Since qemu 8.2, the combination of NBD + TLS + iothread crashes on an
assertion failure:
qemu-kvm: ../io/channel.c:534: void qio_channel_restart_read(void *): Assertion `qemu_get_current_aio_context() == qemu_coroutine_get_aio_context(co)' failed.
It turns out that when we removed AioContext locking, we did so by
having NBD tell its qio channels that it wanted to opt in to
qio_channel_set_follow_coroutine_ctx(); but while we opted in on the
main channel, we did not opt in on the TLS wrapper channel.
qemu-iotests has coverage of NBD+iothread and NBD+TLS, but apparently
no coverage of NBD+TLS+iothread, or we would have noticed this
regression sooner. (I'll add that in the next patch)
But while we could manually opt in to the TLS channel in nbd/server.c
(a one-line change), it is more generic if all qio channels that wrap
other channels inherit the follow status, in the same way that they
inherit feature bits.
CC: Stefan Hajnoczi <stefanha@redhat.com>
CC: Daniel P. Berrangé <berrange@redhat.com>
CC: qemu-stable@nongnu.org
Fixes: https://issues.redhat.com/browse/RHEL-34786
Fixes: 06e0f098 ("io: follow coroutine AioContext in qio_channel_yield()", v8.2.0)
Signed-off-by: Eric Blake <eblake@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-ID: <20240518025246.791593-5-eblake@redhat.com>
Add a new helper function for creating a QIOChannelFile channel with a
duplicated file descriptor. This saves the calling code from having to
do error checking on the dup() call.
Suggested-by: "Daniel P. Berrangé" <berrange@redhat.com>
Signed-off-by: Fabiano Rosas <farosas@suse.de>
Reviewed-by: "Daniel P. Berrangé" <berrange@redhat.com>
Link: https://lore.kernel.org/r/20240311233335.17299-2-farosas@suse.de
Signed-off-by: Peter Xu <peterx@redhat.com>
Commit bc38feddeb ("io: fsync before closing a file channel") added a
fsync/fdatasync at the closing point of the QIOChannelFile to ensure
integrity of the migration stream in case of QEMU crash.
The decision to do the sync at qio_channel_close() was not the best
since that function runs in the main thread and the fsync can cause
QEMU to hang for several minutes, depending on the migration size and
disk speed.
To fix the hang, remove the fsync from qio_channel_file_close().
At this moment, the migration code is the only user of the fsync and
we're taking the tradeoff of not having a sync at all, leaving the
responsibility to the upper layers.
Fixes: bc38feddeb ("io: fsync before closing a file channel")
Reviewed-by: "Daniel P. Berrangé" <berrange@redhat.com>
Signed-off-by: Fabiano Rosas <farosas@suse.de>
Link: https://lore.kernel.org/r/20240305195629.9922-1-farosas@suse.de
Link: https://lore.kernel.org/r/20240305174332.2553-1-farosas@suse.de
[peterx: add more comment to the qio_channel_close()]
Signed-off-by: Peter Xu <peterx@redhat.com>
Make sure the data is flushed to disk before closing file
channels. This is to ensure data is on disk and not lost in the event
of a host crash.
This is currently being implemented to affect the migration code when
migrating to a file, but all QIOChannelFile users should benefit from
the change.
Reviewed-by: "Daniel P. Berrangé" <berrange@redhat.com>
Acked-by: "Daniel P. Berrangé" <berrange@redhat.com>
Signed-off-by: Fabiano Rosas <farosas@suse.de>
Link: https://lore.kernel.org/r/20240229153017.2221-6-farosas@suse.de
Signed-off-by: Peter Xu <peterx@redhat.com>
The upcoming 'mapped-ram' feature will require qemu to write data to
(and restore from) specific offsets of the migration file.
Add a minimal implementation of pwritev/preadv and expose them via the
io_pwritev and io_preadv interfaces.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: "Daniel P. Berrangé" <berrange@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Fabiano Rosas <farosas@suse.de>
Link: https://lore.kernel.org/r/20240229153017.2221-5-farosas@suse.de
Signed-off-by: Peter Xu <peterx@redhat.com>
Introduce basic pwritev/preadv support in the generic channel layer.
