qemu/io/channel.c
Stefan Hajnoczi 06e0f098d6 io: follow coroutine AioContext in qio_channel_yield()
The ongoing QEMU multi-queue block layer effort makes it possible for multiple
threads to process I/O in parallel. The nbd block driver is not compatible with
the multi-queue block layer yet because QIOChannel cannot be used easily from
coroutines running in multiple threads. This series changes the QIOChannel API
to make that possible.

In the current API, calling qio_channel_attach_aio_context() sets the
AioContext where qio_channel_yield() installs an fd handler prior to yielding:

  qio_channel_attach_aio_context(ioc, my_ctx);
  ...
  qio_channel_yield(ioc); // my_ctx is used here
  ...
  qio_channel_detach_aio_context(ioc);

This API design has limitations: reading and writing must be done in the same
AioContext and moving between AioContexts involves a cumbersome sequence of API
calls that is not suitable for doing on a per-request basis.

There is no fundamental reason why a QIOChannel needs to run within the
same AioContext every time qio_channel_yield() is called. QIOChannel
only uses the AioContext while inside qio_channel_yield(). The rest of
the time, QIOChannel is independent of any AioContext.

In the new API, qio_channel_yield() queries the AioContext from the current
coroutine using qemu_coroutine_get_aio_context(). There is no need to
explicitly attach/detach AioContexts anymore and
qio_channel_attach_aio_context() and qio_channel_detach_aio_context() are gone.
One coroutine can read from the QIOChannel while another coroutine writes from
a different AioContext.

This API change allows the nbd block driver to use QIOChannel from any thread.
It's important to keep in mind that the block driver already synchronizes
QIOChannel access and ensures that two coroutines never read simultaneously or
write simultaneously.

This patch updates all users of qio_channel_attach_aio_context() to the
new API. Most conversions are simple, but vhost-user-server requires a
new qemu_coroutine_yield() call to quiesce the vu_client_trip()
coroutine when not attached to any AioContext.

While the API is has become simpler, there is one wart: QIOChannel has a
special case for the iohandler AioContext (used for handlers that must not run
in nested event loops). I didn't find an elegant way preserve that behavior, so
I added a new API called qio_channel_set_follow_coroutine_ctx(ioc, true|false)
for opting in to the new AioContext model. By default QIOChannel uses the
iohandler AioHandler. Code that formerly called
qio_channel_attach_aio_context() now calls
qio_channel_set_follow_coroutine_ctx(ioc, true) once after the QIOChannel is
created.

Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Acked-by: Daniel P. Berrangé <berrange@redhat.com>
Message-ID: <20230830224802.493686-5-stefanha@redhat.com>
[eblake: also fix migration/rdma.c]
Signed-off-by: Eric Blake <eblake@redhat.com>
2023-09-07 20:32:11 -05:00

