This patch is based on the algorithm for the kvm.ko halt_poll_ns
parameter in Linux. The initial polling time is zero.
If the event loop is woken up within the maximum polling time it means
polling could be effective, so grow polling time.
If the event loop is woken up beyond the maximum polling time it means
polling is not effective, so shrink polling time.
If the event loop makes progress within the current polling time then
the sweet spot has been reached.
This algorithm adjusts the polling time so it can adapt to variations in
workloads. The goal is to reach the sweet spot while also recognizing
when polling would hurt more than help.
Two new trace events, poll_grow and poll_shrink, are added for observing
polling time adjustment.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 20161201192652.9509-13-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The AioContext event loop uses ppoll(2) or epoll_wait(2) to monitor file
descriptors or until a timer expires. In cases like virtqueues, Linux
AIO, and ThreadPool it is technically possible to wait for events via
polling (i.e. continuously checking for events without blocking).
Polling can be faster than blocking syscalls because file descriptors,
the process scheduler, and system calls are bypassed.
The main disadvantage to polling is that it increases CPU utilization.
In classic polling configuration a full host CPU thread might run at
100% to respond to events as quickly as possible. This patch implements
a timeout so we fall back to blocking syscalls if polling detects no
activity. After the timeout no CPU cycles are wasted on polling until
the next event loop iteration.
The run_poll_handlers_begin() and run_poll_handlers_end() trace events
are added to aid performance analysis and troubleshooting. If you need
to know whether polling mode is being used, trace these events to find
out.
Note that the AioContext is now re-acquired before disabling notify_me
in the non-polling case. This makes the code cleaner since notify_me
was enabled outside the non-polling AioContext release region. This
change is correct since it's safe to keep notify_me enabled longer
(disabling is an optimization) but potentially causes unnecessary
event_notifer_set() calls. I think the chance of performance regression
is small here.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 20161201192652.9509-4-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The new AioPollFn io_poll() argument to aio_set_fd_handler() and
aio_set_event_handler() is used in the next patch.
Keep this code change separate due to the number of files it touches.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 20161201192652.9509-3-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Polling mode will not call ppoll(2)/epoll_wait(2). Therefore we know
there are no fds ready and should avoid looping over fd handlers in
aio_dispatch().
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 20161201192652.9509-2-stefanha@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
It is simpler and a bit faster, and QEMU does not need the contention
callbacks (and thus the fairness) anymore.
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <1477565348-5458-21-git-send-email-pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
This is the first step towards having fine-grained critical sections in
dataplane threads, which will resolve lock ordering problems between
address_space_* functions (which need the BQL when doing MMIO, even
after we complete RCU-based dispatch) and the AioContext.
Because AioContext does not use contention callbacks anymore, the
unit test has to be changed.
Previously applied as a0710f7995 and
then reverted.
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <1477565348-5458-19-git-send-email-pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
aio_poll is not thread safe; for example bdrv_drain can hang if
the last in-flight I/O operation is completed in the I/O thread after
the main thread has checked bs->in_flight.
The bug remains latent as long as all of it is called within
aio_context_acquire/aio_context_release, but this will change soon.
To fix this, if bdrv_drain is called from outside the I/O thread,
signal the main AioContext through a dummy bottom half. The event
loop then only runs in the I/O thread.
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-Id: <1477565348-5458-18-git-send-email-pbonzini@redhat.com>
Signed-off-by: Fam Zheng <famz@redhat.com>
qemu_bh_delete is already clearing bh->scheduled at the same time
as it's setting bh->deleted. Since it's not using any memory
barriers, there is no synchronization going on for bh->deleted,
and this makes the bh->deleted checks superfluous in aio_compute_timeout,
aio_bh_poll and aio_ctx_check.
Just remove them, and put the (bh->scheduled && bh->deleted) combo
to work in a new function aio_bh_schedule_oneshot. The new function
removes the need to save the QEMUBH pointer between the creation
and the execution of the bottom half.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Parameter **errp of aio_context_setup() is useless, remove it
and clean up the related code.