Specific implementation will follow for the file channel as this is
required in order to support migration streams with fixed location of
each ram page.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: "Daniel P. Berrangé" <berrange@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Fabiano Rosas <farosas@suse.de>
Link: https://lore.kernel.org/r/20240229153017.2221-4-farosas@suse.de
Signed-off-by: Peter Xu <peterx@redhat.com>
Add a generic QIOChannel feature SEEKABLE which would be used by the
qemu_file* apis. For the time being this will be only implemented for
file channels.
Signed-off-by: Nikolay Borisov <nborisov@suse.com>
Reviewed-by: "Daniel P. Berrangé" <berrange@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Fabiano Rosas <farosas@suse.de>
Link: https://lore.kernel.org/r/20240229153017.2221-3-farosas@suse.de
Signed-off-by: Peter Xu <peterx@redhat.com>
All callers of qio_net_listener_set_name() already add some sort of
"listen" or "listener" suffix.
For intance, we currently have "migration-socket-listener-listen" and
"vnc-listen-listen" as ioc names.
Signed-off-by: Fabiano Rosas <farosas@suse.de>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
For SO_EE_ORIGIN_ZEROCOPY the 32-bit notification range is encoded
as [ee_info, ee_data] inclusively, so ee_info should be less or
equal to ee_data.
Signed-off-by: Vladimir Sementsov-Ogievskiy <vsementsov@yandex-team.ru>
Reviewed-by: Maksim Davydov <davydov-max@yandex-team.ru>
Message-id: 20231017125941.810461-7-vsementsov@yandex-team.ru
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
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>
Callers must clean up their coroutines before calling
object_unref(OBJECT(ioc)) to prevent an fd handler leak. Add an
assertion to check this.
This patch is preparation for the fd handler changes that follow.
Signed-off-by: Stefan Hajnoczi <stefanha@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Message-ID: <20230830224802.493686-4-stefanha@redhat.com>
Signed-off-by: Eric Blake <eblake@redhat.com>
The TLS handshake make take some time to complete, during which time an
I/O watch might be registered with the main loop. If the owner of the
I/O channel invokes qio_channel_close() while the handshake is waiting
to continue the I/O watch must be removed. Failing to remove it will
later trigger the completion callback which the owner is not expecting
to receive. In the case of the VNC server, this results in a SEGV as
vnc_disconnect_start() tries to shutdown a client connection that is
already gone / NULL.
CVE-2023-3354
Reported-by: jiangyegen <jiangyegen@huawei.com>
Signed-off-by: Daniel P. Berrangé <berrange@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>
nbd_drained_poll() generally runs in the main thread, not whatever
iothread the NBD server coroutine is meant to run in, so it can't
directly reenter the coroutines to wake them up.
The code seems to have the right intention, it specifies the correct
AioContext when it calls qemu_aio_coroutine_enter(). However, this
functions doesn't schedule the coroutine to run in that AioContext, but
it assumes it is already called in the home thread of the AioContext.
To fix this, add a new thread-safe qio_channel_wake_read() that can be
called in the main thread to wake up the coroutine in its AioContext,
and use this in nbd_drained_poll().
Cc: qemu-stable@nongnu.org
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
Message-Id: <20230517152834.277483-3-kwolf@redhat.com>
Reviewed-by: Eric Blake <eblake@redhat.com>
Signed-off-by: Kevin Wolf <kwolf@redhat.com>
There should be no paths from a coroutine_fn to aio_poll, however in
practice coroutine_mixed_fn will call aio_poll in the !qemu_in_coroutine()
path. By marking mixed functions, we can track accurately the call paths
that execute entirely in coroutine context, and find more missing
coroutine_fn markers. This results in more accurate checks that
coroutine code does not end up blocking.
If the marking were extended transitively to all functions that call
these ones, static analysis could be done much more efficiently.
However, this is a start and makes it possible to use vrc's path-based
searches to find potential bugs where coroutine_fns call blocking functions.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
TLS iochannel will inherit io_shutdown() from the master ioc, however we
missed to do that on the server side.
This will e.g. allow qemu_file_shutdown() to work on dest QEMU too for
migration.