738 lines
21 KiB
C

/*
* QEMU I/O channels
*
* Copyright (c) 2015 Red Hat, Inc.
*
* This library is free software; you can redistribute it and/or
* modify it under the terms of the GNU Lesser General Public
* License as published by the Free Software Foundation; either
* version 2.1 of the License, or (at your option) any later version.
*
* This library is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
* Lesser General Public License for more details.
*
* You should have received a copy of the GNU Lesser General Public
* License along with this library; if not, see <http://www.gnu.org/licenses/>.
*
*/
#include "qemu/osdep.h"
#include "block/aio-wait.h"
#include "io/channel.h"
#include "qapi/error.h"
#include "qemu/main-loop.h"
#include "qemu/module.h"
#include "qemu/iov.h"
bool qio_channel_has_feature(QIOChannel *ioc,
QIOChannelFeature feature)
{
return ioc->features & (1 << feature);
}
void qio_channel_set_feature(QIOChannel *ioc,
QIOChannelFeature feature)
{
ioc->features |= (1 << feature);
}
void qio_channel_set_name(QIOChannel *ioc,
const char *name)
{
g_free(ioc->name);
ioc->name = g_strdup(name);
}
ssize_t qio_channel_readv_full(QIOChannel *ioc,
const struct iovec *iov,
size_t niov,
int **fds,
size_t *nfds,
int flags,
Error **errp)
{
QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
if ((fds || nfds) &&
!qio_channel_has_feature(ioc, QIO_CHANNEL_FEATURE_FD_PASS)) {
error_setg_errno(errp, EINVAL,
"Channel does not support file descriptor passing");
return -1;
}
if ((flags & QIO_CHANNEL_READ_FLAG_MSG_PEEK) &&
!qio_channel_has_feature(ioc, QIO_CHANNEL_FEATURE_READ_MSG_PEEK)) {
error_setg_errno(errp, EINVAL,
"Channel does not support peek read");
return -1;
}
return klass->io_readv(ioc, iov, niov, fds, nfds, flags, errp);
}
ssize_t qio_channel_writev_full(QIOChannel *ioc,
const struct iovec *iov,
size_t niov,
int *fds,
size_t nfds,
int flags,
Error **errp)
{
QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
if (fds || nfds) {
if (!qio_channel_has_feature(ioc, QIO_CHANNEL_FEATURE_FD_PASS)) {
error_setg_errno(errp, EINVAL,
"Channel does not support file descriptor passing");
return -1;
}
if (flags & QIO_CHANNEL_WRITE_FLAG_ZERO_COPY) {
error_setg_errno(errp, EINVAL,
"Zero Copy does not support file descriptor passing");
return -1;
}
}
if ((flags & QIO_CHANNEL_WRITE_FLAG_ZERO_COPY) &&
!qio_channel_has_feature(ioc, QIO_CHANNEL_FEATURE_WRITE_ZERO_COPY)) {
error_setg_errno(errp, EINVAL,
"Requested Zero Copy feature is not available");
return -1;
}
return klass->io_writev(ioc, iov, niov, fds, nfds, flags, errp);
}
int coroutine_mixed_fn qio_channel_readv_all_eof(QIOChannel *ioc,
const struct iovec *iov,
size_t niov,
Error **errp)
{
return qio_channel_readv_full_all_eof(ioc, iov, niov, NULL, NULL, errp);
}
int coroutine_mixed_fn qio_channel_readv_all(QIOChannel *ioc,
const struct iovec *iov,
size_t niov,
Error **errp)
{
return qio_channel_readv_full_all(ioc, iov, niov, NULL, NULL, errp);
}
int coroutine_mixed_fn qio_channel_readv_full_all_eof(QIOChannel *ioc,
const struct iovec *iov,
size_t niov,
int **fds, size_t *nfds,
Error **errp)
{
int ret = -1;
struct iovec *local_iov = g_new(struct iovec, niov);
struct iovec *local_iov_head = local_iov;
unsigned int nlocal_iov = niov;
int **local_fds = fds;
size_t *local_nfds = nfds;
bool partial = false;
if (nfds) {
*nfds = 0;
}
if (fds) {
*fds = NULL;
}
nlocal_iov = iov_copy(local_iov, nlocal_iov,
iov, niov,
0, iov_size(iov, niov));
while ((nlocal_iov > 0) || local_fds) {
ssize_t len;
len = qio_channel_readv_full(ioc, local_iov, nlocal_iov, local_fds,
local_nfds, 0, errp);
if (len == QIO_CHANNEL_ERR_BLOCK) {
if (qemu_in_coroutine()) {
qio_channel_yield(ioc, G_IO_IN);
} else {
qio_channel_wait(ioc, G_IO_IN);
}
continue;
}
if (len == 0) {
if (local_nfds && *local_nfds) {
/*
* Got some FDs, but no data yet. This isn't an EOF