Cc: Stefan Hajnoczi <stefanha@redhat.com>
Cc: Fam Zheng <famz@redhat.com>
Cc: Eric Blake <eblake@redhat.com>
Signed-off-by: Cao jin <caoj.fnst@cn.fujitsu.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-id: 1468578524-23433-1-git-send-email-caoj.fnst@cn.fujitsu.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This has better performance because it executes fewer system calls
and does not use a bottom half per disk.
Originally proposed by Ming Lei.
[Changed #include "raw-aio.h" to "block/raw-aio.h" in win32-aio.c to fix
build error as reported by Peter Maydell <peter.maydell@linaro.org>.
--Stefan]
Acked-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1467650000-51385-1-git-send-email-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
squash! linux-aio: share one LinuxAioState within an AioContext
Commit 57cb38b included qapi/error.h into qemu/osdep.h to get the
Error typedef. Since then, we've moved to include qemu/osdep.h
everywhere. Its file comment explains: "To avoid getting into
possible circular include dependencies, this file should not include
any other QEMU headers, with the exceptions of config-host.h,
compiler.h, os-posix.h and os-win32.h, all of which are doing a
similar job to this file and are under similar constraints."
qapi/error.h doesn't do a similar job, and it doesn't adhere to
similar constraints: it includes qapi-types.h. That's in excess of
100KiB of crap most .c files don't actually need.
Add the typedef to qemu/typedefs.h, and include that instead of
qapi/error.h. Include qapi/error.h in .c files that need it and don't
get it now. Include qapi-types.h in qom/object.h for uint16List.
Update scripts/clean-includes accordingly. Update it further to match
reality: replace config.h by config-target.h, add sysemu/os-posix.h,
sysemu/os-win32.h. Update the list of includes in the qemu/osdep.h
comment quoted above similarly.
This reduces the number of objects depending on qapi/error.h from "all
of them" to less than a third. Unfortunately, the number depending on
qapi-types.h shrinks only a little. More work is needed for that one.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
[Fix compilation without the spice devel packages. - Paolo]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Clean up includes so that osdep.h is included first and headers
which it implies are not included manually.
This commit was created with scripts/clean-includes.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 1454089805-5470-16-git-send-email-peter.maydell@linaro.org
This is the place to initialize platform specific bits of AioContext.
Signed-off-by: Fam Zheng <famz@redhat.com>
Message-id: 1446177989-6702-3-git-send-email-famz@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This patch introduces aio_bh_call function. It is used to execute
bottom halves as callbacks without adding them to the queue.
Signed-off-by: Pavel Dovgalyuk <pavel.dovgaluk@ispras.ru>
Message-Id: <20150917162450.8676.56980.stgit@PASHA-ISP.def.inno>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Pavel Dovgalyuk <Pavel.Dovgaluk@ispras.ru>
All callers pass in false, and the real external ones will switch to
true in coming patches.
Signed-off-by: Fam Zheng <famz@redhat.com>
Reviewed-by: Jeff Cody <jcody@redhat.com>
Reviewed-by: Kevin Wolf <kwolf@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
The notify_me optimization introduced in commit eabc977973
("AioContext: fix broken ctx->dispatching optimization") skips
event_notifier_set() calls when the event loop thread is not blocked in
ppoll(2).
This optimization causes a deadlock if two aio_context_acquire() calls
race. notify_me = 0 during the race so the winning thread can enter
ppoll(2) unaware that the other thread is waiting its turn to acquire
the AioContext.
This patch forces ppoll(2) to return by scheduling a BH instead of
calling aio_notify().
The following deadlock with virtio-blk dataplane is fixed:
qemu ... -object iothread,id=iothread0 \
-drive if=none,id=drive0,file=test.img,... \
-device virtio-blk-pci,iothread=iothread0,drive=drive0
This command-line results in a hang early on without this patch.
Thanks to Paolo Bonzini <pbonzini@redhat.com> for investigating this bug
with me.
Cc: Christian Borntraeger <borntraeger@de.ibm.com>
Cc: Cornelia Huck <cornelia.huck@de.ibm.com>
Cc: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1438101249-25166-4-git-send-email-pbonzini@redhat.com
Message-Id: <1438014819-18125-3-git-send-email-stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
BHs are freed during aio_bh_poll(). This leads to memory leaks if there
is no aio_bh_poll() between qemu_bh_delete() and aio_ctx_finalize().