Acked-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
This leakage can be seen through test-io-channel-tls:
$ ../configure --target-list=aarch64-softmmu --enable-sanitizers
$ make ./tests/unit/test-io-channel-tls
$ ./tests/unit/test-io-channel-tls
Indirect leak of 104 byte(s) in 1 object(s) allocated from:
#0 0x7f81d1725808 in __interceptor_malloc ../../../../src/libsanitizer/asan/asan_malloc_linux.cc:144
#1 0x7f81d135ae98 in g_malloc (/lib/x86_64-linux-gnu/libglib-2.0.so.0+0x57e98)
#2 0x55616c5d4c1b in object_new_with_propv ../qom/object.c:795
#3 0x55616c5d4a83 in object_new_with_props ../qom/object.c:768
#4 0x55616c5c5415 in test_tls_creds_create ../tests/unit/test-io-channel-tls.c:70
#5 0x55616c5c5a6b in test_io_channel_tls ../tests/unit/test-io-channel-tls.c:158
#6 0x7f81d137d58d (/lib/x86_64-linux-gnu/libglib-2.0.so.0+0x7a58d)
Indirect leak of 32 byte(s) in 1 object(s) allocated from:
#0 0x7f81d1725a06 in __interceptor_calloc ../../../../src/libsanitizer/asan/asan_malloc_linux.cc:153
#1 0x7f81d1472a20 in gnutls_dh_params_init (/lib/x86_64-linux-gnu/libgnutls.so.30+0x46a20)
#2 0x55616c6485ff in qcrypto_tls_creds_x509_load ../crypto/tlscredsx509.c:634
#3 0x55616c648ba2 in qcrypto_tls_creds_x509_complete ../crypto/tlscredsx509.c:694
#4 0x55616c5e1fea in user_creatable_complete ../qom/object_interfaces.c:28
#5 0x55616c5d4c8c in object_new_with_propv ../qom/object.c:807
#6 0x55616c5d4a83 in object_new_with_props ../qom/object.c:768
#7 0x55616c5c5415 in test_tls_creds_create ../tests/unit/test-io-channel-tls.c:70
#8 0x55616c5c5a6b in test_io_channel_tls ../tests/unit/test-io-channel-tls.c:158
#9 0x7f81d137d58d (/lib/x86_64-linux-gnu/libglib-2.0.so.0+0x7a58d)
...
SUMMARY: AddressSanitizer: 49143 byte(s) leaked in 184 allocation(s).
The docs for `g_source_add_child_source(source, child_source)` says
"source will hold a reference on child_source while child_source is
attached to it." Therefore, we should unreference the child source at
`qio_channel_tls_read_watch()` after attaching it to `source`. With this
change, ./tests/unit/test-io-channel-tls shows no leakages.
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Matheus Tavares Bernardino <quic_mathbern@quicinc.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Use a close() wrapper instead, so that we don't need to worry about
closesocket() vs close() anymore, let's hope.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Message-Id: <20230221124802.4103554-17-marcandre.lureau@redhat.com>
Until now, a win32 SOCKET handle is often cast to an int file
descriptor, as this is what other OS use for sockets. When necessary,
QEMU eventually queries whether it's a socket with the help of
fd_is_socket(). However, there is no guarantee of conflict between the
fd and SOCKET space. Such conflict would have surprising consequences,
we shouldn't mix them.
Also, it is often forgotten that SOCKET must be closed with
closesocket(), and not close().
Instead, let's make the win32 socket wrapper functions return and take a
file descriptor, and let util/ wrappers do the fd/SOCKET conversion as
necessary. A bit of adaptation is necessary in io/ as well.
Unfortunately, we can't drop closesocket() usage, despite
_open_osfhandle() documentation claiming transfer of ownership, testing
shows bad behaviour if you forget to call closesocket().
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Message-Id: <20230221124802.4103554-15-marcandre.lureau@redhat.com>
A more explicit version of qemu_socket_select() with no events.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Message-Id: <20230221124802.4103554-8-marcandre.lureau@redhat.com>
This is a wrapper for WSAEventSelect, with Error handling. By default,
it will produce a warning, so callers don't have to be modified
now, and yet we can spot potential mis-use.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Berger <stefanb@linux.ibm.com>
Message-Id: <20230221124802.4103554-7-marcandre.lureau@redhat.com>
Because they are actually sockets...