* scenario (yet), so carry on to try to read data
* on next loop iteration
*/
goto next_iter;
} else if (!partial) {
/* No fds and no data - EOF before any data read */
ret = 0;
goto cleanup;
} else {
len = -1;
error_setg(errp,
"Unexpected end-of-file before all data were read");
/* Fallthrough into len < 0 handling */
}
}
if (len < 0) {
/* Close any FDs we previously received */
if (nfds && fds) {
size_t i;
for (i = 0; i < (*nfds); i++) {
close((*fds)[i]);
}
g_free(*fds);
*fds = NULL;
*nfds = 0;
}
goto cleanup;
}
if (nlocal_iov) {
iov_discard_front(&local_iov, &nlocal_iov, len);
}
next_iter:
partial = true;
local_fds = NULL;
local_nfds = NULL;
}
ret = 1;
cleanup:
g_free(local_iov_head);
return ret;
}
int coroutine_mixed_fn qio_channel_readv_full_all(QIOChannel *ioc,
const struct iovec *iov,
size_t niov,
int **fds, size_t *nfds,
Error **errp)
{
int ret = qio_channel_readv_full_all_eof(ioc, iov, niov, fds, nfds, errp);
if (ret == 0) {
error_setg(errp, "Unexpected end-of-file before all data were read");
return -1;
}
if (ret == 1) {
return 0;
}
return ret;
}
int coroutine_mixed_fn qio_channel_writev_all(QIOChannel *ioc,
const struct iovec *iov,
size_t niov,
Error **errp)
{
return qio_channel_writev_full_all(ioc, iov, niov, NULL, 0, 0, errp);
}
int coroutine_mixed_fn qio_channel_writev_full_all(QIOChannel *ioc,
const struct iovec *iov,
size_t niov,
int *fds, size_t nfds,
int flags, Error **errp)
{
int ret = -1;
struct iovec *local_iov = g_new(struct iovec, niov);
struct iovec *local_iov_head = local_iov;
unsigned int nlocal_iov = niov;
nlocal_iov = iov_copy(local_iov, nlocal_iov,
iov, niov,
0, iov_size(iov, niov));
while (nlocal_iov > 0) {
ssize_t len;
len = qio_channel_writev_full(ioc, local_iov, nlocal_iov, fds,
nfds, flags, errp);
if (len == QIO_CHANNEL_ERR_BLOCK) {
if (qemu_in_coroutine()) {
qio_channel_yield(ioc, G_IO_OUT);
} else {
qio_channel_wait(ioc, G_IO_OUT);
}
continue;
}
if (len < 0) {
goto cleanup;
}
iov_discard_front(&local_iov, &nlocal_iov, len);
fds = NULL;
nfds = 0;
}
ret = 0;
cleanup:
g_free(local_iov_head);
return ret;
}
ssize_t qio_channel_readv(QIOChannel *ioc,
const struct iovec *iov,
size_t niov,
Error **errp)
{
return qio_channel_readv_full(ioc, iov, niov, NULL, NULL, 0, errp);
}
ssize_t qio_channel_writev(QIOChannel *ioc,
const struct iovec *iov,
size_t niov,
Error **errp)
{
return qio_channel_writev_full(ioc, iov, niov, NULL, 0, 0, errp);
}
ssize_t qio_channel_read(QIOChannel *ioc,
char *buf,
size_t buflen,
Error **errp)
{
struct iovec iov = { .iov_base = buf, .iov_len = buflen };
return qio_channel_readv_full(ioc, &iov, 1, NULL, NULL, 0, errp);
}
ssize_t qio_channel_write(QIOChannel *ioc,
const char *buf,
size_t buflen,
Error **errp)
{
struct iovec iov = { .iov_base = (char *)buf, .iov_len = buflen };
return qio_channel_writev_full(ioc, &iov, 1, NULL, 0, 0, errp);
}
int coroutine_mixed_fn qio_channel_read_all_eof(QIOChannel *ioc,
char *buf,
size_t buflen,
Error **errp)
{
struct iovec iov = { .iov_base = buf, .iov_len = buflen };
return qio_channel_readv_all_eof(ioc, &iov, 1, errp);
}
int coroutine_mixed_fn qio_channel_read_all(QIOChannel *ioc,
char *buf,
size_t buflen,
Error **errp)
{
struct iovec iov = { .iov_base = buf, .iov_len = buflen };
return qio_channel_readv_all(ioc, &iov, 1, errp);
}
int coroutine_mixed_fn qio_channel_write_all(QIOChannel *ioc,
const char *buf,
size_t buflen,
Error **errp)
{
struct iovec iov = { .iov_base = (char *)buf, .iov_len = buflen };
return qio_channel_writev_all(ioc, &iov, 1, errp);
}
int qio_channel_set_blocking(QIOChannel *ioc,
bool enabled,
Error **errp)
{
QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
return klass->io_set_blocking(ioc, enabled, errp);
}
void qio_channel_set_follow_coroutine_ctx(QIOChannel *ioc, bool enabled)
{
ioc->follow_coroutine_ctx = enabled;
}