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1438101249-25166-3-git-send-email-pbonzini@redhat.com
Message-Id: <1438014819-18125-2-git-send-email-stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
It is pretty rare for aio_notify to actually set the EventNotifier. It
can happen with worker threads such as thread-pool.c's, but otherwise it
should never be set thanks to the ctx->notify_me optimization. The
previous patch, unfortunately, added an unconditional call to
event_notifier_test_and_clear; now add a userspace fast path that
avoids the call.
Note that it is not possible to do the same with event_notifier_set;
it would break, as proved (again) by the included formal model.
This patch survived over 3000 reboots on aarch64 KVM.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Tested-by: Richard W.M. Jones <rjones@redhat.com>
Message-id: 1437487673-23740-7-git-send-email-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
event_notifier_test_and_clear must be called before processing events.
Otherwise, an aio_poll could "eat" the notification before the main
I/O thread invokes ppoll(). The main I/O thread then never wakes up.
This is an example of what could happen:
i/o thread vcpu thread worker thread
---------------------------------------------------------------------
lock_iothread
notify_me = 1
...
unlock_iothread
bh->scheduled = 1
event_notifier_set
lock_iothread
notify_me = 3
ppoll
notify_me = 1
aio_dispatch
aio_bh_poll
thread_pool_completion_bh
bh->scheduled = 1
event_notifier_set
node->io_read(node->opaque)
event_notifier_test_and_clear
ppoll
*** hang ***
"Tracing" with qemu_clock_get_ns shows pretty much the same behavior as
in the previous bug, so there are no new tricks here---just stare more
at the code until it is apparent.
One could also use a formal model, of course. The included one shows
this with three processes: notifier corresponds to a QEMU thread pool
worker, temporary_waiter to a VCPU thread that invokes aio_poll(),
waiter to the main I/O thread. I would be happy to say that the
formal model found the bug for me, but actually I wrote it after the
fact.
This patch is a bit of a big hammer. The next one optimizes it,
with help (this time for real rather than a posteriori :)) from
another, similar formal model.
Reported-by: Richard W. M. Jones <rjones@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Tested-by: Richard W.M. Jones <rjones@redhat.com>
Message-id: 1437487673-23740-6-git-send-email-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This patch rewrites the ctx->dispatching optimization, which was the cause
of some mysterious hangs that could be reproduced on aarch64 KVM only.
The hangs were indirectly caused by aio_poll() and in particular by
flash memory updates's call to blk_write(), which invokes aio_poll().
Fun stuff: they had an extremely short race window, so much that
adding all kind of tracing to either the kernel or QEMU made it
go away (a single printf made it half as reproducible).
On the plus side, the failure mode (a hang until the next keypress)
made it very easy to examine the state of the process with a debugger.
And there was a very nice reproducer from Laszlo, which failed pretty
often (more than half of the time) on any version of QEMU with a non-debug
kernel; it also failed fast, while still in the firmware. So, it could
have been worse.
For some unknown reason they happened only with virtio-scsi, but
that's not important. It's more interesting that they disappeared with
io=native, making thread-pool.c a likely suspect for where the bug arose.
thread-pool.c is also one of the few places which use bottom halves
across threads, by the way.
I hope that no other similar bugs exist, but just in case :) I am
going to describe how the successful debugging went... Since the
likely culprit was the ctx->dispatching optimization, which mostly
affects bottom halves, the first observation was that there are two
qemu_bh_schedule() invocations in the thread pool: the one in the aio
worker and the one in thread_pool_completion_bh. The latter always
causes the optimization to trigger, the former may or may not. In
order to restrict the possibilities, I introduced new functions
qemu_bh_schedule_slow() and qemu_bh_schedule_fast():
/* qemu_bh_schedule_slow: */
ctx = bh->ctx;
bh->idle = 0;
if (atomic_xchg(&bh->scheduled, 1) == 0) {
event_notifier_set(&ctx->notifier);
}
/* qemu_bh_schedule_fast: */
ctx = bh->ctx;
bh->idle = 0;
assert(ctx->dispatching);
atomic_xchg(&bh->scheduled, 1);
Notice how the atomic_xchg is still in qemu_bh_schedule_slow(). This
was already debated a few months ago, so I assumed it to be correct.
In retrospect this was a very good idea, as you'll see later.