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Thomas Huth <thuth@redhat.com>
Message-Id: <20230221124802.4103554-4-marcandre.lureau@redhat.com>
Since the TLS backend can read more data from the underlying QIOChannel
we introduce a minimal child GSource to notify if we still have more
data available to be read.
Signed-off-by: Antoine Damhet <antoine.damhet@shadow.tech>
Signed-off-by: Charles Frey <charles.frey@shadow.tech>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
MSG_PEEK peeks at the channel, The data is treated as unread and
the next read shall still return this data. This support is
currently added only for socket class. Extra parameter 'flags'
is added to io_readv calls to pass extra read flags like MSG_PEEK.
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Daniel P. Berrange <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Suggested-by: Daniel P. Berrange <berrange@redhat.com>
Signed-off-by: manish.mishra <manish.mishra@nutanix.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Random failure was observed when running qtests on Windows due to
"Broken pipe" detected by qmp_fd_receive(). What happened is that
the qtest executable sends testing data over a socket to the QEMU
under test but no response is received. The errno of the recv()
call from the qtest executable indicates ETIMEOUT, due to the qmp
chardev's tcp_chr_read() is never called to receive testing data
hence no response is sent to the other side.
tcp_chr_read() is registered as the callback of the socket watch
GSource. The reason of the callback not being called by glib, is
that the source check fails to indicate the source is ready. There
are two socket watch sources created to monitor the same socket
event object from the char-socket backend in update_ioc_handlers().
During the source check phase, qio_channel_socket_source_check()
calls WSAEnumNetworkEvents() to discover occurrences of network
events for the indicated socket, clear internal network event records,
and reset the event object. Testing shows that if we don't reset the
event object by not passing the event handle to WSAEnumNetworkEvents()
the symptom goes away and qtest runs very stably.
It seems we don't need to call WSAEnumNetworkEvents() at all, as we
don't parse the result of WSANETWORKEVENTS returned from this API.
We use select() to poll the socket status. Fix this instability by
dropping the WSAEnumNetworkEvents() call.
Some side notes:
During the testing, I removed the following codes in update_ioc_handlers():
remove_hup_source(s);
s->hup_source = qio_channel_create_watch(s->ioc, G_IO_HUP);
g_source_set_callback(s->hup_source, (GSourceFunc)tcp_chr_hup,
chr, NULL);
g_source_attach(s->hup_source, chr->gcontext);
and such change also makes the symptom go away.
And if I moved the above codes to the beginning, before the call to
io_add_watch_poll(), the symptom also goes away.
It seems two sources watching on the same socket event object is
the key that leads to the instability. The order of adding a source
watch seems to also play a role but I can't explain why.
Hopefully a Windows and glib expert could explain this behavior.
Signed-off-by: Bin Meng <bin.meng@windriver.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
There is no need to do a type cast on ssource->socket as it is
already declared as a SOCKET.
Suggested-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
In the win32 version qio_channel_create_socket_watch() body there is
no need to do a '#ifdef WIN32'.
Signed-off-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
The initial implementation was changing the pipe state created by GLib
to PIPE_NOWAIT, but it turns out it doesn't work (read/write returns an
error). Since reading may return less than the requested amount, it
seems to be non-blocking already. However, the IO operation may block
until the FD is ready, I can't find good sources of information, to be
safe we can just poll for readiness before.
Alternatively, we could setup the FDs ourself, and use UNIX sockets on
Windows, which can be used in blocking/non-blocking mode. I haven't
tried it, as I am not sure it is necessary.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20221006113657.2656108-6-marcandre.lureau@redhat.com>
Simplify qio_channel_command_new_spawn() with GSpawn API. This will
allow to build for WIN32 in the following patches.
As pointed out by Daniel Berrangé: there is a change in semantics here
too. The current code only touches stdin/stdout/stderr. Any other FDs
which do NOT have O_CLOEXEC set will be inherited. With the new code,
all FDs except stdin/out/err will be explicitly closed, because we don't
set the flag G_SPAWN_LEAVE_DESCRIPTORS_OPEN. The only place we use
QIOChannelCommand today is the migration exec: protocol, and that is
only declared to use stdin/stdout.
