patch.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Reviewed-by: Paul Durrant <paul@xen.org>
Message-ID: <20231205182011.1976568-7-stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
aio_context_acquire()/aio_context_release() has been replaced by
fine-grained locking to protect state shared by multiple threads. The
AioContext lock still plays the role of balancing locking in
AIO_WAIT_WHILE() and many functions in QEMU either require that the
AioContext lock is held or not held for this reason. In other words, the
AioContext lock is purely there for consistency with itself and serves
no real purpose anymore.
Stop actually acquiring/releasing the lock in
aio_context_acquire()/aio_context_release() so that subsequent patches
can remove callers across the codebase incrementally.
I have performed "make check" and qemu-iotests stress tests across
x86-64, ppc64le, and aarch64 to confirm that there are no failures as a
result of eliminating the lock.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Acked-by: Kevin Wolf <kwolf@redhat.com>
Message-ID: <20231205182011.1976568-5-stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
All callers now pass is_external=false to aio_set_fd_handler() and
aio_set_event_notifier(). The aio_disable_external() API that
temporarily disables fd handlers that were registered is_external=true
is therefore dead code.
Remove aio_disable_external(), aio_enable_external(), and the
is_external arguments to aio_set_fd_handler() and
aio_set_event_notifier().
The entire test-fdmon-epoll test is removed because its sole purpose was
testing aio_disable_external().
Parts of this patch were generated using the following coccinelle
(https://coccinelle.lip6.fr/) semantic patch:
@@
expression ctx, fd, is_external, io_read, io_write, io_poll, io_poll_ready, opaque;
@@
- aio_set_fd_handler(ctx, fd, is_external, io_read, io_write, io_poll, io_poll_ready, opaque)
+ aio_set_fd_handler(ctx, fd, io_read, io_write, io_poll, io_poll_ready, opaque)
@@
expression ctx, notifier, is_external, io_read, io_poll, io_poll_ready;
@@
- aio_set_event_notifier(ctx, notifier, is_external, io_read, io_poll, io_poll_ready)
+ aio_set_event_notifier(ctx, notifier, io_read, io_poll, io_poll_ready)
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-Id: <20230516190238.8401-21-stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
* Fix compilation issues under Debian 10
* Update kernel headers to 6.3rc5
* Suppress GCC13 false positive in aio_bh_poll()