Changing thread_pool_completion_bh() to qemu_bh_schedule_fast() didn't
trigger the assertion (as expected). Changing the worker's invocation
to qemu_bh_schedule_slow() didn't hide the bug (another assumption
which luckily held). This already limited heavily the amount of
interaction between the threads, hinting that the problematic events
must have triggered around thread_pool_completion_bh().
As mentioned early, invoking a debugger to examine the state of a
hung process was pretty easy; the iothread was always waiting on a
poll(..., -1) system call. Infinite timeouts are much rarer on x86,
and this could be the reason why the bug was never observed there.
With the buggy sequence more or less resolved to an interaction between
thread_pool_completion_bh() and poll(..., -1), my "tracing" strategy was
to just add a few qemu_clock_get_ns(QEMU_CLOCK_REALTIME) calls, hoping
that the ordering of aio_ctx_prepare(), aio_ctx_dispatch, poll() and
qemu_bh_schedule_fast() would provide some hint. The output was:
(gdb) p last_prepare
$3 = 103885451
(gdb) p last_dispatch
$4 = 103876492
(gdb) p last_poll
$5 = 115909333
(gdb) p last_schedule
$6 = 115925212
Notice how the last call to qemu_poll_ns() came after aio_ctx_dispatch().
This makes little sense unless there is an aio_poll() call involved,
and indeed with a slightly different instrumentation you can see that
there is one:
(gdb) p last_prepare
$3 = 107569679
(gdb) p last_dispatch
$4 = 107561600
(gdb) p last_aio_poll
$5 = 110671400
(gdb) p last_schedule
$6 = 110698917
So the scenario becomes clearer:
iothread VCPU thread
--------------------------------------------------------------------------
aio_ctx_prepare
aio_ctx_check
qemu_poll_ns(timeout=-1)
aio_poll
aio_dispatch
thread_pool_completion_bh
qemu_bh_schedule()
At this point bh->scheduled = 1 and the iothread has not been woken up.
The solution must be close, but this alone should not be a problem,
because the bottom half is only rescheduled to account for rare situations
(see commit 3c80ca1, thread-pool: avoid deadlock in nested aio_poll()
calls, 2014-07-15).
Introducing a third thread---a thread pool worker thread, which
also does qemu_bh_schedule()---does bring out the problematic case.
The third thread must be awakened *after* the callback is complete and
thread_pool_completion_bh has redone the whole loop, explaining the
short race window. And then this is what happens:
thread pool worker
--------------------------------------------------------------------------
<I/O completes>
qemu_bh_schedule()
Tada, bh->scheduled is already 1, so qemu_bh_schedule() does nothing
and the iothread is never woken up. This is where the bh->scheduled
optimization comes into play---it is correct, but removing it would
have masked the bug.
So, what is the bug?
Well, the question asked by the ctx->dispatching optimization ("is any
active aio_poll dispatching?") was wrong. The right question to ask
instead is "is any active aio_poll *not* dispatching", i.e. in the prepare
or poll phases? In that case, the aio_poll is sleeping or might go to
sleep anytime soon, and the EventNotifier must be invoked to wake
it up.
In any other case (including if there is *no* active aio_poll at all!)
we can just wait for the next prepare phase to pick up the event (e.g. a
bottom half); the prepare phase will avoid the blocking and service the
bottom half.
Expressing the invariant with a logic formula, the broken one looked like:
!(exists(thread): in_dispatching(thread)) => !optimize
or equivalently:
!(exists(thread):
in_aio_poll(thread) && in_dispatching(thread)) => !optimize
In the correct one, the negation is in a slightly different place:
(exists(thread):
in_aio_poll(thread) && !in_dispatching(thread)) => !optimize
or equivalently:
(exists(thread): in_prepare_or_poll(thread)) => !optimize
Even if the difference boils down to moving an exclamation mark :)
the implementation is quite different. However, I think the new
one is simpler to understand.
In the old implementation, the "exists" was implemented with a boolean
value. This didn't really support well the case of multiple concurrent
event loops, but I thought that this was okay: aio_poll holds the
AioContext lock so there cannot be concurrent aio_poll invocations, and
I was just considering nested event loops. However, aio_poll _could_
indeed be concurrent with the GSource. This is why I came up with the
wrong invariant.