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20221006113657.2656108-5-marcandre.lureau@redhat.com>
The combined_key[... QIO_CHANNEL_WEBSOCK_GUID_LEN ...] array in
qio_channel_websock_handshake_send_res_ok() expands to a call
to strlen(QIO_CHANNEL_WEBSOCK_GUID), and the compiler doesn't
realize the string is const, so consider combined_key[] being
a variable-length array.
To remove the variable-length array, we provide it a hint to
the compiler by using sizeof() - 1 instead of strlen().
Signed-off-by: Philippe Mathieu-Daudé <philmd@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Message-id: 20220819153931.3147384-5-peter.maydell@linaro.org
For using MSG_ZEROCOPY, there are two steps:
1 - io_writev() the packet, which enqueues the packet for sending, and
2 - io_flush(), which gets confirmation that all packets got correctly sent
Currently, if MSG_ZEROCOPY is used to send packets over IPV6, no error will
be reported in (1), but it will fail in the first time (2) happens.
This happens because (2) currently checks for cmsg_level & cmsg_type
associated with IPV4 only, before reporting any error.
Add checks for cmsg_level & cmsg_type associated with IPV6, and thus enable
support for MSG_ZEROCOPY + IPV6
Fixes: 2bc58ffc29 ("QIOChannelSocket: Implement io_writev zero copy flag & io_flush for CONFIG_LINUX")
Signed-off-by: Leonardo Bras <leobras@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
If flush is called when no buffer was sent with MSG_ZEROCOPY, it currently
returns 1. This return code should be used only when Linux fails to use
MSG_ZEROCOPY on a lot of sendmsg().
Fix this by returning early from flush if no sendmsg(...,MSG_ZEROCOPY)
was attempted.
Fixes: 2bc58ffc29 ("QIOChannelSocket: Implement io_writev zero copy flag & io_flush for CONFIG_LINUX")
Signed-off-by: Leonardo Bras <leobras@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Acked-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Message-Id: <20220711211112.18951-2-leobras@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
This is for code which needs a portable equivalent to a QIOChannelFile
connected to /dev/null.
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Somewhere between v6 and v7 the of the zero-copy-send patchset a crucial
part of the flushing mechanism got missing: incrementing zero_copy_queued.
Without that, the flushing interface becomes a no-op, and there is no
guarantee the buffer is really sent.
This can go as bad as causing a corruption in RAM during migration.
Fixes: 2bc58ffc29 ("QIOChannelSocket: Implement io_writev zero copy flag & io_flush for CONFIG_LINUX")
Reported-by: 徐闯 <xuchuangxclwt@bytedance.com>
Signed-off-by: Leonardo Bras <leobras@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
During implementation of MSG_ZEROCOPY feature, a lot of #ifdefs were
introduced, particularly at qio_channel_socket_writev().
Rewrite some of those changes so it's easier to read.
Also, introduce an assert to help detect incorrect zero-copy usage is when
it's disabled on build.
Signed-off-by: Leonardo Bras <leobras@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Signed-off-by: Juan Quintela <quintela@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
dgilbert: Fixed up thinko'd g_assert_unreachable->g_assert_not_reached
For CONFIG_LINUX, implement the new zero copy flag and the optional callback
io_flush on QIOChannelSocket, but enables it only when MSG_ZEROCOPY
feature is available in the host kernel, which is checked on
qio_channel_socket_connect_sync()
qio_channel_socket_flush() was implemented by counting how many times
sendmsg(...,MSG_ZEROCOPY) was successfully called, and then reading the
socket's error queue, in order to find how many of them finished sending.
Flush will loop until those counters are the same, or until some error occurs.
Notes on using writev() with QIO_CHANNEL_WRITE_FLAG_ZERO_COPY:
1: Buffer
- As MSG_ZEROCOPY tells the kernel to use the same user buffer to avoid copying,
some caution is necessary to avoid overwriting any buffer before it's sent.
If something like this happen, a newer version of the buffer may be sent instead.
- If this is a problem, it's recommended to call qio_channel_flush() before freeing
or re-using the buffer.
2: Locked memory
- When using MSG_ZERCOCOPY, the buffer memory will be locked after queued, and
unlocked after it's sent.