* Add new x86 feature bits
* Coverity fixes
* More steps towards removing qatomic_mb_set/read
* Fix reduced-phys-bits value for AMD SEV
# -----BEGIN PGP SIGNATURE-----
#
# iQFIBAABCAAyFiEE8TM4V0tmI4mGbHaCv/vSX3jHroMFAmRNC0IUHHBib256aW5p
# QHJlZGhhdC5jb20ACgkQv/vSX3jHroNo0wgArWNGKZpbmQ0e5L6ajMvaaPmg4mVL
# a2SJGU0TwTp0fUgZr14z2iwzIpSqQrsqhzTIAzOTs0OICDBPBuNvnRucMa+SVQGO
# Tc89YAwBVDo66dAKhWi+WR9tx7sTFCso0nbsBfczzdnwAw3g1MJ87Ueqc5tlPGBK
# E7YSAD6l4UuogoN5BLU7bSsG/X7bwcyzeUXRB4ik+Z9abWd4DH9qiROnBKLMmBLK
# nAi47h8b8MltWORpO+wf6HtkMKi37SAzl9VLHVuHcRhIdY/JhWCRhYSo0HXhgX66
# JLVkyxFpIndT0dUW/xnqATGez92FRZyTxHbxbAcWM0SoC1jOVfUXB+7Gdw==
# =vxou
# -----END PGP SIGNATURE-----
# gpg: Signature made Sat 29 Apr 2023 01:19:14 PM BST
# gpg: using RSA key F13338574B662389866C7682BFFBD25F78C7AE83
# gpg: issuer "pbonzini@redhat.com"
# gpg: Good signature from "Paolo Bonzini <bonzini@gnu.org>" [undefined]
# gpg: aka "Paolo Bonzini <pbonzini@redhat.com>" [undefined]
# gpg: WARNING: This key is not certified with a trusted signature!
# gpg: There is no indication that the signature belongs to the owner.
# Primary key fingerprint: 46F5 9FBD 57D6 12E7 BFD4 E2F7 7E15 100C CD36 69B1
# Subkey fingerprint: F133 3857 4B66 2389 866C 7682 BFFB D25F 78C7 AE83
* tag 'for-upstream' of https://gitlab.com/bonzini/qemu:
cpus-common: stop using mb_set/mb_read
async: Suppress GCC13 false positive in aio_bh_poll()
tests: vhost-user-test: release mutex on protocol violation
Update linux headers to v6.3rc5
update-linux-headers.sh: Add missing kernel headers.
Fix libvhost-user.c compilation.
target/i386: Add support for PREFETCHIT0/1 in CPUID enumeration
target/i386: Add support for AVX-NE-CONVERT in CPUID enumeration
target/i386: Add support for AVX-VNNI-INT8 in CPUID enumeration
target/i386: Add support for AVX-IFMA in CPUID enumeration
target/i386: Add support for AMX-FP16 in CPUID enumeration
target/i386: Add support for CMPCCXADD in CPUID enumeration
i386/cpu: Update how the EBX register of CPUID 0x8000001F is set
i386/sev: Update checks and information related to reduced-phys-bits
qemu-options.hx: Update the reduced-phys-bits documentation
qapi, i386/sev: Change the reduced-phys-bits value from 5 to 1
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
GCC13 reports an error :
../util/async.c: In function ‘aio_bh_poll’:
include/qemu/queue.h:303:22: error: storing the address of local variable ‘slice’ in ‘*ctx.bh_slice_list.sqh_last’ [-Werror=dangling-pointer=]
303 | (head)->sqh_last = &(elm)->field.sqe_next; \
| ~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~
../util/async.c:169:5: note: in expansion of macro ‘QSIMPLEQ_INSERT_TAIL’
169 | QSIMPLEQ_INSERT_TAIL(&ctx->bh_slice_list, &slice, next);
| ^~~~~~~~~~~~~~~~~~~~
../util/async.c:161:17: note: ‘slice’ declared here
161 | BHListSlice slice;
| ^~~~~
../util/async.c:161:17: note: ‘ctx’ declared here
But the local variable 'slice' is removed from the global context list
in following loop of the same routine. Add a pragma to silent GCC.
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Cc: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Cédric Le Goater <clg@redhat.com>
Reviewed-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Tested-by: Daniel Henrique Barboza <danielhb413@gmail.com>
Message-Id: <20230420202939.1982044-1-clg@kaod.org>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Devices can pass their MemoryReentrancyGuard (from their DeviceState),
when creating new BHes. Then, the async API will toggle the guard
before/after calling the BH call-back. This prevents bh->mmio reentrancy
issues.
Signed-off-by: Alexander Bulekov <alxndr@bu.edu>
Reviewed-by: Darren Kenny <darren.kenny@oracle.com>
Message-Id: <20230427211013.2994127-3-alxndr@bu.edu>
[thuth: Fix "line over 90 characters" checkpatch.pl error]
Signed-off-by: Thomas Huth <thuth@redhat.com>
Explain that aio_context_notifier_poll() relies on
aio_notify_accept() to catch all the memory writes that were
done before ctx->notified was set to true.
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Ever since commit 8c6b0356b5 ("util/async: make bh_aio_poll() O(1)",
2020-02-22), synchronization between qemu_bh_schedule() and aio_bh_poll()
is happening when the bottom half is enqueued in the bh_list; not
when the flags are set. Update the documentation to match.
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Add/remove the AioContext in aio_context_list in graph-lock.c when it is
created/destroyed. This allows using the graph locking operations from
this AioContext.
In order to allow linking util/async.c with binaries that don't include
the block layer, introduce stubs for (un)register_aiocontext().