In the new implementation, "exists" is computed simply by counting how many
threads are in the prepare or poll phases. There are some interesting
points to consider, but the gist of the idea remains:
1) AioContext can be used through GSource as well; as mentioned in the
patch, bit 0 of the counter is reserved for the GSource.
2) the counter need not be updated for a non-blocking aio_poll, because
it won't sleep forever anyway. This is just a matter of checking
the "blocking" variable. This requires some changes to the win32
implementation, but is otherwise not too complicated.
3) as mentioned above, the new implementation will not call aio_notify
when there is *no* active aio_poll at all. The tests have to be
adjusted for this change. The calls to aio_notify in async.c are fine;
they only want to kick aio_poll out of a blocking wait, but need not
do anything if aio_poll is not running.
4) nested aio_poll: these just work with the new implementation; when
a nested event loop is invoked, the outer event loop is never in the
prepare or poll phases. The outer event loop thus has already decremented
the counter.
Reported-by: Richard W. M. Jones <rjones@redhat.com>
Reported-by: Laszlo Ersek <lersek@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Tested-by: Richard W.M. Jones <rjones@redhat.com>
Message-id: 1437487673-23740-5-git-send-email-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This reverts commit a0710f7995.
In qemu-devel email message <556DBF87.2020908@de.ibm.com>, Christian
Borntraeger writes:
Having many guests all with a kernel/ramdisk (via -kernel) and
several null block devices will result in hangs. All hanging
guests are in partition detection code waiting for an I/O to return
so very early maybe even the first I/O.
Reverting that commit "fixes" the hangs.
Reverting this commit for the 2.4 release. More time is needed to
investigate and correct this patch.
Reported-by: Christian Borntraeger <borntraeger@de.ibm.com>
Suggested-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This is the first step towards having fine-grained critical sections in
dataplane threads, which resolves lock ordering problems between
address_space_* functions (which need the BQL when doing MMIO, even
after we complete RCU-based dispatch) and the AioContext.
Because AioContext does not use contention callbacks anymore, the
unit test has to be changed.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 1424449612-18215-4-git-send-email-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
By using thread-local storage, aio_poll can stop using global data during
g_poll_ns. This will make it possible to drop callbacks from rfifolock.
[Moved npfd = 0 assignment to end of walking_handlers region as
suggested by Paolo. This resolves the assert(npfd == 0) assertion
failure in pollfds_cleanup().
--Stefan]
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Message-id: 1424449612-18215-2-git-send-email-pbonzini@redhat.com
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
There are two problems with memory barriers in async.c. The fix is
to use atomic_xchg in order to achieve sequential consistency between
the scheduling of a bottom half and the corresponding execution.
First, if bh->scheduled is already 1 in qemu_bh_schedule, QEMU does
not execute a memory barrier to order any writes needed by the callback
before the read of bh->scheduled. If the other side sees req->state as
THREAD_ACTIVE, the callback is not invoked and you get deadlock.
Second, the memory barrier in aio_bh_poll is too weak. Without this
patch, it is possible that bh->scheduled = 0 is not "published" until
after the callback has returned. Another thread wants to schedule the
bottom half, but it sees bh->scheduled = 1 and does nothing. This causes
a lost wakeup. The memory barrier should have been changed to smp_mb()
in commit 924fe12 (aio: fix qemu_bh_schedule() bh->ctx race condition,
2014-06-03) together with qemu_bh_schedule()'s. Guess who reviewed
that patch?
Both of these involve a store and a load, so they are reproducible on
x86_64 as well. It is however much easier on aarch64, where the
libguestfs test suite triggers the bug fairly easily. Even there the
failure can go away or appear depending on compiler optimization level,
tracing options, or even kernel debugging options.
Paul Leveille however reported how to trigger the problem within 15
minutes on x86_64 as well. His (untested) recipe, reproduced here
for reference, is the following:
1) Qcow2 (or 3) is critical – raw files alone seem to avoid the problem.
2) Use “cache=directsync” rather than the default of
“cache=none” to make it happen easier.
3) Use a server with a write-back RAID controller to allow for rapid
IO rates.