- Depending on the size of each buffer, and how often it's sent, it may require
a larger amount of locked memory than usually available to non-root user.
- If the required amount of locked memory is not available, writev_zero_copy
will return an error, which can abort an operation like migration,
- Because of this, when an user code wants to add zero copy as a feature, it
requires a mechanism to disable it, so it can still be accessible to less
privileged users.
Signed-off-by: Leonardo Bras <leobras@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <20220513062836.965425-4-leobras@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Add flags to io_writev and introduce io_flush as optional callback to
QIOChannelClass, allowing the implementation of zero copy writes by
subclasses.
How to use them:
- Write data using qio_channel_writev*(...,QIO_CHANNEL_WRITE_FLAG_ZERO_COPY),
- Wait write completion with qio_channel_flush().
Notes:
As some zero copy write implementations work asynchronously, it's
recommended to keep the write buffer untouched until the return of
qio_channel_flush(), to avoid the risk of sending an updated buffer
instead of the buffer state during write.
As io_flush callback is optional, if a subclass does not implement it, then:
- io_flush will return 0 without changing anything.
Also, some functions like qio_channel_writev_full_all() were adapted to
receive a flag parameter. That allows shared code between zero copy and
non-zero copy writev, and also an easier implementation on new flags.
Signed-off-by: Leonardo Bras <leobras@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Reviewed-by: Peter Xu <peterx@redhat.com>
Reviewed-by: Juan Quintela <quintela@redhat.com>
Message-Id: <20220513062836.965425-3-leobras@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
The qemu_*block() functions are meant to be be used with sockets (the
win32 implementation expects SOCKET)
Over time, those functions where used with Win32 SOCKET or
file-descriptors interchangeably. But for portability, they must only be
used with socket-like file-descriptors. FDs can use
g_unix_set_fd_nonblocking() instead.
Rename the functions with "socket" in the name to prevent bad usages.
This is effectively reverting commit f9e8cacc55 ("oslib-posix:
rename socket_set_nonblock() to qemu_set_nonblock()").
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Stefan Hajnoczi <stefanha@redhat.com>
Those calls are non-socket fd, or are POSIX-specific. Use the dedicated
GLib API. (qemu_set_nonblock() is for socket-like)
(this is a preliminary patch before renaming qemu_set_nonblock())
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
The function isn't used outside of qio_channel_command_new_spawn(),
which is !win32-specific.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Suggested-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
One less qemu-specific macro. It also helps to make some headers/units
only depend on glib, and thus moved in standalone projects eventually.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Richard W.M. Jones <rjones@redhat.com>
The socket API wrappers were initially introduced in commit
00aa0040 ("Wrap recv to avoid warnings"), but made redundant with
commit a2d96af4 ("osdep: add wrappers for socket functions") which fixes
the win32 declarations and thus removed the earlier warnings.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Reviewed-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
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>
The QAPI schema shouldn't rely on C system headers #define, but on
configure-time project #define, so we can express the build condition in
a C-independent way.
Signed-off-by: Marc-André Lureau <marcandre.lureau@redhat.com>
Message-Id: <20210907121943.3498701-3-marcandre.lureau@redhat.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
The GDateTime APIs provided by GLib avoid portability pitfalls, such
as some platforms where 'struct timeval.tv_sec' field is still 'long'
instead of 'time_t'. When combined with automatic cleanup, GDateTime
often results in simpler code too.
Reviewed-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Signed-off-by: Daniel P. Berrangé <berrange@redhat.com>
Multipath TCP allows combining multiple interfaces/routes into a single
socket, with very little work for the user/admin.
It's enabled by 'mptcp' on most socket addresses:
./qemu-system-x86_64 -nographic -incoming tcp:0:4444,mptcp
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Message-Id: <20210421112834.107651-6-dgilbert@redhat.com>
Acked-by: Markus Armbruster <armbru@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Call the notifier during finalize; it's currently only called
if we change it, which is not the intent.
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
Reviewed-by: Daniel P. Berrangé <berrange@redhat.com>
Message-Id: <20210421112834.107651-3-dgilbert@redhat.com>
Signed-off-by: Dr. David Alan Gilbert <dgilbert@redhat.com>