int qio_channel_close(QIOChannel *ioc,
Error **errp)
{
QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
return klass->io_close(ioc, errp);
}
GSource *qio_channel_create_watch(QIOChannel *ioc,
GIOCondition condition)
{
QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
GSource *ret = klass->io_create_watch(ioc, condition);
if (ioc->name) {
g_source_set_name(ret, ioc->name);
}
return ret;
}
void qio_channel_set_aio_fd_handler(QIOChannel *ioc,
AioContext *read_ctx,
IOHandler *io_read,
AioContext *write_ctx,
IOHandler *io_write,
void *opaque)
{
QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
klass->io_set_aio_fd_handler(ioc, read_ctx, io_read, write_ctx, io_write,
opaque);
}
guint qio_channel_add_watch_full(QIOChannel *ioc,
GIOCondition condition,
QIOChannelFunc func,
gpointer user_data,
GDestroyNotify notify,
GMainContext *context)
{
GSource *source;
guint id;
source = qio_channel_create_watch(ioc, condition);
g_source_set_callback(source, (GSourceFunc)func, user_data, notify);
id = g_source_attach(source, context);
g_source_unref(source);
return id;
}
guint qio_channel_add_watch(QIOChannel *ioc,
GIOCondition condition,
QIOChannelFunc func,
gpointer user_data,
GDestroyNotify notify)
{
return qio_channel_add_watch_full(ioc, condition, func,
user_data, notify, NULL);
}
GSource *qio_channel_add_watch_source(QIOChannel *ioc,
GIOCondition condition,
QIOChannelFunc func,
gpointer user_data,
GDestroyNotify notify,
GMainContext *context)
{
GSource *source;
guint id;
id = qio_channel_add_watch_full(ioc, condition, func,
user_data, notify, context);
source = g_main_context_find_source_by_id(context, id);
g_source_ref(source);
return source;
}
int qio_channel_shutdown(QIOChannel *ioc,
QIOChannelShutdown how,
Error **errp)
{
QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
if (!klass->io_shutdown) {
error_setg(errp, "Data path shutdown not supported");
return -1;
}
return klass->io_shutdown(ioc, how, errp);
}
void qio_channel_set_delay(QIOChannel *ioc,
bool enabled)
{
QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
if (klass->io_set_delay) {
klass->io_set_delay(ioc, enabled);
}
}
void qio_channel_set_cork(QIOChannel *ioc,
bool enabled)
{
QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
if (klass->io_set_cork) {
klass->io_set_cork(ioc, enabled);
}
}
off_t qio_channel_io_seek(QIOChannel *ioc,
off_t offset,
int whence,
Error **errp)
{
QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
if (!klass->io_seek) {
error_setg(errp, "Channel does not support random access");
return -1;
}
return klass->io_seek(ioc, offset, whence, errp);
}
int qio_channel_flush(QIOChannel *ioc,
Error **errp)
{
QIOChannelClass *klass = QIO_CHANNEL_GET_CLASS(ioc);
if (!klass->io_flush ||
!qio_channel_has_feature(ioc, QIO_CHANNEL_FEATURE_WRITE_ZERO_COPY)) {
return 0;
}
return klass->io_flush(ioc, errp);
}
static void qio_channel_restart_read(void *opaque)
{
QIOChannel *ioc = opaque;
Coroutine *co = qatomic_xchg(&ioc->read_coroutine, NULL);
if (!co) {
return;
}
/* Assert that aio_co_wake() reenters the coroutine directly */
assert(qemu_get_current_aio_context() ==
qemu_coroutine_get_aio_context(co));
aio_co_wake(co);
}
static void qio_channel_restart_write(void *opaque)
{
QIOChannel *ioc = opaque;
Coroutine *co = qatomic_xchg(&ioc->write_coroutine, NULL);
if (!co) {
return;
}
/* Assert that aio_co_wake() reenters the coroutine directly */
assert(qemu_get_current_aio_context() ==
qemu_coroutine_get_aio_context(co));
aio_co_wake(co);
}
static void coroutine_fn
qio_channel_set_fd_handlers(QIOChannel *ioc, GIOCondition condition)
{
AioContext *ctx = ioc->follow_coroutine_ctx ?
qemu_coroutine_get_aio_context(qemu_coroutine_self()) :
iohandler_get_aio_context();
AioContext *read_ctx = NULL;
IOHandler *io_read = NULL;
AioContext *write_ctx = NULL;
IOHandler *io_write = NULL;
if (condition == G_IO_IN) {