4) Run a random-access load that (mostly) writes chunks to various
files on the virtual block device.
a. I use ‘diskload.exe c:25’, a Microsoft HCT load
generator, on Windows VMs.
b. Iometer can probably be configured to generate a similar load.
5) Run multiple VMs in parallel, against the same storage device,
to shake the failure out sooner.
6) IvyBridge and Haswell processors for certain; not sure about others.
A similar patch survived over 12 hours of testing, where an unpatched
QEMU would fail within 15 minutes.
This bug is, most likely, also the cause of failures in the libguestfs
testsuite on AArch64.
Thanks to Laszlo Ersek for initially reporting this bug, to Stefan
Hajnoczi for suggesting closer examination of qemu_bh_schedule, and to
Paul for providing test input and a prototype patch.
Reported-by: Laszlo Ersek <lersek@redhat.com>
Reported-by: Paul Leveille <Paul.Leveille@stratus.com>
Reported-by: John Snow <jsnow@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Message-id: 1428419779-26062-1-git-send-email-pbonzini@redhat.com
Suggested-by: Paul Leveille <Paul.Leveille@stratus.com>
Suggested-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This saves about 15% of the clock cycles spent on allocation. Using the
slice allocator does not add a visible improvement; allocation is faster
than malloc, while freeing seems to be slower.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
AioContext can be accessed recursively, in fact that's what we do with
aio_poll. Marking the GSource as recursive avoids that GLib blocks it
and unblocks it around every call to aio_dispatch, which is a pretty
expensive operation.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
g_new(T, 1) is safer than g_malloc(sizeof(T)), because it returns T *
rather than void *, which lets the compiler catch more type errors.
Missed in commit 02c4f26.
Signed-off-by: Markus Armbruster <armbru@redhat.com>
Message-id: 1417697709-13087-1-git-send-email-armbru@redhat.com
Signed-off-by: Max Reitz <mreitz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
On a system with a low limit of open files the initialization
of the event notifier could fail and QEMU exits without printing any
error information to the user.
The problem can be easily reproduced by enforcing a low limit of open
files and start QEMU with enough I/O threads to hit this limit.
The same problem raises, without the creation of I/O threads, while
QEMU initializes the main event loop by enforcing an even lower limit of
open files.
This commit adds an error message on failure:
# qemu [...] -object iothread,id=iothread0 -object iothread,id=iothread1
qemu: Failed to initialize event notifier: Too many open files in system
Signed-off-by: Chrysostomos Nanakos <cnanakos@grnet.gr>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This will be used to implement socket polling on Windows.
On Windows, select() and g_poll() are completely different;
sockets are polled with select() before calling g_poll,
and the g_poll must be nonblocking if select() says a
socket is ready.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
So far, aio_poll's scheme was dispatch/poll/dispatch, where
the first dispatch phase was used only in the GSource case in
order to avoid a blocking poll. Earlier patches changed it to
dispatch/prepare/poll/dispatch, where prepare is aio_compute_timeout.
By making aio_dispatch public, we can remove the first dispatch
phase altogether, so that both aio_poll and the GSource use the same
prepare/poll/dispatch scheme.
This patch breaks the invariant that aio_poll(..., true) will not block
the first time it returns false. This used to be fundamental for
qemu_aio_flush's implementation as "while (qemu_aio_wait()) {}" but
no code in QEMU relies on this invariant anymore. The return value
of aio_poll() is now comparable with that of g_main_context_iteration.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Right now, QEMU invokes aio_bh_poll before the "poll" phase
of aio_poll. It is simpler to do it afterwards and skip the
"poll" phase altogether when the OS-dependent parts of AioContext
are invoked from GSource. This way, AioContext behaves more
similarly when used as a GSource vs. when used as stand-alone.
As a start, take bottom halves into account when computing the
poll timeout. If a bottom half is ready, do a non-blocking
poll. As a side effect, this makes idle bottom halves work
with aio_poll; an improvement, but not really an important
one since they are deprecated.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
In many cases, the call to event_notifier_set in aio_notify is unnecessary.
In particular, if we are executing aio_dispatch, or if aio_poll is not
blocking, we know that we will soon get to the next loop iteration (if
necessary); the thread that hosts the AioContext's event loop does not
need any nudging.
The patch includes a Promela formal model that shows that this really
works and does not need any further complication such as generation
counts. It needs a memory barrier though.