Signed-off-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20221207131838.239125-5-kwolf@redhat.com>
Reviewed-by: Emanuele Giuseppe Esposito <eesposit@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
vCPU execution should be suspended when new BH is scheduled.
This is needed to avoid guest timeouts caused by the long cycles
of the execution. In replay mode execution may hang when
vCPU sleeps and block event comes to the queue.
This patch adds notification which wakes up vCPU or interrupts
execution of guest code.
Signed-off-by: Pavel Dovgalyuk <Pavel.Dovgalyuk@ispras.ru>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
--
v2: changed first_cpu to current_cpu (suggested by Richard Henderson)
v4: moved vCPU notification to aio_bh_enqueue (suggested by Paolo Bonzini)
Message-Id: <165364837317.688121.17680519919871405281.stgit@pasha-ThinkPad-X280>
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>
Adaptive polling measures the execution time of the polling check plus
handlers called when a polled event becomes ready. Handlers can take a
significant amount of time, making it look like polling was running for
a long time when in fact the event handler was running for a long time.
For example, on Linux the io_submit(2) syscall invoked when a virtio-blk
device's virtqueue becomes ready can take 10s of microseconds. This
can exceed the default polling interval (32 microseconds) and cause
adaptive polling to stop polling.
By excluding the handler's execution time from the polling check we make
the adaptive polling calculation more accurate. As a result, the event
loop now stays in polling mode where previously it would have fallen
back to file descriptor monitoring.
The following data was collected with virtio-blk num-queues=2
event_idx=off using an IOThread. Before:
168k IOPS, IOThread syscalls:
9837.115 ( 0.020 ms): IO iothread1/620155 io_submit(ctx_id: 140512552468480, nr: 16, iocbpp: 0x7fcb9f937db0) = 16
9837.158 ( 0.002 ms): IO iothread1/620155 write(fd: 103, buf: 0x556a2ef71b88, count: 8) = 8
9837.161 ( 0.001 ms): IO iothread1/620155 write(fd: 104, buf: 0x556a2ef71b88, count: 8) = 8
9837.163 ( 0.001 ms): IO iothread1/620155 ppoll(ufds: 0x7fcb90002800, nfds: 4, tsp: 0x7fcb9f1342d0, sigsetsize: 8) = 3
9837.164 ( 0.001 ms): IO iothread1/620155 read(fd: 107, buf: 0x7fcb9f939cc0, count: 512) = 8
9837.174 ( 0.001 ms): IO iothread1/620155 read(fd: 105, buf: 0x7fcb9f939cc0, count: 512) = 8
9837.176 ( 0.001 ms): IO iothread1/620155 read(fd: 106, buf: 0x7fcb9f939cc0, count: 512) = 8
9837.209 ( 0.035 ms): IO iothread1/620155 io_submit(ctx_id: 140512552468480, nr: 32, iocbpp: 0x7fca7d0cebe0) = 32
174k IOPS (+3.6%), IOThread syscalls:
9809.566 ( 0.036 ms): IO iothread1/623061 io_submit(ctx_id: 140539805028352, nr: 32, iocbpp: 0x7fd0cdd62be0) = 32
9809.625 ( 0.001 ms): IO iothread1/623061 write(fd: 103, buf: 0x5647cfba5f58, count: 8) = 8
9809.627 ( 0.002 ms): IO iothread1/623061 write(fd: 104, buf: 0x5647cfba5f58, count: 8) = 8
9809.663 ( 0.036 ms): IO iothread1/623061 io_submit(ctx_id: 140539805028352, nr: 32, iocbpp: 0x7fd0d0388b50) = 32
Notice that ppoll(2) and eventfd read(2) syscalls are eliminated because
the IOThread stays in polling mode instead of falling back to file
descriptor monitoring.
As usual, polling is not implemented on Windows so this patch ignores
the new io_poll_read() callback in aio-win32.c.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Stefano Garzarella <sgarzare@redhat.com>
Message-id: 20211207132336.36627-2-stefanha@redhat.com
[Fixed up aio_set_event_notifier() calls in
tests/unit/test-fdmon-epoll.c added after this series was queued.
--Stefan]
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
