Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20220323155743.1585078-33-marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@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>

Simply use qemu_get_current_aio_context().

Signed-off-by: Dietmar Maurer <dietmar@proxmox.com>
Message-Id: <20191024045610.9071-1-dietmar@proxmox.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

In practice the entry argument is always known at creation time, and
it is confusing that sometimes qemu_coroutine_enter is used with a
non-NULL argument to re-enter a coroutine (this happens in
block/sheepdog.c and tests/test-coroutine.c).  So pass the opaque value
at creation time, for consistency with e.g. aio_bh_new.

Mostly done with the following semantic patch:

@ entry1 @
expression entry, arg, co;
@@
- co = qemu_coroutine_create(entry);
+ co = qemu_coroutine_create(entry, arg);
  ...
- qemu_coroutine_enter(co, arg);
+ qemu_coroutine_enter(co);

@ entry2 @
expression entry, arg;
identifier co;
@@
- Coroutine *co = qemu_coroutine_create(entry);
+ Coroutine *co = qemu_coroutine_create(entry, arg);
  ...
- qemu_coroutine_enter(co, arg);
+ qemu_coroutine_enter(co);

@ entry3 @
expression entry, arg;
@@
- qemu_coroutine_enter(qemu_coroutine_create(entry), arg);
+ qemu_coroutine_enter(qemu_coroutine_create(entry, arg));

@ reentry @
expression co;
@@
- qemu_coroutine_enter(co, NULL);
+ qemu_coroutine_enter(co);

except for the aforementioned few places where the semantic patch
stumbled (as expected) and for test_co_queue, which would otherwise
produce an uninitialized variable warning.

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Fam Zheng <famz@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>

Now that QEMU wraps the Win32 sockets methods to automatically
set errno upon failure, there is no reason for callers to use
the socket_error() method. They can rely on accessing errno
even on Win32. Remove all use of socket_error() from general
code, leaving it as a static method in oslib-win32.c only.

Signed-off-by: Daniel P. Berrange <berrange@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: 1454089805-5470-6-git-send-email-peter.maydell@linaro.org

The coroutine files are currently referenced by the block-obj-y
variable. The coroutine functionality though is already used by
more than just the block code. eg migration code uses coroutine
yield. In the future the I/O channel code will also use the
coroutine yield functionality. Since the coroutine code is nicely
self-contained it can be easily built as part of the libqemuutil.a
library, making it widely available.

The headers are also moved into include/qemu, instead of the
include/block directory, since they are now part of the util
codebase, and the impl was never in the block/ directory
either.

Signed-off-by: Daniel P. Berrange <berrange@redhat.com>

In case qemu_co_sendv_recvv() fails without any data read, there is no
reason not to return the perfectly fine error number retrieved from
socket_error().

Signed-off-by: Max Reitz <mreitz@redhat.com>
Message-Id: <1424887718-10800-16-git-send-email-mreitz@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

include/qemu/timer.h has no need to include main-loop.h and
doing so causes an issue for the next patch. Unfortunately
various files assume including timers.h will pull in main-loop.h.
Untangle this mess.

Signed-off-by: Alex Bligh <alex@alex.org.uk>
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>

The RDMA event channel can be made non-blocking just like a TCP
socket. Exporting this function allows us to yield so that the
QEMU monitor remains available.

Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>
Reviewed-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Chegu Vinod <chegu_vinod@hp.com>
Tested-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Michael R. Hines <mrhines@us.ibm.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>

Make it much more understandable, add a missing
iov_cnt argument (number of iovs in the iov), and
add comments to it.

The new implementation has been extensively tested
by splitting a large buffer into many small
randomly-sized chunks, sending it over socket to
another, slow process and verifying the receiving
data is the same.

Also add a unit test for iov_send_recv(), sending/
receiving data between two processes over a socketpair
using random vectors and random sizes.

Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

The same as for non-coroutine versions in previous
patches: rename arguments to be more obvious, change
type of arguments from int to size_t where appropriate,
and use common code for send and receive paths (with
one extra argument) since these are exactly the same.
Use common iov_send_recv() directly.

qemu_co_sendv(), qemu_co_recvv(), and qemu_co_recv()
are now trivial #define's merely adding one extra arg.

qemu_co_sendv() and qemu_co_recvv() callers are
converted to different argument order and extra
`iov_cnt' argument.

Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

Rename arguments and use size_t for sizes instead of int,
from
 int
 qemu_sendv(int sockfd, struct iovec *iov,
            int len, int iov_offset)
to
 ssize_t
 iov_send(int sockfd, struct iovec *iov,
          size_t offset, size_t bytes)

The main motivation was to make it clear that length
and offset are in _bytes_, not in iov elements: it was
very confusing before, because all standard functions
which deals with iovecs expects number of iovs, not
bytes, even the fact that struct iovec has iov_len and
iov_ prefix does not help.  With "bytes" and "offset",
especially since they're now size_t, it is much more
explicit.  Also change the return type to be ssize_t
instead of int.

This also changes it to match other iov-related functons,
but not _quite_: there's still no argument indicating
where iovec ends, ie, no iov_cnt parameter as used
in iov_size() and friends.  If will be added in subsequent
patch/rewrite.

All callers of qemu_sendv() and qemu_recvv() and
related, like qemu_co_sendv() and qemu_co_recvv(),
were checked to verify that it is safe to use unsigned
datatype instead of int.

Note that the order of arguments is changed to: offset
and bytes (len and iov_offset) are swapped with each
other.  This is to make them consistent with very similar
functions from qemu_iovec family, where offset always
follows qiov, to mean the place in it to start from.

Signed-off-by: Michael Tokarev <mjt@tls.msk.ru>

Outside coroutines, avoid busy waiting on EAGAIN by temporarily
making the socket blocking.

The API of qemu_recvv/qemu_sendv is slightly different from
do_readv/do_writev because they do not handle coroutines.  It
returns the number of bytes written before encountering an
EAGAIN.  The specificity of yielding on EAGAIN is entirely in
qemu-coroutine.c.

Reviewed-by: MORITA Kazutaka <morita.kazutaka@lab.ntt.co.jp>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>