The generation counts are not needed because any change to
ctx->dispatching after the memory barrier is okay for aio_notify.
If it changes from zero to one, it is the right thing to skip
event_notifier_set. If it changes from one to zero, the
event_notifier_set is unnecessary but harmless.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
qemu_bh_schedule() is supposed to be thread-safe at least the first time
it is called. Unfortunately this is not quite true:
bh->scheduled = 1;
aio_notify(bh->ctx);
Since another thread may run the BH callback once it has been scheduled,
there is a race condition if the callback frees the BH before
aio_notify(bh->ctx) has a chance to run.
Reported-by: Stefan Priebe <s.priebe@profihost.ag>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Tested-by: Stefan Priebe <s.priebe@profihost.ag>
It can be useful to run an AioContext from a thread which normally does
not "own" the AioContext. For example, request draining can be
implemented by acquiring the AioContext and looping aio_poll() until all
requests have been completed.
The following pattern should work:
/* Event loop thread */
while (running) {
aio_context_acquire(ctx);
aio_poll(ctx, true);
aio_context_release(ctx);
}
/* Another thread */
aio_context_acquire(ctx);
bdrv_read(bs, 0x1000, buf, 1);
aio_context_release(ctx);
This patch implements aio_context_acquire() and aio_context_release().
Note that existing aio_poll() callers do not need to worry about
acquiring and releasing - it is only needed when multiple threads will
call aio_poll() on the same AioContext.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Calculate the timeout in aio_ctx_prepare taking into account
the timers attached to the AioContext.
Alter aio_ctx_check similarly.
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Add a notify pointer to QEMUTimerList so it knows what to notify
on a timer change.
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Add a QEMUTimerListGroup each AioContext (meaning a QEMUTimerList
associated with each clock is added) and delete it when the
AioContext is freed.
Signed-off-by: Alex Bligh <alex@alex.org.uk>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
The .io_flush() handler no longer exists and has no users. Drop the
io_flush argument to aio_set_fd_handler() and related functions.
The AioFlushEventNotifierHandler and AioFlushHandler typedefs are no
longer used and are dropped too.
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
BH will be used outside big lock, so introduce lock to protect
between the writers, ie, bh's adders and deleter. The lock only
affects the writers and bh's callback does not take this extra lock.
Note that for the same AioContext, aio_bh_poll() can not run in
parallel yet.
Signed-off-by: Liu Ping Fan <pingfank@linux.vnet.ibm.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
This patch adds a ThreadPool to AioContext. It's possible that some
AioContext instances will never use the ThreadPool, so defer creation
until aio_get_thread_pool().
The reason why AioContext should have the ThreadPool is because the
ThreadPool is bound to a AioContext instance where the work item's
callback function is invoked. It doesn't make sense to keep the
ThreadPool pointer anywhere other than AioContext. For example,
block/raw-posix.c can get its AioContext's ThreadPool and submit work.
Special note about headers: I used struct ThreadPool in aio.h because
there is a circular dependency if aio.h includes thread-pool.h.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
AioHandler already has a GPollFD so we can directly use its
events/revents.
Add the int pollfds_idx field to AioContext so we can map g_poll(3)
results back to AioHandlers.
Reuse aio_dispatch() to invoke handlers after g_poll(3).
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Laszlo Ersek <lersek@redhat.com>
Message-id: 1361356113-11049-10-git-send-email-stefanha@redhat.com
Signed-off-by: Anthony Liguori <aliguori@us.ibm.com>
Commit ed2aec4867f0d5f5de496bb765347b5d0cfe113d changed the return
value of aio_ctx_prepare from false to true when only idle bottom
halves are available. This broke PC old-style DMA, which uses them.
Fix this by making aio_ctx_prepare return true only when non-idle
bottom halves are scheduled to run.
Reported-by: malc <av1474@comtv.ru>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: malc <av1474@comtv.ru>
Some cleanups can now be made, now that the main loop does not anymore need
hooks into the bottom half code.
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
With this change async.c does not rely anymore on any service from
main-loop.c, i.e. it is completely self-contained.
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
This will be used when polling the GSource attached to an AioContext.
Reviewed-by: Anthony Liguori <aliguori@us.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