ioc->read_coroutine = qemu_coroutine_self();
ioc->read_ctx = ctx;
read_ctx = ctx;
io_read = qio_channel_restart_read;
/*
* Thread safety: if the other coroutine is set and its AioContext
* matches ours, then there is mutual exclusion between read and write
* because they share a single thread and it's safe to set both read
* and write fd handlers here. If the AioContext does not match ours,
* then both threads may run in parallel but there is no shared state
* to worry about.
*/
if (ioc->write_coroutine && ioc->write_ctx == ctx) {
write_ctx = ctx;
io_write = qio_channel_restart_write;
}
} else if (condition == G_IO_OUT) {
ioc->write_coroutine = qemu_coroutine_self();
ioc->write_ctx = ctx;
write_ctx = ctx;
io_write = qio_channel_restart_write;
if (ioc->read_coroutine && ioc->read_ctx == ctx) {
read_ctx = ctx;
io_read = qio_channel_restart_read;
}
} else {
abort();
}
qio_channel_set_aio_fd_handler(ioc, read_ctx, io_read,
write_ctx, io_write, ioc);
}
static void coroutine_fn
qio_channel_clear_fd_handlers(QIOChannel *ioc, GIOCondition condition)
{
AioContext *read_ctx = NULL;
IOHandler *io_read = NULL;
AioContext *write_ctx = NULL;
IOHandler *io_write = NULL;
AioContext *ctx;
if (condition == G_IO_IN) {
ctx = ioc->read_ctx;
read_ctx = ctx;
io_read = NULL;
if (ioc->write_coroutine && ioc->write_ctx == ctx) {
write_ctx = ctx;
io_write = qio_channel_restart_write;
}
} else if (condition == G_IO_OUT) {
ctx = ioc->write_ctx;
write_ctx = ctx;
io_write = NULL;
if (ioc->read_coroutine && ioc->read_ctx == ctx) {
read_ctx = ctx;
io_read = qio_channel_restart_read;
}
} else {
abort();
}
qio_channel_set_aio_fd_handler(ioc, read_ctx, io_read,
write_ctx, io_write, ioc);
}
void coroutine_fn qio_channel_yield(QIOChannel *ioc,
GIOCondition condition)
{
AioContext *ioc_ctx;
assert(qemu_in_coroutine());
ioc_ctx = qemu_coroutine_get_aio_context(qemu_coroutine_self());
if (condition == G_IO_IN) {
assert(!ioc->read_coroutine);
} else if (condition == G_IO_OUT) {
assert(!ioc->write_coroutine);
} else {
abort();
}
qio_channel_set_fd_handlers(ioc, condition);
qemu_coroutine_yield();
assert(in_aio_context_home_thread(ioc_ctx));
/* Allow interrupting the operation by reentering the coroutine other than
* through the aio_fd_handlers. */
if (condition == G_IO_IN) {
assert(ioc->read_coroutine == NULL);
} else if (condition == G_IO_OUT) {
assert(ioc->write_coroutine == NULL);
}
qio_channel_clear_fd_handlers(ioc, condition);
}
void qio_channel_wake_read(QIOChannel *ioc)
{
Coroutine *co = qatomic_xchg(&ioc->read_coroutine, NULL);
if (co) {
aio_co_wake(co);
}
}
static gboolean qio_channel_wait_complete(QIOChannel *ioc,
GIOCondition condition,
gpointer opaque)
{
GMainLoop *loop = opaque;
g_main_loop_quit(loop);
return FALSE;
}
void qio_channel_wait(QIOChannel *ioc,
GIOCondition condition)
{
GMainContext *ctxt = g_main_context_new();
GMainLoop *loop = g_main_loop_new(ctxt, TRUE);
GSource *source;
source = qio_channel_create_watch(ioc, condition);
g_source_set_callback(source,
(GSourceFunc)qio_channel_wait_complete,
loop,
NULL);
g_source_attach(source, ctxt);
g_main_loop_run(loop);
g_source_unref(source);
g_main_loop_unref(loop);
g_main_context_unref(ctxt);
}
static void qio_channel_finalize(Object *obj)
{
QIOChannel *ioc = QIO_CHANNEL(obj);
/* Must not have coroutines in qio_channel_yield() */
assert(!ioc->read_coroutine);
assert(!ioc->write_coroutine);
g_free(ioc->name);
#ifdef _WIN32
if (ioc->event) {
CloseHandle(ioc->event);
}
#endif
}
static const TypeInfo qio_channel_info = {
.parent = TYPE_OBJECT,
.name = TYPE_QIO_CHANNEL,
.instance_size = sizeof(QIOChannel),
.instance_finalize = qio_channel_finalize,
.abstract = true,
.class_size = sizeof(QIOChannelClass),
};
static void qio_channel_register_types(void)
{
type_register_static(&qio_channel_info);
}
type_init(qio_channel_register_types);